From dubreus at csit.fsu.edu  Tue Jun 11 11:32:24 2002
From: dubreus at csit.fsu.edu (mitsch)
Date: 11 Jun 2002 08:32:24 -0700
Subject: compiling error: undefined symbol: __gxx_personality_v0
Message-ID: <93523ebb.0206110732.791e745@posting.google.com>

Hi,
I try to wrap C functions in Python.
when I compile, here's the result:
Traceback (most recent call last):
  File "./boost.py", line 2, in ?
    import getting_started1
ImportError: ./getting_started1.so: undefined symbol:
__gxx_personality_v0


can somebody tell why this error?

here's my files:
getting_stated!.cpp :
#include <boost/python/class_builder.hpp>
namespace python = boost::python;
#include <string>

namespace { // Avoid cluttering the global namespace.

  // A couple of simple C++ functions that we want to expose to
Python.
  std::string greet() { return "hello, world"; }
  int square(int number) { return number * number; }
}

BOOST_PYTHON_MODULE_INIT(getting_started1)
{
    // Create an object representing this extension module.
    python::module_builder this_module("getting_started1");

    // Add regular functions to the module.
    this_module.def(greet, "greet");
    this_module.def(square, "square");
}





my compile command:
g++3 -fpic  -g  -I/usr/include/python2.1
-I/usr/local/lib/python/boost_1_28_0/  -c ./getting_started1.cpp
gcc -shared -lc  getting_started1.o  -o getting_started1.so


From rdacker at pacbell.net  Thu Jun 13 21:39:56 2002
From: rdacker at pacbell.net (rdack)
Date: 13 Jun 2002 18:39:56 -0700
Subject: urllib question
References: <644f6688.0206101835.5dc793b1@posting.google.com>
Message-ID: <644f6688.0206131739.6178e286@posting.google.com>

rdacker at pacbell.net (rdack) wrote in message news:<644f6688.0206101835.5dc793b1 at posting.google.com>...
> this retrieves data from my router:
>   params = urllib.urlencode({'RC': '@D', 'ACCT' : "root", 'PSWD' :
> "71:29:26", 'URL': 'admin' })
>   ipurl = "http://" + iphost + "/cgi-bin/logi"
>   urlfp = urllib.urlopen(ipurl, params)
> 
is there some way i can what exactly this is posting?
a way to get the post object out of urlib?
it should be something like:

POST http://192.168.0.1/cgi-bin/logi
Content-Length: 47
Content-Type: application/x-www-form-urlencoded

RC=%40D&PSWD=71%3A29%3A26&URL=admin&ACCT=root


From jpapa at websense.com  Fri Jun 14 16:06:53 2002
From: jpapa at websense.com (J. Papa)
Date: 14 Jun 2002 13:06:53 -0700
Subject: Graceful socket shutdown
Message-ID: <9e180f21.0206141206.1f4d8c78@posting.google.com>

I am posting this because I just spend too many hours
debugging and I wanted to save someone else the angst.

According to MSDN the proper way to close a winsock connection
is to:

1. shutdown( SD_SEND );
2. call recv in a loop until it returns 0 or SOCKET_ERROR
3. close()


So in Python:

# This is not robust.
# s.close()

# This is, hideous, but correct for Win32
s.shutdown( 1 )
while 1:
	try:
		byte = s.recv( 1 )
		if "" == byte:
			break
	except:
		break
s.close()


The only time you will see this bug is if your recv buffer still has
data in it
when you try and close() the socket. This happened to me because my
client
was a wininet app. Wininet, for some mysterious reason, appends an
extra "\r\n" to POSTed data. (I swear I sniffed it, its true.) This
extra "\r\n" is not included in the Content-Length of the POST. Hence,
there were two extra bytes in the recv buffer when I went to close()
my socket. Instead of sending a FIN a RST is sent instead that gets
ugly.

Regards,
J. Papa


From quinn at retch.ugcs.caltech.edu  Thu Jun 27 16:50:59 2002
From: quinn at retch.ugcs.caltech.edu (Quinn Dunkan)
Date: 27 Jun 2002 20:50:59 GMT
Subject: RE strings (was: Variable Interpolation - status of PEP 215)
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <4PfQ8.44245$n4.10307683@newsc.telia.net> <KhiQ8.44269$n4.10312432@newsc.telia.net> <29e28c51.0206201156.7e61a7f5@posting.google.com> <3D127B89.EE1C17F4@engcorp.com> <29e28c51.0206210944.400eaedc@posting.google.com>
Message-ID: <slrnahmulj.cqr.quinn@retch.ugcs.caltech.edu>

On 21 Jun 2002 10:44:50 -0700, Cimarron Taylor <cimarron+google at taylors.org>
wrote:
>   import re
>   def parseline3(line):
>      m = re.match(r'^EMP:([^,]*),([^,]*),([^,]*),([^,]*)$', line)
>      if m:
>          ...
>
>
>Unless I'm mistaken about how the re module works, parseline3 will
>recompile the regex each time it is called.  parseline2 avoids this

You are.  Mistaken, that is :)

Take a look at sre.py

Of course, if you did 20 regexes in the meanwhile, it would be recompiled.

I generally take the parseline2 approach anyway, though.


From mgerrans at mindspring.com  Sun Jun 23 04:13:15 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Sun, 23 Jun 2002 01:13:15 -0700
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <3D14B7B1.7040602@ix.netcom.com>
Message-ID: <af3vv6$qr9$1@slb6.atl.mindspring.net>

> ...That's the hard lesson that has been learned about
> safety systems, is that building in six layers of safety to protect
> workers from the consequences of their actions will in fact train
> workers to get all six wrong at once.

Ah, but that is also what gives us the most entertaining Darwin Awards!  :-)




From peter at engcorp.com  Sun Jun  2 02:12:00 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 02 Jun 2002 02:12:00 -0400
Subject: What does sys.exit(1) means?
References: <adcb22$100q9d$1@ID-49758.news.dfncis.de>
Message-ID: <3CF9B730.4D160FDB@engcorp.com>

Ken wrote:
> 
> What does sys.exit(1) means?

You can find this if you Read the Fine Manual at 
http://www.python.org/doc/current/lib/module-sys.html .  There is similar
documentation for all functions in all standard modules, in case you have
similar questions in the future...

sys.exit([arg]) 
     Exit from Python. This is implemented by raising the SystemExit exception, 
     so cleanup actions specified by finally clauses of try statements are honored, and 
     it is possible to intercept the exit attempt at an outer level. The optional 
     argument arg can be an integer giving the exit status (defaulting to zero), or 
     another type of object.  If it is an integer, zero is considered ``successful 
     termination'' and any nonzero value is considered ``abnormal termination'' by shells 
     and the like. Most systems require it to be in the range 0-127, and produce 
     undefined results otherwise. Some systems have a convention for assigning specific 
     meanings to specific exit codes, but these are generally underdeveloped; Unix programs 
     generally use 2 for command line syntax errors and 1 for all other kind of errors. 
     If another type of object is passed,  None is equivalent to passing zero, and any 
     other object is printed to sys.stderr and results in an exit code of 1. In 
     particular, sys.exit("some error message") is a quick way to exit a program when 
     an error occurs. 

-Peter


From h.baumgartl at chello.NOSPAM.nl  Fri Jun 21 09:56:10 2002
From: h.baumgartl at chello.NOSPAM.nl (Henry Baumgartl)
Date: Fri, 21 Jun 2002 13:56:10 GMT
Subject: application development - separating content from function
References: <yroQ8.260949$i6.17986369@amsnews02.chello.com>
Message-ID: <_fGQ8.280952$i6.18906908@amsnews02.chello.com>

Thanks very much for the input!

I do believe it has helped me to take off my blinders, look around and
realise what it is all about.

Going for the ultimate design and functionality prize may well have
(indefinitely) delayed actually getting the project under way.

How was it again, about the importance of the journey rather than the
destination? ;-?

Best regards,

Henry




From brian at sweetapp.com  Mon Jun 24 22:03:43 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Mon, 24 Jun 2002 19:03:43 -0700
Subject: Suggestions for good programming practices?
In-Reply-To: <3D17CC8E.ABB72D14@engcorp.com>
Message-ID: <013901c21bec$8a950910$bd5d4540@Dell2>

Peter Hansen wrote: 
> Brian Quinlan wrote:
> >
> > Peter Hansen wrote:
> > > You _are_ returning them "in a struct".  It's spelled "tuple"
> > > but other than that what's the difference?
> >
> > struct members are access by name while tuple elements are accessed
> > positionally.
> 
> Okay.  So why is that a problem?

I'm not saying it is. I'm saying that tuples and structs are different.

Cheers,
Brian





From tchur at optushome.com.au  Sat Jun  8 17:51:23 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Sun, 09 Jun 2002 07:51:23 +1000
Subject: Confidence intervals, stats modules?
References: <ug2r3aac4etnef@corp.supernews.com> <adt0hs$3bl$1@news7.svr.pol.co.uk>
Message-ID: <3D027C5A.E5D8B61E@optushome.com.au>

Duncan Smith wrote:
> 
> "Stephen Boulet" <stephen at theboulets.net> wrote in message
> news:ug2r3aac4etnef at corp.supernews.com...
> > Are there any stats modules to calculate things like confidence intervals,
> > mean, standard deviation? Thanks.
> >
> > -- Stephen
> 
> http://www.nmr.mgh.harvard.edu/Neural_Systems_Group/gary/python.html
> 

Gary's stats module is excellent and is probably sufficient for the uses
you
describe. There are also some elementary statistical functions in Konrad
Hinsen's
equally excellent Scientific Python package - see
http://starship.python.net/crew/hinsen/scientific.html

In fact, you could just use NumPy (Numerical Python) - have a look at
the MLab 
module in NumPy (which is at http://www.pfdubois.com/numpy/ ).

However, if you are interested in more advanced statistical analysis
and/or in 
drawing statistical graphs, then have a look at RPy, by Walter Moreira, 
at http://rpy.sourceforge.net

RPy embeds the R stats environment within Python. R is good for most
things
statistical, from the elementary to the advanced and experimental, and
it does
truly beautiful graphics. It is a mature project backed by a number of
eminent 
statisticians in five continents.

Note that R is multi-platform (see http://www.r-project.org ) but RPy
only
works under Linux and Unix (and Mac Os X, perhaps) at this stage. I
believe 
a Windows version is in the works.

Tim C




From senux at senux.com  Sun Jun  2 01:09:09 2002
From: senux at senux.com (Brian Lee)
Date: Sun, 2 Jun 2002 14:09:09 +0900
Subject: writing to a file (atomic operation)
Message-ID: <20020602050909.GA7192@mercury.senux.com>

Is this a atomic operation?

open('/tmp/test.log', 'a').write('test\n')

from py newbie.

-- 
 ____       
 |o | i   / Brian Lee    <senux at senux.com>
 #######|<  Web - http://www.senux.com/en/
(x-x-x-x) \ GnuPG public key - 0x46C763A3




From peter at engcorp.com  Tue Jun 18 21:09:09 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 18 Jun 2002 21:09:09 -0400
Subject: xml fo->pdf processor wanted
References: <slrnagu96r.9jc.johannes@orbital.zellner.org>
Message-ID: <3D0FD9B5.5A91B1F4@engcorp.com>

Johannes Zellner wrote:
> 
> Hi,
> 
> is there a fo (formatting object) to pdf converter in python?

I'm pretty sure there is not (yet).  You could probably just
call out to Java and run FOP though.  http://xml.apache.org/fop

It would be nice to see something put together from PyXML and
ReportLab.  I'm a little surprised no enterprising (and totally
psychotic :-) soul has announced such a thing yet...

-Peter


From logiplexsoftware at earthlink.net  Mon Jun 24 13:29:46 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 24 Jun 2002 10:29:46 -0700
Subject: How to represent the infinite ?
In-Reply-To: <3d160e45@news.swissonline.ch>
References: <3d11ee9b$1@news.swissonline.ch>
	<mailman.1024600723.6989.python-list@python.org>
	<20020620.124436.1266235189.32516@software1.logiplex.internal>
	<3d160e45@news.swissonline.ch>
Message-ID: <20020624102946.5dbf92f5.logiplexsoftware@earthlink.net>

On Sun, 23 Jun 2002 20:06:37 +0200
erreur wrote:

> Thanks.... excelent !!

Thanks for replying.  Sometimes when no one responds to the bits of code I post
I worry that perhaps people don't say anything because my code was just too
lame to comment on, but then I remember that it's really because I'm a total
<expletive/> and I feel better ;)


> "Cliff Wells" <logiplexsoftware at earthlink.net> wrote in message
> news:20020620.124436.1266235189.32516 at software1.logiplex.internal...
> > In article <mailman.1024600723.6989.python-list at python.org>, "Cliff Wells"
> > <logiplexsoftware at earthlink.net> wrote:
> >
> >
> >
> > > On Thu, 20 Jun 2002 17:02:33 +0200
> > > erreur wrote:
> > >
> > >> Will somebody have an idea, to represent the infinite one?  I have
> > >> variables to initialize with is +inf (or - inf).  To be sure that
> > >> later, all will be smaller (or larger) than my variables. I tried to
> > >> redefine the operators on an object.  It goes for Inf>10 but I do not
> > >> arrive for 10>Inf (because that takes > of Int).  --------------------
> > >> def __gt__(self, val):

[snip]

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From dreed at capital.edu  Sun Jun 23 12:06:05 2002
From: dreed at capital.edu (Dave Reed)
Date: Sun, 23 Jun 2002 12:06:05 -0400
Subject: (Newbie)  Counting Instances ("Hits") with Regular Expressions
In-Reply-To: <nbmR8.300036$cQ3.15150@sccrnsc01> (emile@fenx.com)
References: <3d15e215.43448155@news.texas.net> <nbmR8.300036$cQ3.15150@sccrnsc01>
Message-ID: <200206231606.g5NG65117471@localhost.localdomain>

> From: "Emile van Sebille" <emile at fenx.com>
> 
> "Ben Fairbank" <baf at texas.net> wrote in message
> news:3d15e215.43448155 at news.texas.net...
> > I am new both to Python and to regular expressions, which may account
> > for my difficulty.  I must count the frequenies of certain words in
> > files of moderate length (about150 k bytes).  I have been reading
> > files and then using count(s,sub), which is fast and easy.  I now have
> > to allow for punctuation and eliminate words within words, etc, and so
> > am trying to use regular expressions instead of simple words as
> > targets.  I do not, however, find a similarly easy to use count
> > function in the re module.  Yet this is such common operation it must
> > be there, or easy to implement.  What is the usual way of simply
> > counting "hits" in the re module?  (And what have I missed in the
> > documentation; where is this to be found?  I have looked through Lutz
> > and Ascher)
> >
> 
> >>> s = 'spam and eggs and ham and eggs and spam and eggs'
> >>> s.count('am')
> 3
> >>>

I don't think he wants to count those, only ' am '

There is a much easier solution than what I originally posted:

import re
s = 'the word is the word but not a keyword'
re.findall('\sword\s')

outputs: [' word ', ' word ']
so len(re.findall('\sword\s') will do what he wants

Dave




From bokr at oz.net  Sat Jun 22 14:43:50 2002
From: bokr at oz.net (Bengt Richter)
Date: 22 Jun 2002 18:43:50 GMT
Subject: Web templating/db tool with best designer/coder separation?
Message-ID: <af2gh6$2c2$0@216.39.172.122>

E.g., if you were hoping for an improvement over current practice,
which do you consider to be the tool(s) to beat, and what would
the killer improvement(s) be? TIA.

Regards,
Bengt Richter


From gerhard at bigfoot.de  Fri Jun  7 10:09:16 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 7 Jun 2002 16:09:16 +0200
Subject: Native code compiler for python?
In-Reply-To: <baf24bd4.0206070550.84f1376@posting.google.com>
References: <baf24bd4.0206070550.84f1376@posting.google.com>
Message-ID: <20020607140916.GA11037@lilith.my-fqdn.de>

* nobody <d2002xx at softhome.net> [2002-06-07 06:50 -0700]:
> Is there any existing native code compiler for python?

<hint>Have you checked groups.google.com?</hint>

> Or Python to C translator? Or someone is doing this?

Yes, several, Pyrex among others. But they only compile a subset of
Python.

Is this question just motivated by interest or do you have any problem
with just using the Python interpreter?

Do you only want to have a standalone app without requiring an installed
interpreter? freeze, McMillan's Installer and py2exe are solutions for
_this_ problem.

The solution for slow Python code is using a better algorithm, and if
this is not possible, writing the innermost loops in C/C++/Fortran/Ada
(even Ocaml is possible, IIRC). Using Pyrex might become a possible
solution, too.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From edream at tds.net  Tue Jun 25 11:04:56 2002
From: edream at tds.net (Edward K. Ream)
Date: Tue, 25 Jun 2002 15:04:56 GMT
Subject: ConfigParser & converting strings to lists
References: <3D152A28.2177B054@tds.net> <3D159F06.2154D3F3@tds.net> <Xns9236975A47F67cliechtigmxnet@62.2.16.82> <af5p6l0pf2@enews4.newsguy.com>
Message-ID: <3D188696.1694EE9F@tds.net>

> P.S. Do you have an Erdos number? I've never talked to anyone who actually
> has one (that I know of).

One of my teachers, Chuck MacCluer, http://www.mth.msu.edu/~maccluer/,
has an Erdos number of 1, I think. He once distributed a small proof of
mine, but it almost certainly never got published.  Thanks for asking.

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


From whisper at oz.net  Mon Jun 17 19:15:40 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 17 Jun 2002 16:15:40 -0700
Subject: Medusa Questions
Message-ID: <GCEDKONBLEFPPADDJCOEKEMBDJAA.whisper@oz.net>

I'm using the latest Sourceforge distro of Medusa and have some questions.

Is there any documentation on writing extensions?

Is there any documentation on using the demos, in particular the
"script_server"?

Is the asynccore C extension still usable?

BTW, I would have asked on the Yahoo group for Medusa, but didn't want to
subscribe to a group that's literally 95%+ spam. You ought to set your group
to subscribers-only posting.

TIA,

David LeBlanc
Seattle, WA USA





From aahz at pythoncraft.com  Fri Jun 28 19:11:55 2002
From: aahz at pythoncraft.com (Aahz)
Date: 28 Jun 2002 19:11:55 -0400
Subject: squared functions--most Pythonic way?
References: <ballabio-E54F22.21380327062002@newssrv.fastweb.it>
Message-ID: <afiqfr$9re$1@panix1.panix.com>

In article <ballabio-E54F22.21380327062002 at newssrv.fastweb.it>,
Luigi Ballabio  <ballabio at mac.com> wrote:
>
>      suppose we want to define a function which, given f(x), returns 
>its square. There are a number of ways which come to mind, such as:
>
>object-oriented programming:
>
>    def fsquare:
>        def __init__(self,f):
>            self.f = f
>        def __call__(self,x):
>            return self.f(x)**2
>
>functional programming:
>
>    # with an internal named function
>    def fsquare(f):
>        def f2(x):
>            return f(x)**2
>        return f2

I'm sensing an ill-defined spec here.  To me, your English sentence
implies that what's wanted is

    def fsquare(f, x):
        return f(x)**2

but both of your examples define closures of some sort.

Assuming that your code does what you really want, I don't think either
of these two is "more Pythonic"; it depends on your circumstances and
who your likely code readers are going to be.  I don't have much
functional experience, so the OO mechanism works better for me
(particularly given that it works back to 1.5.2).
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From gerhard.haering at gmx.de  Fri Jun 21 03:59:38 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 21 Jun 2002 09:59:38 +0200
Subject: compile python witout tkinter
In-Reply-To: <Pine.LNX.4.44.0206210139380.585-100000@aesaeion.com>
References: <20020621074746.GA765@lilith.my-fqdn.de> <Pine.LNX.4.44.0206210139380.585-100000@aesaeion.com>
Message-ID: <20020621075938.GA896@lilith.my-fqdn.de>

* kosh at aesaeion.com <kosh at aesaeion.com> [2002-06-21 01:40 -0600]:
> On Fri, 21 Jun 2002, Gerhard [iso-8859-15] H?ring wrote:
> 
> > * kosh at aesaeion.com <kosh at aesaeion.com> [2002-06-21 01:26 -0600]:
> > > On a machine I need to build python on I need to build it without tkinter.
> >
> > Why? Even if Tkinter gets compiled, it doesn't change anything for the
> > rest. You could even delete the tkinter modules afterwards, if you're
> > short on disk (quota).
> 
> Because the libraries it depends on are not installed on that machine and
> there is no point installing them since they are not needed.

No problem, then. It won't get built if the dependencies are not met.
As seen in Modules/Setup.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 23.7 ?C      Wind: 4.0 m/s




From mjackson at wrc.xerox.com  Thu Jun 27 09:35:21 2002
From: mjackson at wrc.xerox.com (Mark Jackson)
Date: 27 Jun 2002 13:35:21 GMT
Subject: question od default args
References: <o3vjhuguen6gfm8vnq7rbk5volmi80shl5@4ax.com> <YynS8.13571$p56.4313990@newsb.telia.net>
Message-ID: <aff4ap$aqm$2@news.wrc.xerox.com>

"Fredrik Lundh" <fredrik at pythonware.com> writes:
> Gon?alo Rodrigues wrote:
> 
> > When I want default args I usually do
> >
> > def (arg = None):
> >     ...
> >
> > But what if I want to make None a reasonable argument too? That is, I
> > want to know if the user has in fact passed an argument, and if not to
> > do something special, with None (or whatever object) being a reasonable
> > arg to be passed.
> 
> hmm.  wasn't this just discussed in some thread on a
> newsgroup near you?
> 
> three alternatives:
> 
>     __UNDEF__ = [] # guaranteed to have a unique id
> 
>     def myfunc(arg=__UNDEF__):
>         if arg is __UNDEF__:
>             print "no argument"
>         else:
>             ...

More quickly:

    def myfunc(arg=[]):
        if arg is myfunc.func_defaults[0]:
            print "no argument"
        else:
            ...

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	Until we test our beliefs, we can't say for sure if we have
	leeches or we have aspirin.	- David Faigman




From sholden at holdenweb.com  Thu Jun  6 09:55:54 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 6 Jun 2002 09:55:54 -0400
Subject: Pipe return mesage from spawnv to string?
References: <3ebec3ea.0206060443.d2cc5bb@posting.google.com>
Message-ID: <cWJL8.313202$%u2.288245@atlpnn01.usenetserver.com>

"Stefhan Stj?rn?s" <stefhan at hotmail.com> wrote in message
news:3ebec3ea.0206060443.d2cc5bb at posting.google.com...
> Hi!
>
> Anyone who knows how to pipe from spawnv?
> I start isql.exe with os.spawnv and must see the result.
>
> run ("isql.exe"," -s ....")
>
> def run (program, *args):
>     """Executes program with arguments, using os.spawnv"""
>     try:
>         return os.spawnv(os.P_WAIT,program, (program,) +args)
>     except (os.error), why:
>         raise SErr, str(why)
>

Well in that case you want popen*(), not spawn*(). Whether you use popen(),
popen2(), popen3() or popen4() will depend on your needs.

if-you-need-a-pipe-then-open-one-ly y'rs  - steve
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From ark at research.att.com  Fri Jun  7 13:47:37 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 7 Jun 2002 13:47:37 -0400 (EDT)
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
	(message from Skip Montanaro on Fri, 7 Jun 2002 11:42:19 -0500)
References: <yu99vg8v6orb.fsf@europa.research.att.com> <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
Message-ID: <200206071747.g57HlbH01057@europa.research.att.com>

PPS: I tried building one of the .so files by hand, essentially by
snarfing the output from the Makefile log and executing it manually:

    $ gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/tmp/build-python/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /tmp/build-python/Python-2.2.1/Modules/structmodule.c -o build/temp.solaris-2.8-sun4u-2.2/structmodule.o
    $ gcc -shared build/temp.solaris-2.8-sun4u-2.2/structmodule.o -L/usr/local/lib -o build/lib.solaris-2.8-sun4u-2.2/struct.so
    $ ldd build/lib.solaris-2.8-sun4u-2.2/struct.so
        libgcc_s.so.1 =>         /usr/gnu/lib/libgcc_s.so.1
        libc.so.1 =>     /usr/lib/libc.so.1
        libdl.so.1 =>    /usr/lib/libdl.so.1
        /usr/platform/SUNW,Ultra-5_10/lib/libc_psr.so.1

So apparently it is building a .so file with references correctly resolved.

How do I go about finding out the cause of the failure?




From tdelaney at avaya.com  Sun Jun 23 21:50:06 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 24 Jun 2002 11:50:06 +1000
Subject: Python hits the spot
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A3B3@natasha.auslabs.avaya.com>

> From: Peter Hansen [mailto:peter at engcorp.com]
> 
> Aahz wrote:
> > 
> > In article <3d149f19$0$53133$edfadb0f at dspool01.news.tele.dk>,
> > Frihtiof Andreas Jensen <nomailhere at lookitup.yourself.com> wrote:
> > >
> > >My bet is that once you cool off and sits down with some 
> debugging tools
> > >like GDB, Boundschecker or Electric Fence you will indeed find the
> > >programming error and fix it.
> > >
> > >My bet is also that you will not tell anyone either.
> > 
> > Ya know, you could have saved yourself a lot of typing by 
> simply saying
> > PBCAK.  ;-)
> 
> To save us from those who can't Google:  Problem Between 
> Chair And Keyboard.

That's always been PEBCAK to me ...
                    ^
Problem Exists Between Chair and Keyboard

Tim Delaney




From donn at u.washington.edu  Thu Jun 20 13:52:38 2002
From: donn at u.washington.edu (Donn Cave)
Date: 20 Jun 2002 17:52:38 GMT
Subject: How to represent the infinite ?
References: <3d11ee9b$1@news.swissonline.ch> <20020620193230.53d58f8f.christophe.delord@free.fr>
Message-ID: <aet4p6$33kk$1@nntp6.u.washington.edu>

Quoth Christophe Delord <christophe.delord at free.fr>:
|
| There is another simple solution. Just use a float that is bigger than any 64-bit float. For example 10^1000. This float has a special encoding meaning +oo !
| So no need to redefine anything else ;-)
|
| >>> x=1e1000
| >>> x
| inf
...
| >>> x-x
| nan

$ python
Python 2.0 (#1, Dec 18 2000, 10:19:52) [C] on osf1V4
Type "copyright", "credits" or "license" for more information.
>>> x = 1e1000
>>> x
1.7976931348623157e+308
>>> x > 10
1
>>> x - x
0.0
>>>

I missed the original post, hope this helps.

	Donn Cave, donn at u.washington.edu

| > Not need to say to me that I can use 99999999999999 or -99999999999999 or I
| > do not know what.
| > Nor from a MAX_INT of the machine... bus Python can go higher.
| > 
|
|
|
|
| -- 
|
| (o_   Christophe Delord                   _o)
| //\   http://christophe.delord.free.fr/   /\\
| V_/_  mailto:christophe.delord at free.fr   _\_V


From benjl at cse.unsw.edu.au  Sat Jun  1 20:39:07 2002
From: benjl at cse.unsw.edu.au (Benno)
Date: 1 Jun 2002 17:39:07 -0700
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org>
Message-ID: <332c7c11.0206011639.a396f84@posting.google.com>

Andrei Kulakov <ak at silmarill.org> wrote in message news:<slrnafhmoa.389.ak at ak.silmarill.org>...
> Hello,
> 
>     "hello world" type program from wxPython tutorial takes 6
> seconds to start on my cel 366 128mb system (running Debian). Is
> this typical? Why is it so slow? I'm looking for a graphic
> toolkit that'd be fairly quick. I don't want to use Tk because of
> the looks.

I was using wxPython for a while. The main advantage being that it was
going to be cross platform, however I found its speed too slow. (Not startup
speed, which is fairly irrelevant to me, but just general responsiveness.)
I'm not sure if this was just wxWindows itself or the python bindings.

I have started using pygtk <http://www.daa.com.au/~james/pygtk/> (easily
apt-getable). The only problem I've had with this is the lack of docs.
http://www.moeraki.com/pygtktutorial/gtk-tut.html, seems to be the best 
available, followed by reading the source.

Cheers,

Benno


From quinn at retch.ugcs.caltech.edu  Thu Jun 27 13:51:47 2002
From: quinn at retch.ugcs.caltech.edu (Quinn Dunkan)
Date: 27 Jun 2002 17:51:47 GMT
Subject: Questiion on list
References: <pan.2002.06.27.14.36.03.151659.5728@linuxmail.org> <slrnahl7j8.fd3.quinn@upchuck.ugcs.caltech.edu> <pan.2002.06.27.15.32.28.404968.5728@linuxmail.org>
Message-ID: <slrnahmk5g.brh.quinn@retch.ugcs.caltech.edu>

On Thu, 27 Jun 2002 15:32:28 +0000, SiverFish <occeanlinux at linuxmail.org> wrote:
>On Thu, 27 Jun 2002 05:11:06 +0000, Quinn Dunkan wrote:
>
>> On Thu, 27 Jun 2002 14:36:03 +0000, SiverFish
>> <occeanlinux at linuxmail.org> wrote:
>>>I got the list of words now i want to calculate how many word in that
>>>list which is the same word (insensitive case),can anyone show me how to
>>>do it for example ls = ['abc','cd','abc','adf',abc','dwc','cd'] it will
>>>show abc = 3 ,cd = 2, adf = 1, dwc=1
>> 
>> import string
>> d = {}
>> for e in map(string.lower, ls):
>>     d[e] = d.get(e, 0) + 1
>> freqs = [ (count, w) for w, count in d.items() ] freqs.sort()
>> freqs.reverse()
>> print ', '.join([ '%s = %d' %(count, w) for count, w in freqs ])
>
>Could you  do it in python 1.5.2 and explain clearly to me please 

The English version is "use a dictionary".  Go through all the words and put
each one in the dictionary with the value 1.  If the word is already in the
dictionary, increment the value.  When you are done, you have a dict mapping
words to frequency of occurrance.

>How about we have to do the same problem here when read in the multiple
>files,Thank a lot

The words can come from anywhere.  Just write a function that takes a list
of words and returns a histogram dict as above.  Then all you need to do is
convert your list of files into a list of words.

If you have large files and are concerned about memory usage, you could
upgrade to 2.2 and investigate iterators, and pass an appropriate iterator
to the function, or stay with 1.5.2 and rewrite the function to modify a
dict, and then pass in smaller lists consecutively.  Or just lose the function
and inline the whole thing.  Up to you.

d = {}
for fp in map(open, files):
    for line in fp.readlines():
        for w in line.split():
            d[w] = d.get(w, 0) + 1
    fp.close()

Of course, if you have to worry about punctuation and hyphenation, things get
more complicated...


From Backflip at gmx.de  Tue Jun  4 04:12:35 2002
From: Backflip at gmx.de (Backflip)
Date: 4 Jun 2002 01:12:35 -0700
Subject: Printing a text centered, it would be a newbie prob
Message-ID: <c7bfc769.0206040012.20aa0a10@posting.google.com>

Hello World (= 
I?ve got a little problem, i want to put out an import string centert. For example:
for word in raw_input("bitte hier ihren text eingeben:").split():
   print word      

text=dies ist ein test

now the output looks like this

dies
ist
ein
test

but i want that it looks like this one

dies ist
ein test 

that both rows are simular to each other, could smb help me? 
Thanks
Backflip


From tim.one at comcast.net  Fri Jun 21 18:19:25 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 21 Jun 2002 18:19:25 -0400
Subject: How to represent the infinite ?
In-Reply-To: <20020621183459.63fe2e4d.christophe.delord@free.fr>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEDIPPAA.tim.one@comcast.net>

[Christophe Delord]
> ...
> But addition, multiplication and exponentiation seem to be implemented
> differently (1e300+1e300 is inf and 1e300**2 is an overflow).

Python float + and * are done via C + and * on C doubles.  Python float **
is done via C libm pow(), and

    x**y

strives to act the same as

    import math
    math.pow(x, y)


> Is there a reason for these different behaviours?

Of course.  The question is whether they're rational reasons <0.9 wink>.





From syver-en+usenet at online.no  Mon Jun  3 19:19:03 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Mon, 03 Jun 2002 23:19:03 GMT
Subject: python and windows and pydoc and favorite editor == True
References: <un0ugnxa3.fsf@online.no> <adggk2$psv$00$1@news.t-online.com>
Message-ID: <uznycc5om.fsf@online.no>

"J?rgen Hermann" <jh at web.de> writes:

> "Syver Enstad" <syver-en+usenet at online.no> schrieb im Newsbeitrag
> news:un0ugnxa3.fsf at online.no...
> > Post if there is any interest.
> 
> OK, post. :)
> 
> 

http://home.online.no/~syver-en/software.html/files/PyDocEditorStarter.py




-- 

Vennlig hilsen 

Syver Enstad


From donn at u.washington.edu  Tue Jun  4 17:35:34 2002
From: donn at u.washington.edu (Donn Cave)
Date: 4 Jun 2002 21:35:34 GMT
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <3CFCF3E7.AE891EBE@noaa.gov> <adiumj$22g2$1@nntp6.u.washington.edu> <Mx9L8.239704$%u2.17455@atlpnn01.usenetserver.com>
Message-ID: <adjbr6$1j7k$1@nntp6.u.washington.edu>

Quoth "Steve Holden" <sholden at holdenweb.com>:
| "Donn Cave" <donn at u.washington.edu> wrote ...
...
|> Not really - the current semantics are actually better than the fixed
|> version, inasmuch as they are at least consistent.  As I interpret the
|> proposal, the fixed += would sometimes work without rebinding (if the
|> object supports it) and sometimes would be an assignment (falling back
|> to +.)  To me, to allow an object to dictate its relationship to its
|> computational environment like that is anathema.
|
| >>> a = b = (1, 2, 3)
| >>> a += ('boo!',)
| >>> a, b
| ((1, 2, 3, 'boo!'), (1, 2, 3))
| >>> a = b = [1, 2, 3]
| >>> a += ['boo!']
| >>> a, b
| ([1, 2, 3, 'boo!'], [1, 2, 3, 'boo!'])
|
| Doesn't look that consistent to me. Add in the fact that implementors of
| mutable objects are free to choose either semantic and what you have is ...
| exactly what Guido intended. I regard it as a wart, but not one that worried
| me.

Well, it is consistent.  Maybe not in a way that you can see
and still remain sane.  "Mutable" or not, it must always work
like "a = iadd(a, b)", where iadd attempts to perform some
computation like __iadd__() or eventually "+".  The fact that
__iadd__ may return any imaginable value is no queerer than
the same fact about __add__.

I'm not saying it's good, on the contrary, it's terrible.  I just
don't think it sounds very good to fix it by changing the rules
so it sometimes means the above, and sometimes omits the assignment.
As long as it has to work on values that don't support it as a
mutating __iadd__ or equivalent, it should probably stay the way
it is.

	Donn Cave, donn at u.washington.edu


From johnroth at ameritech.net  Mon Jun 24 07:36:42 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 24 Jun 2002 07:36:42 -0400
Subject: Generic Python
References: <3D16E465.F0D2509E@ifib.uni-karlsruhe.de>
Message-ID: <uhe1069gs1f4fc@news.supernews.com>

"Uwe Mayer" <Uwe.Mayer at ifib.uni-karlsruhe.de> wrote in message
news:3D16E465.F0D2509E at ifib.uni-karlsruhe.de...
> Hi,
>
> is it possible to to write a class which, f.e. takes an argument in
> __init__() and that doesn't return an instance object but a new class
> object?

Sure, but you have to do it closer to the source level. One
quibble here: you can't do it with the __init__() method, that's
defined to always return an instance. You have to do it with some
kind of generator or factory function.

The way you do it is to create the source in a string, and
then use exec.

However, as another poster mentions, there may be better
ways of solving your specific problem. Dynamically generating
pieces of program is useful in some highly specialized
circumstances, but this might not be one of them.

John Roth





From mwh at python.net  Fri Jun 14 06:09:31 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 14 Jun 2002 10:09:31 GMT
Subject: __subclasses__
References: <a9bcagho74.22t.jimrowe@optilink.com>
Message-ID: <lk4rg6matg.fsf@pc150.maths.bris.ac.uk>

jimrowe at optilink.com (James Rowe) writes:

> A search for __subclasses__ on the www.python.org
> documentation search pages turns up empty. I can
> find not mention of it in my Python books or even
> in the "What's new in Python 2.2" document.  But:
> 
> class Foo(object):
>     def __init__(self):
>         object.__init__(self)
> 
> class Bar(Foo):
> 	pass
> 
> dir(Foo) # returns:
> 
>  ['__class__', '__delattr__', '__dict__', '__doc__',
>  '__getattribute__', '__hash__', '__init__', '__module__',
>  '__new__', '__reduce__', '__repr__', '__setattr__',
>  '__str__', '__weakref__']

Whether things show up in dir() is a bit of a lottery (think about
__getattr__ methods!).  Notice __bases__ isn't in there either, for
example.

> But there is a __subclasses__ method:
> 
> Foo.__subclasses__() # returns:
> 
>  [<class '__main__.Bar'>]
> 
> 
> So I suppose the question is: Why is __subclasses__ not documented
> or listed?

Dunno.  Lack of time would be my guess.  Lots of this stuff isn't
documented.

> Can we safely use __subclasses__ and not worry about it "going away"
> in the future?

Well the implementation needs to have access to the functionality it
provides (I think this is for situtations like:

class a(object):
 pass
class b(a):
 pass

a.__getitem__ = some_meth

b()[1] should now invoke some_meth

but I'm not sure).  This doesn't mean that the current exposure of the
functionality to Python code is not going to disappear or change in
spelling, but I'd say it was pretty unlikely.

Cheers,
M.

-- 
  It's actually a corruption of "starling".  They used to be carried.
  Since they weighed a full pound (hence the name), they had to be
  carried by two starlings in tandem, with a line between them.
                 -- Alan J Rosenthal explains "Pounds Sterling" on asr


From rkr1410 at poczta.onet.pl  Sun Jun 23 12:32:58 2002
From: rkr1410 at poczta.onet.pl (hellboy)
Date: Sun, 23 Jun 2002 18:32:58 +0200
Subject: Newbie question (anyone programming with curses?)
References: <af4ovd$1bt3$1@pingwin.acn.pl> <ummR8.290538$352.27434@sccrnsc02>
Message-ID: <af4t70$1igl$1@pingwin.acn.pl>

Emile van Sebille <emile at fenx.com> wrote:

> No one has made a Windows port of the curses module. On a Windows
> platform, try the Console module

8<--cut--8<

So there is no way to get fast console input/output + colors in text mode
and code portability?

btw. there actually is a windows implemetation of curses
(http://pdcurses.sourceforge.net), but it is written in c (of course) and I
am too inexperienced to say if there is some way of incorporating it into
python, and how to do it....




From pyth at devel.trillke.net  Fri Jun  7 19:46:58 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 8 Jun 2002 01:46:58 +0200
Subject: Efficient python programming...
In-Reply-To: <3D013F11.F25FFA72@engcorp.com>; from peter@engcorp.com on Fri, Jun 07, 2002 at 07:17:37PM -0400
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com> <mailman.1023452349.14713.python-list@python.org> <3D013F11.F25FFA72@engcorp.com>
Message-ID: <20020608014658.G6609@prim.han.de>

Peter Hansen wrote:
> holger krekel wrote:
> > Peter Hansen wrote:
> > > You forgot the even more important first thing for a beginner:
> > > get it correct!  The only good way to do this is to write unit
> > > tests that prove it.  If you don't have tests, you don't have
> > > working code (or how can you prove it?).
> > 
> > yeah. writing tests is *the way* to go if you want other people use
> > your code. Although successfull unittests theoretically don't prove
> > anything<wink> they tend to do in reality<doublewink>.
> 
> Could you explain why you think unittests don't prove anything 
> when they are successful?

I meant to play with the academic notion ('theoretically') of 
proving algorithms to be correct.  From a stricter perspective
unittests are just examples and examples do not prove anything. 
But 'in reality' they do prove that intended usage of an API 
basically works and nothing broke while fixing bugs or adding 
new features. 
 
> Are you pointing out that if you write a test after the code
> and the test passes, you have learned almost nothing?  I agree
> with that.
> 
> With Test-Driven Development, however, you write the test first,
> run it immediately, and ensure that it fails.  Only after you
> have code that makes it pass are you done.
> 
> In this case, making sure the test fails when the code is wrong,
> wouldn't you say that the test mostly proves that the code works?

*Mostly*, yes.  Unfortunately mathematicians impose a boolean
view regarding proofs on us poor coders: proven or not proven. 
 
> (Not quibbling about the fact you are still not 100% sure...)

which was the whole point of my posting :-)

    holger




From jepler at unpythonic.net  Tue Jun 25 10:10:36 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 25 Jun 2002 09:10:36 -0500
Subject: Thread Memory Leak
In-Reply-To: <pan.2002.06.25.12.37.12.535288.17223@pobox.com>
References: <pan.2002.06.25.12.37.12.535288.17223@pobox.com>
Message-ID: <20020625141035.GC4435@unpythonic.net>

On Tue, Jun 25, 2002 at 09:37:12AM -0300, Marco Catunda wrote:
> Hello,
> 
> I've developed a python program with some threads. 
> This program is a daemon server. I have seen the
> memory of that program growing very fast in a short 
> period of time. So I decided to make a little program
> to see what's happen.

I see memory grow slowly over time on a RedHat 7.3 system with Python
2.3a0 CVS.  The memory numbers are erratic, but at the start they are in
a range 3000-13000 VSZ.  At the end of the run I did, the numbers seemed
to have grown to 6000-18000 VSZ.

I did not see this growth on a RedHat 6.2 system with Python 2.1.

Since both the OS and Python versions varied between these two systems,
I can't say which is at fault.

Jeff




From d2002xx at myrealbox.com  Wed Jun 19 02:54:21 2002
From: d2002xx at myrealbox.com (d2002xx)
Date: 18 Jun 2002 23:54:21 -0700
Subject: Create a JIT compiler for python
Message-ID: <8d3f4438.0206182254.203fce42@posting.google.com>

Hello!

I want to create a jit compiler for python. Although I don't yet
register the CVS and website, if you want to join or give me any
opinions, please e-mail to me: mailto:d2002xx at myrealbox.com

*NOT* send to the USENET, because my ISP doesn't provide me NNTP
server, using Google is too slow.


From theller at python.net  Wed Jun 19 10:28:11 2002
From: theller at python.net (Thomas Heller)
Date: Wed, 19 Jun 2002 16:28:11 +0200
Subject: imp.loadmodule creates compiled file
References: <zv%P8.332187$o66.856939@news-server.bigpond.net.au>
Message-ID: <aeq4hu$92bi4$1@ID-59885.news.dfncis.de>

"Neil Hodgson" <nhodgson at bigpond.net.au> wrote in message news:zv%P8.332187$o66.856939 at news-server.bigpond.net.au...
>    I've been using imp.loadmodule with Python 2.2 (on both Windows and
> Linux) to load changed versions of source code into a long running process
> where they will run alongside objects already created from earlier versions
> of the source code. An unexpected result of the call is to create a compiled
> module on disk with a name based on the filename argument with "c" appended.
>

There has been a thread named "Importing a module without creating a .pyc/.pyo file?"
recently. I'm not sure this link will work, otherwise search google yourself.
http://groups.google.com/groups?hl=de&lr=&ie=UTF8&oe=UTF8&safe=off&th=5a6c36264f7b6d07&rnum=65

IIRC, there have been several useful suggestions.

Thomas




From david.abrahams at rcn.com  Fri Jun 14 21:53:01 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Fri, 14 Jun 2002 21:53:01 -0400
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de>
Message-ID: <aee6m9$qop$1@bob.news.rcn.net>

"Michael 'Mickey' Lauer" <mickey at tm.informatik.uni-frankfurt.de> wrote in
message news:3d0a8b6e at nntp.server.uni-frankfurt.de...


> I had a quick look at the other wrappers (SILOON, GRAD, CXX, etc.)
> but most look either outdated or not appropriate for this task.
> Do you know an (automatic) wrapper generator which has this feature?
>
> I know sip can do this for c++ libraries - can sip also be used
> to wrap plain c-libs without "handwriting" code?

On most platforms, Boost.Python would work for you. The determining issue is
whether or not the C++ compiler treats "C" and "C++" linkage function
pointers the same way. If the appended program will compile with your C++
compiler, you're probably OK:

-Dave

extern "C" void f() {}
void g(void (*)());
int main()
{
    g(f);
    return 0;
}





From opus at value.net  Sun Jun 30 03:40:22 2002
From: opus at value.net (Opus)
Date: Sun, 30 Jun 2002 00:40:22 -0700
Subject: Python needs better error reporting
In-Reply-To: <3D1EAF12.84CE5914@cascade-sys.com>
Message-ID: <3D1E5376.32716.DF5C94@localhost>

I would normally support a more verbose parser for a language.  I fully believe 
that if the thing can tell you there is an error, it should be able to show 
some proof, and help you out.  This being said, the fact that this is not a 
compiled language (it still needs to read the source file to make sure it is up 
to date), a verbose parser would slow it down.

At the most, I might suggest that there be an option to run just the parser, 
with more verbose output.  Maybe even a different parser would be in order.  
Idea for anyone????

True, the language definition of Python is nice and clean, it would be nice to 
see a pre-parser written for it, in it.  If it passes, it could even call 
Python to write out a .pyc file.  Maybe even an option for the parser to spot 
bad-practice code.  I.E. Is is faster to call 'var += 1' or 'var = var + 1'.

Maybe even one that gives you O analysis of your code.  That would be really 
nice.  Could not only play with some code to opomize speed before you use it.

On 30 Jun 2002 at 0:11, James J. Besemer wrote:

> 
> I rather think Dave has a good point.
> 
> A better parser could give a more informative diagnostic in this case and I
> suspect others.
> 
> Tim Peters wrote:
> 
> > Well, "xor" isn't legit there, but other tokens that are OK after
> >
> >     if s == ' '
> >
> > include
> >
> >    [a ong list]
> 
> The question is NOT be what all may follow the string literal.  Rather it should
> be what tokens are legal between the string literal and end of line?  Very
> different questions.
> 
> If I'm not mistaken, only ":" or "\" are legal in that narrow context.  While
> additional tokens may be legal in terms of following the string literal, they
> still result in an error, so there's no point in listing them.  Line
> continuation is a special case all its own.  Either it's illegal for some
> reason, in which that situation constitutes a diagnostic all its own.  Otherwise
> it can be completely ignored, as any errors/diagnostics would involve the next
> following token, not the line continuation itself.
> 
> So it would be perfectly reasonable for the parser to say "semicolon missing" or
> "semicolon expected" in this case.
> 
> Whether or not it's worth the effort to implement is, as always, debatable and
> furthermore, I suppose, moot unless somebody is offering to do the work.
> 
> Regards
> 
> --jb
> 
> --
> James J. Besemer  503-280-0838 voice
> http://cascade-sys.com  503-280-0375 fax
> mailto:jb at cascade-sys.com
> 
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list


--Opus--

You can't tell me what sucks!
                            - Beavis, a true Objectivist.

--------------------------------------------------------
Get added to my Humor list:
mailto:opus at value.net?subject=ADD_HUMOR
Get added to my Neat list:
mailto:opus at value.net?subject=ADD_NEAT
Get my PGP public key:
mailto:opus at value.net?subject=PSEND&body=send%20PublicKEY.asc
Visit My Home Page:
http://value.net/~opus/







From johnroth at ameritech.net  Sun Jun 30 20:05:23 2002
From: johnroth at ameritech.net (John Roth)
Date: Sun, 30 Jun 2002 20:05:23 -0400
Subject: Python needs to explain parsing? was Re: Python needs better error reporting
References: <mailman.1025315859.18149.python-list@python.org> <3D1F8843.8080602@ix.netcom.com>
Message-ID: <uhv73gr2pvf23f@news.supernews.com>

"Antaeus Feldspar" <feldspar at ix.netcom.com> wrote in message
news:3D1F8843.8080602 at ix.netcom.com...
> After reading the thread spun out of all of this, this brings up an
idea
> that I've had for quite some time:  can the Python interpreter be made
> to report more information about *how* it is parsing the syntax it
finds?
>
> For instance, given the valid line "if s == ' ':" it might return
> something like the following (or at least this might be what is
> displayed to the user based on information the parser displays):
>
> if <condition>: ---- [next 8 lines of code]
> |
> ==
>        / \
>       /   \
>      s    ' '
>
> thus confirming that what was meant to be a condition is being parsed
as
> a condition.  Clearly if the user sees that the parser is putting the
> block of code to be conditionally executed in where the condition
should
> be, that's a very visual sign of trouble...
>
> It's not quite the solution David seems to be looking for, but I think
> it would address the same problem; is it possible?

Frankly, "expecting <foorap> found <bletch>" for reasonably
intuitive values of <foorap> and <bletch> can tell a moderately
experienced programmer a great deal.

It does very little for novices, however. Except that with a little
experience, most people discover that they make the same mistake,
over and over again. Then the same error message is helpful, even
if it means little to anyone else.

John Roth




From mcherm at destiny.com  Mon Jun 24 09:40:31 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Mon, 24 Jun 2002 09:40:31 -0400
Subject: I'd give up Perl tomorrow if only...
Message-ID: <3D17214F.5000902@destiny.com>

Michael Chermside wrote:
 > Where can I find the pyperl module?

Gerhard H?ring responded:
> Google - Enter "pyperl" - then hit "I feel lucky" :-)
> 
> This will bring you to
> http://www.cpan.org/modules/by-module/LWP/GAAS/pyperl-1.0.readme which looks to
> a non-Perl guy like me that its available as a CPAN module.

Well, I had tried that, but (not being a perl kind of guy) I'm not 
really sure what a CPAN module is or how to obtain or download it. I DID 
find some version of pyperl at ActiveState (yeah, ActiveState!), but 
couldn't (in the brief amount of time I spent on it) figure out how to 
set it up.

What I WAS able to figure out has been written up at 
http://www.faqts.com/knowledge_base/view.phtml/aid/17202/fid/1102
If anyone else can help more, let me know.

-- Michael Chermside





From sholden at holdenweb.com  Mon Jun  3 07:51:56 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 3 Jun 2002 07:51:56 -0400
Subject: socket module and broadcasts ?
References: <3CF780D6.DFEDCFB7@ipm.fhg.de> <hrLJ8.80417$%u2.72437@atlpnn01.usenetserver.com> <3CFB14B9.CF8FAA4B@ipm.fhg.de>
Message-ID: <GPIK8.179391$%u2.141705@atlpnn01.usenetserver.com>

OK, in that case a read of

        http://www.mcmillan-inc.com/sock1.html

will be helpful for general background if you aren't used to socket
programming. But that's more useful for TCP. What you need is for your
server to listen on a UDP port, and have the clients do a local broadcast to
that UDP port at IP address 255.255.255.255 (which most routers will ignore,
since that address is intended for LAN only communications).

The sequence of actions required at the server is

    1. Create the socket
    2. Bind it to appropriate address(es)
    3. Use a recvfrom() call to wait for client broadcasts.

The sequence of actions required at the client is

    1. Create the socket.
    2. Bind it to appropriate address.
    3. Use sendto() to broadcast for the server.

Once the server sees the client broadcast it know the client's IP address,
and can use sendto() to return a resposne which will, in turn, give the
client the server's address when it executes a recvfrom().

Hope this helps.

regards
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------


"Markus von Ehr" <markus.vonehr at ipm.fhg.de> wrote in message
news:3CFB14B9.CF8FAA4B at ipm.fhg.de...
> Hi Steve,
>
>
> Steve Holden wrote:
>
> > Do you really want your server to have top
> > listen for broadcasts from potential clients?
>
> that is exactly what I want to do.
>
> > What if the server and the
> > client are of different IP networks?
>
> I assume that they are both in the same network, but the server is
> on an embedded system and initially it can be that its IP address is
> unknown, so I want to query for the server and assign an IP address.
>
> Markus
>





From sholden at holdenweb.com  Wed Jun 19 11:02:42 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 19 Jun 2002 11:02:42 -0400
Subject: simple socket redirector app question
References: <5DrP8.25$OT1.34057@newsfeed.slurp.net> <mailman.1024360162.13732.python-list@python.org>
Message-ID: <kQjQ8.213354$Fp1.92793@atlpnn01.usenetserver.com>

"Michael Gilfix" <mgilfix at eecs.tufts.edu> top-posted the code ...
>   I think SocketServer is a good choice for this. SocketServer is
> aimed at providing a easy to drop-in infrastructure for apps that
> don't require any complicated server setup. SocketServer would make
> this quick and compact. Essentially you just want to read from one
> socket and write to the other. Via inhertiance you could construct
> a bridge class that contains two socket servers and talks between
> them...
>
>                    -- Mike
>
> On Mon, Jun 17 @ 13:27, Bruce Edge wrote:
> > I have an app which talks via sockets to local host clients only.
> > I'd like to get at it remotely.
> >
> > Basically it would accept remote connections from a remote client,
> > relay them the the local app, and forward the replies back to the
> > remote client.
> >
> > What's the simplest way to write this type of a socket redirector?
> >
> > Use SocketServer, or is that overkill?
> >

There's also a suitable proxy in the Medusa distribution, available from Sam
Rushing's nightmare.com site. Here's a slightly-modified copy I used in
"Python Web Programming". It's perhaps a bit more general than you really
need, but I founf it an excellent demonstration of asyncore's operation.

regards
 Steve
----------------------------- Code Follows ----------------------------
import asynchat
import asyncore
import socket
import string

class proxy_server(asyncore.dispatcher):

    def __init__(self, host, port):
        asyncore.dispatcher.__init__(self)
        self.create_socket(socket.AF_INET, socket.SOCK_STREAM)
        self.set_reuse_addr()
        self.there =(host, port)
        here =('', port + 8000)
        self.bind(here)
        self.listen(5)

    def handle_accept(self):
        proxy_receiver(self, self.accept())

class proxy_sender(asynchat.async_chat):

    def __init__(self, receiver, address):
        asynchat.async_chat.__init__(self)
        self.receiver = receiver
        self.set_terminator(None)
        self.create_socket(socket.AF_INET, socket.SOCK_STREAM)
        self.buffer = ''
        self.set_terminator('\n')
        self.connect(address)

    def handle_connect(self):
        print 'Connected'

    def collect_incoming_data(self, data):
        self.buffer = self.buffer + data

    def found_terminator(self):
        data = self.buffer
        self.buffer = ''
        print '==>(%d) %s' %(self.id, repr(data))
        self.receiver.push(data + '\n')

    def handle_close(self):
        self.receiver.close()
        self.close()

class proxy_receiver(asynchat.async_chat):

    channel_counter = 0

    def __init__(self, server,(conn, addr)):
        asynchat.async_chat.__init__(self, conn)
        self.set_terminator('\n')
        self.server = server
        self.id = proxy_receiver.channel_counter
        proxy_receiver.channel_counter = proxy_receiver.channel_counter + 1
        self.sender = proxy_sender(self, server.there)
        self.sender.id = self.id
        self.buffer = ''

    def collect_incoming_data(self, data):
        self.buffer = self.buffer + data

    def found_terminator(self):
        data = self.buffer
        self.buffer = ''
        print '<==(%d) %s' %(self.id, repr(data))
        self.sender.push(data + '\n')

    def handle_close(self):
        print 'Closing'
        self.sender.close()
        self.close()

if __name__ == '__main__':
    import sys
    import string
    if len(sys.argv) < 3:
        print 'Usage: %s <server-host> <server-port>' % sys.argv[0]
    else:
        ps = proxy_server(sys.argv[1], string.atoi(sys.argv[2]))
        asyncore.loop()

-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From bokr at oz.net  Wed Jun 26 22:01:35 2002
From: bokr at oz.net (Bengt Richter)
Date: 27 Jun 2002 02:01:35 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com>
Message-ID: <afdrlv$r83$0@216.39.172.122>

On Thu, 27 Jun 2002 01:28:22 GMT, candiazoo at attbi.com wrote:

>I assume some sort of block i/o or preallocating a large block of file space
>would be the best way to do it?  I am outputting about 700,000 records to a text
>file and my program only spits out about 20 records per second.  This is using
>standard file open and write...
>
>This is on a W2K machine... perhaps I should use the win32file interface?  Any
>thoughts/opinions?
>
I doubt that that is the solution. Where are your "records" coming from and/or
how are they being generated/mdified, and how big are they?

If you are building string records by s='' followed by s+='chunk' or the like many times,
this is a typical newbie performance killer. Try accumulating your chunks in a list,
like sl=[] followed by sl.append('chunk') instead, and then do f.write(''.join(sl))
instead of the f.write(s) you would have done.

Just guessing...

Try timing generation of output data vs writing it, and see what's happening.
import time and use clock() to get an accurate floating point seconds time reading. E.g.,

from time import clock
tstart = clock()
# ... generate an output record here
tendgen=clock()
# ... write the output  something like f.write(record) here
tendwrite=clock()
print """\
generating took %f sec
outputting took %f sec""" % (tendgen-tstart, tendwrite-tendgen)

Regards,
Bengt Richter


From phlip_cpp at yahoo.com  Fri Jun 14 00:47:14 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 14 Jun 2002 04:47:14 GMT
Subject: More Tkinter Help please...
References: <mailman.1024026701.13559.python-list@python.org>
Message-ID: <aebsgi$jp2@dispatch.concentric.net>

David LeBlanc wrote:

> Code comes up for me, but has other bugs I'm not able to diagnose.

My theory may be right, and the OP's User-Agent, 
Microsoft-Entourage/10.0.0.1309, might be converting spaces to (4) tabs at 
paste time.

-- 
  Phlip
   http://www.greencheese.org/EvolutionaryPsychology
  --  It's a small Web, after all...  --



From johnroth at ameritech.net  Wed Jun  5 17:15:59 2002
From: johnroth at ameritech.net (John Roth)
Date: Wed, 5 Jun 2002 17:15:59 -0400
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023234674.21472.python-list@python.org> <ufrupva3edm6fd@news.supernews.com> <3CFE6B13.79CA4A8C@noaa.gov>
Message-ID: <ufsvrhe8m64ee3@news.supernews.com>

"Chris Barker" <Chris.Barker at noaa.gov> wrote in message
news:3CFE6B13.79CA4A8C at noaa.gov...
> John Roth wrote:
> > I don't see a compatability issue here, because
> > the current behavior raises an exception.
>
> Unless someone did the perverse thing I proposed in a earlier post,
and
> took advantage of the fact that the operation did, in fact, work, even
> though it raised an exception!

That's perverse enough that whoever does it deserves for their
program to fail on an upgrade.

> While I completely agree with Huaiyu's argument (though I'm not sure
the
> "increment with re-binding" is really all that useful for mutables),
and
> I think the current semantics is a minor wart, I am convinced that the
> original behavior that started this thread is a MAJOR WART (I'd call
it
> a bug). I think I have a pretty good understanding now of why it
> happens, and why it isn't trivial to change, but I think it is a very
> bad idea for an operation to raise an exception, but also be
successful!
>
> Chris
>
I thoroughly agree. Operations like += should be atomic -
they should either succeed or fail. The remaining question is
simply whether it goes to the bug list or to a PEP.

John Roth




From warlock at eskimo.com  Fri Jun  7 20:35:51 2002
From: warlock at eskimo.com (Jim Richardson)
Date: Fri, 7 Jun 2002 17:35:51 -0700
Subject: Detecting OS and Window Manager
References: <mailman.1023304388.9659.python-list@python.org> <ufvsco8mlhf213@corp.supernews.com>
Message-ID: <7hjrda.ilm.ln@127.0.0.1>

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

On Fri, 07 Jun 2002 00:34:30 +0100,
 Darren Winsper <dw133 at cs.york.ac.uk.go.away> wrote:
> In <mailman.1023304388.9659.python-list at python.org>, Sean 'Shaleh' Perry
> wrote:
> 
>> the only fool proof way to detect KDE, GNOME, etc is to check for X atoms.  The
>> bigger question is why do you care?  As long as they can display a GUI you
>> should not be bothered about which one it is.  Especially under the X Window
>> System it was designed so that random applications DIDN'T know what they were
>> running under.
> 
> As the person writing the KDE front-end to the mentioned application, let
> me answer your question.  The idea is that if the user is running KDE,
> then they get a KDE front-end by default.  I also hope that should
> somebody write a GTK/GNOME front-end, if the user was running GNOME they
> would get a GTK-based front-end by default.  It's good UI design.
> 
>> Also remember KDE is not a window manager, it is a desktop environment.  It is
>> perfectly reasonable to run KDE and use fvwm (or even twm) as the window
>> manager.  Same applies to GNOME.
> 
> OK, Andrew meant DE instead of window manager, but it's still a valid
> question.
> 
>> Also, what about someone who only runs say
>> Window Maker and never uses GNOME or KDE?
> 
> Well, in that case we just choose a default.  The thing is, we want to
> detect if PyKDE is installed.  If it is, then we can do "import kdecore"
> without risk of it throwing an exception.  If it isn't, I'd rather not
> have to try something along the lines of attempting "import kdecore" and
> then having to catch an exception if it's not found.
> 


But the only way to see if PyKDE is installed, is to check it, KDE
doesn't require PyKDE, you can't assume that PyKDE will be present
simply because someone is running KDE. About the only way I can see you
toing this, is with a whole lot of try/fail checks, and even then, if
someone is running something other than KDE/Gnome, you will be back to
the default. Why not start the default front end first time, then ask
the user what front end they would prefer, then save that to a dotfile? 


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)

iD8DBQE9AVFmd90bcYOAWPYRAnkSAKC/z+7HbHqAUEQj53vSdOTUc2QNqACeOArx
jTNFcrwH8pAs57SSjsYHiA4=
=Ooz6
-----END PGP SIGNATURE-----

-- 
Jim Richardson
	Anarchist, pagan and proud of it
http://www.eskimo.com/~warlock
Linux, from watches to supercomputers, for grandmas and geeks. 


From djc at object-craft.com.au  Sun Jun 23 05:31:42 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 23 Jun 2002 19:31:42 +1000
Subject: Web templating/db tool with best designer/coder separation?
References: <af2gh6$2c2$0@216.39.172.122> <mailman.1024782097.17337.python-list@python.org>
Message-ID: <m34rfufiip.fsf@ferret.object-craft.com.au>

>>>>> "Tim" == Tim Churches <tchur at optushome.com.au> writes:

Tim> Bengt Richter wrote:
>>  E.g., if you were hoping for an improvement over current practice,
>> which do you consider to be the tool(s) to beat, and what would the
>> killer improvement(s) be? TIA.

If there was a clear definition of what was required in this area I
suspect that there would be less attempts at building toolkits.

Albatross is a design that evolved from my experience in building a
collection of small but data rich intranet applications.  I discovered
that I was doing the same things over and over.  It does not show from
the release history of Albatross, but the toolkit was several years in
the making.

Tim> Albatross by Object Craft (see http://www.object-craft.com.au )
Tim> does an excellent job of separating the presentation layer from
Tim> the underlying logic - my understanding is that such separation
Tim> was one of its main design goals. However, it does not take a
Tim> pedantic position on this - it still allows Python code to be
Tim> embedded in HTML templates - such practice is discouraged, but
Tim> Albatross provides mechanisms to allow you to do it if you
Tim> insist. But more usefully, Albatross provides tags which allow
Tim> conditional processing (if-elif-else) and most usefully,
Tim> iteration, to be placed in the HTML template. So your Python
Tim> business logic code can assemble a Python sequence of, say,
Tim> database query results, but it is the HTML template which
Tim> iterates through that list to create the final HTML which is sent
Tim> to the user's browser. Such a feature is essential, because
Tim> otherwise the business logic Python code ends up needing to
Tim> creating HTML output itself, which is then merely substituted
Tim> into the HTMl template. Thus, the Albatross model is really more
Tim> like: "all HTML-related stuff, including conditional processing
Tim> and iteration, is looked after in the HTML templates, and all
Tim> business logic stuff, divorced from the manner in which it will
Tim> be presented, is handled by Python modules." Of course, this is
Tim> just the well-known model-view-controller approach, but Albatross
Tim> implements this very nicely, IMO.  See
Tim> http://www.object-craft.com.au/projects/albatross/albatross/fig-presimpexec.html
Tim> for more information about this (and to get an idea of the depth
Tim> of the Albatross documentation).

I think the biggest area which requires improvement in Albatross is
still the documentation.

Tim> I suspect that the Object Craft guys would welcome help with
Tim> further development of Albatross. What the world really needs are
Tim> more Pythonic Web applications, rather than more Pythonic Web
Tim> application frameworks.

People who have tried Albatross could really help us by writing
beginner level instructions and by pointing out where the existing
documentation is misleading.  We are all too close to the code to see
clearly.

- Dave

-- 
http://www.object-craft.com.au


From tdelaney at avaya.com  Sun Jun  2 23:33:18 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 3 Jun 2002 13:33:18 +1000 
Subject: prototyping good OOdesign in Python?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A340@natasha.auslabs.avaya.com>

> From: Roman Suzi [mailto:rnd at onego.ru]
>
> >Since most of the code never /actually/ gets rewritten in C++,
> >the need for special C++ "discipline" in the python code goes
> >away.  Weeks are slashed from the schedule, and the program is
> >more maintainable.
> >
> >Another nice advantage is that the python programmer gets to ask
> >for a raise, and has excellent job security!
> 
> ...because only he knows the magic powers of Python?

Raise (hopefully) for completing projects massively under time and budget.

Security ... because they will want to hold onto someone so good.

Unfortunately, this will only work for 1 or 2 projects ... after that they
will *expect* you to bring everything in under time and budget ... but by
then the time and budget is massively reduced because of the great results
you've brought in ...

:)

Tim Delaney




From jdhunter at nitace.bsd.uchicago.edu  Sun Jun 30 10:17:33 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Sun, 30 Jun 2002 09:17:33 -0500
Subject: can't import generators
References: <m2hejlc060.fsf@mother.paradise.lost> <afl5je$4iu$1@panix1.panix.com>
 <m2d6u9by8e.fsf@mother.paradise.lost> <aflnal$aak$1@panix1.panix.com>
Message-ID: <m21yaox342.fsf@mother.paradise.lost>

>>>>> "Aahz" == Aahz  <aahz at pythoncraft.com> writes:

    Aahz> Try re-downloading 2.2.1 and building in a completely new
    Aahz> directory.  It's looking to me like you have cross-version
    Aahz> contamination, and that's probably the best way to test it.

Thanks for your help -- I eventually figured out the problem (while
working in the kitchen on something else, naturally).  My PYTHONPATH
still pointed to the python 2.1 libraries.

Thanks again,
John Hunter


From dguo at ux1.cso.uiuc.edu  Wed Jun  5 19:13:01 2002
From: dguo at ux1.cso.uiuc.edu (Carolyn Guo)
Date: Wed, 5 Jun 2002 18:13:01 -0500
Subject: alignment excel written in Python
In-Reply-To: <3CFE9861.5090200@skippinet.com.au>
References: <6bd9f01b.0206050736.f469082@posting.google.com> <3CFE9861.5090200@skippinet.com.au>
Message-ID: <Pine.GSO.4.31.0206051810190.22089-100000@staff1.cso.uiuc.edu>

Dear all,
I am new one to write something in Python. Currently I am working on
writing something to Excel from Python.

I have a questions now: How can I deal with the alignment of Excel from
Python? for example, put the value in a cell with alignment right, or
center.

Thanks a lot.
Carolyn




From mis6 at pitt.edu  Wed Jun 19 15:27:16 2002
From: mis6 at pitt.edu (Michele Simionato)
Date: 19 Jun 2002 12:27:16 -0700
Subject: Tkinter and the Menu Widget
References: <2259b0e2.0206190415.1f232447@posting.google.com> <xw%P8.257292$cQ3.11997@sccrnsc01>
Message-ID: <2259b0e2.0206191127.455ba17d@posting.google.com>

"Emile van Sebille" <emile at fenx.com> wrote in message news:<xw%P8.257292$cQ3.11997 at sccrnsc01>...
>
> You've got a scope issue.  Try:
> 
>         smn[key].add_command(label=submenu,command=lambda key=key,
> submenu=submenu:
>                              dosomething(key+'='+submenu))

Thanks ! This is really magic and it works !!
Actually I had found another method to fix the problem by using
of exec() and eval() but your suggestion is much more elegant.
Thanks again,

                           Michele


From op73418 at mail.telepac.pt  Sun Jun 23 10:12:29 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Sun, 23 Jun 2002 15:12:29 +0100
Subject: a Tree data-structure class?
References: <b16e4ef7.0206210510.77a956b6@posting.google.com>
Message-ID: <aqlbhucj12s0h8hkg7usafcn4puher3ho5@4ax.com>

On 21 Jun 2002 06:10:30 -0700, steve at ferg.org (Stephen Ferg) wrote:

>I do a lot of processing of tree-structured data, so I'm thinking it
>would be useful to develop a generic Tree class that I could adapt to
>various applications.  Or perhaps a Tree class and a generic Node
>class.
>
>But before I start trying to re-invent the wheel, I thought I'd ask if
>anybody knows of something like this that has already been developed.
>
>Note that I'm not talking about a Tree GUI widget, but about a tree
>data structure, like a list, a queue, etc.
>
>-- Steve Ferg (steve at ferg.org)

I already have a module with implementations of binary and generalized
trees. It is not finished yet, but I can send you what I have right now.
I will try to finish it in a few days if time allows and then post the
whole thing in the ActiveState cookbook.

All the best,
Gon?alo Rodrigues


From BPettersen at NAREX.com  Tue Jun 25 12:53:44 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 25 Jun 2002 10:53:44 -0600
Subject: Suggestions for good programming practices?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158EFB@admin56.narex.com>

> From: Aahz [mailto:aahz at pythoncraft.com] 
> 
> In article <hd_R8.306964$352.31513 at sccrnsc02>,
> Emile van Sebille <emile at fenx.com> wrote:
> >Aahz
> >> In Python 2.1.2 or higher, fire up the interactive interpreter and 
> >> type "import this".
> >
> >Except ActiveState's
> >
> >F:\Python22as>python
> >Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] 
> on win32 
> >Type "help", "copyright", "credits" or "license" for more 
> information.
> >>>> import this
> >Traceback (most recent call last):
> >  File "<stdin>", line 1, in ?
> >ImportError: No module named this
> 
> Interesting.  That's an ActiveState bug, pure and simple; 
> this.py is a normal module in Lib/.

It's been fixed in the latest version.

-- bjorn




From emile at fcfw.fenx.com  Sun Jun 23 12:17:28 2002
From: emile at fcfw.fenx.com (Emile van Sebille)
Date: Sun, 23 Jun 2002 09:17:28 -0700
Subject: (Newbie)  Counting Instances ("Hits") with Regular Expressions
References: <3d15e215.43448155@news.texas.net> <nbmR8.300036$cQ3.15150@sccrnsc01> <200206231606.g5NG65117471@localhost.localdomain>
Message-ID: <2f9701c21ad1$78616110$0a06a8c0@FENX.COM>

Dave Reed
> 
> I don't think he wants to count those, only ' am '
> 
> There is a much easier solution than what I originally posted:
> 
> import re
> s = 'the word is the word but not a keyword'
> re.findall('\sword\s')
> 
> outputs: [' word ', ' word ']
> so len(re.findall('\sword\s') will do what he wants
> 

>>> s = 'the word is the word, but not a keyword'
>>> re.findall(r'\b(word)\b', s.lower())
['word', 'word']
>>>

--

Emile van Sebille
emile at fenx.com

---------





From mgilfix at eecs.tufts.edu  Sun Jun  9 09:45:25 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Sun, 9 Jun 2002 09:45:25 -0400
Subject: Simple pychecker question
In-Reply-To: <3D02FDE9.19198F33@engcorp.com>; from peter@engcorp.com on Sun, Jun 09, 2002 at 03:04:09AM -0400
References: <mailman.1023482712.8801.python-list@python.org> <pan.2002.06.09.00.07.47.445563.6814@metaslash.com> <3D02FDE9.19198F33@engcorp.com>
Message-ID: <20020609094525.H9486@eecs.tufts.edu>

On Sun, Jun 09 @ 03:04, Peter Hansen wrote:
> For a routine named and structured like this:
> 
>   def transmit(data, timeout=1.0):
> 
> which of the following would you say was the better call
> in some code that might be remote from that definition
> above?
> 
> 1.  transmit('this is a string', 5)
> 
> 2.  transmit('this is a string', timeout=5)
> 
> My point: named arguments can increase readability and 
> maintainability.  That's not superfluous, to me.

  I tend to opt for the hidden defaults. Sometimes, I want to provide
the users of my classes the option of intializing the class with data,
or setting the data later (as in the case of a graph). So something
like:

   def __init__ (self, data=None)
   def set_data (self, data)

  is pretty useful for me. My version of pychecker complained, which
got me wondering. But if Neil says it's fixed, then I'm sure it is :)

                       -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From catunda at pobox.com  Tue Jun 25 14:30:53 2002
From: catunda at pobox.com (Marco Catunda)
Date: Tue, 25 Jun 2002 15:30:53 -0300
Subject: Thread Memory Leak
References: <mailman.1025018254.27029.python-list@python.org>
Message-ID: <pan.2002.06.25.18.30.51.286352.18707@pobox.com>

Sorry, I forgot it.

Plataform: Linux RedHat 7.3
Glibc:     glibc-2.2.5-34 (libpthread)
Python:    2.2.1

Thank you
-- Marco Catunda


On Tue, 25 Jun 2002 12:14:22 -0300, Tim Peters wrote:

> [Marco Catunda]
>> I've developed a python program with some threads. This program is a
>> daemon server. I have seen the memory of that program growing very fast
>> in a short period of time. So I decided to make a little program to see
>> what's happen.
>> ...
> 
> Which version of Python, which OS, and which thread package?  There's no
> leak evident when running your program on Win2K using any of Python
> 2.1.3, 2.2.1, or current CVS, so whatever you're seeing is specific to
> something you haven't told us.


From gleki at gol.ge  Fri Jun  7 05:01:26 2002
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Fri, 07 Jun 2002 11:01:26 +0200
Subject: map
Message-ID: <3D007666.2B6D3582@gol.ge>

Hi all!

Is there a way to map a function with several arguments to a list? E.g.,

>>>import math
>>>import MLab
>>>a=MLab.rand(3,5)
>>>b=map(math.pow,a)

of course, this doesn't work.

What to do? My task is to optimize the power (i.e., y in pow(x,y)) to
achieve maximal performance.


thanx,
Giorgi

PS: Perhaps I have missed the needed functionality, that's presented in
NumPy?





From merkosh at hadiko.de  Fri Jun 21 09:07:35 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Fri, 21 Jun 2002 15:07:35 +0200
Subject: why extending copy.*
Message-ID: <MPG.177d438123ae240598968d@news.rz.uni-karlsruhe.de>

hi,

i need shallow and deep copies of objects and I found the copy module in 
the reference library which does quite a good job.
however, they write that one could implement the copy protocol by 
extending/ overwriting the __copy__() and __deepcopy__() method.

since __copy__() and __deepcopy__() already duplicate the instance 
dictionary, __deepcopy__() even recursively. For what reasons would I 
want to extend /overwrite the capabilities of __copy__() or __deepcopy__
()?


From huaiyu at gauss.almadan.ibm.com  Thu Jun 13 21:38:17 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 14 Jun 2002 01:38:17 +0000 (UTC)
Subject: filter2
References: <mailman.1023975618.6474.python-list@python.org> <lkptyvmdfz.fsf@pc150.maths.bris.ac.uk> <mailman.1023984463.10662.python-list@python.org> <lkhek7m9c8.fsf@pc150.maths.bris.ac.uk>
Message-ID: <slrnagii89.55q.huaiyu@gauss.almadan.ibm.com>

Michael Hudson <mwh at python.net> wrote:
>Jeff Epler <jepler at unpythonic.net> writes:
>> 
>> def filter2(test, list):
>>     l1 = []
>>     l2 = []
>>     funcs = [l2.append, l1.append]
>>     map(lambda x: funcs[bool(test(x))](x), list)
>>     return l1, l2
>> 
>> >>> print filter2(lambda x: x%3==0, range(16))
>> ([0, 3, 6, 9, 12, 15], [1, 2, 4, 5, 7, 8, 10, 11, 13, 14])
>
>Hey, that's pretty neat!  Still three function calls per element,
>though.  Well, four including bool, but that could be gotten rid of by
>gross hacks (e.g. not not test(x)).

It's not that gross, but it only gives 14% speedup, though.

def filter2a(test, list):
    l1 = []
    l2 = []
    funcs = [l1.append, l2.append]
    map(lambda x: funcs[not test(x)](x), list)
    return l1, l2

BTW, what's the reason that using a list for funcs is faster than using a
tuple (as observed under Python 2.2.1 on Linux)?

Huaiyu


From gerhard.haering at gmx.de  Wed Jun 19 21:25:40 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 20 Jun 2002 01:25:40 GMT
Subject: procmail replacement in Python
References: <20020618235901.GA2197@lilith.my-fqdn.de> <oqit4grq03.fsf@carouge.sram.qc.ca> <20020619022110.GA6996@lilith.my-fqdn.de> <oq660froj9.fsf@carouge.sram.qc.ca> <mailman.1024521813.14158.python-list@python.org>
Message-ID: <slrnah2bon.cpt.gerhard.haering@lilith.my-fqdn.de>

Skip Montanaro wrote in comp.lang.python:
> I strongly recommend SA. 

ACK. I've also never needed to hack the code (I wouldn't know Perl,
anyway), as it can be configured very nicely via its config file.

> > > [Gerhard H?ring]
> >> I'm also investingating Pyzor, as I found the Razor author to be too
> >> clueless (also the razor servers aren't open-source).

That needs some explanation:

http://sf.net/tracker/index.php?func=detail&aid=532357&group_id=3978&atid=103978

He doesn't get mbox parsing, and continues to use his own b0rken
implementation. Having described the bug in a very detailed way, and
even uploaded a test case, and still having it marked as "invalid"
made be /a little/ bitter.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From djc at object-craft.com.au  Thu Jun 20 08:41:49 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 20 Jun 2002 22:41:49 +1000
Subject: Python to MS SQL Server
References: <mailman.1024428323.19642.python-list@python.org> <nQjQ8.213357$Fp1.114650@atlpnn01.usenetserver.com>
Message-ID: <m3vg8e9l6a.fsf@ferret.object-craft.com.au>

>>>>> "Steve" == Steve Holden <sholden at holdenweb.com> writes:

Steve> "Bjorn Pettersen" <BPettersen at NAREX.com> wrote ...
>> > From: Mark McEahern [mailto:marklists at mceahern.com]
>> >
>> > > I am interested in using Python in a MS SQL Server > >
>> environment, but do > > not see any database modules out there,
>> especially for > > native versus > > using ODBC. Is there such a
>> module?
>> >
>> > This is a total guess, because I haven't done this myself:
>> >
>> > Have you tried using win32com to use ADO?

>> AFAIK, there is no Python database module that goes directly to MS
>> SQL Server db lib. ADO does work however, so that's certainly an
>> option.  Personally, I would probably investigate e.g. mxODBC first
>> (the interface is much easier to use from Python).


Steve> I agree that ADO is a total pain (although hesitantly, since
Steve> Alex Martelli is known to favour it) in Python. I just never
Steve> got used to it in VBScript, I guess.

Steve> Don't I remember that Dave Col's Sybase module is supposed to
Steve> be usable with SQL Server too?

I also have a module which uses the MS db library.

    http://www.object-craft.com.au/projects/mssql/

It needs more work, but might be sufficient for your needs.

- Dave

-- 
http://www.object-craft.com.au


From fgeiger at datec.at  Thu Jun  6 02:14:16 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Thu, 6 Jun 2002 08:14:16 +0200
Subject: constructive critisism on pythonwin IDE
References: <6bd9f01b.0206050736.f469082@posting.google.com>
Message-ID: <admuke$gu8$1@newsreaderg1.core.theplanet.net>

This a Python issue, not a PythonWin IDE issue.

Cheers
Franz

"sameer" <sameer_ at email.com> schrieb im Newsbeitrag
news:6bd9f01b.0206050736.f469082 at posting.google.com...
> First of all, let me start of by saying that Pythonwin is a great IDE.
>  Now for the constructive critisism.  The thing I hate about PythonWin
> is, that despite the fact that it's a development environment, there
> is a lot of caching done of modules, and the cached copy is not
> checked against a potentialy modified copy.  Whenever I change the
> contents of a file and try to run a module that depends on that file,
> I always end up with the old copy.  I have to then import and then
> reload the changed module in the interactive window for it to refresh.
>  It would greatly increase my productivity, if I can specify which
> modules can be cached.  I understand that there is a speed and
> development time issue, but I would like a little more fine grained
> control on how caching works on PythonWin, as the way it's currently
> implemented now is not acceptable.




From chandan_mathur at hotmail.com  Sat Jun  1 02:14:49 2002
From: chandan_mathur at hotmail.com (Chandan Mathur)
Date: Sat, 1 Jun 2002 02:14:49 -0400
Subject: How to install Python without root priviledges ?
References: <acvmin$svjfb$1@ID-146704.news.dfncis.de>
Message-ID: <3cf866fa@news.poly.edu>

you have done the right thing.., only without SU priviliges, I think you can
run python only on the directory Python-2.2.1 as follows,  ./python
I have the same problem... and it really sucks :-(

"Shagshag13" <shagshag13 at yahoo.fr> wrote in message
news:acvmin$svjfb$1 at ID-146704.news.dfncis.de...
>
> I'm a real newbie in unix's world and i would like to install python on a
> machine where *i can't have* any root priviledges, how should i do that ?
>
> (it's on a HP Unix, but don't even know how to check the exact os name or
> version...)
>
> i do this :
>
> mkdir /home/shagshag/local
> download Python-2.2.1.tgz
> gunzip
> tar -xvf Python-2.2.1.tgz
> cd Python-2.2.1
> ./configure --prefix=/home/shagshag/local
> make
>
> And i get nothing !!!!
>
> Is it the good way to install ???
>
> Thanks in advance,
>
> S13.
>
> please, don't forget i'm not a skilled unix guy...
>
>




From cliechti at gmx.net  Sun Jun  2 13:58:00 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 2 Jun 2002 19:58:00 +0200
Subject: win32com under Cygwin
References: <3CF9846B.579606ED@wiencko.com>
Message-ID: <Xns9221CB7B32FDcliechtigmxnet@62.2.16.82>

Tom Wiencko <tew at wiencko.com> wrote in news:3CF9846B.579606ED at wiencko.com:

> There may be an easy answer for this, but I have not yet found it...
> 
> I noticed that the Python distribution with Cygwin does not include the
> win32 extensions.  The only distribution I can find for them is a
> Windows executible, which does not seem to know about Cygwin.

thats the official page of win32all:
http://starship.python.net/crew/mhammond/

the binary version is compiled for the official python.org interpreter.

> Can these libraries be loaded under Cygwin?  Do they even work under
> Cygwin?  Can somebody point me toward how to load them if it is
> possible?

the sources are available from the page above. maybe you can manage to 
compile it with cygwin for cygwin python. don't know if that is easy...

chris

-- 
Chris <cliechti at gmx.net>



From james.kew at btinternet.com  Thu Jun 27 04:04:23 2002
From: james.kew at btinternet.com (James Kew)
Date: Thu, 27 Jun 2002 09:04:23 +0100
Subject: background-processes with win-python?
References: <mailman.1025157935.3443.python-list@python.org>
Message-ID: <afegu8$dlvqk$1@ID-71831.news.dfncis.de>

"Klaus Reinhardt" <K.Rdt at TU-Berlin.DE> wrote in message
news:mailman.1025157935.3443.python-list at python.org...

> Is there a possibility to execute system-processes
> under windows-python in the background? So
> I can do this
> system("plink  .. tu-berlin.de")
> without waiting for the real connection.

system("start plink ... tu-berlin.de") ?

James





From DLNXPEGFQVEB at spammotel.com  Thu Jun 20 10:38:43 2002
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Thu, 20 Jun 2002 17:38:43 +0300
Subject: CR+LF problem
References: <3d11bc33.1129927479@news.mad.ttd.net>
Message-ID: <12q3hukse2cepcm9tpern92jc71a5vrhvi@4ax.com>

On Thu, 20 Jun 2002 11:33:55 GMT, rumours say that mruiz at safa.es (Manuel
Ruiz) might have written:

>char * txt = "print 123\r\n";
>result = PyRun_Simple( txt, Py_file_input, PyDict_New(), NULL );

did you try
	char *txt= "print 123\n";
by any chance?
I am not very sure if I do help, but in C, newlines are just "\n".
"\r\n" is by convention the text line terminator of VMS IIRC, MS-DOS &
window derivatives etc.

>this code fails to exec txt, I got a SyntaxError exception,
>however I if save to file the statement "print 123\r\n" and try to run
>it ussing python interprete (version 2.2.1) It run fine.
>
>Can anyone tell me anything about this rare behaviour

-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')


From achim.domma at syynx.de  Fri Jun  7 05:11:23 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Fri, 7 Jun 2002 11:11:23 +0200
Subject: Distributing Packages / os independent installation folder
Message-ID: <adpsrg$m02$03$1@news.t-online.com>

Hi,

I'm building a python interface to ImageMagick. I already succeded in
building an installer with distutils, but have some questions regarding
portability:

- where should my package and shared librarie go to? On windows I put the
dll into python22/DLLs and the rest into a folder
python22/Lib/site-packages/ImageMagick. What to do on other plattforms? I
don't expect a DLLs folder on Linux. ;-)

- In my __init__.py I have to change one environment variable. Currently I
do it like this:

import os
os.environ['MAGICK_HOME'] = r'C:\Python22\Lib\site-packages\ImageMagick'
from _ImageMagick import *

Is there an plattform independent way to get the installation folder of my
package? On windows I can get the python folder from registry, but this will
also break on linux.

greetings
Achim




From cliechti at gmx.net  Mon Jun 10 19:50:31 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 11 Jun 2002 01:50:31 +0200
Subject: getting files from a windows network
References: <6bd9f01b.0206101523.7af66fc8@posting.google.com>
Message-ID: <Xns922A1317B34C7cliechtigmxnet@62.2.16.82>

sameer_ at email.com (sameer) wrote in 
news:6bd9f01b.0206101523.7af66fc8 at posting.google.com:
> is there a way to read files in python that are found on a windows
> network, i.e. files that users choose to share on their computers on a
> windows network.

do you mean by e.g.
>>> open(r'\\computername\share\test.txt').read()

yes :-)

( r'' is a raw string that doesn't touches backslashes unlike normal 
strings)

if you want a list of computers you can use:
>>> map(str.strip, os.popen("net view").readlines()[3:-2])

chris

-- 
Chris <cliechti at gmx.net>



From jblazi at hotmail.com  Mon Jun 24 16:16:05 2002
From: jblazi at hotmail.com (JB)
Date: Mon, 24 Jun 2002 22:16:05 +0200
Subject: converting an array of chars to a string
References: <3d12dda2_6@news.newsgroups.com> <aevj41$fg2$0@216.39.172.122> <3d17038e_6@news.newsgroups.com> <af7jei$3kr$0@216.39.172.122> <3d1757c7_1@news.newsgroups.com> <af7q85$lit$0@216.39.172.122>
Message-ID: <3d177c4b_9@news.newsgroups.com>

Bengt Richter wrote:

If I had told you, what I was after, you would have pointed 
out, that there is a library function that does exactly 
this. Then I should have explained, why I thought, that the 
library function was not sufficient (why it was actualy 
unusable). All this was not necessary and you helped me a 
lot.

I simply did not understand, why those arrays are 
implemented with overflow checking, which involves 
additional overhead and is absolutely unnecessary.

So thank you very much.
-- 
Janos Blazi


-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From mickey at tm.informatik.uni-frankfurt.de  Tue Jun 18 17:00:32 2002
From: mickey at tm.informatik.uni-frankfurt.de (Michael 'Mickey' Lauer)
Date: 18 Jun 2002 23:00:32 +0200
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de> <aeftmj$m29$1@bob.news.rcn.net> <3d0c6d67@nntp.server.uni-frankfurt.de> <3D0CE698.1050806@SPAMnwinternet.com> <aeiu3a$4ff$1@bob.news.rcn.net> <3D0D3223.5070500@SPAMnwinternet.com> <aekrd1$4ao$1@bob.news.rcn.net>
Message-ID: <3d0f9f70@nntp.server.uni-frankfurt.de>

David Abrahams <david.abrahams at rcn.com> wrote:
> Heh, yeah. Unfortunately, documentation is often harder than coding. Even
> the current official release of Boost.Python is not very well-documented.
> We're working hard to correct that for v2, which is in pre-release.

Is it completed enough to try it out for wrapping a small library? If not,
could you estimate when it will be ?

Yours,

:M:
--
|----------------------------------------------------------------------------|
| Dipl.-Inf. Michael 'Mickey' Lauer    mickey at tm.informatik.uni-frankfurt.de |
|   Raum 10b - ++49 69 798 28358        Fachbereich Informatik und Biologie  |
|----------------------------------------------------------------------------|


From Sascha.Ferley at infineon.net  Sun Jun 23 19:37:27 2002
From: Sascha.Ferley at infineon.net (Sascha Ferley)
Date: Sun, 23 Jun 2002 17:37:27 -0600
Subject: PIL, Python and CGI
References: <3D1626FD.DA934B09@infineon.net> <1HrR8.44835$n4.10519234@newsc.telia.net>
Message-ID: <3D165BB7.B901F064@infineon.net>

Fredrik Lundh wrote:

> Sascha Ferley wrote:
>
> >     raise ImportError, "The _imaging C module is not installed"
> > ImportError: The _imaging C module is not installed
> >
> > Anyone have any ideas?
>
> when running your CGI script, the Python interpreter
> cannot find the _imaging extension.
>
> chances are that sys.path isn't what you think it is.
>
> </F>

Well I wrote a quick cgi script to get the path .. here is the output:
['/export/server/web/CPSC461/cgi-bin', '/usr/local/lib/python2.2',
'/usr/local/lib/python2.2/plat-sunos5',
'/usr/local/lib/python2.2/lib-tk',
'/usr/local/lib/python2.2/lib-dynload',
'/usr/local/lib/python2.2/site-packages',
'/usr/local/lib/python2.2/site-packages/PIL']

Program used:
#!/usr/local/bin/python

import sys
import cgi

print "Content-Type: text/plain \n\n"
print sys.path

As one can see the PIL library is in the path .. but still for some
strange reason it can't find the _imaging extention..
It is sort of strange that it wouldn't pick it up then.. any more ideas?

Sascha




From donn at u.washington.edu  Tue Jun 18 16:23:29 2002
From: donn at u.washington.edu (Donn Cave)
Date: 18 Jun 2002 20:23:29 GMT
Subject: socket.inet_aton - bug?
References: <MPG.1771a71d361193fc989683@news.rz.uni-karlsruhe.de> <TaMP8.165343$Fp1.44927@atlpnn01.usenetserver.com>
Message-ID: <aeo4s1$2avk$1@nntp6.u.washington.edu>

Quoth "Steve Holden" <sholden at holdenweb.com>:
| "Uwe Mayer" <merkosh at hadiko.de> wrote ...
|> socket.inet_ntoa() accepts the string '\xff\xff\xff\xff' which is
|> excactly the same i'd expect socket.inet_aton('255.255.255.255') to
|> produce if it accepted the string '255.255.255.255' as an IP adress.
|> socket.inet_aton() dies with socket.error "illegal IP adress".
|>
|> i know this is no valid ip adress, but it is a valid netmask. the
|> purpose of converting the netmask to a binary format was that i wanted
|> to isolate the subnet of an ip adress.
|> i wanted to use test_ip & netmask == ip_range to allow subnets to access
|> a tcp server.
|> is there another way of controlling this or why does socket.inet_aton()
|> behave like that?
|
| I can't really tell you why socket does that, but it seems to be a feature
| of the underlying C socket library, which is where inet_aton() is defined.

Yes (you mean "C socket module", I suppose - at first I thought you were
pointing to the C library inet_aton.)  The good news is that it's only
this one value, which happens to be identical to INADDR_NONE, and that
this value is reliably -1.  So at worst, there's a work-around.

| The address 255.255.255.255 is, in fact, a valid destination for a
| restricted broadcast (one which should be sent out only to hosts on the
| local LAN).

	Donn Cave, donn at u.washington.edu


From jonathan at onegoodidea.com  Wed Jun 26 09:00:48 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Wed, 26 Jun 2002 14:00:48 +0100
Subject: Python 2.2 __slots__ Bug?
References: <mailman.1025089474.17780.python-list@python.org>
Message-ID: <B93F7990.CECB%jonathan@onegoodidea.com>

On 26/6/2002 12:03, in article
mailman.1025089474.17780.python-list at python.org, "Glyph Lefkowitz"
<glyph at twistedmatrix.com> wrote:

> I get "I'm fine" and then a segmentation fault.  This only happens when
> __slots__ is defined on a new-style class.  Seriously, i'm not making those
> numbers up -- reliably, 43551 objects will crash the interpreter; fewer will
> not.

Interestingly I get the construction working fine, but the 'del' crashes the
interpreter with (on Mac OS X):

-----
Date/Time:  2002-06-26 13:53:19 +0100
OS Version: 10.1.5 (Build 5S66)

Command:    python
PID:        18472

Exception:  EXC_BAD_ACCESS (0x0001)
Codes:      KERN_INVALID_ADDRESS (0x0001) at 0xbff7ffe0

Thread 0 Crashed:
 #0   0x00037bbc in lookdict_string
 #1   0x0003804c in PyDict_GetItem
 #2   0x00018994 in _PyType_Lookup
 #3   0x00017348 in lookup_maybe
 #4   0x00016f38 in call_finalizer
 #5   0x00017054 in subtype_dealloc
 #6   0x00017118 in subtype_dealloc
 #7   0x00017118 in subtype_dealloc
[etc...]
-----

This seems to fail for me at 5430 objects. I couldn't get the allocation to
crash at all (or at least up to a million objects, after which I got bored).

Jonathan



From cliechti at gmx.net  Mon Jun 24 18:32:48 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 25 Jun 2002 00:32:48 +0200
Subject: AF_UNIX + SOCK_DGRAM
References: <aes1he$gv2$1@news.idiom.com> <aet3ai$1l0o$2@nntp6.u.washington.edu> <af84af$1dkt$1@news.idiom.com>
Message-ID: <Xns92386002B71cliechtigmxnet@62.2.16.82>

[posted and mailed]

"James T. Dennis" <jadestar at idiom.com> wrote in
news:af84af$1dkt$1 at news.idiom.com: 

>> Actually think you have run into a bug in Python's socket module.
>> If it's convenient for you to build your own socket module, change
>> it like this and see what happens (I'm looking at version 2.1 here.)
> 
>> *** socketmodule.c.dist Sun Apr 15 17:21:33 2001
>> --- socketmodule.c      Thu Jun 20 10:10:12 2002
>> ***************
>> *** 597,602 ****
>> --- 597,603 ----
>>         case AF_UNIX:
>>         {
>>                 struct sockaddr_un *a = (struct sockaddr_un *) addr;
>> +               a->sun_path[addrlen - (sizeof(*a) -
>> sizeof(a->sun_path))] = 0; 
>>                 return PyString_FromString(a->sun_path);
>>         }
>>   #endif

why not simply use PyString_FromStringAndSize(const char *v,  int len)?

couln't you just file a patch on SF? should i do it?

>> The address returned from recvfrom is (addr, len), but we've been
>> ignoring len on the theory that the string would be NUL-terminated.
>> It isn't, so there can be some garbage on the end.




-- 
Chris <cliechti at gmx.net>



From skip at pobox.com  Tue Jun 11 09:50:18 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 11 Jun 2002 08:50:18 -0500
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
In-Reply-To: <Xns922AD3C74280gustafl@195.100.94.182>
References: <Xns9229EE722467Fgustafl@195.100.94.182>
        <mailman.1023746539.13373.python-list@python.org>
        <Xns922AD3C74280gustafl@195.100.94.182>
Message-ID: <15622.26.84312.207270@12-248-41-177.client.attbi.com>

    Gustaf> Cliff's solution was really fast, but it took a while to
    Gustaf> understand. :)

One thing I frequently forget about is the else: clause for try/except
statements.  This lets you isolate as small a piece of code as possible in
the try: clause, making it a bit easier to understand what triggers the
exception:

    def isin(n, ranges):
        for r in ranges:
            try: 
                s, e = r
            except TypeError:
                if n == r:
                    return 1
            else:
                if s <= n <= e:
                    return 1
        return 0

You could also fold the int case into the tuple case (making it a bit more
uniform), but at a slight cost in performance if you have lots of ints.  In
this case, all the try/except statement is responsible for is setting s and
e:

    def isin(n, ranges):
        for r in ranges:
            try: 
                s, e = r
            except TypeError:
                s, e = n, n
            if s <= n <= e:
                return 1
        return 0

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html





From kseehof at neuralintegrator.com  Tue Jun 11 21:55:31 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Tue, 11 Jun 2002 18:55:31 -0700
Subject: (no subject)
Message-ID: <EFEFIHAINIINNELLHFAIEECACJAA.kseehof@neuralintegrator.com>

Where are the windows binaries for python22_d.lib?
Do I need to build the sources to get it?

Thanks,
- Ken





From dryose at juno.com  Tue Jun  4 14:57:34 2002
From: dryose at juno.com (yose)
Date: 4 Jun 2002 11:57:34 -0700
Subject: mySQL on Solaris .8
References: <79d1869.0206030646.512af52c@posting.google.com>        <79d1869.0206031202.228c4387@posting.google.com> <mailman.1023135185.13149.python-list@python.org>
Message-ID: <79d1869.0206041057.2c73a2c7@posting.google.com>

Skip Montanaro <skip at pobox.com> wrote in message news:<mailman.1023135185.13149.python-list at python.org>...
> >> I am trying to install a mySQL database in solaris 2.8. I have a copy
>     >> of the distro for mySQL-python-0.9.1. When I try to run setup.py, I
>     >> am told that I am missing distutils.core
>  
>     >> Where do I get this? What package is it part of?
> 
> Joe,
> 
> It's part of the core Python distribution (since 2.0 or 2.1 I think).
> Sounds like either you're running an older version of Python or the one you
> have is misinstalled in some way.  Assuming you have MySQL's client libs and
> a recent enough version of Python correctly installed, Installing
> mysql-python should take less than 30 seconds from the time you execute
> 
>     python setup.py install


Interesting. It seems like my distro (2.0; the latest that cleanly
supports sockets, etc.) DOES have distutils. However, python itself
doesnt seem to notice. What system variable do I need to set to point
to the Lib/ directory?

Joe


From BPettersen at NAREX.com  Sun Jun 30 18:07:18 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Sun, 30 Jun 2002 16:07:18 -0600
Subject: How to get rid the new line
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158F15@admin56.narex.com>

> From: Frank Tobin [mailto:ftobin at neverending.org] 
> 
> Bjorn Pettersen, on 2002-06-30, wrote:
> 
> > I submitted a patch for this (that handled mac, win, and unix line
> > endings) but it was declined by the BDFL. I suspect one of 
> > the reasons were that I named it chomp...
> 
> sed s/chomp/bite/ < patch | mail python-bugs-list at python.org

I believe in CVS Python you can say 

  line = line.rstrip('\n')

to get a sufficiently similar effect to chomp.

2.3-will-be-out-any-day-now-<wink>'ly y'rs
-- bjorn




From jiba at tuxfamily.org  Thu Jun 20 15:18:24 2002
From: jiba at tuxfamily.org (Lamy Jean-Baptiste)
Date: Thu, 20 Jun 2002 21:18:24 +0200
Subject: [ANN] EventObj
Message-ID: <mailman.1024601109.12981.clpa-moderators@python.org>

EventObj 0.2

http://oomadness.tuxfamily.org/p-eventobj.php

EventObj is a (hackish) pure Python module that allows to add
modification callback (=event) to any Python object (including
lists and dicts).

EventObj requires at least Python 2.2 (tested with 2.2.1).

EventObj can be usefull for debugging, for IDE (to spy object
modification ?), for OO databases, or just for maitaining
anything up-to-date with an object.

Python hack-lovers may be interesting at looking the pure Python
source code (I use __setattr__ on ).

EventObj is Free Software and is available under the GNU GPL.

Enjoy the hack !

Jiba -- Jean-Baptiste LAMY -- jiba at tuxfamily.org
                              http://oomadness.tuxfamily.org




From rnd at onego.ru  Sat Jun  1 14:54:49 2002
From: rnd at onego.ru (Roman Suzi)
Date: Sat, 1 Jun 2002 22:54:49 +0400 (MSD)
Subject: prototyping good OOdesign in Python?
Message-ID: <Pine.LNX.4.33.0206012225290.16212-100000@rnd.onego.ru>

Maybe this was already discussed, but I'd liked to raise this topic again.

Python has superb OOProgramming support. Lets not dispute this for a
moment. Now let's consider that we need to make well-designed OOP solution
first prototyped in Python.

What troubles are there? Python is too dynamic and all that. And all
those nifty features need to be translated properly into, say, C++.

Recently I encountered the following piece of code
(from Arts++ project):

//===========================================================================
//  Copyright Notice
//
//  By accessing this software, arts++, you are duly informed
//  of and agree to be bound by the conditions described below in this
//  notice:
//
//  This software product, arts++, is developed by Daniel W. McRobb, and
//  copyrighted(C) 1998 by the University of California, San Diego
//  (UCSD), with all rights reserved.  UCSD administers the CAIDA grant,
//  NCR-9711092, under which part of this code was developed.
//
//  There is no charge for arts++ software.  You can redistribute it
//  and/or modify it under the terms of the GNU Lesser General Public
//  License, Version 2.1, February 1999, which is incorporated by
//  reference herein.
//
//  arts++ is distributed WITHOUT ANY WARRANTY, IMPLIED OR EXPRESS, OF
//  MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE or that the use
//  of it will not infringe on any third party's intellectual
//  property rights.
//
//  You should have received a copy of the GNU Lesser General Public
//  License along with arts++.  Copies can also be obtained from:
//
//    http://www.gnu.org/copyleft/lesser.html
//
//  or by writing to:
//
//  Free Software Foundation, Inc.
//  59 Temple Place, Suite 330
//  Boston, MA 02111-1307
//  USA
//
//  Or contact:
//
//    info at caida.org
//===========================================================================

//  constructor

............skip

Arts::Arts()
{
  this->_data._ipPath = (ArtsIpPathData *)0;
  this->_data._asMatrix = (ArtsAsMatrixData *)0;
  this->_data._netMatrix = (ArtsNetMatrixData *)0;
  this->_data._portTable = (ArtsPortTableData *)0;
  this->_data._portMatrix = (ArtsPortMatrixData *)0;
  this->_data._protocolTable = (ArtsProtocolTableData *)0;
  this->_data._selectedPortTable = (ArtsSelectedPortTableData *)0;
  this->_data._interfaceMatrix = (ArtsInterfaceMatrixData *)0;
  this->_data._nextHopTable = (ArtsNextHopTableData *)0;
  this->_data._bgp4RouteTable = (ArtsBgp4RouteTableData *)0;
  this->_data._rttTimeSeriesTable = (ArtsRttTimeSeriesTableData *)0;
  this->_data._tosTable = (ArtsTosTableData *)0;
  
  #ifndef NDEBUG
    ++_numObjects;
  #endif
}

Arts::Arts(const Arts & arts)
{
  this->_header = arts.Header();
  this->_attributes = arts.Attributes();

  switch (this->_header.Identifier()) {
    case artsC_OBJECT_IP_PATH:
      this->_data._ipPath = new ArtsIpPathData;
      assert(this->_data._ipPath != (ArtsIpPathData *)0);
      *(this->_data._ipPath) = *(arts.IpPathData());
      break;
      
    case artsC_OBJECT_AS_MATRIX:
      this->_data._asMatrix = new ArtsAsMatrixData;
      assert(this->_data._asMatrix != (ArtsAsMatrixData *)0);
      *(this->_data._asMatrix) = *(arts.AsMatrixData());
      break;

    case artsC_OBJECT_NET:
      this->_data._netMatrix = new ArtsNetMatrixData;
      assert(this->_data._netMatrix != (ArtsNetMatrixData *)0);
      *(this->_data._netMatrix) = *(arts.NetMatrixData());
      break;

    case artsC_OBJECT_PORT:
      this->_data._portTable = new ArtsPortTableData;
      assert(this->_data._portTable != (ArtsPortTableData *)0);
      *(this->_data._portTable) = *(arts.PortTableData());
      break;

    case artsC_OBJECT_SELECTED_PORT:
      this->_data._selectedPortTable = new ArtsSelectedPortTableData;
      assert(this->_data._selectedPortTable != (ArtsSelectedPortTableData *)0);
      *(this->_data._selectedPortTable) = *(arts.SelectedPortTableData());
      break;
      
    case artsC_OBJECT_PORT_MATRIX:
      this->_data._portMatrix = new ArtsPortMatrixData;
      assert(this->_data._portMatrix != (ArtsPortMatrixData *)0);
      *(this->_data._portMatrix) = *(arts.PortMatrixData());
      break;

    case artsC_OBJECT_PROTO:
      this->_data._protocolTable = new ArtsProtocolTableData;
      assert(this->_data._protocolTable != (ArtsProtocolTableData *)0);
      *(this->_data._protocolTable) = *(arts.ProtocolTableData());
      break;

    case artsC_OBJECT_TOS:
      this->_data._tosTable = new ArtsTosTableData;
      assert(this->_data._tosTable != (ArtsTosTableData *)0);
      *(this->_data._tosTable) = *(arts.TosTableData());
      break;

    case artsC_OBJECT_INTERFACE_MATRIX:
      this->_data._interfaceMatrix = new ArtsInterfaceMatrixData;
      assert(this->_data._interfaceMatrix != (ArtsInterfaceMatrixData *)0);
      *(this->_data._interfaceMatrix) = *(arts.InterfaceMatrixData());
      break;
      
    case artsC_OBJECT_NEXT_HOP:
      this->_data._nextHopTable = new ArtsNextHopTableData;
      assert(this->_data._nextHopTable != (ArtsNextHopTableData *)0);
      *(this->_data._nextHopTable) = *(arts.NextHopTableData());
      break;

    case artsC_OBJECT_BGP4:
      this->_data._bgp4RouteTable = new ArtsBgp4RouteTableData;
      assert(this->_data._bgp4RouteTable != (ArtsBgp4RouteTableData *)0);
      *(this->_data._bgp4RouteTable) = *(arts.Bgp4RouteTableData());
      break;

    case artsC_OBJECT_RTT_TIME_SERIES:
      this->_data._rttTimeSeriesTable = new ArtsRttTimeSeriesTableData;
      assert(this->_data._rttTimeSeriesTable != (ArtsRttTimeSeriesTableData *)0);
      *(this->_data._rttTimeSeriesTable) = *(arts.RttTimeSeriesTableData());
      break;
      
    default:
      break;
  }

  #ifndef NDEBUG
    ++_numObjects;
  #endif
}

..........skip the rest of file - there are 2-3 more similar
.......... switch-constructs.

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

This code looks like... automatically generated. Rather than manually
written. I do not know why it was written in C++ at all. Of course, I
imagine 2-3 variants of how to refactor it if I were writing the code
in Python. I am pretty sure there were no select statement (switches)
in Python code, because that is what polymorphism is about.

But then again, translating back to C++ could be difficult. The
mentioned example is all about reading a specially formatted binary
file to produce certain slices of information. And for each
class it uses separate .cc/.hh file! While all the difference
between different classes - structure of corresponding data entries.

My approach in this case could be one resource file, with format
specification (like DTD). Plus one univarsal class to handle big
binary file according to "DTD"). Plus maybe separate resource file
with data query specifications. And it could be at most 80 Kb Python
project, not that monster Arts++ is in both source and binary forms...

Well, if you still follow this thread, the question is - how well
Python could serve as an ARCHITECTURAL prototype if it has too rich
OOP facilities? That is, is Python of help when trying to
prototype certain design? Developing in C++ looks so unnecessary hard
after things were done in Python... 
Or, it may be put this way: what discipline a Python programmer must
obey to allow it's prototype to be conveniently rewriten in C++?

The above thoughts aren't probably well-formed. But I hope
my concern is understood.


Sincerely yours, Roman Suzi
-- 
\_ Russia \_ Karelia \_ Petrozavodsk \_ rnd at onego.ru \_
\_ Saturday, June 01, 2002 \_ Powered by Linux RedHat 7.2 \_
\_ "Tried to play my shoehorn... all I got was footnotes!" \_





From hwcowan at hotmail.com  Fri Jun  7 00:16:22 2002
From: hwcowan at hotmail.com (Hugh Cowan)
Date: 6 Jun 2002 21:16:22 -0700
Subject: Creating Dynamic Web Pages (ZOPE)
Message-ID: <46ca81a0.0206062016.789d3af6@posting.google.com>

Hello,

I am not exactly sure where to post this question -- the only
reference that I could find for ZOPE in the Newsgroups is here.  I
appologize if this is not the correct place.

I have the situation where I need to create a simple Intranet for my
company to display dynamic / data driven web-pages.  We are running a
Windows NT network with MS-SQL Server and IIS.

I had originally thought of using ASP (Server Side Scripting) to
generate the dynamic web-pages.  While I know how to create the
Intranet using ASP, I have yet to find any RAD type Web-Tools that
will allow me to quickly and easily generate the ASP pages required,
and I don't fancy using the Notepad approach either.

My scenario is typical where I need to put together a company Intranet
with a few front end Web-Pages where users can select information,
various links, and enter search criteria.  The information is then
retrieved from the MS-SQL Server and the results are sent back to the
user.  It's nothing fancy, no shopping carts, or remembering sessions
states, etc.. just really a simple front-end interface for
non-technical people to easily retrieve information.
 
I don't know if ZOPE would fit my situation or not.  I know that it is
/ can be used to create complex E-commerce sites, but I am wondering
if it is overkill for what I need done.  Would I be better off trying
to find some sort of RAD Web Tool instead (anyone have any good
suggestions) ?

Thanks so much,

Hugh


From djc at object-craft.com.au  Tue Jun 18 17:40:52 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 19 Jun 2002 07:40:52 +1000
Subject: Python to MS SQL Server
References: <mailman.1024428323.19642.python-list@python.org>
Message-ID: <m38z5c6za3.fsf@ferret.object-craft.com.au>

>>>>> "Bjorn" == Bjorn Pettersen <BPettersen at NAREX.com> writes:

>> From: Mark McEahern [mailto:marklists at mceahern.com]
>> 
>> > I am interested in using Python in a MS SQL Server > environment,
>> but do > not see any database modules out there, especially for >
>> native versus > using ODBC. Is there such a module?
>> 
>> This is a total guess, because I haven't done this myself:
>> 
>> Have you tried using win32com to use ADO?

Bjorn> AFAIK, there is no Python database module that goes directly to
Bjorn> MS SQL Server db lib. ADO does work however, so that's
Bjorn> certainly an option.  Personally, I would probably investigate
Bjorn> e.g. mxODBC first (the interface is much easier to use from
Bjorn> Python).

        http://www.object-craft.com.au/projects/mssql/

The page is a bit alarmist, but is it worth trying the module - what
can you lose?

I stopped doing development when I realised that MS have effectively
abandoned the DB library in favour of ADO.

- Dave

-- 
http://www.object-craft.com.au


From johnroth at ameritech.net  Mon Jun 17 07:44:16 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 17 Jun 2002 07:44:16 -0400
Subject: [development doc updates]
References: <mailman.1024057266.26182.python-list@python.org> <ugk341aivqpl27@news.supernews.com> <lku1o2l062.fsf@pc150.maths.bris.ac.uk>
Message-ID: <ugriqqbhlt4u15@news.supernews.com>

"Michael Hudson" <mwh at python.net> wrote in message
news:lku1o2l062.fsf at pc150.maths.bris.ac.uk...
> "John Roth" <johnroth at ameritech.net> writes:
>
> > "Fred L. Drake" <fdrake at acm.org> wrote in message
> > news:mailman.1024057266.26182.python-list at python.org...
> > > The development version of the documentation has been updated:
> > >
> > >     http://www.python.org/dev/doc/devel/
> > >
> > > Updated to reflect recent changes.
> >
> > I'm not certain I understand item 5. Extended Slices.
>
> Where specifically?  In the whatsnew document?  That section isn't
> finished yet.
>
> > Does this mean that the standard sequence types now support
> > the stride operand,
>
> Yes.  Well, at least lists, tuples, strings (both flavours), arrays.
>
> > or that the stride operand is being depreciated and will eventually
> > vanish?
>
> No.  Assuming you mean "deprecated" -- a common typo!.

Possibly I didn't make myself clear. Item 5 in the What's New
document contains two paragraphs. It's not clear that the second
paragraph is talking about something completely different than
the topic in the header (Extended Slices). Hence my original question.

John Roth
>
> Cheers,
> M.
>
> --
> 93. When someone says "I want a programming language in which I
>     need only say what I wish done," give him a lollipop.
>   -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html




From printers at sendme.cz  Thu Jun 27 02:24:37 2002
From: printers at sendme.cz (A)
Date: Thu, 27 Jun 2002 08:24:37 +0200
Subject: How to find out operating system
Message-ID: <3D1ACBC5.22482.D4313A4@localhost>

Hi,
What is the best way of  finding out the kind operating system?
I can use os.name but if the system is Windows I would like also 
know if the system is Windows98 or Windows ME or W2K or 
Windows XP.
Thanks for help.
Ladislav






From siegfried.gonzi at kfunigraz.ac.at  Sat Jun 22 09:59:10 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Sat, 22 Jun 2002 15:59:10 +0200
Subject: Python hits the spot
Message-ID: <3D1482AE.8FA299D8@kfunigraz.ac.at>

A few weeks ago I complaint here that Python on Windows XP sucks. I had
a problem which would took theoretically about 1.5 hours; but in reality
it takes 6 hours on Windows XP and Python 2.2. After that 6 hours I must
reboot my system in order to work on.

In the following I got the advice to install a real user system on my
laptop. Okay, I bought Linux SuSE 8.0 and installed it on my laptop (256
MB RAM, 1000 MHz Celeron). I also installed Python 2.2 (the ActivePython
version, otherwise I couldn't install idle due to the annoying
"free_software_garbage_2_weeks_installation_nightmare").

SWIG helped me to make my external Python C function (with gcc); and
F2PY helped me to create my external Python Fortran 90 function (with
the Intel Fortran 95 compiler).

The directories and all the other programs and data are the same as on
Windows.


1. After starting the process from within idle the complete system
freezes after the first wavelength (after about 20 minutes). I couldn't
kill a process or open a different session (via F2) in order to kill
python. The last resort: turn off the power supply.

2. I thought it is better to start the process from the command line and
in a X terminal: python forcing.py
The same behavior as aforementioned: last resort: power supply. That is
really good and fine because a Linux system really is greedy for "shut
down immediately".

3. I thought it has to do with Gnome and I logged in to a KDE session.
Same behavior as in point 1. and 2.


My conclusion: I will rewrite my simulation in Fortran 90 and will
abadon Python. Python is maybe good for some small scripts but not
appropiate for serious programming.

Thank you Billy Boy for a forgiving user system Windows XP which in turn
relieves Python's "wreak havoc" behavior.

I cannot believe that Linux is responsible for Python's ill behavior
under Linux. Before you start ranting about note the following: In a
first step it has nothing to do with my programming style (admittedly
not very good and famous) it has to do that the same calculation cannot
be performed on Linux and Python 2.2!


S. Gonzi


From Mitch.Chapman at bioreason.com  Tue Jun 25 17:57:38 2002
From: Mitch.Chapman at bioreason.com (Mitch Chapman)
Date: Tue, 25 Jun 2002 15:57:38 -0600
Subject: Memory Monitor
References: <pan.2002.06.25.12.14.06.994352.17223@pobox.com> <3D1894B5.720EB7F7@bioreason.com> <200206252139.QAA88038@starbase.neosoft.com>
Message-ID: <3D18E752.7A5A4EAC@bioreason.com>

Cameron Laird wrote:
> Mitch Chapman  <Mitch.Chapman at bioreason.com> wrote:
> >        os.system("cat /proc/self/status | grep VmSize")
> Rather than
>   os.system("grep VmSize /proc/self/status")
> ?

No, that's also concise =8)

-- 
Mitch


From maxm at mxm.dk  Fri Jun 14 03:56:33 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 14 Jun 2002 09:56:33 +0200
Subject: Date formats
Message-ID: <3D09A1B1.6040705@mxm.dk>

I have a project where I have defined the following date format:

     gDateFormat = '%d-%m-%Y'

This give dates of the type:

     01-01-2002

For some reason my customer wants dates of the type:

     1-1-2002

Without the leading zero in single digit dates. But when I look in the 
docs for date time formats there is no way to specify days and months in 
that format.

Off course it's easy to write a function which returns the correct date 
format but can this really be true that I have to do it?

I want batteries dam'n it!

;-)


regards Max M



From lailian98 at hotmail.com  Sat Jun  1 06:01:48 2002
From: lailian98 at hotmail.com (Hazel)
Date: 1 Jun 2002 03:01:48 -0700
Subject: Extracting data from HTML
References: <82096df4.0205311152.5891a17b@posting.google.com> <1022883953.6150.5802.camel@localhost> <mailman.1022908262.30070.python-list@python.org>
Message-ID: <82096df4.0206010201.736c4b2c@posting.google.com>

Geoff Gerrietts <geoff at gerrietts.net> wrote in message news:<mailman.1022908262.30070.python-list at python.org>...
> Quoting Ian Bicking (ianb at colorstudy.com):
> > On Fri, 2002-05-31 at 14:52, Hazel wrote:
> > > how do i write a program that
> > > will extract info from an HTML and print
> > > of a list of TV programmes, its Time, and Duration
> > > using urllib?
> > 
> > You can get the page with urllib.  You can use htmllib to parse it, but
> > I often find that regular expressions (the re module) are an easier way
> > -- since you aren't looking for specific markup, but specific
> > expressions.  You'll get lots of false negatives (and positives), but
> > when you are parsing a page that isn't meant to be parsed (like most web
> > pages) no technique is perfect.
> 
> Definitely agree with this sentiment.
> 
> I'll go a step farther, and do a little compare/contrast.
> 
> Once upon a time, I wanted to grab data from the
> weatherunderground.com website. I know there are lots of better ways
> to go about getting this information, these days, but I was not so
> well-informed back then.
> 
> So I wanted to grab this information, and I tried using regular
> expressions to mangle the page. But truthfully, it was just too hard
> to do. I could guess about where in the file the table with all the
> info would appear, but getting a regular expression that was inclusive
> enough to catch all the quirks, yet exclusionary enough to filter out
> all the other embedded tables, proved a very large challenge.
> 
> That's when the idea of a parser made a lot of sense.
> 
> I could push the whole page through a parser, looking for one
> particular phrase in a <TH> element, and from that point forward, map
> <TH> elements to <TD> elements effectively. It became a very simple
> exercise, because I knew how to find that info.
> 
> But as Ian rightly points out, htmllib and a real parser can be very
> heavy if you're just looking to grab unformatted info -- or if you
> can't rely on the formatting to be reliable.
> 
> Both techniques are worth knowing -- but better than either would be
> finding a way to get the information you're after via XML-RPC or some
> other protocol that's designed to carry data rather than rendering
> instructions.
> 
> Best of luck,
> --G.
Dear Geoff, Ian
I'm relying onsgmllib to do the work....
since htmllib requires heavy coding.
Here, an instance of what I want to extract....
the time of the TV programme >> 12:15:00AM

"<TR>
   <TD align=right bgColor=#000033><FONT color=#ffffff 
   face="verdana, arial, helvetica" size=1>12:15:00 
   AM</FONT></TD>"

So what do u think?

-Thanx
Hazel


From mike at radiomag.com  Mon Jun 17 23:35:16 2002
From: mike at radiomag.com (Michael Steuer)
Date: Tue, 18 Jun 2002 13:35:16 +1000
Subject: Listing all available (not just loaded) modules
Message-ID: <aem9om$888fb$1@ID-140151.news.dfncis.de>

I've been searching the newsgroup for information on this but so far have been
unsuccessful:

Is there a way to dynamically list all available modules (not just those that
are already loaded) in python? By available I mean those modules, which are
installed on the machine executing python (in the lib path(s)).

Cheers,

Mik.e




From in-call at gmx.net  Thu Jun 20 20:28:03 2002
From: in-call at gmx.net (Pablo Ripolles)
Date: 20 Jun 2002 17:28:03 -0700
Subject: Passing inner functions as arguments
Message-ID: <d72779c7.0206201628.33d6a657@posting.google.com>

Hello all!

I am new to python, so sorry if i am asking the obvious. Here it
comes. Is there anything wrong with this code?

# Module to define math tools.
def momentum(f, k, a, b):
# Momentum of kth order of the given function f in [a, b].
    from Quadrature import quadrature
    def integrand(x):
    # Momentum's integrand in terms of the passed function f and k.
        return (f(x)) ** k
    return quadrature(integrand, a, b)

I am not sure if it is correct. It runs OK in Python 2.2.1 but not at
all in Jython 2.1 (java1.4.0). I am not sure about the scoping rules.
I there's something strange, please let me know!

Thanx in advance,
Pablo.


From jdhunter at nitace.bsd.uchicago.edu  Sat Jun 22 10:03:03 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Sat, 22 Jun 2002 09:03:03 -0500
Subject: text search again??????????
References: <mailman.1024681168.18086.python-list@python.org>
Message-ID: <m2fzzfza08.fsf@mother.paradise.lost>

>>>>> "jubafre" == jubafre  <jubafre at zipmail.com.br> writes:

    jubafre> self.text.search(st, 0.0, regexp='true') is it correct???
    jubafre> i?m not a exact string, i?m searchin for winzip, but in
    jubafre> the text appears WinZip and the search doesn?t work, how
    jubafre> use regexp flag?????

What kind of object is self.text?  You say, 'I'm not a exact
string'. I'm not either, but this confuses me in the context of your
question.

My guess is that self.text is a string, and you'll be fine with the
string 'find' method.  You don't need a regex to search for 'winzip'.

self.text.lower().find('winzip') will do a case insensitive search for
winzip and return the index into the string, or '-1' if nothing is
found.  self.text.find('WinZip') will do a case sensitive search.

If self.text is a rgx object, then you could use it's search method,
but the syntax would be:

self.text = re.compile('WinZip')  # see the docs for case insensitive
self.text.search(s)               # s is the string you want to search.

But if this is the case, text would be a bad name.

self.rgxWinzip would be better than self.text.


Good luck,
John Hunter



The re search method:


`search(string[, pos[, endpos]])'
     Scan through STRING looking for a location where this regular
     expression produces a match, and return a corresponding
     `MatchObject' instance.  Return `None' if no position in the
     string matches the pattern; note that this is different from
     finding a zero-length match at some point in the string.

     The optional POS and ENDPOS parameters have the same meaning as
     for the `match()' method.


The string find method:  

`find(s, sub[, start[,end]])'
     Return the lowest index in S where the substring SUB is found such
     that SUB is wholly contained in `S[START:END]'.  Return `-1' on
     failure.  Defaults for START and END and interpretation of
     negative values is the same as for slices.



From sgt at outline.ru  Tue Jun  4 08:49:19 2002
From: sgt at outline.ru (Sergei Barbarash)
Date: Tue, 04 Jun 2002 16:49:19 +0400
Subject: sharing data in list sub-classes
Message-ID: <87znybyzsg.fsf@zaraza.fep.ru>

Hello,

Is there a way to share list data in classes that inherit the "list" class?

Example:

,----
| class A(list):
|     <what should I put here?>
| 
| mem = [1,2,3,4,5]
| 
| a1 = A(mem)
| a2 = A(mem)
| 
| a1[0] = 1024
| if a2[0] == 1024:
|     print "Everything's fine"
`----

With obsolete UserList module it works fine, because it stores the data in
self.data variable. Is there a "right" 2.2-ish way to do the same thing?

Thanks,
-- 
Sergei


From dryose at juno.com  Mon Jun  3 10:46:25 2002
From: dryose at juno.com (yose)
Date: 3 Jun 2002 07:46:25 -0700
Subject: mySQL on Solaris .8
Message-ID: <79d1869.0206030646.512af52c@posting.google.com>

I am trying to install a mySQL database in solaris 2.8. I have a copy
of the distro for mySQL-python-0.9.1. When I try to run setup.py, I am
told that I am missing distutils.core

Where do I get this? What package is it part of?

Thanks!


From tdelaney at avaya.com  Sun Jun 16 20:59:58 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 17 Jun 2002 10:59:58 +1000
Subject: filter2
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A390@natasha.auslabs.avaya.com>

# Baseline 1
def filter2a (test, l):
    l1 = filter(test, l)
    l2 = filter(lambda x, test=test: not test(x), l)
    return l1, l2

# Baseline 2
def filter2b (test, l):
    l1 = []
    l2 = []
    l1append = l1.append
    l2append = l2.append

    for x in l:
        if test(x):
            l1append(x)
        else:
            l2append(x)

    return l1, l2
 
# Based on Jeff Epler's - won't work < Python 2.0
def filter2t (test, l):
    l1 = []
    l2 = []
    funcs = (l1.append, l2.append,)
    map(lambda x: funcs[not test(x)](x), l)
    return l1, l2

# Based on Jeff Epler's - won't work < Python 2.0
def filter2l (test, l):
    l1 = []
    l2 = []
    funcs = [l1.append, l2.append,]
    map(lambda x: funcs[not test(x)](x), l)
    return l1, l2

# Based on Jeff Epler's
def filter3t (test, l):
    l1 = []
    l2 = []
    map(lambda x, test=test, funcs=(l1.append, l2.append,): funcs[not
test(x)](x), l)
    return l1, l2

# Based on Jeff Epler's
def filter3l (test, l):
    l1 = []
    l2 = []
    map(lambda x, test=test, funcs=[l1.append, l2.append,]: funcs[not
test(x)](x), l)
    return l1, l2

# Based on Bengt Richter's - won't work < Python 2.0
def filter4 (test, l):
    l2 = []
    append = l2.append
    return [x for x in l if test(x) or append(x)], l2
 
# Based on Bengt Richter's
def filter5 (test, l):
    l2 = []
    return filter(lambda x, test=test, append=l2.append: test(x) or
append(x), l), l2

import operator

def test (x, truth=operator.truth):
    return truth(x)

def bench (f, r, l, t):
    """"""
    import time
    clock = time.clock

    start = clock()
    for i in r:
        f(t, l)
    t = clock() - start
    print '%-10s %s' % (f.__name__ + ':', t,)

r = range(1000)
l = range(1000)
filters = (filter2a, filter2b, filter2t, filter2l, filter3t, filter3l,
filter4, filter5,)

print 'Python function'
print
for func in filters:
    bench(func, r, l, test)

print
print 'C function'
print
for func in filters:
    bench(func, r, l, operator.truth)

---
Python 2.2.1 Windows 2000

Python function

filter2a:  6.18176173589
filter2b:  3.81115916127
filter2t:  4.99432009837
filter2l:  4.9499274483
filter3t:  4.8721376543
filter3l:  4.75815178722
filter4:   3.86890220203
filter5:   3.73075037996

C function

filter2a:  3.6114880356
filter2b:  2.18989099755
filter2t:  3.73375104874
filter2l:  3.60997196065
filter3t:  3.68825664697
filter3l:  3.58173029097
filter4:   2.21596360152
filter5:   2.64008410331

So, the most naive version (with common optimisations) works fastest for C
functions (operator.truth *is* a C function, isn't it?) but the difference
is minimal, Bengt's works fastest for Python functions with a fair
difference. I actually expected it to be the other way around ... 

Tim Delaney




From pyth at devel.trillke.net  Thu Jun 13 14:58:52 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 13 Jun 2002 20:58:52 +0200
Subject: why not "'in' in 'in'"?
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E20192158ED1@admin56.narex.com>; from BPettersen@NAREX.com on Thu, Jun 13, 2002 at 11:35:09AM -0600
References: <60FB8BB7F0EFC7409B75EEEC13E20192158ED1@admin56.narex.com>
Message-ID: <20020613205852.E6609@prim.han.de>

Bjorn Pettersen wrote:
> > From: Grant Griffin [mailto:not.this at seebelow.org] 
> > That's not a bad suggestion, Mark, but if I did that I guess 
> > I would be forever wondering if I was testing whether the 
> > first one was in the second or the second one was in the 
> > first.  Again, it doesn't read nearly as well as:
> > 
> >    if x in subject:
> >        ...
> > 
> > which leaves no doubt.
> 
> Ok, how about:
> 
> >>> class mystr(str):
> ...   def __contains__(self, s):
> ...     return self.find(s) >= 0
> ...
> >>> 'hello' in mystr('hello world')
> 1
> >>> 'foo' in mystr('hello world')
> 0
> >>>

that's fine. but usually you don't want to make up an 
instance for every string you get from somewhere. 
I still think that patching 'stringobject.c' to get the
reasonable behaviour might be worth a try for python2.3. 

    holger




From shagshag13 at yahoo.fr  Sat Jun  1 09:34:42 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Sat, 1 Jun 2002 15:34:42 +0200
Subject: Which is fastest dict or list.index() ?
Message-ID: <adaih2$vtb8c$1@ID-146704.news.dfncis.de>

I need a structure to hold a correspondence between toons and id like

{'Bugs' : 0, 'Mickey' : 1, 'Babs' : 2, 'Jerry' : 3, 'Tom' : 4, 'Babs' : 5}

I must be able to alpha sort these toons (without loosing their id), and
access a toon from an id or an id from a toon.

By now a use a dict and a list : the dict gives index from a toon, and the
list gives me a toon from an index.

{'Bugs' : 0, 'Mickey' : 1, 'Babs' : 2, 'Jerry' : 3, 'Tom : 4, 'Babs' : 5}
['Bugs', 'Mickey', 'Babs', 'Jerry', 'Tom', 'Babs']

(the id is used somewhere else a key to a collection of elements)

My question is which is the best way (best = quickest and memory safe) to
handle this ?
Should i subclasse from dict or list ?
Should i consider using only one list with toons (and use their index as id)
?
Should i use 2 dicts ?

Many thanks to you for helping

s13.





From aahz at pythoncraft.com  Thu Jun  6 20:45:20 2002
From: aahz at pythoncraft.com (Aahz)
Date: 6 Jun 2002 20:45:20 -0400
Subject: Medium to Large Scale Python Deployments
References: <mailman.1023405789.30972.python-list@python.org>
Message-ID: <adovn0$i9$1@panix1.panix.com>

In article <mailman.1023405789.30972.python-list at python.org>,
Sam Penrose  <spenrose at intersight.com> wrote:
>
>I'm also curious about large amounts of Python, period. Who knows of 
>~50K Python lines-of-code (LOC) systems? 100K LOC? More?

Well, there's Zope.  My last formal employment was at a job where we had
something like 30-40K LOC.  I'd imagine it's moderately rare to get much
larger than that, simply because Python is so compact.  (A good
refactoring run at my last job ought to have shrunk the code by at least
10K, IMO.)
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From rnd at onego.ru  Sat Jun  8 01:57:54 2002
From: rnd at onego.ru (Roman Suzi)
Date: Sat, 8 Jun 2002 09:57:54 +0400 (MSD)
Subject: Efficient python programming...
In-Reply-To: <3D013FB1.3D7477EA@engcorp.com>
Message-ID: <Pine.LNX.4.33.0206080953280.23456-100000@rnd.onego.ru>

On Fri, 7 Jun 2002, Peter Hansen wrote:

>Roman Suzi wrote:
>> 
>> On Fri, 7 Jun 2002, Peter Hansen wrote:
>> > You forgot the even more important first thing for a beginner:
>> > get it correct!  The only good way to do this is to write unit
>> > tests that prove it.  If you don't have tests, you don't have
>> > working code (or how can you prove it?).
>> 
>> To prove that code is correct is very difficult (practically impossible).
>> Unittests are there to prove that you have errors, not the other way
>> around.
>> 
>> Of course, they can increase your confidence in that optimized code is
>> equivalent to what you had before.
>
>I disagree, if the tests are written first (and run, proving that
>they will fail when the code is wrong or missing).
>
>The tests are executable requirement specifications.  Running
>them "proves" (yes, not 100%... nothing could) that the code
>meets the specs.  The specs could be wrong of course, but that
>doesn't mean the code is broken...

I can agree that tests give practical confidence in the 
tested code, but I was speaking about rigourous 
verification that code is 'correct'.

And tests can test only some subset of input-output
mapping. Making tests is an art, is not formal act.

I had no intention to say tests are useless, only that
they are not a method of proving the correctness of code
or correctenss of code modifications.


Sincerely yours, Roman Suzi
-- 
\_ Russia \_ Karelia \_ Petrozavodsk \_ rnd at onego.ru \_
\_ Saturday, June 08, 2002 \_ Powered by Linux RedHat 7.2 \_
\_ "Things are getting worse. Please send chocolate." \_





From peter at engcorp.com  Mon Jun 24 11:19:00 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 24 Jun 2002 11:19:00 -0400
Subject: sys.exc_info() into string
References: <KZoR8.44805$n4.10504231@newsc.telia.net> <3D1635B2.99BCD944@engcorp.com> <00GR8.44905$n4.10556369@newsc.telia.net>
Message-ID: <3D173864.EF3A6AE9@engcorp.com>

Magnus wrote:
> 
> Peter Hansen wrote:
> > try:
> >     1/0
> > except Exception, ex:
> >     print "Error: %s" % ex
> >
> > Good enough??
> 
> Thanks both of you. I'm still learning Python and issues that are trivial to
> solve for some isn't trivial for me yet. I'm converting from C++ and Java
> and perhaps in a few months I might be a more experienced Python
> programmer.

Hi Magnus, don't think just because our replies are sometimes very
concise that we think the solutions are trivial.  I remember having
to go through the same learning curve as you (short and fun though it
is with Python) and not knowing how to do the above.  

This newsgroup operates by throwing a dozen small answers at people,
and the odd monster response by one of the bots.  Usually, but not
always, we try to teach you how to catch fish rather than just handing
you the bowl of chowder.

Welcome to Python. :)

-Peter


From mwh at python.net  Tue Jun 18 07:31:42 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 18 Jun 2002 11:31:42 GMT
Subject: "Fatal Python error: GC object already in linked list"
References: <yfsfzzleg6s.fsf@black132.ex.ac.uk>
Message-ID: <lkhek0lt79.fsf@pc150.maths.bris.ac.uk>

Alexander Schmolck <a.schmolck at gmx.net> writes:

> Just got the following in an interactive session:
> 
>     >>> reload(awmstools)
>     Exception exceptions.RuntimeError: 'maximum recursion depth exceeded' in  ignored
>     Fatal Python error: GC object already in linked list
>     
>     Process Python aborted
>     
> Any ideas as to what might be the cause?

Is this the lastest version of Python?  Is it reproducible?

Cheers,
M.

-- 
  You can lead an idiot to knowledge but you cannot make him 
  think.  You can, however, rectally insert the information, 
  printed on stone tablets, using a sharpened poker.        -- Nicolai
               -- http://home.xnet.com/~raven/Sysadmin/ASR.Quotes.html


From aahz at pythoncraft.com  Fri Jun 14 23:47:30 2002
From: aahz at pythoncraft.com (Aahz)
Date: 14 Jun 2002 23:47:30 -0400
Subject: Using Python to script a game?
References: <cdc2589b.0206141718.7dc2ae13@posting.google.com> <mailman.1024106376.21850.python-list@python.org>
Message-ID: <aeedci$29e$1@panix1.panix.com>

In article <mailman.1024106376.21850.python-list at python.org>,
Gerhard =?iso-8859-15?Q?H=E4ring?=  <gerhard at bigfoot.de> wrote:
>
>Yes. And it would be a pretty good match, too. There's even been a
>commercial game that uses Python for internal scripting, its name
>escapes me, though.

Blade of Darkness.  It's a first-person shooter, too, so that ought to
be all the evidence needed to prove that Python is fast enough.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From bowman at montana.com  Sat Jun  8 21:03:29 2002
From: bowman at montana.com (bowman)
Date: Sat, 08 Jun 2002 19:03:29 -0600
Subject: Program mysteriously STOPS!
References: <yZuM8.146064$ux5.186314@rwcrnsc51.ops.asp.att.net>
Message-ID: <zMxM8.4$r_1.714@newsfeed.slurp.net>

LJD wrote:
> 
> The following code is designed (at least thats my intent) to run forever.
> The program should sit in a loop until it connects with the server.??It
> should just loop and loop and loop.

where do you close the socket, other than in the exception handler? you may 
be using up all the file descriptors allowed for a process. 


From nookieNO_SP_AM at online.no  Wed Jun  5 12:00:19 2002
From: nookieNO_SP_AM at online.no (Erlend J. Leiknes)
Date: Wed, 05 Jun 2002 16:00:19 GMT
Subject: distutils and distributing "scripts"
References: <adl4e4$19ep$1@news.adamastor.ac.za>
Message-ID: <nAqL8.10168$fG3.343408@news2.ulv.nextra.no>

do not call the files .pyw, just use py

when you then use py2exe, do the following
python setup.py py2exe -w

the -w means --window or --windows (dont remember), anyway, this will use
the pythonw.exe.


"Hugo van der Merwe % str (33) >" <<s1%s61562 at bach.sun.ac.za> wrote in
message news:adl4e4$19ep$1 at news.adamastor.ac.za...
> I have recently started using distutils. I have a .pyw file for the
windows
> users, how do I get this to be a part of my ... sdist, for example?
>
> I have added this to the scripts list passed to "setup()", this seems to
me
> the "correct" place for it (I want to eventually play with py2exe).
>
> It seems "python setup.py sdist" does not add "scripts" by default though,
> just the py_modules. The docs seem to confirm this. How do I get my .pyw
> file distributed, do I really have to go and do all that MANIFEST.in
stuff,
> just for this one file?
>
> Thanks,
> Hugo van der Merwe




From jimmy at retzlaff.com  Tue Jun  4 00:24:44 2002
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Mon, 3 Jun 2002 21:24:44 -0700
Subject: Fatal Python error: GC object already in linked list
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F569D@MAYOR.retzlaff.com>

Amit Mongia [mailto:mongiaamit at yahoo.com] writes:
> I am running a program using pycurl, to upload files on a server. I
> am using Python 2.2.1. I am using threads and lists in my program. the
> application just keeps waiting for new files to arrive and uploads
> them. There are times my program errors out and crashes with this
> error message.
>
> Fatal Python error: GC object already in linked list

I had a similar problem where a background thread was periodically
checking if the main thread had added anything to a list and my program
would crash with this same message (the crash didn't occur anywhere near
the list however). It was quite random and many different things would
make it disappear. One of those things was using the Queue class from
the standard library instead of a list to communicate between my two
threads. Maybe something to try. Good luck.

Jimmy




From skip at pobox.com  Tue Jun 11 10:09:41 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 11 Jun 2002 09:09:41 -0500
Subject: MySQLdb warnings problem
In-Reply-To: <Xns922AA23E3DE4Dandy47halfcookedcom@203.109.252.31>
References: <mailman.1023595512.23403.python-list@python.org>
        <Xns922AA23E3DE4Dandy47halfcookedcom@203.109.252.31>
Message-ID: <15622.1189.268647.346102@12-248-41-177.client.attbi.com>

    Andy> There isn't (as far as I know) a MySQLdb mailing list, but
    Andy> questions of this type are very welcome at the db-sig mailing list
    Andy> (db-sig at python.org or
    Andy> http://mail.python.org/mailman/listinfo/db-sig).

The mysql-python project on SourceForge has two forums (help and open
discussion):

    http://sourceforge.net/forum/?group_id=22307

Andy Dustman's pretty prompt at answering questions.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html





From bokr at oz.net  Mon Jun 24 13:04:18 2002
From: bokr at oz.net (Bengt Richter)
Date: 24 Jun 2002 17:04:18 GMT
Subject: converting an array of chars to a string
References: <3d12dda2_6@news.newsgroups.com> <aevj41$fg2$0@216.39.172.122> <3d17038e_6@news.newsgroups.com>
Message-ID: <af7jei$3kr$0@216.39.172.122>

On Mon, 24 Jun 2002 13:40:51 +0200, JB <jblazi at hotmail.com> wrote:

>Bengt Richter wrote:
>
>> Have a look at the array module. It can construct an array
>> of signed or unsigned character elements optionally
>> initialized with values from a list or string. 
>[...]
>
>Thank you very much. In principle, this was exactly, for 
>what I was looking. But most unforunately if I declare such 
>an array a, then
>
>a[0]=210
>a[0] += 46
>
>produces an error instead of the correct result 0. Now I am 
>a bit desperate.
No cause for desperation ;-)

Are you saying that you want modulo 256 arithmetic on unsigned
8-bit data? No problem. But what does your data represent? And
what does adding a number to such a data item mean in the real world?

Let's look at the problem:

 >>> import array
 >>> a = array.array('B',[210])
  >>> a
 array('B', [210])
 >>> a[0]
 210
 >>> a[0] += 46
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
 OverflowError: unsigned byte integer is greater than maximum

It's just saying you tried to store 256 in a byte. You can't
expect the array module to know what policy you want to use
in dealing with oversize numbers. For another person's application
it might be just as reasonable to clamp >255 at the max value of 255,
and clamp <0 at 0.

If you want to enforce modulo 256, you can:
 >>> a[0]
 210
 >>> a[0] = (a[0]+46)%256
 >>> a[0]
 0

Or do it with a bit mask and bitwise '&' operation:
 >>> a[0]=210
 >>> a[0]
 210
 >>> a[0] = (a[0]+46)&255
 >>> a[0]
 0

Or you could write yourself a wrapper to hide that stuff just for
indexed single item access:

 >>> class A256:
 ...     def __init__(self, byte_arr): self.a = byte_arr
 ...     def __getitem__(self, i): return self.a[i]
 ...     def __setitem__(self, i, v):
 ...         self.a[i] = v&255
 ...     def __len__(self): return len(self.a)
 ...

Now make the array as usual
 >>> import array
 >>> a = array.array('B',[210,46,0,255])

Wrap it
 >>> w = A256(a)

Try some indexed sequence things
 >>> w[0]
 210
 >>> len(w)
 4
 >>> [x for x in w]
 [210, 46, 0, 255]

Note that we can see the original
array as an attribute of the wrapper
 >>> w.a
 array('B', [210, 46, 0, 255])

And the same thing directly, since we
kept a separate 'a' binding
 >>> a
 array('B', [210, 46, 0, 255])

Add some recognizable stuff using array methods
(that aren't available as methods of this wrapper)
 >>> w.a.fromstring('<ABC>')

Notice that 'a' refers to the same
 >>> a
 array('B', [210, 46, 0, 255, 60, 65, 66, 67, 62])
 >>> a.tostring()
 '\xd2.\x00\xff<ABC>'

Now your complaint:
 >>> a[0] +=46
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
 OverflowError: unsigned byte integer is greater than maximum

Which is what it was designed to do, but we now have a wrapper
 >>> w[0] +=46
 >>> w[0]
 0

Which gave the "correct" result, and is also visible through 'a'
 >>> a[0]
 0

Look at another item both ways
 >>> a[1],w[1]
 (46, 46)

Add with no expected overflow
 >>> w[1]+=209
 >>> a[1],w[1]
 (255, 255)

Now expect overflow and "correction"
 >>> w[1]+=1
 >>> a[1],w[1]
 (0, 0)

See it in the array both ways
 >>> a
 array('B', [0, 0, 0, 255, 60, 65, 66, 67, 62])
 >>> w.a
 array('B', [0, 0, 0, 255, 60, 65, 66, 67, 62])
 >>>

Wrapping like that  will cost a little speed, but
you'll have to balance benefits vs ease etc. Unless
you're processing megabytes regularly, you probably
don't have to worry. If speed is a problem, there
will be ways to improve that.
>
>This is like many things in the Python library: they look 
>good at first sight but when you take a closer look at 
>them, you see, that...
... you have a great basis for solving almost any problem,
and all you need is some imagination to tailor them
to your specific requirements? ;-)

Regards,
Bengt Richter


From peter at engcorp.com  Fri Jun  7 08:17:26 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 07 Jun 2002 08:17:26 -0400
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk>
Message-ID: <3D00A456.4315EDA3@engcorp.com>

Eddie Corns wrote:
> 
> A friend is fond of quoting this to me:
> 
>   Rules of Optimization:
>     Rule 1: Don't do it.
>     Rule 2 (for experts only): Don't do it yet.
>   -- M.A. Jackson
> 
> I wasn't going to bother contributing since others have said the same thing
> but the more people say the more you'll believe it :)
> 
> Especially for a beginner the important thing is to get clear and obvious
> code.  Later, IF IT ISN'T FAST ENOUGH, then you think about ways to speed it
> up.

You forgot the even more important first thing for a beginner: 
get it correct!  The only good way to do this is to write unit
tests that prove it.  If you don't have tests, you don't have 
working code (or how can you prove it?).

-Peter


From martin at v.loewis.de  Thu Jun  6 16:35:32 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 06 Jun 2002 22:35:32 +0200
Subject: 64-bit Python?
References: <mailman.1023393193.10285.python-list@python.org>
Message-ID: <m3zny8uovf.fsf@mira.informatik.hu-berlin.de>

"Bjorn Pettersen" <BPettersen at NAREX.com> writes:

> Are you using Win64 on IA-64? Last time I checked (using the MS
> compiler) a long was still 32 bits, and you had to use __int64 to
> get a 64 bit integer...

Yes, Win64 is known for not using LP64. Many people consider this a
mistake. Of those, a majority is not surprised since it comes from
Microsoft :-)

Regards,
Martin


From mkelly2002 at adelphia.net  Sat Jun  8 13:21:17 2002
From: mkelly2002 at adelphia.net (Michael Kelly)
Date: Sat, 08 Jun 2002 13:21:17 -0400
Subject: Removing ^M
References: <mailman.1023549912.12643.python-list@python.org>
Message-ID: <adteeb$250i1$4@ID-100555.news.dfncis.de>

Michael Hall wrote:

> Python ... I love it! I knew about dos2unix but it's not on my box for
> some reason, and downloading and installing it seemed like overkill.
> This script worked like a charm.
> 

If you're on a unix flavor this works pretty well

tr -d '\015' <infile >outfile

-- 

"Only choice is an oxymoron."



From holger at trillke.net  Mon Jun  3 17:43:13 2002
From: holger at trillke.net (holger at trillke.net)
Date: Mon, 3 Jun 2002 23:43:13 +0200
Subject: optimize lambda usages with strings
Message-ID: <20020603234313.A22545@prim.han.de>

Hello fellow-minimalists,

While skimming the python-standard library (2.3a0) i noticed 
some inefficient usages of lambdas, e.g.: 

asyncore.py:    info = '[' + '] ['.join(map(lambda x: '|'.join(x), tbinfo)) + ']'
cgi.py:     plist = map(lambda x: x.strip(), line.split(';'))
inspect.py:                  formatvarargs=lambda name: '*' + name,
inspect.py:                  formatvarkw=lambda name: '**' + name,
inspect.py:                    formatvarargs=lambda name: '*' + name,
inspect.py:                    formatvarkw=lambda name: '**' + name,

For example the last line can be rewritten as:

    formatvarkw='**'.__add__

or the 'asyncore.py' line as

    .... map ('|'.join, tbinfo)

resulting in shorter and faster code. 
Besides i have often seen and used myself the idiom:

    filter(lambda x: x=='whatever', list)

which can be rewritten as

    filter('whatever'.__eq__, list)    (*)

I didn't bother to actually time the speed-increase. 
But i guess it's significant for these simple cases. Anyway,  i 
wanted to share these little thoughts just in case someone
finds it interesting.

have fun,

    holger


(*) although i somewhat dislike the '.__eq__, line noise
    and would prefer to write '.equals, 




From dm_top at hotmail.com  Fri Jun 21 12:04:10 2002
From: dm_top at hotmail.com (Dmitri)
Date: 21 Jun 2002 09:04:10 -0700
Subject: Python+Cygwin+XML SAX build failure of PyXML 0.7.1
Message-ID: <59d8a5f6.0206210804.25353aa3@posting.google.com>

Hello,
(possibly this is the second message, I have had a little prob., so
sorry
if it is, but this message contains the source info :)

Source from 
http://sourceforge.net/project/showfiles.php?group_id=6473:
PyXML-0.7.1.tar.gz 606480  2060  Platform-Independent Source .gz 

Target:
I need the XML SAX parser for Python under Cygwin.
(And I can not find precompiled version for Cygwin)
---------------------------------------------------------------
Precondition:

*Windows 2000 Prof.
*CygWin ver.:  2.194.2.22
*GCC ver.:     2.95.3-5
    Reading specs from /usr/lib/gcc-lib/i686-pc-cygwin/2.95.3-5/specs
    gcc version 2.95.3-5 (cygwin special)
*Python ver.:  2.2.1
    $ python
    Python 2.2.1 (#1, Jun  6 2002, 16:53:43)
    [GCC 2.95.3-5 (cygwin special)] on cygwin

---------------------------------------------------------------
Description of the problem:

After unpacking of the tarball, the issuing the python setup.py build
cause the following compiling error during building the expat
extention:

.
. -- copying ---
.
.
copying xml/xslt/_4xslt.py ->
build/lib.cygwin-1.3.10-i686-2.2/_xmlplus/xslt
copying xml/xslt/__init__.py ->
build/lib.cygwin-1.3.10-i686-2.2/_xmlplus/xslt
running build_ext
building '_xmlplus.parsers.pyexpat' extension
creating build/temp.cygwin-1.3.10-i686-2.2
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -DUSE_DL_IMPORT
-DHAVE_EXPAT_H -DV
ERSION="1.95.2" -DXML_NS=1 -DXML_DTD=1 -DXML_BYTE_ORDER=12
-DXML_CONTEXT_BYTES=1
024 -Iextensions/expat/lib -I/usr/local/include/python2.2 -c
extensions/pyexpat.
c -o build/temp.cygwin-1.3.10-i686-2.2/pyexpat.o
extensions/pyexpat.c:1709: initializer element is not constant
extensions/pyexpat.c:1709: (near initialization for
`handler_info[2].setter')
extensions/pyexpat.c:1712: initializer element is not constant
extensions/pyexpat.c:1712: (near initialization for
`handler_info[3].setter')
extensions/pyexpat.c:1715: initializer element is not constant
extensions/pyexpat.c:1715: (near initialization for
`handler_info[4].setter')
extensions/pyexpat.c:1718: initializer element is not constant
extensions/pyexpat.c:1718: (near initialization for
`handler_info[5].setter')
extensions/pyexpat.c:1727: initializer element is not constant
extensions/pyexpat.c:1727: (near initialization for
`handler_info[8].setter')
extensions/pyexpat.c:1736: initializer element is not constant
extensions/pyexpat.c:1736: (near initialization for
`handler_info[11].setter')
extensions/pyexpat.c:1739: initializer element is not constant
extensions/pyexpat.c:1739: (near initialization for
`handler_info[12].setter')
extensions/pyexpat.c:1742: initializer element is not constant
extensions/pyexpat.c:1742: (near initialization for
`handler_info[13].setter')
extensions/pyexpat.c:1745: initializer element is not constant
extensions/pyexpat.c:1745: (near initialization for
`handler_info[14].setter')
extensions/pyexpat.c:1754: initializer element is not constant
extensions/pyexpat.c:1754: (near initialization for
`handler_info[17].setter')
extensions/pyexpat.c:1757: initializer element is not constant
extensions/pyexpat.c:1757: (near initialization for
`handler_info[18].setter')
extensions/pyexpat.c:1760: initializer element is not constant
extensions/pyexpat.c:1760: (near initialization for
`handler_info[19].setter')
extensions/pyexpat.c:1763: initializer element is not constant
extensions/pyexpat.c:1763: (near initialization for
`handler_info[20].setter')
error: command 'gcc' failed with exit status 1

---------------------------------------------------------------
Workaround:

the simple code review and primitive test like:
void funcA(int a, int b)
{
  printf("from A: %d, %d\n", a, b);
}

void funcB(int a, int b)
{
  printf("from B: %d, %d\n", a, b);
}

typedef void (*PF)(int, int);

struct S {
  const char * _F;
  PF           _pf;
};


extern struct S ars[2];

static struct S ars[] =
  {
    {"funcA", funcA },
    {"funcB", funcB },
  };

int main()
{
  printf("%s ", ars[0]._F);
  ars[0]._pf(1, 2);
  printf("%s ", ars[1]._F);
  ars[1]._pf(3, 4);

  return 0;
}

and 
$ gcc -o test main.c

shows that everything should be all right. But it is not...

---------------------------------------------------------------
Searching the net:
1. I did not find the precompiled version of the PyXML for cygwin.
   Does someone know where can I get it, it it does exist?
2. The problem like "initializer element is not constant" is shown but
the recommendations goes to the updating of the gcc compiler with
unclear direction. That's possibly true, but instalation of the GNU's
projects under cygwin is a different story. And it's hard to beleive
that the upgrating compiler would fix this problem...

Thank you.

With the best regards,
  Dmitri


From K.Rdt at TU-Berlin.DE  Fri Jun 28 02:27:29 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Fri, 28 Jun 2002 08:27:29 +0200
Subject: crypt for windows?
In-Reply-To: <7x6604vsj3.fsf@ruckus.brouhaha.com>
Message-ID: <YT63T4ZW52YTZUE921SPKYX08RMG.3d1c01d1@FRITZweb>

Am 28.06.02 08:14:40, schrieb Paul Rubin <phr-n2002b at NOSPAMnightsong.com>:

>I don't understand your question then.
-------------------------------------------------------------------
Hi
Well I have some batch-files with my login
and password and want to encrypt them
with another password.
Meanwhile I got fcrypt .. 
		K at Rdt
---------------------------------------------------------------------







From james.kew at btinternet.com  Sat Jun 29 15:22:48 2002
From: james.kew at btinternet.com (James Kew)
Date: Sat, 29 Jun 2002 20:22:48 +0100
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com> <mailman.1025336262.12245.python-list@python.org> <3d1dc547.468619140@netnews.attbi.com> <afkv73$f2274$1@ID-139865.news.dfncis.de>
Message-ID: <afl1ei$ffa93$1@ID-71831.news.dfncis.de>

"Dave Kuhlman" <dkuhlman at rexx.com> wrote in message
news:afkv73$f2274$1 at ID-139865.news.dfncis.de...

> candiazoo at attbi.com wrote:
> > 4) I localized constants I was using in boolean
> > tests (there were declared at the beginning of the module, I moved
> > them to the function call just before the loop).
>
> Is finding things in globals slower than finding them in locals?

Yes: you have to look, and fail to find, in the locals before you look in
the globals.

James





From jubafre at zipmail.com.br  Mon Jun 24 10:44:44 2002
From: jubafre at zipmail.com.br (jubafre at zipmail.com.br)
Date: Mon, 24 Jun 2002 11:44:44 -0300
Subject: =?iso-8859-1?Q?HELP=2DME=20pLEASE?=
Message-ID: <3D172BFB00000167@www.zipmail.com.br>

i don?t know how to make a .exe,i?m atachin the file .py

SOmebody can compile for me please a need the .exe urgently!!!!


Juliano
www.gebrasil.hpg.com.br



------------------------------------------
Use o melhor sistema de busca da Internet
Radar UOL - http://www.radaruol.com.br




-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: lfa.py
URL: <http://mail.python.org/pipermail/python-list/attachments/20020624/c5349665/attachment.ksh>

From mfranklin1 at gatwick.westerngeco.slb.com  Mon Jun 17 08:47:39 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Mon, 17 Jun 2002 12:47:39 +0000
Subject: unhashable type when using pickle.load()
In-Reply-To: <aekdei$n7e$1@helios.herts.ac.uk>
References: <aekdei$n7e$1@helios.herts.ac.uk>
Message-ID: <200206171151.g5HBpMB12546@helios.gatwick.geco-prakla.slb.com>

On Monday 17 Jun 2002 10:03 am, you wrote:
> I wonder if anyone can shed some light on a troublesome intermittent
> problem I am having.
>
> Firstly the necessary:
> I am running on windows XP, using jython 2.1 and java 1.3.1_03.
>
> A pickled file is created as follows:
>     pickle.dump((self.seq.getMotifs(), self.s[self.seqIdx]),
>                     newFile)
>
> self.seq.getMotifs() simply returns a dictionary containing various complex
> python objects (which will have been created and used sucessfully)
>
> self.s[self.seqIdx] is a python sequence object.
>
>
> Intermittently, when I try and reload from this pickle file I get the
> following exception:
>
>
> Exception occurred during event dispatching:
> Traceback (innermost last):
>   File "pars.py", line 129, in OpenFile
>   File "D:\program files\jython-2.1\Lib\pickle.py", line 947, in load
>   File "D:\program files\jython-2.1\Lib\pickle.py", line 567, in load
>   File "D:\program files\jython-2.1\Lib\pickle.py", line 885, in
> load_setitem
> TypeError: unhashable type
>
> I have read that this error typically can occur when a pyDictionary or
> pyList is inserted into a hashtable. Is this a problem that anyone else is
> familiar with, is it a limitation of the jython implementation of pickle?
>
> Any help will be appreciated
>
> Blobby


Mark,

Could you be falling foul of a file close() problem IOW are you closing the 
file explicitly?  This could be a GC thing (discussed recently) where java 
rather than python's GC is not closing the file when it goes out of scope....

Could be way off but worth saying.
Cheers,
Martin









From oren-py-l at hishome.net  Sat Jun 15 09:59:15 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Sat, 15 Jun 2002 16:59:15 +0300
Subject: Pronouncing '__init__'
In-Reply-To: <20020615134410.A15079@prim.han.de>; from pyth@devel.trillke.net on Sat, Jun 15, 2002 at 01:44:10PM +0200
References: <20020615132519.A14452@hishome.net> <20020615134410.A15079@prim.han.de>
Message-ID: <20020615165915.A17174@hishome.net>

On Sat, Jun 15, 2002 at 01:44:10PM +0200, holger krekel wrote:
> for oral language:
> 
>     init constructor
> 
> and if the context allows
> 
>     init
> 
> for written languange __init__
> 
> Isn't it obvious? :-)
> 
>     holger

I used __init__ just as an example.  I meant how do you pronounce names 
with leading and trailing double underscore in general.  The magic name 
__init__ may be by far the one most commonly encountered in everyday 
programming but __there__ __are__ __many__ __others__.

	Oren




From jimmy at retzlaff.com  Mon Jun  3 16:08:51 2002
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Mon, 3 Jun 2002 13:08:51 -0700
Subject: Version strings in executable created by Py2exe
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F569B@MAYOR.retzlaff.com>

Robert L. Oelschlaeger [mailto:roelsch at comtrak.com] wrote:
> Running with Python 2.2. I've created myAppName.exe using Py2exe, with
> the "version_companyname" and "version_legalcopyright" specified in
> the setup .cfg file as follows:
...
> Why are the version_companyname and version_legalcopyright strings not
> installed for the "Company Name" and "Copyright"?

"version" should be followed by a dash rather than an underscore. So,
for example, you should use "version-companyname" rather than
"version_companyname".

Jimmy




From giulio.agostini.remove.this at libero.it  Thu Jun  6 17:39:11 2002
From: giulio.agostini.remove.this at libero.it (Giulio Cespuglio)
Date: Thu, 6 Jun 2002 21:39:11 +0000 (UTC)
Subject: Extracting data from HTML
References: <82096df4.0205311152.5891a17b@posting.google.com> <1022883953.6150.5802.camel@localhost> <mailman.1022908262.30070.python-list@python.org> <82096df4.0206010201.736c4b2c@posting.google.com> <23891c90.0206050321.d6a64c3@posting.google.com>
Message-ID: <4slvfucpesg4ikfs3dfrjo8deio5de24ph@4ax.com>

>  http://www.boddie.org.uk/python/HTML.html

Hi Paul,

Thanks a lot for this resource. To be honest, I wonder how you could
work out how to use htmllib, IMHO the documentation is very poor.
Actually, I am happily using the flexibility of regular expressions to
do these things ATM, but I'm willing to give this library a try.

Cheers,
Giulio


From eppstein at ics.uci.edu  Thu Jun 20 19:39:23 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Thu, 20 Jun 2002 16:39:23 -0700
Subject: Are generators in nested functions possible?
References: <mailman.1024614876.11270.python-list@python.org> <eppstein-9035CF.16382220062002@news.service.uci.edu>
Message-ID: <eppstein-317C17.16392320062002@news.service.uci.edu>

In article <eppstein-9035CF.16382220062002 at news.service.uci.edu>,
 David Eppstein <eppstein at ics.uci.edu> wrote:

> This is a no-op.  Did you mean
>                       for x in genInd(res+[val],length-1): return x
> ?

Oops, that should be yield, not return, of course...

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From donn at u.washington.edu  Wed Jun 19 17:01:10 2002
From: donn at u.washington.edu (Donn Cave)
Date: 19 Jun 2002 21:01:10 GMT
Subject: What If.....    Strong Types
References: <mailman.1024498968.2603.python-list@python.org> <3D10D17B.6070004@bgb.cc>
Message-ID: <aeqrem$1ipu$1@nntp6.u.washington.edu>

Quoth Don Garrett <garrett at bgb.cc>:
| brueckd at tbye.com wrote:
|>   Worse, to make it be "good code" I'll
|> probably define -1 to be NOT_WEIGHED or something, and check against it so
|> I'm not using magic numbers. 
|
|    I might suggest the use of None over -1, since it has a real meaning of 
| 'not really there'. But not defining the member would have the same meaning as
| well. I do get your point here.

The statically typed languages I've been fooling around with, like Haskell,
allow for this with a type that's basically a union of whatever value, or
Nothing.  You can have a non-value, but you can't write code that treats
it as a value without resolving that ambiguity.

|> It's sometimes hard for me to model my objects this way because of having
|> had to always do it the other (completely statically defined) way, but
|> when this approach works the solutions are often quite elegant and
|> straightforward (btw - I believe that a good portion of bugs and program
|> cruftiness is related to forcing the programmer to clearly define, up
|> front, what the complete type of the object is. In many cases this forces
|> the programmer to make decisions without enough information, in others it
|> is too constricting, and in others it's just too tedious to make the
|> programmer worry about all those details, and in all cases general
|> hackiness ensues).

|    The point about programmers having to make decisions about the data before 
| they are ready is a very good one. I've certainly run into that myself, but 
| I've never heard it phrased that way, so it never clicked the way it just did.
|
|    Don't you find that you miss the advantages of a clear definition of what a
| class is and/or does? That's more important with group development, but even 
| when it's just me.... I need memory aids.

Some languages, again like Haskell, manage to infer types and don't
require declarations.  In this case, you might actually add declarations
later in the development process, for the kind of clarity you're talking
about - Haskell software already tends to be pretty damned abstract.
It also can help clarify the source of the problem when there's an error.

I'm not in a position to be writing big projects in Haskell, but in my
casual experiments with static type inferring languages so far, I like it.
I'm no big picture architect, any software more than the simplest thing
usually takes me several general rewrites.  Static typing catches a very
large number of errors in my program, and in fact that helps me rewrite
more rapidly, because the compiler is doing work that I would otherwise
have to do myself.  Type inference makes the compiler do a lot more of
that work, and allows me to express functions at an appropriate level
of abstraction.  Put together, the result is much stricter and more type
safe than a language like C, which allows all kinds of casting and promotion
and so forth, and yet more convenient to work with.  (Note that I'm talking
strictly about typing here, and not making any more general claim about
Haskell or other similar languages.)

The place I find dynamic typing particularly annoying is in exception
handling.  The handler receives this parameter that is of some type,
whose value may or may not be a tuple, may or may not have an 'errno'
attribute, etc.  Even qualified by exception, for example the structure
of a socket exception is somewhat variable.  And of course one would
like to avoid raising an exception from the damned exception handler.
Wouldn't it be nice if there were some way for the socket module to
export the complete type information for its exception, and for the
interpreter to check your operations on the exception value against
that type?  Like,

    except socket.error, v:
        # socket.error :: { args :: (Int, String) | (String,) }
        # phrase = v.args[1]
        # ^ oops!  Here v.args is (a, String, ...)
        #          but was declared as (Int, String) | (String,)
        phrase = v.args[-1]
        # ... ahh, that's better.
        sys.stderr.write('ouch: ' + phrase + '\n')

Maybe an interpreter flag meaning "compile this module to byte-code
and check against exported type signatures."

	Donn Cave, donn at u.washington.edu


From whisper at oz.net  Wed Jun  5 05:35:31 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 5 Jun 2002 02:35:31 -0700
Subject: writing a chat program
In-Reply-To: <d4c18195.0206040840.7ece1b8c@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEGELIDGAA.whisper@oz.net>

If you want control of the console layout, you might care to have a look at
PySlang and sLang, which are curserish. Should be possible to split the
console into two sub-windows for sending/receiving.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Shaun Whelden
> Sent: Tuesday, June 04, 2002 9:41
> To: python-list at python.org
> Subject: writing a chat program
>
>
> Hi all,
> I'm trying to throw together a simple text-based chat program in
> Python.  Here's the problem I have right now: I need a way to allow
> the user to be entering in a message they want to send, while at the
> same time waiting for any data that may come through the socket.  If
> data does come through the socket, I need the user to be able to
> continue typing while the data is printed to the screen.
>
> I'm writing this on a Windows machine, so curses is out of the
> question.  Also, I want this to be a text based program, so I don't
> want to have to mess around with Tkinter or any other forms of GUI
> unless I have to.  I'm pretty sure that I need to do something
> involving thread.py and/or threading.py, but I've never used a thread
> before in my life.  Can anyone help?
>
> Shaun Whelden
> shaunw at liberateschool.com
> --
> http://mail.python.org/mailman/listinfo/python-list





From mcfletch at rogers.com  Fri Jun  7 01:15:32 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Fri, 07 Jun 2002 01:15:32 -0400
Subject: How do you read unicode files?
References: <adpd7h$6sv$1@slb3.atl.mindspring.net>
Message-ID: <3D004174.5070200@rogers.com>

Well, on my win2k machine, I can create a text file using notepad and 
specify that it's in "Unicode" format (utf16) and load it with:

unicode( open( filename, 'r').read(), 'utf16' )

That works even if there's ANSI characters > 128, and even if I specify 
"big-endian unicode" in notepad.

unicode( open( filename,'r').read(), 'utf8' )

works if I specify "UTF-8" format in notepad.

Depending on what format you want for the "standard string", you'd then 
just call, for instance .encode( 'utf8') on the resulting unicode object.

Here's a sample session:
 >>> data = open( filename,'r').read()
 >>> data
'\xff\xfeT\x00e\x00s\x00t\x00i\x00n\x00g\x00 
\x00u\x00n\x00i\x00c\x00o\x00d\x00e\x00\r\x00\n\x00\xe1\x00\xed\x00'
 >>> u = unicode( data, 'utf16' )
 >>> u
u'Testing unicode\r\n\xe1\xed'
 >>> u.encode( 'utf8')
'Testing unicode\r\n\xc3\xa1\xc3\xad'
 >>> u.encode( 'iso8859-1' )
'Testing unicode\r\n\xe1\xed'
 >>>

That last is a plain, windows-native-encoding (well, my windows-native 
encoding ;) ) of the unicode as a simple Python string.

HTH,
Mike

Matt Gerrans wrote:
> How do you read in a unicode file and convert it to a standard string?
> 
> It seems that when you open a file and read it, what you get is a string of
> single-byte characters.   I've tried all kinds of permutations of calls to
> unicode(), decode(), encode(), etc. with different flavors of encoding
> ('utf-8',  'utf-16' and so on).
> 
> I could parse the data myself (skipping the initial two bytes and then every
> other one -- I'm only working with ASCII in double byte format, so the high
> order byte is always 0), but I imagine there must be a way to get the
> existing tools to work.
> 
> What I want to be able to do is write a search and replace tool that will
> work equally well on ANSI and Unicode (or double-byte) text files (without
> changing the file type, of course)...
> 
> 


-- 
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From candiazoo at attbi.com  Wed Jun 26 21:28:22 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Thu, 27 Jun 2002 01:28:22 GMT
Subject: Most efficient way to write data out to a text file?
Message-ID: <3d1a6993.248534625@netnews.attbi.com>

I assume some sort of block i/o or preallocating a large block of file space
would be the best way to do it?  I am outputting about 700,000 records to a text
file and my program only spits out about 20 records per second.  This is using
standard file open and write...

This is on a W2K machine... perhaps I should use the win32file interface?  Any
thoughts/opinions?

Thank you!

Mike J.


From peter at engcorp.com  Wed Jun 19 08:40:08 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 19 Jun 2002 08:40:08 -0400
Subject: Python+ASP->import problem
References: <slrnah0pat.fi1.jarvin24@lehtori.cc.tut.fi>
Message-ID: <3D107BA8.87696860@engcorp.com>

J?rvinen Petri wrote:
> 
> When importing my modules, I always get AttributeError for attributes that
> surely exist on imported module.
> 
> <clip>
> Response.Write(enactHTML.defineCSS('Enact Main'))
> AttributeError: 'module' object has no attribute 'defineCSS'
> </clip>

Not familiar with ASP, but is it possible to troubleshoot
this by using Response.Write(dir(enactHTML)) instead?
Then you'll know exactly what is available in that
module... if it works.

-Peter


From gerhard at bigfoot.de  Wed Jun 19 11:04:12 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 19 Jun 2002 17:04:12 +0200
Subject: procmail replacement in Python
In-Reply-To: <20020619181953.E4127@phd.pp.ru>
References: <oqelf3s7n4.fsf@carouge.sram.qc.ca> <NCBBLFCOHHDIKCAFGCFBAEGILFAA.marklists@mceahern.com> <20020619181953.E4127@phd.pp.ru>
Message-ID: <20020619150411.GB11789@lilith.my-fqdn.de>

* Oleg Broytmann <phd at phd.pp.ru> [2002-06-19 18:19 +0400]:
> On Wed, Jun 19, 2002 at 09:08:53AM -0500, Mark McEahern wrote:
> >   http://www.qcc.sk.ca/~charlesc/software/getmail-2.0/
> 
>    "A POP3 mail retriever..." Hence, it is not fetchmail replacement, as
> fetchamil also does IMAP, and does it very well.

I don't need to retrieve emails from an IMAP server, and the whole point
of the exercise for me is to put them into my IMAP server at home. It
also looks an easy task to enhance getmail to include imap support. Hey,
it's written in Python, not C, where you'd need to reinvent to protocol
handling for the n-th time.

> Fetchmail also has good GUI configuration program, written in Python
> and Tkinter.

Pointless. fetchmail (and getmail even more) configuration is very easy.
If there were a GUI tool to write procmail rules, now that would make
sense. But even then only if it included a regex-wizard ;-)

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From phr-n2002b at NOSPAMnightsong.com  Fri Jun 28 03:35:41 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jun 2002 00:35:41 -0700
Subject: crypt for windows?
References: <mailman.1025246004.28511.python-list@python.org>
Message-ID: <7xk7ojq2ia.fsf@ruckus.brouhaha.com>

Klaus Reinhardt <K.Rdt at TU-Berlin.DE> writes:
> Hi
> Well I have some batch-files with my login
> and password and want to encrypt them
> with another password.
> Meanwhile I got fcrypt .. 

Here's a module you can try:

  http://www.nightsong.com/phr/crypto/p2.py

This version still has a date check in it but I I'll try to a real
version with no date check in the next few days (I couldn't mess with
it last month because I was away).  I may make one further change,
removing the nonce from the authentication key, which might speed up
some as yet nonexistent alternate implementaitons.  Comments are
welcome and anyone who hasn't checked this thing for bugs, please do
so now.


From dfackrell at DELETETHIS.linuxmail.org  Tue Jun 25 10:38:52 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Tue, 25 Jun 2002 08:38:52 -0600
Subject: GOTO w/ Python?
References: <mailman.1024964672.10880.python-list@python.org>
Message-ID: <3d18807d$1_1@hpb10302.boi.hp.com>

"Delaney, Timothy" <tdelaney at avaya.com> wrote in message
news:mailman.1024964672.10880.python-list at python.org...
> > From: Daniel Fackrell [mailto:dfackrell at DELETETHIS.linuxmail.org]
> >
> > After skimming the linked page to get a feel for the CAME
> > FROM statement, is
> > it conceptually different from short procedures?  Does the
> > only difference
> > lie in the fact that for compiled languages one of the
> > instances will be
> > inline?  I do understand that it would have some effect on
> > variable scoping,
> > but this can be easily handled in other ways.
> >
> > So what's the benefit that might outweigh the opacity of the
> > code created?
>
> I hope you're not serious ... :)
>
> COME FROM is a joke. Literally. It is a parody of GOTO whose only
redeeming
> feature is that it is not GOTO.
>
> Tim Delaney

*WHOOSH*

I missed that one completely.  Maybe I'm trying too hard.

I was actually trying to figure out any possible reason that one would want
to use such a construct.  I got thinking about compiler-level optimizations,
but the overhead involved would outweigh any possible savings in time or
compiled code size in almost all cases.

Thanks for straightening me out.

--
Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.




From peter at engcorp.com  Sun Jun  2 10:37:31 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 02 Jun 2002 10:37:31 -0400
Subject: How to delete file?
References: <add9ai$107clv$1@ID-49758.news.dfncis.de>
Message-ID: <3CFA2DAB.140ABF0C@engcorp.com>

Ken wrote:
> 
> How do you delete file?

os.remove()

> If you write to file from beginning position of an existing file, do you
> overwrite the existing data or insert from it?

You overwrite the data.  It's not possible to insert data except by
writing the data to a new file, appending the contents of the old
file, deleting the old file, then renaming the new one to the old name.

-Peter


From mhammond at skippinet.com.au  Mon Jun  3 21:53:38 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Tue, 04 Jun 2002 01:53:38 GMT
Subject: using getopt
References: <adh37d$dj8$1@news.asu.edu> <adh59t$k4g$1@peabody.colorado.edu>
Message-ID: <3CFC1E18.4070205@skippinet.com.au>

Fernando P?rez wrote:


> getopt sucks. Period. Take a look at DPyGetOpt, which implements the 
> Getopt::Long perl facilities. It's very very nice. I've also heard good 
> things about optik, but haven't used it myself.

It looks like optik will be incorporated into the core Python library 
(but possibly inside the getopt module).  However it actually appears, 
using optik now will almost certainly make your future transition to the 
standard library simpler.

Mark.



From occeanlinux at linuxmail.org  Mon Jun  3 17:32:03 2002
From: occeanlinux at linuxmail.org (Gold Fish)
Date: Mon, 03 Jun 2002 21:32:03 GMT
Subject: Time
Message-ID: <3cfbdf0c_1@news.iprimus.com.au>

Anyone know how to determine the time in python. If the mtime in the file 
filename.py is Mon Jun 3 03:25:57 2002 ,if i want to set the time forward 
or backwards in 5 days time, how can i do that. 


From jadestar at idiom.com  Mon Jun 10 21:31:12 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 11 Jun 2002 01:31:12 GMT
Subject: Another newbie question
References: <B916FCB7.57AC%sarmstrong13@mac.com>
Message-ID: <ae3jt0$2hdt$1@news.idiom.com>

SA <sarmstrong13 at mac.com> wrote:

> I am trying to find a way to list the contents of a directory into a list so
> that the contents of the list are then read into an html doc as hrefs.

> What I'm looking for is a way to ls a directory. Is there a way to access
> the shell command ls from within a python script and dumping the output into
> a list for further manipulation? Or does this have to be built from the
> ground up inside the python script?

> Thanks.
> SA

  You could use popen():

	import os
	listing = os.popen('ls -l').readlines()
	for eachLine in listing:
		htmlify(eachLine[:-1])

  ... where "htmlify()" is a hypothetical function that you provide
  to wrap each line in your preferred HTML tags.  I'm assuming you
  DON'T want the trailing linefeeds, so I'm trimming them using the
  slice ([:-1] is sort of like perl's chop() function).

  Of course this comes with the stock caveats about using popen()
   --- the string you pass to popen() is passed to a shell, parsed
  and executed by that shell and thus subject to many exploits
  (you *DON'T* want to pass foreign/alien string variables to popen()
  and you probably don't want to try "sanitizing" them yourself; you 
  probably can't account for all of the ways that an attacker could
  trick your sanitizer into passing a "dirty" string to the shell).


  It's probably much safer to use os.listdir() with os.stat(), 
  and some processing of the stat() info using the pwd and grp 
  modules (to look up user and group names given UIDs and GIDs) and
  the time module to convert the st_mtime field into a readable 
  date/time stamp.  You can write your own function to convert octal
  st_mode values into the usual '-rwxr-x---' format and you'd want
  to use the stat module to help with that (especially to set the 
  "type" character (- for regular files, l for symlinks, s for UNIX
  domain sockets, etc) and you'd have to fuss with lstat vs stat
  for symlinks.

  In other words re-implementing a basic "ls -al" in Python is likely
  to take about five or six modules and probably about 100 lines of 
  code.  Supporting any significant number of ls options would add 
  quite a bit more code (and maybe require the glob and/or fnmatch
  modules).

  On the one hand that's not really alot of work (and you can take 
  shortcuts if you don't need the output to be "just like" ls' --
  for example if you don't care to support non-regular files, don't
  care about the file modes and/or don't care about ls' handling of 
  dates, where it shows one form for files within the last year and 
  another for older files.)  That might bring it down to 50 lines of
  code; which might be considerably less then you'd spend "sanitizing"
  data to be suitable for passing to popen().

  On the other hand, it is alot more work than two lines that it takes 
  to get the ls -l from popen().  

  Alot also depends on what else you're going to do and how robust
  you want the app. to be (how you want to to exception handling).  
  using popen(), popen2(), popen3(), and popen4() provides limited
  exception and error handling).  In fact you might have to use the 
  popen2.Popen3() or Popen4() *classes* in the popen2 module in order to
  get exit code values from your ls commands.  That's assuming that you want
  to gracefully handle some of the errors that you know *could* happen
  under the circumstances that exist on your system(s) when your script is 
  running.

  [As usually, writing sometthing that works for optimal input and
  conditions is reasonably straightforward, accounting for degenerate 
  input and handling exceptional conditions is more challenging].


From peter at engcorp.com  Sun Jun 23 22:48:16 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 23 Jun 2002 22:48:16 -0400
Subject: Python hits the spot
References: <mailman.1024883491.10691.python-list@python.org>
Message-ID: <3D168870.FFC96651@engcorp.com>

"Delaney, Timothy" wrote:
> 
> > From: Peter Hansen [mailto:peter at engcorp.com]
> > Aahz wrote:
> > > Ya know, you could have saved yourself a lot of typing by
> > > simply saying PBCAK.  ;-)
> >
> > To save us from those who can't Google:  Problem Between
> > Chair And Keyboard.
> 
> That's always been PEBCAK to me ...
>                     ^
> Problem Exists Between Chair and Keyboard

Ah, /that/ one has 100 times as many google hits...

-Peter


From merkosh at hadiko.de  Sun Jun  9 06:09:41 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Sun, 9 Jun 2002 12:09:41 +0200
Subject: xml.* getting the hang...
Message-ID: <MPG.176d47ced6b87f6d989681@news.rz.uni-karlsruhe.de>

hi,

i am working myself into the xml.* packages and am just about to start 
getting an idea of what works together with what in wich order.

basically i want to read in xml packages from a byte stream and from 
these "request" objects i want to create frame-response xml objects, 
just taking the attributes of a request object and copy them into a 
response object, leaving the rest blank.
i figured that xml.sax and xml.parsers.expat is where to look for. 
problem number first:

- in order to build an xml dom object i have to get started on a 
Document(); and this is the problem: i couldn't find anything; with one 
exception, i can create xml documents using xml.dom.minidom.Document(), 
if this is the regular case then i don't quite see how xml.sax.* and 
xml.dom.minidom relate to eachother. it feels strange if i use functions 
from various packages when i'd expect them to be of importance for all 
sub packages. any expenations?

Secondly i found the package xml.dom.pulldom which seems to be just what 
i'm looking for, however the documentation is very poor on these classe. 
what, f.e. is a "documentFactory"?

I don't want to do xml validating, so xml.parsers.expat could be used, 
too. So now i don't know what to use at all:

1. xml.parsers.expat: i don't do validating

2. xml.dom.pulldom: i want to create partial dom trees from sax events

3. xml.sax.xmlreader.IncrementalParser(): i read xml from a byte stream

Any idea or comments on this?

Thanks in advance
Uwe


From merkosh at hadiko.de  Fri Jun 14 23:17:25 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Sat, 15 Jun 2002 05:17:25 +0200
Subject: file.tell()  --  obscure IOError
References: <MPG.1774c0f11ff79749989686@news.rz.uni-karlsruhe.de> <MPG.1774c753f4d1f6fe989687@news.rz.uni-karlsruhe.de>
Message-ID: <MPG.1774d03280868110989688@news.rz.uni-karlsruhe.de>

In article <MPG.1774c753f4d1f6fe989687 at news.rz.uni-karlsruhe.de>, 
merkosh at hadiko.de says...
[...]

now I did it. I isolated a piece of code which produces the same error 
in the command line:

----------------------------------
>>> f = file('test.riff.txt')
>>> f.seek(0,2)
>>> f.tell()
109L
>>> f.seek(0,0)
>>> while f.tell() < 109L:
...   f.read(3)
...
'//d'
'ies'
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
IOError: (0, 'Error')
>>>
----------------------------------


test.riff.txt contains the following:
--------------------------
//dies ist ein beispielkommentar

hier nichtmehr\r
hier auch nicht //hier schon

und hier ist schluss
//ende
--------------------------

<--- some time later: 5:17 o'clock in the morning --->

i ound the mistake. 
that is: i know what went wrong, but i still don't know why:
f was opened in text mode. calling 
f = file('test.riff.txt','rb')

works fine.
assumingly read() took \r and the new line characters as control 
characters... grml...

Ciao
Uwe


From shagshag13 at yahoo.fr  Sun Jun  2 05:44:32 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Sun, 2 Jun 2002 11:44:32 +0200
Subject: Pop a list from beginning ? and memory saving...
Message-ID: <adcpdg$vui53$1@ID-146704.news.dfncis.de>

I had a huge list containing consuming data, as i only need a one pass in a
for statement,
i'm wondering if by doing this with a pop i could save memory ? (as i think
that python
free it after the pop...)
and is there a way to pop from the beginning  of the list ? (i think i could
use reverse but
i don't know if this memory consuming)

>>> a = [1,2,3,4,5,6,7]
>>> while 1:
 try:
  f = a.pop()
 except IndexError:
  break
 print f

thanks,

s13.





From logiplexsoftware at earthlink.net  Tue Jun 18 14:02:25 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 18 Jun 2002 11:02:25 -0700
Subject: python version?
In-Reply-To: <9YBP8.44418$GY.13865843@twister.nyroc.rr.com>
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com>
	<aegfck$6tkre$1@ID-114079.news.dfncis.de>
	<wEPO8.16066$uk2.8447171@twister.nyroc.rr.com>
	<3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com>
	<gYgP8.42356$GY.12361719@twister.nyroc.rr.com>
	<aekcug$7js8n$1@ID-114079.news.dfncis.de>
	<9YBP8.44418$GY.13865843@twister.nyroc.rr.com>
Message-ID: <20020618110225.5e83167b.logiplexsoftware@earthlink.net>

On Tue, 18 Jun 2002 08:12:53 GMT
George Hester wrote:
> I think you guys don't how to do it.

I think you wouldn't know an answer if it bit you:

[You]
What is a simple script that will tell me the version of python I am currently
running?  IIS 5.0 Windows 2000.

[Jeremy Yallop]
Look up sys.version and sys.version_info.

[You]
Traceback (innermost last): File "<Script Block >", line 5, in ?
Response.Write(sys.version) NameError: name 'sys' is not defined

[Gerhard]
You did not import the sys module.

[You]
I think you guys don't how to do it.

[Duncan Booth, doing your job for you]
Try this:
<html>
<head>
  <title>Python Version</title>
</head
<body>
<SCRIPT Language="Python" RUNAT=Server>
import sys
Response.Write(sys.version)
</SCRIPT>
</body>
</html>

Perhaps you'd like someone to come out and type it for you as well. 

Please remember that the people on this list aren't being paid to help you
write your program.  They expect you to do at least /a little/ of the research
yourself.  It's understandable if you don't know where to look, but when
someone points you to the relevant portion of the docs, you should at least
read it before asking again.  It's clear from your postings that you haven't
even read the tutorial.
If you can't do that then you should pay someone to write your program for you.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From gcordova at hebmex.com  Mon Jun  3 18:47:52 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Mon, 3 Jun 2002 17:47:52 -0500 
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233010A5AE1@mail.hebmex.com>

> 
> The fact that it shows the change is most likely due
> to the list not having moved in memory. I hope that
> the reference count was handled properly!
> 
> AFIC, this is clearly wrong - operators should
> be atomic - no side effects. Either they work, or
> they don't.
> 

The thing is that += MUST work also for immutable objects,
as well as mutables. So, it does a rebind of the final
value obtained --which in a list or dictionary, being mutable,
is the same as the initial reference, but in the case of
immutable objects, it's a different one than before--.

BUT, since it's trying to rebind a tuple's item, and it's
not allowed, an exception is raised; this is correct
behaviour.

So, it's not actually a bug, but an artifact of the hoops
Python needs to jump through in order to maintain total
dynamism in data types.

Instead of this:

   a = ([1],[2],[3])
   a[0] += [3]

do this:

   a = ([1],[2],[3])
   t = a[0]
   t += [3]

there won't be any exception, because 't' *can* be rebinded
to the value emmited by the += operation; *and* 'a' will
also contain the same reference.

It's not *totally* equivalent, because it's missing a
single, last step:

   a[0] = t

*that* is what's raising the exception, which is what
the += operator is trying to do in you're original version.

Happy hunting :-)

-gustavo




From syver-en+usenet at online.no  Thu Jun 20 12:15:18 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Thu, 20 Jun 2002 16:15:18 GMT
Subject: python cgi + IIS
References: <3D0F1813.3010304@gmx.de> <x0NP8.165989$Fp1.91529@atlpnn01.usenetserver.com> <3D102AF2.5010204@gmx.de>
Message-ID: <u4rfyc4gj.fsf@online.no>

Axel Grune <NineOfSix at gmx.de> writes:

> Steve Holden schrieb:
> 
> >     http://www.e-coli.net/pyiis.html
> 
> Thats what i read and did ;-(
> 

Hi Axel, you really should put the %s in double quotes ("%s") because
the python interpreter won't find the file if you don't (if there are
spaces in the path name that is).

If you would tell us exactly what you have done, (where you put your
.py file, which web application you have modified the app mappings
for, and so forth) give an account of all the steps you have performed
and it will be easier to tell where you went wrong, we should be able
to help you out and get python cgi under IIS on the road.

-- 

Vennlig hilsen 

Syver Enstad


From syver-en+usenet at online.no  Thu Jun 20 19:29:26 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Thu, 20 Jun 2002 23:29:26 GMT
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <aer2fn$frc$1@peabody.colorado.edu> <3D1214B6.E9BB77D2@noaa.gov>
Message-ID: <u660dcyxf.fsf@online.no>

Chris Barker <Chris.Barker at noaa.gov> writes:
> 
> On that note: does anyone know if there is a way to turn a Python
> script
> 
> into something that acts like an application on Windows...without
> resorting to Py2exe and the like. What I want is something just like:
> 
> chmod +x scriptname
> and a #! line at the top 
> 
> on Unix. I want it to work on the windows command line.
> 

Make a text file, save it as something.bat, put the line below in the
first line of the file.

@python -x "%~f0" %* & goto :EOF
or 
@python -x -OO "%~f0" %* & goto :EOF

Put python code in the lines underneath.

If you're using emacs and want syntax highlighting:

Put
# Local Variables: ***
# mode:python ***
# End: ***

at the end of the file


I think that this only works on win2k because it uses extensions to 
the batch language that is not part of win9x family of os'es.
-- 

Vennlig hilsen 

Syver Enstad


From brueckd at tbye.com  Tue Jun  4 18:23:16 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Tue, 4 Jun 2002 15:23:16 -0700 (PDT)
Subject: urllib, urllib2, httplib -- Begging for consolidation?
In-Reply-To: <Pine.SOL.4.30.0206042145300.10073-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <Pine.LNX.4.44.0206041520070.5687-100000@ns0.tbye.com>

On Tue, 4 Jun 2002, John J. Lee wrote:

> On 4 Jun 2002, Dave Brueck wrote:
> [...]
> > together. For example, what is the correct way to do a HTTP HEAD
> > request that follows redirects? It's not hard, but it's silly to have
> > to code it yourself if somebody already did the work for you. Well,
> > httplib doesn't know how to follow an HTTP redirect, but lo and
> > behold, urllib does. Unfortunately, there's no way to use it because
> > in this case because urllib is for opening URLs (the GET is hardcoded
> > and not easily overridable).
> 
> I suppose the standard answer applies here: it isn't there because
> nobody has written it yet.

Argh... this misses the whole point of the thread. Never mind, I repent
for having revived this topic.

-Dave





From gcordova at hebmex.com  Fri Jun  7 12:52:53 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Fri, 7 Jun 2002 11:52:53 -0500 
Subject: Detect events positively
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062330116FCA6@mail.hebmex.com>

> 
> Hello,
> 
> The following code display two button:
> 
> from Tkinter import *
> def Calculate(event = None):
>         while 1:
>                 pass

This loop will never exit, hence, will never
return control to Tkinter (and your GUI).

> def Exit(event = None):
>         import sys
>         sys.exit(0)
> root = Tk()
> btn_cal = Button(root, text="calculate", command = Calculate)
> btn_cal.pack()
> btn_exit = Button(root, text="exit", command = Exit)
> btn_exit.pack()
> root.mainloop()
> 
> if clicked the "calculate" button, the program will fall in a 
> loop that will never end. So program will not response to clicking
> to the "exit" button, so it cannot be terminated normally. If we
> don't use multithread, can we add some code in the loop of the
> function Calculate to let it exit the loop after the user clicks
> the "exit" button or press a "Esc" key?
> 

Well, you're little program is doing what you asked
it to do, :-)

You can use some multithreading, like:

# This function does all the hard work.
def Do_Calculations():
   do_lotsa_calculations()
   save_results_somewhere()
   set_finished_flag()
   raise SystemExit()

def Calculate(event=None):
   if calculation_thread_not_set():
      import thread
      set_calculation_thread_flag()
      thread.start_new_thread(Do_Calculations, (,))

Now the calculations are done independantly from the GUI
management thread; this is good. But, you should NOT DO ANY
GUI MANIPULATION from the thread doing the calculations,
because Tkinter WILL CRASH AND BURN. Hence, the set_finished_flag()
thing, where you inform the main thread by any means you
like that the calculations are done; that flag can be
picked up by your main thread, which can update the GUI.

Good luck :-)

> Many thanks from, 
> Xiao-Qin Xia
> 

-gustavo




From merkosh at hadiko.de  Thu Jun 13 11:40:02 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Thu, 13 Jun 2002 17:40:02 +0200
Subject: modifying string objects
Message-ID: <MPG.1772db35d89f7b9c989685@news.rz.uni-karlsruhe.de>

[This followup was posted to comp.lang.python and a copy was sent to the 
cited author.]

hi,

strings are immutable in python. so what do i do if i need to pass a 
function a string as parameter and need it to be modified within the 
function so that changes are reflected outside of the function body?

f.e. a class constructor must return None

class foobar:
  def __init__(text):
    text = text[-1]			#this is what i actually want

however:

teststr = 'dummy'
foobar(teststr)
print(teststr)

prints "dummy" to the screen.

i _could_ use the globals() function in the object. how 'bad' is this 
style considered to be? can i do it or will i be flamed by other python 
programmers afterwards? ;-)


another question. from perl i konw that f.e.

foo(text = 'some value')

works in the way that first "text" is assigned 'some value' which is 
then passed to the function "foo". i know that the above syntax is used 
to assign the named parameter "text" the value 'some value' and if there 
is no such parameter an error message is produce by the interpreter, but 
why do such language constructs not work in python?
another example:

if (value = myFile.read(10)) == '': ...

?

Thanks in advance 
Uwe


From guido-clpy at a-nugget.de  Sat Jun 22 09:00:22 2002
From: guido-clpy at a-nugget.de (Guido Goldstein)
Date: 22 Jun 2002 15:00:22 +0200
Subject: urllib POST question
References: <644f6688.0206200930.48110837@posting.google.com> <slrnah4dbq.r77.insyte-clp@petra.bitstream.net> <644f6688.0206210900.68dd930c@posting.google.com>
Message-ID: <m2elezcvtl.fsf@nathan.mail.a-nugget.de>

Hi!

On 21 Jun 2002 10:00:44 -0700
  rdacker at pacbell.net (rdack) wrote:
> Ben Beuchler <insyte-clp at emt-p.org> wrote in message news:<slrnah4dbq.r77.insyte-clp at petra.bitstream.net>...
> > In article <644f6688.0206200930.48110837 at posting.google.com>, rdack wrote:
> > > i am doing a POST to a server using urllib and it works.
> > > is there some way i can see exactly what is being posted?
> > > a way to get the post object out of urllib - the exact lines it is sending?
> > 
> > I'm a big fan of Ethereal...
> > "Sniffing the glue that holds the internet together!"
> > 
> > http://www.ethereal.com/
> 
> i am looking into it. seems overkill for wanting to see a few bytes
> that python is generating.
> No way within python to see what urllib intends to send out for the
> POST?

If you're on *NIX (like GNU/Linux) there might be a program call
netcat (or nc on some machines).

>From the man-page:
-- cut --
NAME
       nc - TCP/IP swiss army knife

SYNOPSIS
       nc [-options] hostname port[s] [ports] ...
       nc -l -p port [-options] [hostname] [port]

DESCRIPTION
       netcat  is a simple unix utility which reads and writes data across network connec?
       tions, using TCP or UDP protocol. It is designed to be a reliable  "back-end"  tool
       that  can  be used directly or easily driven by other programs and scripts.  At the
       same time, it is a feature-rich network debugging and exploration  tool,  since  it
       can create almost any kind of connection you would need and has several interesting
       built-in capabilities.  Netcat, or "nc" as the actual program is named, should have
       been supplied long ago as another one of those cryptic but standard Unix tools.
[...]
-- cut --

When you start it in listening mode it will sit and wait for
connections. So when you connect with your program, nc will show what
it has received.

BTW: nc is a *must* on all machines <0.5 wink>

HIH
  Guido -- no, not this one...


From hesterloli at hotmail.com  Wed Jun 19 04:47:03 2002
From: hesterloli at hotmail.com (George Hester)
Date: Wed, 19 Jun 2002 08:47:03 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <D7NP8.46301$GY.14487785@twister.nyroc.rr.com> <slrnagvdp4.1sg.gerhard@lilith.my-fqdn.de>
Message-ID: <byXP8.47929$GY.15272487@twister.nyroc.rr.com>

You did it again.  Why are you spamming me?

--
George Hester
_________________________________
"Gerhard H?ring" <gerhard at bigfoot.de> wrote in message
news:slrnagvdp4.1sg.gerhard at lilith.my-fqdn.de...
> George Hester wrote in comp.lang.python:
> > I just cannot believe what I ran into in this newsgroup. [...]
>
>
>                                                 .:\:/:.
>              +-------------------+             .:\:\:/:/:.
>              |   PLEASE DO NOT   |            :.:\:\:/:/:.:
>              |  FEED THE TROLLS  |           :=.' -   - '.=:
>              |                   |           '=(\ 9   9 /)='
>              |   Thank you,      |              (  (_)  )
>              |       Management  |              /`-vvv-'\
>              +-------------------+             /         \
>                      |  |        @@@          / /|,,,,,|\ \
>                      |  |        @@@         /_//  /^\  \\_\
>        @x@@x@        |  |         |/         WW(  (   )  )WW
>        \||||/        |  |        \|           __\,,\ /,,/__
>         \||/         |  |         |      jgs (______Y______)
>     /\/\/\/\/\/\/\/\//\/\\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
> --
> mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
> web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
> public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
> reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From gerhard at bigfoot.de  Thu Jun 13 12:00:48 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 13 Jun 2002 16:00:48 GMT
Subject: modifying string objects
References: <MPG.1772db35d89f7b9c989685@news.rz.uni-karlsruhe.de>
Message-ID: <slrnaghga9.2iv.gerhard@lilith.my-fqdn.de>

Uwe Mayer wrote in comp.lang.python:
> hi,
> 
> strings are immutable in python. so what do i do if i need to pass a 
> function a string as parameter and need it to be modified within the 
> function so that changes are reflected outside of the function body?

You _could_ use a mutable wrapper object, like a custom class or
simply a list, like in:

def add_an_a(swrap):
    swrap[0] = swrap[0] + "a"

x = ["b"]
add_an_a(x)
print x[0]

But it's generally much better to just let the function return the new
value, as in:

def add_an_a(s):
    return s + "a"

print add_an_a("b")

HTH,

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From skip at pobox.com  Fri Jun 14 22:31:55 2002
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 14 Jun 2002 21:31:55 -0500
Subject: file.tell()  --  obscure IOError
In-Reply-To: <MPG.1774c0f11ff79749989686@news.rz.uni-karlsruhe.de>
References: <MPG.1774c0f11ff79749989686@news.rz.uni-karlsruhe.de>
Message-ID: <15626.42779.744955.97247@12-248-41-177.client.attbi.com>

    Uwe> when does tell() raise an IOError?

When the underlying ftell() call returns -1 for the position, that is, an
error was detected.  My ftell man page suggests the only possible error
condition that ftell detects directly is

       EBADF  The stream specified is not a seekable stream.

however it also says

       The  function  fgetpos, fseek, fsetpos, and ftell may also
       fail and set errno for any of the errors specified for the
       routines fflush(3), fstat(2), lseek(2), and malloc(3).

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html






From sarmstrong13 at mac.com  Thu Jun 13 23:04:18 2002
From: sarmstrong13 at mac.com (SA)
Date: Thu, 13 Jun 2002 22:04:18 -0500
Subject: More Tkinter Help please...
Message-ID: <B92EC762.70EC%sarmstrong13@mac.com>

Hi Everyone-

    I have another Tkinter question for you. I'm still trying to translate
"that previously mentioned" Tcl/Tk program into Python/Tkinter. The
following is all of the code I've managed so far:

from Tkinter import *
import sys
import os

class PyShell:
    def __init__(self, top):
    
        f = Frame(top)
        f.pack()
        
        self.t1 = Text(f, height="12", width="84", font="Courier 12")
        self.t1.pack(side=TOP, pady=2)
        
        self.t2 = Text(f, height="12", width="84", bg="lightblue",
font="Courier 12")
        self.t2.pack(side=TOP, pady=2)
        
        self.b1 = Button(f, text="Execute", command=self.expyth)
        self.b1.pack(side=LEFT)
        
        self.b2 = Button(f, text="Clear Input", command=self.clearin)
        self.b2.pack(side=LEFT)
        
        self.b3 = Button(f, text="Clear Output", command=self.clearout)
        self.b3.pack(side=LEFT)
        
        self.b4 = Button(f, text="Save Input", command=self.savin)
        self.b4.pack(side=LEFT)
        
    def clearin(self):
        self.t1.delete(0,END)
        
    def clearout(self):
        self.t2.delete(0,END)
        
    def expyth(self):
        output = os.popen("python -c").t1()
        self.t2.delete(0,END)
        sys.stdout.write.t2(output)
        
    def savin(self):
        pass
        
root = Tk()

app = PyShell(root)

root.mainloop()

When I run this I get the following error:

Traceback (most recent call last):
  File "PyShell.py", line 45, in ?
    app = PyShell(root)
  File "PyShell.py", line 17, in __init__
    self.b1 = Button(f, text="Execute", command=self.expyth)
AttributeError: PyShell instance has no attribute 'expyth'


Any ideas on what I'm doing wrong here?

Thanks.
SA



From emile at fenx.com  Sat Jun  8 18:57:55 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 08 Jun 2002 22:57:55 GMT
Subject: Program mysteriously STOPS!
References: <yZuM8.146064$ux5.186314@rwcrnsc51.ops.asp.att.net>
Message-ID: <TZvM8.134595$352.6795@sccrnsc02>

LJD
> The following code is designed (at least thats my intent) to run
forever.
> The program should sit in a loop until it connects with the server.
It
> should just loop and loop and loop.
>
> It does, for a while, maybe five or ten minutes, but then the process
just
> seems to vanish.  It's in the c1==0 loop when it quits. When I do a PS
AUX,
> the process is just gone.  Obviously it quit, but more importantly
WHY???,
> and what can I do so it doesn't?


If you get rid of all the 'c1=...'  lines after the first does it run
longer?

HTH,


--

Emile van Sebille
emile at fenx.com

---------



From sholden at holdenweb.com  Wed Jun 19 15:02:10 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 19 Jun 2002 15:02:10 -0400
Subject: Caching query resuls
References: <aentug$m8d$1@news.onet.pl>
Message-ID: <lQjQ8.213356$Fp1.198722@atlpnn01.usenetserver.com>

"Pawel Lewicki" <lewy0lewy at poczta.onet.pl> wrote in message
news:aentug$m8d$1 at news.onet.pl...
> Hallo,
> Is there any good solution to cache the results of SQL queries? I found
the
> great cache manager in dmtools (http://csl.anu.edu.au/ml/dm/), but I
wonder
> if there is any alternative. I will start with MySQL database, but it
would
> be great to be able to plug it into other RDBMS's. I am interested in
both -
> file and memory caching. I also know ZSQL Method in Zope.
>
I developed this (primitive) solution for "Python Web PRogramming". It might
be a starting point for you.

regards
 Steve

import mx.DateTime
import dtuple

#################################################################
# $Revision: 7 $
# $Date: 10/19/01 1:37p $
#################################################################

class CacheQuery:

    """Defines a database query that caches database row sets.

    This object is initialized with

    tbl         table name in the database
    colnames    list of field names retrieved
    keynames    list of keys used for retrieval
    conn        database connection
    refresh     caching refresh interval

    Individual results are read by calling the object with a
    tuple of key values as an argument. If the row set associated
    with this particular set of keys is not present, or was read
    longer than the refresh interval ago, it is re-read from the
    database and stored in the content table as a list of database
    tuples, which allow columns to be accessed by name.

    Otherwise the already-cached database tuple set is returned.

    Refinements might be added, such as registering with an
    observer that might clear down all cache entries periodically
    to force a global database refresh, and using a second SQL query
    on record modified timestamps to determine whether a refresh is
    really required (which may or may not be a win for a given set
    of columns).
    """

    def __init__(self, tbl, colnames, keynames, conn, refresh=0,
ORDER=None):
        """Create a caching data set for the given table, columns and
keys."""
        self._flush()
        self.tbl = tbl
        self.keynames = keynames
        self.refresh = refresh
        self.cursor = conn.cursor()
        self.sql = "SELECT %s FROM %s" % (",".join(colnames), tbl)
        if keynames:
            condition = " AND ".join(["%s=?" % f for f in keynames])
            self.sql += " WHERE %s" % condition
        if ORDER:
            self.sql += " ORDER BY " + ", ".join(ORDER)
        self.desc = dtuple.TupleDescriptor([[n, ] for n in colnames])
        print "Descriptor:", self.desc
        print "SQL:", self.sql

    def _flush(self):
        """Remove all trace of previous caching."""
        self.recs = {}
        self.when = {}

    def __call__(self, keyvals=(), debug=0):
        """Return the data set associated with given key values."""
        assert len(keyvals) == len(self.keynames)
        now = mx.DateTime.now()
        if self.recs.has_key(keyvals) and self.refresh and (now -
self.when[keyvals] < self.refresh):
            if debug: print "Interval:", now - self.when[keyvals]
            return self.recs[keyvals]
        else:
            self.cursor.execute(self.sql, keyvals)
            rows = self.cursor.fetchall()
            result = [dtuple.DatabaseTuple(self.desc, row) for row in rows]
            if self.refresh:
                if debug: print "Caching", self.tbl, keyvals, " at", now
                self.recs[keyvals] = result
                self.when[keyvals] = now
            return result

    def close(self):
        self.recs = None
        self.when = None
        self.cursor.close()


if __name__ == "__main__":
    #
    # Sorry, you'll need your own database details in here
    #
    import mx.ODBC.Windows as odbc
    conn = odbc.connect("prom2000")
    s1 = CacheQuery("department",                               # table
                    "DptName DptWelcome DptLnksTxt".split(),    # columns
                    ("DptCode",),                               # key
columns
                    conn, refresh=0)                            # other
stuff
    while 1:
        dc = raw_input("Department Code: ")
        if not dc:
            break
        rows = s1((dc, ), debug=1)
        if len(rows) == 0:
            print "No such department"
        else:
            for row in rows:
                print """
Department: %s Full Name: %s
Welcome Text:
%s
Links Text:
%s
""" % (dc, row.DptName, row.DptWelcome, row.DptLnksTxt)
    s1.close()
    conn.close()



--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From skip at pobox.com  Wed Jun 12 11:52:38 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 12 Jun 2002 10:52:38 -0500
Subject: JIT Complier for python?
In-Reply-To: <20020612120936.W6609@prim.han.de>
References: <8d3f4438.0206111905.7934664b@posting.google.com>
        <20020612120936.W6609@prim.han.de>
Message-ID: <15623.28230.744872.342655@12-248-41-177.client.attbi.com>

    holger> d2002xx wrote:
    >> Is there any JIT Compiler for Python?

    holger> you might want to take a look at

    holger> http://homepages.ulb.ac.be/~arigo/psyco/

I think the Psyco project on Sourceforge might be more up-to-date:

    http://sourceforge.net/projects/psyco

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html






From bowman at montana.com  Tue Jun  4 00:08:34 2002
From: bowman at montana.com (bowman)
Date: Mon, 03 Jun 2002 22:08:34 -0600
Subject: getting windows processes list
References: <mailman.1023143106.7818.python-list@python.org>
Message-ID: <80XK8.386$ZM.143854@newsfeed.slurp.net>

David LeBlanc wrote:
>
> I'm not sure that mere mortal programs are allowed access to that
> information due to Window's (NT's) arcane idea of what constitutes
> security. I think you're doing well to get your own PID.

A User cannot get quite as much information as an Administrator, but there 
is quite a bit available. If WMI is installed, that is probably the easiest 
way. The psapi.dll offers another handy interface. The last ditch is to 
open the PerformanceData and paw around, but this tends to be a little slow 
since it really isn't a static part of the Registry, but is built on the 
fly. 





From leaf00 at linuxfreemail.com  Thu Jun 27 18:49:43 2002
From: leaf00 at linuxfreemail.com (coh)
Date: Thu, 27 Jun 2002 23:49:43 +0100
Subject: Newbie: GUI
Message-ID: <3D1B9687.8060606@linuxfreemail.com>

Is there a native gui, widget set, for python, or any in the works?

coh



From ark at research.att.com  Fri Jun  7 14:05:20 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 7 Jun 2002 14:05:20 -0400 (EDT)
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <15616.62758.469825.704384@12-248-41-177.client.attbi.com>
	(message from Skip Montanaro on Fri, 7 Jun 2002 13:02:14 -0500)
References: <yu99vg8v6orb.fsf@europa.research.att.com>
        <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
        <200206071649.g57GnmC00708@europa.research.att.com>
        <15616.62131.521803.383860@12-248-41-177.client.attbi.com>
        <200206071756.g57HulT01147@europa.research.att.com> <15616.62758.469825.704384@12-248-41-177.client.attbi.com>
Message-ID: <200206071805.g57I5KW01219@europa.research.att.com>

Andrew> No -- LD_LIBRARY_PATH tells the runtime system ...  LD_RUN_PATH
Andrew> tells the linker ...

Skip> Like I said, it's been ages... ;-)

I'm just glad for the assistance.

Somehow, I doubt I'm going to be the the only person who encounters this
problem...




From jarvin24 at lehtori.cc.tut.fi  Wed Jun 19 07:05:01 2002
From: jarvin24 at lehtori.cc.tut.fi (=?iso-8859-1?Q?J=E4rvinen?= Petri)
Date: Wed, 19 Jun 2002 11:05:01 +0000 (UTC)
Subject: Python+ASP->import problem
Message-ID: <slrnah0pat.fi1.jarvin24@lehtori.cc.tut.fi>

Hey,

I'm doing little project in combination Python+ASP.

Now I have mystical problem which I cannot solve on my own.

When importing my modules, I always get AttributeError for attributes that 
surely exist on imported module.

<clip>
Response.Write(enactHTML.defineCSS('Enact Main'))
AttributeError: 'module' object has no attribute 'defineCSS'
</clip>


Here is my .asp page.
<clip>
<-- select Python as scripting engine for ASP -!>
<%@Language="Python" %>

<-- Import enact modules -->
<% 
import enactHTML
%>

<%
Response.Write(enactHTML.defineCSS('Enact Main'))
Response.Write("Python gets ready to rumble inside a server-side scripting 
block!")
%>
</clip>


And here I always got error on enactHTML.defineCSS('Enact Main')

And when I import same module (located in root of python directory) form 
the interactive shell all works fine.

Could you guys help me a bit.

-Petri


From candiazoo at attbi.com  Fri Jun 14 06:45:18 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Fri, 14 Jun 2002 10:45:18 GMT
Subject: Help with ADO and dynamic insert statement...
References: <3d0826ec.240888125@netnews.attbi.com> <c5b3dbee.0206132203.793dc5a3@posting.google.com>
Message-ID: <3d09c924.347952062@netnews.attbi.com>

Path:
rwcrnsc51.ops.asp.att.net!rwcrnsc53!attbi_slave52!attbi_master51!attbi_feed3!attbi.com!rwcrnsc52.ops.asp.att.net.POSTED!not-for-mail
From: candiazoo at attbi.com
Newsgroups: comp.lang.python
Subject: ADO Command object/Parameters problem...
Message-ID: <3d09426d.313465031 at netnews.attbi.com>
X-Newsreader: Forte Free Agent 1.21/32.243
Lines: 62
NNTP-Posting-Host: 24.128.120.33
X-Complaints-To: abuse at attbi.com
X-Trace: rwcrnsc52.ops.asp.att.net 1024017522 24.128.120.33 (Fri, 14 Jun 2002
01:18:42 GMT)
NNTP-Posting-Date: Fri, 14 Jun 2002 01:18:42 GMT
Organization: AT&T Broadband
Date: Fri, 14 Jun 2002 01:18:42 GMT
Xref: attbi_master51 comp.lang.python:38572
X-Received-Date: Fri, 14 Jun 2002 01:18:42 GMT (rwcrnsc51.ops.asp.att.net)

I am still having problems using dynamic sql statements to perform lookups and
inserts into a MS Access database and an Oracle database.  I've looked up the
error at Microsoft and the solution is to find the OLEDB provider and add
"OLEDB_SERVICES" DWORD value to the registry entry... did this (actually, it
already HAD it)... no go.  I used MakePy... no go.  The code I used works in a
VB application.  I used the errors collection from the connection object and it
only gives me a more concise version of what you see below.  Does anyone have
any idea why I cannot create a command object with parameter objects??  I am
using only adVarChar and adNumeric values... no dates, no blobs, nothing odd at
all...!!!

Has anyone else run into this (note: I have seen other entries for the same
error, which is rather generic, but not for my specific reason).

Thank you, in advance...

Mike J.

-----------------------------------------------------------------------------------------------------------------------------------------
begin updating recType[s] on Thu Jun 13 12:15:39 2002
Traceback (most recent call last):
  File "C:\Documents and Settings\mjessop\My Documents\Python Development\Catalo
g\BuildCatalog.py", line 1300, in ?
    setRecordType()
  File "C:\Documents and Settings\mjessop\My Documents\Python Development\Catalo
g\BuildCatalog.py", line 506, in setRecordType
    rstOracle, status = cmdOracle.Execute()
  File "C:\Python22\Lib\site-packages\win32com\gen_py\00000205-0000-0010-8000-00
AA006D2EA4x0x2x5.py", line 1698, in Execute
    return self._ApplyTypes_(0x5, 1, (9, 0), ((16396, 18), (16396, 17), (3, 49))
, 'Execute', '{00000556-0000-0010-8000-00AA006D2EA4}',RecordsAffected, Parameter
s, Options)
  File "C:\Python22\Lib\site-packages\win32com\client\__init__.py", line 341, in
 _ApplyTypes_
    return self._get_good_object_(apply(self._oleobj_.InvokeTypes, (dispid, 0, w
Flags, retType, argTypes) + args), user, resultCLSID)
pywintypes.com_error: (-2147352567, 'Exception occurred.', (0, 'Microsoft OLE DB
 Provider for ODBC Drivers', 'Multiple-step OLE DB operation generated errors. C
heck each OLE DB status value, if available. No work was done.', None, 0, -21472
17887), None)

HERE is an example of what I am doing... (note, I added the long() typecasting
to make sure the value was coerced into a long integer)...

sql = "select pty.orgname " + \
      "from ccc_party pty, ccc_right_rightsholder rgh, ccc_right rgt " + \
      "where pty.pty_inst = rgh.rgh_inst " + \
      "and rgh.rgt_inst = rgt.rgt_inst " + \
      "and rgt.rgt_inst = ?"

cmdOracle.ActiveConnection = cnnOracle
cmdOracle.CommandType = 1
cmdOracle.CommandText = sql

trsInst = long(rstTempDB.Fields("trs_inst").Value)
aasInst = long(rstTempDB.Fields("aas_inst").Value)
anyInst = trsInst or aasInst
        
cmdOracle.Parameters.Append(cmdOracle.CreateParameter("orgname", 200, 2, 80))
cmdOracle.Parameters.Append(cmdOracle.CreateParameter("rgt_inst", 3, 1, 4,
long(anyInst)))
rstOracle, status = cmdOracle.Execute()


From jjl at pobox.com  Tue Jun  4 17:12:49 2002
From: jjl at pobox.com (John J. Lee)
Date: Tue, 4 Jun 2002 22:12:49 +0100
Subject: telnet functionality
In-Reply-To: <9013d0b4.0206040622.f78f95a@posting.google.com>
References: <9013d0b4.0206040622.f78f95a@posting.google.com>
Message-ID: <Pine.SOL.4.30.0206042211470.10073-100000@mimosa.csv.warwick.ac.uk>

On 4 Jun 2002, Ben wrote:
[...]
> the script to the point where I can open a telnet account, change to
> new directory, and list the directory's contents...but, I can't seem
> to figure out how to grab the last file in the directory and append it
> to the list.  It seems that telnetlib doesn't offer that ability.
[...]

Why should it?  Telnet isn't ftp.  Just parse the output of ls yourself.


John



From rjones at ekit-inc.com  Sun Jun 23 23:08:33 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Mon, 24 Jun 2002 13:08:33 +1000
Subject: PIL, Python and CGI
In-Reply-To: <Pine.GSO.4.33.0206232101300.27090-100000@infinity.infineon.net>
References: <Pine.GSO.4.33.0206232101300.27090-100000@infinity.infineon.net>
Message-ID: <200206241308.33270.rjones@ekit-inc.com>

On Mon, 24 Jun 2002 13:02, Sascha Ferley wrote:
> On Mon, 24 Jun 2002, Richard Jones wrote:
> > Your web server environment most likely doesn't have /usr/local/lib in
> > its LD_LIBRARY_PATH.
>
> You probly mean Apache... I'll take a look for at the conf files for it..
> Any ideas where to statically define the lib paths for apache.. i doubt in
> the httpd.conf file..

You've basically got three choices as I see it:

1. recompile PIL to statically link libjpeg etc.
2. recompile PIL to use -R (or whatever the flag is) so it remembers where the 
dynamic libs are
3. modify your apache start script to set the environment var appropriately.

1 or 2 are going to be the easiest, and safest options.


    Richard





From huaiyu at gauss.almadan.ibm.com  Fri Jun 21 22:20:42 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Sat, 22 Jun 2002 02:20:42 +0000 (UTC)
Subject: How to represent the infinite ?
References: <mailman.1024698451.7353.python-list@python.org>
Message-ID: <slrnah7nnp.27i.huaiyu@gauss.almadan.ibm.com>

Tim Peters <tim.one at comcast.net> wrote:
>
>I intended that x**y act the same as math.pow(x, y).

OK, here it is:

>>> from math import *
>>> pow (1e200, 2)
inf
>>> 1e200**2
inf

Here's the new patch (without any testing for other effects).  Are there
documentations of the circumstances where the commented-out code is supposed
to be useful?  Are there unit tests for them?

Huaiyu


diff -rcN Python-2.2.1/Include/pyport.h Python-2.2.1.i/Include/pyport.h
*** Python-2.2.1/Include/pyport.h	Mon Mar 11 02:16:23 2002
--- Python-2.2.1.i/Include/pyport.h	Wed May  8 15:20:30 2002
***************
*** 301,311 ****
   */
  #define Py_ADJUST_ERANGE1(X)						\
  	do {								\
! 		if (errno == 0) {					\
  			if ((X) == Py_HUGE_VAL || (X) == -Py_HUGE_VAL)	\
  				errno = ERANGE;				\
  		}							\
! 		else if (errno == ERANGE && (X) == 0.0)			\
  			errno = 0;					\
  	} while(0)
  
--- 301,311 ----
   */
  #define Py_ADJUST_ERANGE1(X)						\
  	do {								\
! 		/*if (errno == 0) {					\
  			if ((X) == Py_HUGE_VAL || (X) == -Py_HUGE_VAL)	\
  				errno = ERANGE;				\
  		}							\
! 		else */ if (errno == ERANGE && (X) == 0.0)			\
  			errno = 0;					\
  	} while(0)
  
***************
*** 313,320 ****
  	do {								\
  		if ((X) == Py_HUGE_VAL || (X) == -Py_HUGE_VAL ||	\
  		    (Y) == Py_HUGE_VAL || (Y) == -Py_HUGE_VAL) {	\
! 				if (errno == 0)				\
! 					errno = ERANGE;			\
  		}							\
  		else if (errno == ERANGE)				\
  			errno = 0;					\
--- 313,320 ----
  	do {								\
  		if ((X) == Py_HUGE_VAL || (X) == -Py_HUGE_VAL ||	\
  		    (Y) == Py_HUGE_VAL || (Y) == -Py_HUGE_VAL) {	\
! 			/*	if (errno == 0)				\
! 				errno = ERANGE; */			\
  		}							\
  		else if (errno == ERANGE)				\
  			errno = 0;					\
diff -rcN Python-2.2.1/Modules/mathmodule.c Python-2.2.1.i/Modules/mathmodule.c
*** Python-2.2.1/Modules/mathmodule.c	Thu Sep  6 01:16:17 2001
--- Python-2.2.1.i/Modules/mathmodule.c	Fri Jun 21 19:06:48 2002
***************
*** 35,44 ****
  		 * overflow, so testing the result for zero suffices to
  		 * distinguish the cases).
  		 */
! 		if (x)
! 			PyErr_SetString(PyExc_OverflowError,
! 					"math range error");
! 		else
  			result = 0;
  	}
  	else
--- 35,44 ----
  		 * overflow, so testing the result for zero suffices to
  		 * distinguish the cases).
  		 */
! 		/* if (x)
! 		   PyErr_SetString(PyExc_OverflowError,
! 			 "math range error");
! 		else */
  			result = 0;
  	}
  	else
diff -rcN Python-2.2.1/configure.in Python-2.2.1.i/configure.in
*** Python-2.2.1/configure.in	Mon Mar 11 02:14:23 2002
--- Python-2.2.1.i/configure.in	Wed May  8 15:47:07 2002
***************
*** 1777,1786 ****
  # (none yet)
  
  # Linux requires this for correct f.p. operations
! AC_CHECK_FUNC(__fpu_control,
!   [],
!   [AC_CHECK_LIB(ieee, __fpu_control)
! ])
  
  # Check for --with-fpectl
  AC_MSG_CHECKING(for --with-fpectl)
--- 1777,1787 ----
  # (none yet)
  
  # Linux requires this for correct f.p. operations
! #AC_CHECK_FUNC(__fpu_control,
! #  [],
! #  [AC_CHECK_LIB(ieee, __fpu_control)
! #])
! AC_CHECK_LIB(ieee, __fpu_control)
  
  # Check for --with-fpectl
  AC_MSG_CHECKING(for --with-fpectl)
diff -rcN Python-2.2.1/pyconfig.h.in Python-2.2.1.i/pyconfig.h.in
*** Python-2.2.1/pyconfig.h.in	Wed Oct 24 10:10:49 2001
--- Python-2.2.1.i/pyconfig.h.in	Wed May  8 15:53:51 2002
***************
*** 748,754 ****
  #undef HAVE_LIBDLD
  
  /* Define if you have the ieee library (-lieee).  */
! #undef HAVE_LIBIEEE
  
  #ifdef __CYGWIN__
  #ifdef USE_DL_IMPORT
--- 748,754 ----
  #undef HAVE_LIBDLD
  
  /* Define if you have the ieee library (-lieee).  */
! #define HAVE_LIBIEEE
  
  #ifdef __CYGWIN__
  #ifdef USE_DL_IMPORT


From baf at texas.net  Sun Jun 23 11:08:17 2002
From: baf at texas.net (Ben Fairbank)
Date: Sun, 23 Jun 2002 15:08:17 GMT
Subject: (Newbie)  Counting Instances ("Hits") with Regular Expressions
Message-ID: <3d15e215.43448155@news.texas.net>

I am new both to Python and to regular expressions, which may account
for my difficulty.  I must count the frequenies of certain words in
files of moderate length (about150 k bytes).  I have been reading
files and then using count(s,sub), which is fast and easy.  I now have
to allow for punctuation and eliminate words within words, etc, and so
am trying to use regular expressions instead of simple words as
targets.  I do not, however, find a similarly easy to use count
function in the re module.  Yet this is such common operation it must
be there, or easy to implement.  What is the usual way of simply
counting "hits" in the re module?  (And what have I missed in the
documentation; where is this to be found?  I have looked through Lutz
and Ascher)

Thanks for any help.

BAF


From from-020625011948-4b56 at secretlab.mine.nu  Mon Jun 24 19:21:53 2002
From: from-020625011948-4b56 at secretlab.mine.nu (Walter Hofmann)
Date: Tue, 25 Jun 2002 01:21:53 +0200
Subject: working with secure shell
References: <ILbQ8.1749$CZ4.555208@news1.news.adelphia.net>
Message-ID: <af89ih$fcg$04$2@news.t-online.com>

Jon J. Morin schrieb im Artikel <ILbQ8.1749$CZ4.555208 at news1.news.adelphia.net>:
> One question: Is there a python library or module for working with the 
> secure shell (ssh)?

Yes, see http://pyssh.sourceforge.net/

Walter



From g at g.com  Mon Jun  3 03:52:54 2002
From: g at g.com (d)
Date: Mon, 03 Jun 2002 07:52:54 GMT
Subject: test
Message-ID: <qfFK8.5771$J51.371780@news1.telusplanet.net>

test








From cliechti at gmx.net  Mon Jun 24 17:34:58 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 24 Jun 2002 23:34:58 +0200
Subject: Embed terminal into a Tkinter application
References: <3D1758B4.8FA63EBE@club-internet.fr>
Message-ID: <Xns9237F055C1CABcliechtigmxnet@62.2.16.82>

Fabien H?non <ffjhenon at club-internet.fr> wrote in 
news:3D1758B4.8FA63EBE at club-internet.fr:

> The title says it all.
> Has someone already managed to include a Linux terminal into a Tkinter
> GUI ?
> 
> Redirecting the output of a console into the application does not do
> what I want.

well, what do you want? let us help you!

there is also popen[2,3,4] - you can also redirect stdin.

chris

-- 
Chris <cliechti at gmx.net>



From phr-n2002b at NOSPAMnightsong.com  Sun Jun 30 02:51:42 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Jun 2002 23:51:42 -0700
Subject: yield equivalent in C/JavaScript?
References: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk> <7xy9cxhlvx.fsf@ruckus.brouhaha.com> <w3xdKUAuljH9EwgH@jessikat.fsnet.co.uk>
Message-ID: <7xit41i7i9.fsf@ruckus.brouhaha.com>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:
> the generators here are things  associated with compression or
> decompression methods. The complexity is such that a severe
> reformulation will be tough. I understand the concept, but I think it's
> easier to just try and consume the whole input at one go and pass it on.

If you are trying to implement data compression in javascript with
large enough inputs that memory consumption is a problem, you probably
have worse problems than absence of "yield".  Javascript is so slow
your program will probably be near unuseable.  What's the JS
environment and the application?  If it's client side JS in a web
browser, you're probably using the wrong approach to the problem.  If
it's something like Netscape server side JS, use the LiveConnect stuff
to implement the compression in C.



From sholden at holdenweb.com  Sun Jun  2 14:08:41 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 2 Jun 2002 14:08:41 -0400
Subject: semi-concatenated strings
References: <ad5vp6$17kc@drn.newsguy.com>        <mailman.1022861511.22700.python-list@python.org>        <ad8e9m015hd@drn.newsguy.com> <mailman.1022874291.30340.python-list@python.org>
Message-ID: <LetK8.147895$%u2.31463@atlpnn01.usenetserver.com>

"Skip Montanaro" <skip at pobox.com> wrote in message
news:mailman.1022874291.30340.python-list at python.org...
>     >> rows = self.executesql(("select cities.city, state, country"
+
>     >>                         "  from cities, venues, events, addresses"
+
>     >>                         "  where cities.city like %s"
+
>     >>                         "    and events.active = 1"
+
>     >>                         "    and venues.address = addresses.id"
+
>     >>                         "    and addresses.city = cities.id"
+
>     >>                         "    and events.venue = venues.id"),
>     >>                        (city,))
>
>     Grant> I dunno...I don't know anything about SQL, but your code looks
>     Grant> sortta Pythonic.
>
> Yeah it does.  It serves two purposes.  One it makes it clear to me where
> the various clauses of the select statement begin ("from ...", "where
...").
> It also guarantees I have at least one space between the different
> fragments.  This:
>
>     ("select cities.city, state, country"
>      "from cities, venues, events, addresses"
>      "where cities.city like %s"
>      "and events.active = 1"
>      "and venues.address = addresses.id"
>      "and addresses.city = cities.id"
>      "and events.venue = venues.id")
>
> might appear correct, but there are no spaces between the end of one
> substring and the start of the next...
>
I would personally much prefer to see this written as

    ("""select cities.city, state, country
     from cities, venues, events, addresses
     where cities.city like %s
         and events.active = 1
         and venues.address = addresses.id
         and addresses.city = cities.id
         and events.venue = venues.id""",
    (city,))

Perhaps the reason that Python's conventional wisdom states "there should be
oine obvious way to do it" is because, given alternatives, people like you
and me will spend endless time debating which is to be preferred!

regards
 Steve
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From xah at xahlee.org  Sat Jun 29 19:19:16 2002
From: xah at xahlee.org (Xah Lee)
Date: 29 Jun 2002 16:19:16 -0700
Subject: newbie Q: delete list few lines in file
Message-ID: <7fe97cc4.0206291519.1c95ce31@posting.google.com>

i have a folder full of individual email text file, whose ending may
be the following:

_______________________________________________
some mailing list
some at list.com
http://some.com/ttt

I want to delete them, but only if they are the last few lines in the
file.

if you can show me a script that'll be great otherwise my question can
be broken down as:

* how to write to file inplace? (or, do i write to a new file, then
delete original, and rename the new file?)

* as far as a newbie, i use "xreadlines" module. I wasn't able to get
the total number of lines in a file.
li=xreadlines.xreadlines(f)
num=len(li)
fails.


I'm much more interested in learning python idiom and correctness
here. I'm not interested in just "get it to work". For the latter, i
can do well with perl. (also, is there a site or mailing list that
collect very small trivial python programs for learning purposes?)

Thanks.

 Xah
 xah at xahlee.org
 http://xahlee.org/PageTwo_dir/more.html


From timr at probo.com  Sat Jun  1 01:23:49 2002
From: timr at probo.com (Tim Roberts)
Date: Fri, 31 May 2002 22:23:49 -0700
Subject: semi-concatenated strings
References: <ad5vp6$17kc@drn.newsguy.com>
Message-ID: <dgmgfug60vlv3meh9jc13l9qj7055mh6pk@4ax.com>

Grant Griffin <Grant_member at newsguy.com> wrote:
>
>I discovered today that strings can sometimes be concatenated without using a
>"+":
>
>   ActivePython 2.2.1 Build 222 (ActiveState Corp.) based on
>   Python 2.2.1 (#34, Apr 15 2002, 09:51:39) [MSC 32 bit (Intel)] on win32
>   Type "help", "copyright", "credits" or "license" for more information.
>   >>> a = 'one' ' plus ' 'two'
>   >>> a
>   'one plus two'

It is interesting (but only marginally relevant) to note that this exact
same feature is part of ISO standard C:

  #include <stdio.h>
  int main()
  {
      puts( "one" " plus " "two" );
  }
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From maxm at mxm.dk  Tue Jun 25 11:04:42 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 25 Jun 2002 17:04:42 +0200
Subject: Suggestions for good programming practices?
References: <mailman.1024971992.26429.python-list@python.org> <roy-2CF124.07480625062002@reader2.panix.com> <mailman.1025012263.2719.python-list@python.org>
Message-ID: <3D18868A.4050002@mxm.dk>

Jeff Epler wrote:

>>class myData:
>>   def __init__ (self, foo, bar):
>>      self.foo = foo
>>      self.bar = bar
>>
>>you've added quite a bit of functionality.  First, the code becomes a 
>>little more self-documenting.  "power = ckt.volts * ckt.amps" is a lot 
>>easier to read and understand than "power = ckt[0] * ckt[1]".
> 
> 
> (the following applies most obviously to modern versions of python.  I
> use subclassing of builtin objects, __slots__ and zip().  Even so, the
> myStruct class is useful in any version of Python back to at least 1.5,
> if you make myStruct a regular class, and include a definition
> of zip())
> 
> You can do even better (lower memory overhead) with
>     class myData(object):
> 	__slots__ = ['foo', 'bar']
>         def __init__ (self, foo, bar):
>             self.foo = foo
>             self.bar = bar


You can also make it even more general:

class MyData:

     def __init__(self, **args):
         self.__dict__.update(args)

d = MyData(firstName='Max M', lastName='Rasmussen')

print d.firstName, d.lastName

 >>> Max M Rasmussen


regards Max M



From kragen at pobox.com  Sat Jun  1 16:08:55 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 01 Jun 2002 16:08:55 -0400
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org> <3CF8E382.3A44F0DA@engcorp.com>
Message-ID: <83ptzavk14.fsf@panacea.canonical.org>

Peter Hansen <peter at engcorp.com> writes:
> How fast do you need it to be?  (That's a serious question.
> Are you planning to open and close the application so frequently
> that six seconds is unacceptable?)

>From a human interface perspective, having to wait six seconds for a
response to a command (including "start program") is annoying; this is
the reason for "splash screens".  You don't want people to be annoyed
every time they start your program.

Worse, six seconds is plenty of time for a user's thoughts to wander;
it means that the user may have forgotten what they wanted to do by
the time the program starts.



From tim.one at comcast.net  Tue Jun 11 21:26:18 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 11 Jun 2002 21:26:18 -0400
Subject: Sorting list (maybe stupid)
In-Reply-To: <EFEFIHAINIINNELLHFAIOEBLCJAA.kseehof@neuralintegrator.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEEDPMAA.tim.one@comcast.net>

[Ken Seehof]
> >>> a = [1,1.0,3.0,3,2.0,2,0,0.0]
> >>> a
> [1, 1.0, 3.0, 3, 2.0, 2, 0, 0.0]
> >>> a.sort()
> >>> a
> [0, 0.0, 1, 1.0, 2.0, 2, 3.0, 3]
>
> So yes, sort() is stable.

No, it's not.

> I suppose it's theoretically possible for a different
> python implementation to use a different sort algorithm,
> but practically speaking, that is very unlikely.

In fact it's almost certain <wink>.  Small lists are handled by binary
insertion sort, for peak speed, and the way that was coded just so happens
to be stable.  That's why small examples will always lead you to believe
.sort() is stable.  The full-blown samplesort used for non-trivial lists
isn't stable and can't be made stable efficiently.





From lewy0lewy at poczta.onet.pl  Wed Jun 19 10:11:00 2002
From: lewy0lewy at poczta.onet.pl (Pawel Lewicki)
Date: Wed, 19 Jun 2002 16:11:00 +0200
Subject: Caching query resuls
References: <aentug$m8d$1@news.onet.pl> <mailman.1024470864.21143.python-list@python.org>
Message-ID: <aeq3sn$8nh$1@news.onet.pl>

> * Pawel Lewicki <lewy0lewy at poczta.onet.pl> [2002-06-18 20:25 +0200]:
> > Hallo,
> > Is there any good solution to cache the results of SQL queries?
>
> Your database has one. It should save at least the query execution plan.
>
> > I found the great cache manager in dmtools
(http://csl.anu.edu.au/ml/dm/), but I wonder
> > if there is any alternative. I will start with MySQL database, but it
would
> > be great to be able to plug it into other RDBMS's. I am interested in
both -
> > file and memory caching. I also know ZSQL Method in Zope.
>
> I'm sceptical. How does this help? How does the cache manager get
> notified of changes in the database? How is it different from just
> storing the queries I'm interested in away on disk if they deliver large
> resultsets?

Probably not much. :) But dmtools give the out-of-the-box solution. It takes
the function (sql query execution in that case), arguments (query) and
checks given dependency files, which in case of MySQL are table files. It
deals perfectly with my needs and I wondered if there was any other option.

Pawel Lewicki





From BPettersen at NAREX.com  Thu Jun 20 18:57:21 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 20 Jun 2002 16:57:21 -0600
Subject: do...while
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158EEC@admin56.narex.com>

> From: Michael P. Soulier [mailto:msoulier at nortelnetworks.com] 
> 
> On Thu, Jun 20, 2002 at 10:42:56AM -0700, Sean 'Shaleh' Perry wrote:
> > 
> > On 20-Jun-2002 Michael P. Soulier wrote:
> > >     Greetings.
> > > 
> > >     How do you people handle the lack of a do...while in 
> Python? I 
> > > find that at times, I must initialize variables for a loop with 
> > > identical code to the loop, which feels like a waste to me.
> > > 
> > > ie.
> > > 
> > >     line = filehandle.readline()
> > >     while len(line) > 5:
> > >         line = filehandle.readline()
> > > 
> > >     A do...while here would be nice.
> > > 
> > >     Just curious.
> > > 
> > 
> > while 1:
> >     line = filehandle.readline()
> >     if len(line) <= 5: break
> >     ....
> >     ....
> > 
> > is a typical way to implement it.
> 
>     Right, I've used that in the past. It doesn't seem nearly 
> as clean though, as making the conditions for terminating the 
> loop plain in the loop declaration. Considering the Python 
> community's leaning towards readable code, necessitating a 
> break for code factoring goes against that philosophy. 

Doing a quick search over our C++ libraries (~900KLoc), "do {} while()"
seems to be used in right around 3% of the loops (including while, for;
excluding recursion).

Seems like overkill to add special syntax for a feature used so
rarely...

-- bjorn




From phr-n2002b at NOSPAMnightsong.com  Thu Jun 20 03:44:33 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 20 Jun 2002 00:44:33 -0700
Subject: Variable Interpolation - status of PEP 215
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <aermlt$in5$1@panix1.panix.com>
Message-ID: <7xptym4cny.fsf@ruckus.brouhaha.com>

aahz at pythoncraft.com (Aahz) writes:
> Peter already told you what you really wanted to know, but for the
> benefit of other people, PEP 215 is dead and PEP 292 is currently the
> subject of a LARGE thread on python-dev.

Another suggestion: rather than adding the clumsy sub() method,
how about overloading a unary prefix operator on strings?  For
example the unary minus sign: say

   print -"my name is ${name}"

instead of

   print "my name is ${name}".sub()

Note that unlike PEP 215, this is not a syntax change.  It's just
operator overloading.  I think it can already be implemented
in Python 2.2 using a string subclass.  In fact perhaps PEP 215
can be implemented the same way, using "-" as the prefix instead of "$".


From aoeu at go.com  Mon Jun 17 07:17:03 2002
From: aoeu at go.com (Thinkit)
Date: 17 Jun 2002 04:17:03 -0700
Subject: coding python by web?
Message-ID: <3eeda89d.0206170317.100549db@posting.google.com>

Is there any way to test out simple python code with only a web
browser (no input or files, just printing text)?  I'm guessing a
simple cgi setup could do this, but has it been implemented anywhere?


From bokr at oz.net  Sun Jun 23 00:42:47 2002
From: bokr at oz.net (Bengt Richter)
Date: 23 Jun 2002 04:42:47 GMT
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at>
Message-ID: <af3jk7$c4h$0@216.39.172.122>

On Sat, 22 Jun 2002 21:27:17 +0200, Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote:

>Bengt Richter wrote:
>
>> 
>> You apparently have a lot of different stuff glued together. Have you tested the
>> separate pieces separately? Do you have some test problems that you can use to validate
>> numerical results *exactly* (trailing decimal differences need an explanation. They
>> can be benign/acceptable or not).
>
>The Mie code (C function) works, because on Windows I wrote a program 
>under DISLIN in order to depict the phase function, and I can use the 
>same program under Linux without problems.
[my line break] "without problems" presumably means "without apparent problems".
IOW, you could have had a memory leak that didn't cause enough of a problem for
you to notice. Or it may have worked because you interfaced to it differently,
or you may have linked all compatible libraries, versions, etc.

>The Fortran function works 
>too (the Fortran function is really a tiny piece and straightforward in 
>contrast to the C function which is a third party insane pointer hell).
What is this C function?

>The C code itself is essentially the same:  on Windows I compiled it 
>with SWIG and Microsoft Visual C++ and on Linux I used gcc and SWIG; but 
>the name of the function is actually the same. The same goes for all the 
>other functions and directories.
>
What _didn't_ you compile? How many different compilers and run time support libraries
were involved? Could there have been name clashes or incompatibilities?

>
>> I hate to see Python (or anything) bashed without proof. So far I haven't seen the proof,
>> nor an explanation.
>> 
>> If your code is open source, maybe someone will be interested in discovering why
>> it bogs Python down (or Python bogs it down ;-)
>
>The code is open source but not very well documented. But I can send it 
>to you if you like. You can also get the data if you like.
>
I'm not exactly chomping at the bit, but I'd be curious to see the Python if
you want to zip it up and send it attached. How many bytes of python etc.
does it add up to? And how much data? Just curious.

>First, I thought there is something wrong with my data because I have it 
>converted  from DOS to UNIX with: dos2unix.
>
Have you hex-dumped a little to make sure?

>There is an interesting thing to observe: On Windows I am likely to hit 
>the 100MB threshold, but on Unix I only hit 10MB (and sometimes 20MB). 
>At least "top" shows me that.
>
>
>You can get all the functions if you like. I would only have to get it 
>from my laptop back to my stationary machine.
>
>Drop me a note if you want the functions and data. I can prepare it that 
>way that it will be ready to use for you.
>
Maybe you will have it solved before I have that amount of time ;-)

>What I further noticed. On Windows the Task manager tells me that in the 
>first 20 minutes Python uses constantly 99% of the processor power. The 
>same goes under Linux, but after that 20 minutes it sucks down to 87% on 
>Linux and on Windows! Note: same functions and that 8 times!
Unless you are doing a lot of i/o from the program, sounds like maybe that
13% (100-87) may be waiting for swapping to get more virtual space? Or it could be
coincidental background stuff? (Do you trigger any other processes at some point?)

Regards,
Bengt Richter


From look at replyto.address.invalid  Tue Jun  4 01:21:57 2002
From: look at replyto.address.invalid (Greg Ewing)
Date: Tue, 04 Jun 2002 17:21:57 +1200
Subject: Why does this work: 5<"five" ?
References: <3CF80435.689AE006@icc-computer.de> <nig9da.7s9.ln@10.0.0.1> <831ybprot3.fsf@panacea.canonical.org>
Message-ID: <3CFC4E75.BA7EC3DE@replyto.address.invalid>

Kragen Sitaker wrote:
> 
> So if you put complex numbers into your binary search tree, you will
> have problems.
> 
> Oh, enjoy this:
> >>> (1,) < (2,)
> 1
> >>> (1j,) < (1j,)
> 0
> 
> I'm not sure what's going on there.

I think what's going on is that the tuple elements
are being compared for equality first, and if that
fails, compared again for ordering:

>>> (1j,) < (2j,)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: cannot compare complex numbers using <, <=, >, >=

Incidentally, I believe that Python ought to have
a separate notion of compare-for-arbitrary-ordering
that can be explicitly invoked for things like the
b-tree situation. In fact, I think that's what
cmp() should do. But currently it doesn't -- it
seems that if the inequality operators don't work
on a type, cmp() won't work either.

-- 
Greg Ewing, Computer Science Dept, 
University of Canterbury,	  
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg


From loredo at astro.cornell.edu  Fri Jun 28 15:41:33 2002
From: loredo at astro.cornell.edu (Tom Loredo)
Date: Fri, 28 Jun 2002 15:41:33 -0400
Subject: C API for Python's RNG?
References: <3D1CBB3A.E1C43D8A@astro.cornell.edu>
Message-ID: <3D1CBBED.58EAC61E@astro.cornell.edu>

Tom Loredo wrote:
> 
> I'll dig in the source to try to puzzle
> it out, but I thought I should ask publicly:  

Oops!  I should have dug first---the standard lib RNG is pure
Python.  So I guess it's Numeric or some third-party RNG if
I want to do this quickly.  Any suggestions?

Thanks,
Tom


From hamish_lawson at yahoo.co.uk  Tue Jun 11 04:51:40 2002
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: 11 Jun 2002 01:51:40 -0700
Subject: Iterators vs. Generators
References: <ae3a6f$9d9$1@panix1.panix.com>
Message-ID: <915a998f.0206110051.5605201d@posting.google.com>

aahz wrote:

> So when would one actually write an iterator instead of a generator?

When you need to ensure that any allocated resources get cleared up
even if the iteration is abandoned early. This can be handled by an
iterator that has a clear-up method and a __del__ method that calls
that clear-up method. That gives you the option of explicitly
initiating the clear-up immediately on abandoning an iteration, or
else leaving the clear-up to be done automatically when the iterator
object is destroyed.

Hamish Lawson


From gerhard at bigfoot.de  Tue Jun 18 18:34:40 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 18 Jun 2002 22:34:40 GMT
Subject: tomorrow in yyyymmdd format
References: <qpbvguscmblone2r92o8sfc7evo5709mq8@4ax.com>
Message-ID: <slrnagvdbv.1sg.gerhard@lilith.my-fqdn.de>

Giulio Cespuglio wrote in comp.lang.python:
> Hello everybody,
> 
> We all know that if I want to get the present GMT date in yyyymmdd
> format I should say
> 
> 	print time.strftime("%Y%m%d", time.gmtime() )
> 
> Now, what about tomorrow? Note that a simple string manipulation is
> not the solution I'm looking for, as tomorrow might be next month,
> next year, the 29th of February and such.
> Do you confirm that there's nothing better than the following
> silly-looking expression?
> 
> 	time.strftime("%Y%m%d", time.gmtime(time.time() + 60*60*24))

No, there's nothing better. Unless you're using the DateTime type from
the eGenix mxExtensions:

>>> import mx.DateTime as DateTime
>>> DateTime.today()
<DateTime object for '2002-06-19 00:00:00.00' at 81274e0>
>>> print DateTime.today()
2002-06-19 00:00:00.00
>>> print DateTime.today() + 1
2002-06-20 00:00:00.00
>>> 

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From emile at fenx.com  Thu Jun 27 10:54:42 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 27 Jun 2002 14:54:42 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com> <3D1A8CD9.C8B8EC28@engcorp.com> <3d1b1bd7.294171125@netnews.attbi.com>
Message-ID: <SGFS8.12990$Uu2.1087@sccrnsc03>

candiazoo at attbi.com
> I am a newbie... this is my first Python "project" so I am probably
writing
> horribly inefficient code...  I have never used the profiler but I'll
try it!
>
> I am not opening/closing the file each time.  I am extracting 700000
rows from a
> mysql database, extracting additional data from our primary, Oracle
database per
> row, then stuffing each piece of data into a class which preformats
the data (I
> need to output the data into a fixed format string/record for another
> application which reads them) and returns a single string... which I
write out
> to the file.
>

This is not the most efficient way, but for comparison purposes this
wrote almost 3 million records using ~400Mb in a minute:

>>> rec = 'this is a test'*10
>>> import time
>>> def test(rec):
...     global count
...     t = time.time()+60
...     while time.time() < t:
...             outfile.write(rec)
...             count += 1
...
>>> count = 0
>>> outfile = open(r'f:\test.out', 'w')
>>> test(rec)
>>> print count
2982620


--

Emile van Sebille
emile at fenx.com

---------



From brueckd at tbye.com  Tue Jun 25 10:39:05 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Tue, 25 Jun 2002 07:39:05 -0700 (PDT)
Subject: Suggestions for good programming practices?
In-Reply-To: <Xns923818184AEC3cliechtigmxnet@62.2.16.82>
Message-ID: <Pine.LNX.4.44.0206250721250.27456-100000@ns0.tbye.com>

On 25 Jun 2002, Chris Liechti wrote:

> >> * Avoid exec, execfile, eval, and input.
> > 
> > Might one ask why? What do you have to know to use them successfully?
> 
> where does the string you exec come from? from the user: is he a python 
> programmer? could he make subtle errors that break your code and then blame 
> you? could it be somebody that like to play tricks with others and insert 
> harmful expression or satements (like os.system('rm -fR * /*')? How do you 
> handle exceptions, as any exception could be raised by the unknown code 
> (hint: 'try: ... except: pass' is bad too...)?
> 
> if you don't want to deal with such problems, avoid exec* and eval. they 
> are fine for a throw away script, and they're needed when you embedd a 
> python interpreter in your app (for this case there is code.interact and 
> even better reexec). for all the other cases they're a risk.

This view is a overly extreme. Rather than teaching people to fear certain 
features like eval/exec, it's better to explain the risks so that they can 
make informed decisions as to when it's wise to use them.

For example, eval/exec are incredibly useful in building mini-languages
based on Python, in which you execute user code in some dictionary that
includes your set of custom functions (you basically end up with a
context-specific superset of Python). I've found this to be very useful in
several production systems (not 'throw away script[s]') as well as in
creating programmable test harnesses for the QA department (the
non-programmer QA engineers aren't scared off because they don't realize
it's programming, and the more technical QA engineers are pleased that
you've built them such a "rich" test harness language...hehehe). Can
malicious users do malicious things? Of course! But that's like saying
Guido should disable os.system!

So... rather than teaching "avoid W!", let's say "be careful with W 
because of X, Y, and Z". I still wouldn't use eval/exec on code posted 
through a web form, for example, but there are times when they are very 
useful and I can use them in good confidence because I understand their 
risks.

-Dave





From observer at NOSPAM.space.pl  Tue Jun 18 04:28:52 2002
From: observer at NOSPAM.space.pl (Johann)
Date: Tue, 18 Jun 2002 10:28:52 +0200
Subject: PythonWorks Pro problem!
References: <q3nrgu019nlnc2bc19jaouhfo0833fknrj@4ax.com> <3lwP8.112157$pw3.6716@sccrnsc03>
Message-ID: <8nrtgusniqrs9r3fs7l4nt8qka5fm7imi3@4ax.com>

On Tue, 18 Jun 2002 01:49:52 GMT, "Emile van Sebille" <emile at fenx.com>
wrote:
>> I registered some examples from above link. The problem is, I cannot
>> find  this "small, coloured icon" , so I cannot cycle through avalable
>> settings. :-(
>>
>> I tried to clink on everything, but without results :( Anybody can
>> help?
>
>Hmmm... that icon appears to be missing from 1.3...
>
>It's there in 1.2

So, it is a bug. :(

I wrote to authors, but without response. How to get 1.2 if they allow
to d-l 1.3 only?

--
Johann



From m.faassen at vet.uu.nl  Thu Jun 20 15:53:03 2002
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 20 Jun 2002 19:53:03 GMT
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com>
Message-ID: <aetbqv$kgq$1@newshost.accu.uu.nl>

John Roth <johnroth at ameritech.net> wrote:
[snip]
> There's no equivalent of a goto command in Python, for good reason,
> which I won't repeat here. Prof. Djikstra did it very well in 1974,
> in his letter "Goto Considered Harmful."

The good old Djikstra! You know, some years ago in another newsgroup I saw 
so many posts referring to some computer scientist called 'Djikstra', I
started to doubt myself and had to confirm that he really is called 
'Dijkstra'.

Now I realize that there is a Djikstra, and a Dijkstra as well. Djikstra 
is an evil clone of Dijkstra, which the Python Secret Underground is
keeping on ice in case they need any evil clones. 

Regards,

'Martjin'
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From skip at pobox.com  Sat Jun 15 16:52:59 2002
From: skip at pobox.com (Skip Montanaro)
Date: Sat, 15 Jun 2002 15:52:59 -0500
Subject: IDE recommendations??
In-Reply-To: <b9veea.l85.ln@majestix.gallien.de>
References: <b9veea.l85.ln@majestix.gallien.de>
Message-ID: <15627.43307.695989.311162@12-248-41-177.client.attbi.com>

    Detlev> I have googled this list for IDE recommendations, but have found
    Detlev> none that satisfied me.

Two suggestions.  One, check the editors page on the Python website:

    http://www.python.org/editors/

Two, I guess you should take a look at PyUnitTestBrowser.  The announcement
yesterday said

    This build bonds with PyChecker. That means you can click on a file, get
    a list of warnings, click on a warning, and your editor will pop up with
    the file open and the cursor on the line with the warning.

    Spin:       http://www.c2.com/cgi/wiki?PyUnitTestBrowser

    Source:     http://flea.sourceforge.net/browser005.zip

    That means PUB now tests your code with PyUnit, lints your code with 
    PyChecker, and edits your code with PyShell. 

    It's an IDE in a deceptively small package.

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html




From gerhard at bigfoot.de  Sat Jun 15 15:24:30 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 15 Jun 2002 21:24:30 +0200
Subject: IDE recommendations??
In-Reply-To: <83lm9gtla9.fsf@panacea.canonical.org>
References: <b9veea.l85.ln@majestix.gallien.de> <3d0b06bd$0$52043$e4fe514c@dreader3.news.xs4all.nl> <83lm9gtla9.fsf@panacea.canonical.org>
Message-ID: <20020615192430.GA1923@lilith.my-fqdn.de>

* Kragen Sitaker <kragen at pobox.com> [2002-06-15 15:05 -0400]:
> Boudewijn Rempt <boud at valdyas.org> writes:
> > I've tried to use most IDE's out there -- from Pythonwin to Pythonworks. 
> > None was really comfortable, except for Wing IDE, which is really powerful. 
> > It doesn't include integration of a GUI designer, though, which makes it 
> > little more than a glorified editor/debugger (but _very_ glorified). At one
> > point there was a version of KDevelop that supported Python development, but
> > that has disappeared, it seems.
> > 
> > I use bash & XEmacs & Designer nowadays, and that gives me a fair 
> > productivity.
> 
> I'd like to improve XEmacs to be a better Python IDE.

Does it have an integrated user-friendly debugger (not gdb/pdb-like),
something like right-click on a line and set a breakpoint?

If that GUD thing is only a crude interface like gdb/pdb, then no
thanks, I don't need it.

Otherwise, it would be really useful to have a web page that described
how to use it for Python development. Would be a nice addition to
www.python.org's Emacs page.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From mikeb at mitre.org  Sun Jun 16 13:03:35 2002
From: mikeb at mitre.org (Mike Brenner)
Date: Sun, 16 Jun 2002 13:03:35 -0400
Subject: MS Word -- finding text
Message-ID: <3D0CC4E7.5460F610@mitre.org>

The COM objects (like Project, Word, Excel, etc.) sometimes return stuff in Unicode format. When they do, the python str() function dies when converting non-ASCII unicode characters. 

To avoid this problem, I use the following conversion routine. After making the necessary check for None, it attempts a quick conversion str() first. When necessary, it slowly goes through each character, handling the exceptions that are raised. 

The default is a prime because that is the most common character that hits me in Word and Excel documents. Instead of coding it as an ASCII single-quote characters, these applications code it as a more "beautiful" character, so it kills the python str() function.

You may or may not wish to change the return to eliminate the string.strip there, depending on your needs.

You could make a separate function that has just the TRY and the EXCEPT in it, in order to use the MAP function instead of the for loop.

Mike Brenner


def phrase_unicode2string(message):
    """
    phrase_unicode2string works around the built-in function str(message)
    which aborts when non-ASCII unicode characters are given to it.
    """
    if type(message)==types.NoneType:
       return ""
    try: st=str(message)
    except: # untranslatable unicode character
       list=[]
       for uc in message:
          try:
             c=str(uc)
          except:
             c="`"
          list.append(c)
       # Note: because it raises exception instead of returning
       # a default characters, we cannot use map() here.
       st=string.join(list,"")
    return string.strip(st)

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

Mike Prema wrote: 

#######
from win32com.client import Dispatch
W=Dispatch('Word.Application')
D=W.Documents.Open('c:\\windows\\Desktop\\TOR.doc') ## Test Doc
FindRange=D.Content
F=FindRange.Find.Execute('Conman','True')
print FindRange.Text
#######
str() doesn't seem to work in this case
I tried using the codecs library but I think I am missing something





From mlh at vier.idi.ntnu.no  Sat Jun  8 18:22:05 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 8 Jun 2002 22:22:05 +0000 (UTC)
Subject: [ANN] constraint-0.2
References: <mailman.1023468735.788.python-list@python.org>
Message-ID: <slrnag50sd.gac.mlh@vier.idi.ntnu.no>

In article <mailman.1023468735.788.python-list at python.org>, Alexandre wrote:
>Logilab has released constraint-0.2.
>
>The constraint package is an extensible Constraint Satisfaction Problem
>(CSP) solver using constraint propagation algorithms written in 100%
>pure Python (tested with Python 2.1)
[snip]
>How well does it perform?
> * 8-queens problems solved in 15 seconds on my 1GHz PC

Interesting... The package looks really cool, but the performance in
this example is truly abysmal. The simple queens.py found in the
standard Python distribution solves the 8-queens problem (finds all
solutions) in 0.11 seconds on an 800MHz PC. I guess maybe that't the
price you have to pay for generality... On the other hand, a speedup
factor of well over 100 seems to indicate a potential for
streamlining... Maybe?

- Magnus (who hasn't looked at the internals, and who looks forward to
  playing around with the package :)

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From christophertavares at earthlink.net  Sun Jun  2 17:10:17 2002
From: christophertavares at earthlink.net (Chris Tavares)
Date: Sun, 02 Jun 2002 21:10:17 GMT
Subject: Python Turing machine
References: <d9akfu0c7ro0sebio1vd9c6ik2i0fg59le@4ax.com> <Xns9221CF0CC998Dcliechtigmxnet@62.2.16.82> <ade0mc$j4r$1@newsg3.svr.pol.co.uk>
Message-ID: <ZQvK8.1098$Qy4.62@newsread2.prod.itd.earthlink.net>

"G. Willoughby" <never at mind.info> wrote in message
news:ade0mc$j4r$1 at newsg3.svr.pol.co.uk...
> > hint: post some URLs of those implementations to the list. that way we
> > don't need to look up what it is and people here have fun trying to
solve
> > things pythonic...
>
> I belive he means a chat bot that tries to fool humans into beliving its a
> real person.
>
> --G. Willoughby
>

No, that's the Turing TEST. The universal Turing Machine (both were created
by Alan Turing) is the theoretically minimal general purpose computer. It's
an interesting bit of CS theory, but not particularly useful in day-to-day
life.

-Chris





From edream at tds.net  Mon Jun  3 07:57:28 2002
From: edream at tds.net (Edward K. Ream)
Date: Mon, 03 Jun 2002 11:57:28 GMT
Subject: Using tcl/tk 8.3.4.2 with Python 2.2 on XP
Message-ID: <3CFB5994.5B2F9839@tds.net>

Hi all,

This may be something trivial, but the answer doesn't seem to jump out
at me from the FAQ or the Tkinter page at:
http://python.org/topics/tkinter/download.html

I would like to use tcl/tk 8.3.4.2 with Python/tkinter 2.2 on Windows
XP. Apparently the Python 2.2 distribution uses tcl/tk 8.3.2.  Is there
some way to use tcl/tk 8.3.4.2 with the binary Python 2.2 distribution
on XP?  The thoughts I have:

1.  set an environment variable somewhere or
2.  just move the 8.3.4.2 dll's into the Python22/tcl folder.

Neither seems likely to work well, as there may be problems with header
compatibility, and there are more 8.3.4.2 dll's than dll's in
Python22/tcl.

I hope the next version of Python will come with tcl/tk 8.3.4 or higher
right out of the box.  This will allow us to use wm_iconbitmap(bitmap)
to set (at last!) the default icon for tk frames.

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


From btjenkins at att.net  Tue Jun  4 10:22:41 2002
From: btjenkins at att.net (Ben)
Date: 4 Jun 2002 07:22:41 -0700
Subject: telnet functionality
Message-ID: <9013d0b4.0206040622.f78f95a@posting.google.com>

I am trying to write a script that combines telnet, FTP, and gzip
functionalities.  Here is the process: open telnet connection, change
to new directory, list directory, grab last file in directory (the
file most recently added), and append it to a list.  Then, close the
telnet connection and open an FTP connection and place the list I
created in a folder specified by user (through FTP connection).  I
will then use gzip to open file and perform other actions.

Initially, I designed this to work with FTP only, but the project has
changed and I need to add telnet functionality.  I have worked with
the script to the point where I can open a telnet account, change to
new directory, and list the directory's contents...but, I can't seem
to figure out how to grab the last file in the directory and append it
to the list.  It seems that telnetlib doesn't offer that ability.

Here is what I am working with currently:
from ftplib import FTP
import getpass
import string, sys
import telnetlib
import gzip

HOST = sys.argv[1]
user = sys.argv[2]
password = sys.argv[3]
pw = 'sample'
tn = telnetlib.Telnet(HOST)

tn.read_until("login: ")
tn.write(user + "\n")
if password:
    tn.read_until("Password: ")
    tn.write(password + "\n")

tn.read_until('$' )
tn.write('su')
tn.write(pw + "\n")
tn.write('cd /home/dir\n')
tn.write("ls\n")                        

#flist = []                     # I used these lines from my previous
script
#sdr = []
#tn.ls(flist.append)

tn.write("exit\n")
print tn.read_all()

___________________________________________

Any help would be greatly appreciated.  My efforts to find advice has
not been very successful to date.

Thank you, in advance.

- Ben


From sandeep182 at hotmail.com  Tue Jun 25 16:20:43 2002
From: sandeep182 at hotmail.com (Sandeep Gupta)
Date: 25 Jun 2002 13:20:43 -0700
Subject: Pychecker and /var/tmp/python2-2.2-root/usr/lib/python2.2/.....
References: <d71cc7f6.0206241327.17ebabfa@posting.google.com> <pan.2002.06.25.01.46.31.138959.7184@metaslash.com>
Message-ID: <d71cc7f6.0206251220.4ae17aa0@posting.google.com>

Neal Norwitz <neal at metaslash.com> wrote in message news:<pan.2002.06.25.01.46.31.138959.7184 at metaslash.com>...
> On Mon, 24 Jun 2002 17:27:30 -0400, Sandeep Gupta wrote:
> 
> > I'm using pychecker on my application, which uses BaseHTTPServer.py and
> > unittest.py.  Both those files have "No method" warnings that are
> > displayed as:
> > /var/tmp/python2-2.2-root/usr/lib/python2.2/BaseHTTPServer.py and
> > /var/tmp/python2-2.2-root/usr/lib/python2.2/unittest.py
> > 
> > Does anyone know why /var/tmp/python2-2.2-root gets prepended to the
> > file name?
> > How can I remove the warnings for these files?
> 
> I had this same problem.  From what I remember, Redhat compiles
> the .py -> .pyc/.pyo files in the /var/tmp/.... directory.
> 
> I didn't find a great way to fix it in pychecker, but to
> fix your problem, you should just have to do:
> 
> 	cd /usr/lib/python2.2
> 	python compileall.py .
> 
> Neal

I tried the above steps replacing the second line with:
sudo python2 compileall.py .
Successful compile.  I verified that
/usr/lib/python2.2/BaseHTTPServer.pyo and unittest.pyo exist.  But I'm
still receiving the same error message from pychecker.  Any other
ideas?
Thanks
Sandeep


From fredrik at pythonware.com  Thu Jun 20 15:06:17 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 20 Jun 2002 19:06:17 GMT
Subject: do...while
References: <slrnah40l1.12v.msoulier@pmerd071.ca.nortel.com> <XFMail.20020620104256.shalehperry@attbi.com> <mailman.1024596866.20597.python-list@python.org>
Message-ID: <JIpQ8.44319$n4.10345719@newsc.telia.net>

Michael P. Soulier wrote:

> Right, I've used that in the past. It doesn't seem nearly as clean though,
> as making the conditions for terminating the loop plain in the loop
> declaration.

this has been debated hundreds of times (google can show you all
arguments for and against adding yet another control structure).

unless you really don't have something better to do with your time,
I suggest you get over it and write some code...

</F>




From lhudson at geminidataloggers.com  Thu Jun 27 11:19:21 2002
From: lhudson at geminidataloggers.com (Lawrence Hudson)
Date: Thu, 27 Jun 2002 16:19:21 +0100
Subject: How to find out operating system
References: <mailman.1025159195.9033.python-list@python.org>
Message-ID: <3D1B2CF9.1000605@geminidataloggers.com>

If you have PyWin32, try this:

d:\>python
Python 2.0 (#8, Jun 12 2002, 10:20:16) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
 >>> import win32api
 >>> win32api.GetVersionEx()
(5, 0, 2195, 2, 'Service Pack 2')
 >>>

majorVersion:   Identifies the major version number of the operating system.
minorVersion:   Identifies the minor version number of the operating system.
buildNumber:    Identifies the build number of the operating system in 
the low-order word. (The high-order word contains the major and minor 
version numbers.)
platformId:     Identifies the platform supported by the operating 
system.  May be one of VER_PLATFORM_WIN32s, VER_PLATFORM_WIN32_WINDOWS 
or VER_PLATFORM_WIN32_NT
version:        Contains arbitrary additional information about the 
operating system.

Thanks,
Lawrence


A wrote:
> Hi,
> What is the best way of  finding out the kind operating system?
> I can use os.name but if the system is Windows I would like also 
> know if the system is Windows98 or Windows ME or W2K or 
> Windows XP.
> Thanks for help.
> Ladislav
> 
> 
> 
> 




From mdk at mdk.com  Thu Jun 20 13:37:59 2002
From: mdk at mdk.com (mdk)
Date: 20 Jun 2002 17:37:59 GMT
Subject: Function to add commas in numbers?
References: <Xns923372EFBBC5Dmdkmdkmdkmdk@130.133.1.4> <3d11f8d2$1_1@hpb10302.boi.hp.com>
Message-ID: <Xns92338AAF39BD6mdkmdkmdkmdk@130.133.1.4>

"Daniel Fackrell" <dfackrell at DELETETHIS.linuxmail.org> wrote in
news:3d11f8d2$1_1 at hpb10302.boi.hp.com: 

> mdk,
> 
> The following thread is about formatting numbers for Danish, but it
> should give the info you need (paste it as one line):
> 
> http://groups.google.com/groups?hl=en&lr=&ie=UTF8&oe=UTF8&threadm=3D0A1
> 791.3 
> 060307%40mxm.dk&rnum=5&prev=/groups%3Fas_q%3Dnumber%2520format%26ie%3DU
> TF8%2 6oe%3DUTF8%26as_ugroup%3Dcomp.lang.python%26lr%3D%26hl%3Den

This worked great, thanks.


From radix at twistedmatrix.com  Sun Jun  9 20:16:00 2002
From: radix at twistedmatrix.com (Christopher Armstrong)
Date: 09 Jun 2002 20:16:00 -0400
Subject: Source formatting for long expressions.
In-Reply-To: <874rgcxy99.fsf@twistedmatrix.com>
References: <mailman.1023570972.27430.python-list@python.org>
	<m3it4srhq5.fsf@mira.informatik.hu-berlin.de>
	<874rgcxy99.fsf@twistedmatrix.com>
Message-ID: <87wut8vvi7.fsf@twistedmatrix.com>

>>>>> "ca" == Christopher Armstrong <radix at twistedmatrix.com> writes:

>>>>> "mvl" == Martin v Loewis <martin at v.loewis.de> writes:

    mvl> As you can see, it tries to align further arguments to a function
    mvl> together with the opening parenthesis. I would assume that any other
    mvl> auto-formatters use the same style, so you won't get any "sane"
    mvl> indentation until you drop the nesting level (perhaps in favour of
    mvl> method invocations).

[snip..]
    ca> I'm probably going to end up doing this myself, and I thought of
    ca> something while reading your message. I could have two steps for
    ca> formatting: template- generation and indentation. the first step would
    ca> generate something like a repr() of all of my objects, but with \ns and
    ca> \ts embedded at strategic locations. Then I could replace all \ts with
    ca> current_indent_level, calculating current_indent_level by counting (s,
    ca> {s, and [s. I'm going to try this out now.  :-)


I got this working. It was fairly easy, once I got a templated source 
representation::

def indentify(s):
    out = []
    stack = []
    for ch in s:
        if ch in ['[', '(', '{']:
            stack.append(ch)
        elif ch in [']', ')', '}']:
            stack.pop()
        if ch == '\t':
            out.append('  '*len(stack))
        else:
            out.append(ch)
    return string.join(out, '')


This is a fairly stupid indenter, but it works for my needs::

>>> print aot.indentify("Instance('twisted.internet.app.Application', \
\n\tfoo=bar, \n\tbaz={\n\t'quux': 1, \n\t'spam': \n\tInstance('Eggs', \
\n\tmore='state')})") #this source will really be generated by my __repr__s.
Instance('twisted.internet.app.Application', 
  foo=bar, 
  baz={
    'quux': 1, 
    'spam': 
    Instance('Eggs', 
      more='state')})


-- 
                                Chris Armstrong
                         << radix at twistedmatrix.com >>
                http://twistedmatrix.com/users/carmstro.twistd/





From prema at prema.co.nz  Sat Jun 15 15:40:40 2002
From: prema at prema.co.nz (Prema)
Date: 15 Jun 2002 12:40:40 -0700
Subject: MS Word -- finding text
References: <mailman.1024124539.13851.python-list@python.org>
Message-ID: <ae58a20e.0206151140.82a5f71@posting.google.com>

Hi David ! 
Thanks for your response.  
The script I am building is intended to provide a search facility over
a group of word docs. The strategy is to pull plain text from the doc
in numbered paragraphs then search each para in Python and return any
found ones.

Simplified code so far .....

#######
from win32com.client import Dispatch
W=Dispatch('Word.Application')
D=W.Documents.Open('c:\\windows\\Desktop\\TOR.doc') ## Test Doc
FindRange=D.Content
F=FindRange.Find.Execute('Conman','True')
print FindRange.Text
#######
str() doesn't seem to work in this case
I tried using the codecs library but I think I am missing something

Thanks for your help with this !!
Kind regards 
Mike



"David LeBlanc" <whisper at oz.net> wrote in message news:<mailman.1024124539.13851.python-list at python.org>...
> Please post some snippets. That's always more likely to get useful
> assistance then stating a general problem.
> 
> Where are you encountering the unicode decoding problem?
> 
> Regards,
> 
> David LeBlanc
> Seattle, WA USA
> 
> > -----Original Message-----
> > From: python-list-admin at python.org
> > [mailto:python-list-admin at python.org]On Behalf Of Prema
> > Sent: Friday, June 14, 2002 23:50
> > To: python-list at python.org
> > Subject: MS Word -- finding text
> >
> >
> > Hi All !
> > Just wondering if anyone has some sample code around how to open a
> > document, find a string and return the paragraph(s) containing the
> > string.
> >
> > I have got a good part of the way but running up against Unicode
> > decoding
> >
> > Can post some snippets if needed
> > Thanks very much in advance
> >
> > Kind regards
> > Mike
> > --
> > http://mail.python.org/mailman/listinfo/python-list


From peter at engcorp.com  Sun Jun 23 17:51:32 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 23 Jun 2002 17:51:32 -0400
Subject: Zope
References: <3D163DC7.739A83FD@NOSPAM_hotmail.com>
Message-ID: <3D1642E4.C85BDE4E@engcorp.com>

Marc Bigler wrote:
> 
> I would like to know how I can generate a navigation bar in zope ?
> 
> Something like that: 
[...] 
> Home > Flowers > Orchidee
> 
> Is that somehow possible examining the BASE0, BASE1,... and so on
> variables or is there a better method ?

The Zope mailing list is the best place to ask this sort of
question.  There is little Zope discussion here.

You might try doing a Zope.org or Google search for "breadcrumbs"
since I think that's what they call the thing you're trying to do.

-Peter


From duncan at NOSPAMrcp.co.uk  Wed Jun 19 06:56:59 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Wed, 19 Jun 2002 10:56:59 +0000 (UTC)
Subject: CPU Usage Windows
References: <ad381f5b.0206190233.4fa1bf58@posting.google.com>
Message-ID: <Xns9232788EA19ADduncanrcpcouk@127.0.0.1>

zaka07 at hotmail.com (Rad) wrote in 
news:ad381f5b.0206190233.4fa1bf58 at posting.google.com:

> Hi,
> 
> I've noticed that when I run Python script on 
> Windows 2k Pro
> Dual Intel Xeon (2.2GHz)
> CPU Usage for pythonw.exe is only 50%?!
<snip>
> Is it something to do with python not being able to use the power of
> both processors?
Yes, although if your scripts are suitably threaded, and some time is 
being spent outside of Python (reading files/querying databases) I think 
you should be able to get some usage above 50%.

If you can run two of your programs at the same time then CPU usage should 
go up to about 100%. Running 3 will probably be able to soak up any idle 
time still left waiting for disc or network responses. The single processor 
limitation only applies within each process.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From tlirobot at tlirobotics.cc  Wed Jun 12 19:09:40 2002
From: tlirobot at tlirobotics.cc (Dylan Appleman)
Date: Wed, 12 Jun 2002 23:09:40 GMT
Subject: Compiling to .exe
Message-ID: <UwQN8.43539$zh2.13069449@twister.neo.rr.com>

Is it possible to compile Python to .exe? If so, what program(s) do I need?
Windows XP, no other programming experience but Python.




From aahz at pythoncraft.com  Wed Jun 19 13:25:33 2002
From: aahz at pythoncraft.com (Aahz)
Date: 19 Jun 2002 13:25:33 -0400
Subject: any users of ExtensionClass out there?
References: <mailman.1024440443.17628.python-list@python.org> <8b04704.0206190855.5696145b@posting.google.com>
Message-ID: <aeqeqd$k62$1@panix1.panix.com>

In article <8b04704.0206190855.5696145b at posting.google.com>,
Rodrigo Senra <rodrigo.senra at ic.unicamp.br> wrote:
>Jeremy Hylton <jeremy at zope.com> wrote in message news:<mailman.1024440443.17628.python-list at python.org>...
>>
>> There has been little perceived interest in ExtensionClass releases
>> because Zope is moving away from ExtensionClass.  I wonder, though, if
>> there are other users of ExtensionClass who would be interested in a
>> new release.  Please let me know if you are.
>
> Even though I have no production code based in ExtensionClass, I'd
> used it sometimes to explore Python reflection capabilities over C
> extensions.  Therefore, I'm interested in any changes you & your team
> may have come up with.

In that case, you're better off continuing any future explorations in
Python 2.2+; new-style classes are intended to replace ExtensionClass,
and the dev team is *far* more interested in hearing about issues with
new-style classes.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From hjwidmaier at web.de  Fri Jun 21 03:56:04 2002
From: hjwidmaier at web.de (Hans-Joachim Widmaier)
Date: 21 Jun 2002 00:56:04 -0700
Subject: Advice sought on ps module
References: <6e990e29.0206200325.355f0413@posting.google.com> <aesoqg$kc9$1@scotsman.ed.ac.uk>
Message-ID: <6e990e29.0206202356.544eb696@posting.google.com>

eddie at holyrood.ed.ac.uk (Eddie Corns) wrote in message news:<aesoqg$kc9$1 at scotsman.ed.ac.uk>...

> I'm not sure that I understand completely what you're trying to do so maybe
> I'm barking up the wrong gum tree here but ...

I admit that my writing wasn't very clear ...
 
> Surely you want to use postscript itself to do all the work.  PS already has
> an operator called stringwidth and a complete set of programming constructs so
> the trick usually is to generate the appropriate PS commands.

That's how I do it today. Well, I should have given a more concrete
example. I've written a program that creates good looking timing
diagrams for documentation purposes. These consist of a drawing part
and signal names to the left. While the python program knows exactly
how large the drawing itself is, it has no way of knowing how far the
signal names extend to the left, making it all but impossible to give
a valid boundig box for the whole thing -- which is required, as it's
going to be an EPS file. Sure, I can get at the BBox by piping the
finished file through one of the ps utils programs, but this is kinda
cumbersome. And it still doesn't allow me to adjust the right-hand
drawing width so the combined width is of some fixed size.

> As an example,
> to draw a border around some arbitrary text I guess you could generate code to
> do something like:  save current position, for each string you draw calculate
> the bounding box and remember the highest X value and at the end use the new
> position along with the saved position and highest X value to define the four
> corners of a rectangle.  It does require a fair degree of competence in PS but
> you wouldn't tackle such a project without it (or a desire to achieve it)
> would you :)

Another example: In my CD cover/inlay generator I have to do insert
line breaks if the Title+Singer+Time line doesn't fit. I do this in
PostScript, as I don't know the width in the Python program. This has
the drawback that the height of the title list changes and my careful
layout (not to be taken too seriously) is disrupted. And no, I do
*not* want to write a layout engine in PostScript.

Despite having read most of the PostScript Reference Manual, the AFM
spec, the DSC specs, etc., ... and having written a bunch of programs
that output PostScript, I still consider myself a novice (at best!).
This reverse notation thing with all those stacks gives me the creeps.
;-)

Hope these explanations make the misery I'm in somewhat more clear.

Hans-Joachim


From peter at engcorp.com  Mon Jun  3 22:14:03 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 03 Jun 2002 22:14:03 -0400
Subject: Interfacing to 1-Wire LAN, especially Dallas temperature sensors?
References: <a7iefuk650noamrqcmpvroqj60mtgbpf9v@4ax.com> <3CF82E8E.23C6FA10@engcorp.com> <3cfb83e2$1_4@goliath.newsgroups.com>
Message-ID: <3CFC226B.B6CB8995@engcorp.com>

Brad Clements wrote:
> 
> There is a technique where you can use a UART at some whacky baud rate and
> bits/byte setting to communicate. There's an APPNOTE somewhere, I don't have
> it handy.

If someone can dig that out (I searched before and couldn't find it
in all the noise) I'd appreciate it.  I would also be happy to try
to implement it under Python, since we'd have some uses for such a
beast.

(But I still don't believe it can be done reliably.  Maybe I'm wrong.
Would be cool.)

-Peter


From cbrown at metservice.com  Sun Jun 23 22:19:02 2002
From: cbrown at metservice.com (Colin Brown)
Date: Mon, 24 Jun 2002 14:19:02 +1200
Subject: Socket Disconnection
References: <3d1304d0$0$224$4d4ebb8e@news.nl.uu.net>
Message-ID: <3d16817e$1@news.nz.asiaonline.net>

Hi Guyon

I assume here that you are talking TCP/IP.

If you are waiting on a receive [ data = connection.recv() ] then you will
have an empty data string returned when the client disconnects.

If you are sending you will get an error telling you the link was shut down.

Colin Brown
PyNZ


"Guyon Mor?e" <gumuz at looze.net> wrote in message
news:3d1304d0$0$224$4d4ebb8e at news.nl.uu.net...
> Hi!
>
> I have written a small and simpel server and I was wondering how do you
find
> out that a client has disconnected?
>
> thnax,
>
> Guyon
>
>




From deathtospam43423 at altavista.com  Fri Jun  7 11:17:35 2002
From: deathtospam43423 at altavista.com (netvegetable)
Date: 7 Jun 2002 15:17:35 GMT
Subject: Newby: How do I strip HTML tags?
Message-ID: <Xns92278CC566FEAusYourStandingInIt@130.133.1.4>

I'm mucking around with cgi, and I'm trying to work out a way to strip the 
html tags of a string. e.g, I want to convert this...

><font size = 12><b><big>Really Big String</big></b></font>

to this this ...

>Really Big String

... and store it as a value.

I worked out a crude, but effective way of doing it (see code below), but I 
can't escape the feeling there must be a built in way of doing it more. If 
nothing else, I'm sure somebody who knows their regular expressions could 
neaten it up (please?).

def strip_html_tags(it):
	left = it[:(len(it)/2)]
	right = it[(len(it)/2):]
	final = left[left.rfind('>')+1:] + right[:right.find('<')]
	return final

-- 
netvegetable at excite.com


From johnroth at ameritech.net  Fri Jun 21 08:13:56 2002
From: johnroth at ameritech.net (John Roth)
Date: Fri, 21 Jun 2002 08:13:56 -0400
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com> <lkznxqjqx2.fsf@pc150.maths.bris.ac.uk> <3D11A21D.9010809@mxm.dk> <uh3e1usajle42c@news.supernews.com> <3D11DA9B.163A7BD0@info.unicaen.fr> <uh446elev7af68@news.supernews.com> <3D12EEAA.9C3A3D29@info.unicaen.fr>
Message-ID: <uh6626e7pb5lc6@news.supernews.com>

"Jerzy Karczmarczuk" <karczma at info.unicaen.fr> wrote in message
news:3D12EEAA.9C3A3D29 at info.unicaen.fr...
> John Roth comments my observation concerning the relation between
gotos
> and continuations:
>
> > >    ... in order to speak about genuine gotos *you don't need to
think
> > >    assembler*.
> > >    What you need is to understand correctly the notion of
continuation
> > >    (something quite standard - again - in the Functional
Programming
> > >    world). This is a high-level model of "flat" branching, but may
be
> > >    used to more exquisite constructions, such as backtracking.
> >
> > That's still not a goto in the classical sense. In the classical
sense,
> > a goto allows you to go to any statement. The point you're missing
is
> > that the goto statement is a denial of structure - since it can go
> > anywhere, you can't make any assumptions about the program
> > structure. The piece of code you're looking at can be mysteriously
> > activated from somewhere else at any time.
>
> I am not sure I am "missing a point". Perhaps I have a different
vision
> of the computational process, hm? What is the "classical sense" for
one
> may be just an orthodox, restricted view for somebody from another
galaxy.
> I tend to apply often a Functionalist, sometimes too formal view on
the
> semantics of computations, you don't need to accept it, but I assure
you
> that for many people gotos *ARE* continuations. Continuation in a
broad
> sense is just the *specification* of what your <<future>> is.

Well, yes. My vision of what a GOTO is comes from my experiance in
the '60s, with early versions of Fortran, COBOL, various assemblers,
and so forth. This is what Dijkstra was talking about when he wrote
"GOTO considered harmful." It's completely unstructured. To put it
bluntly (because I can't think of a way to be more diplomatic this early
in the morning, sorry) attempting to redefine GOTO in terms of some
other structure misses the point completely: a major thread of
programming
language development since Dijkstra's letter has been to find language
structures that can do the job while eliminating the completely
unstructured
GOTO. Functional programming, continuations, structured exceptions,
and similar stuff are all attempts to provide structures that are
powerful
enough to get the job done, and structured enough that you can
reason about them.

To delve into the history of language development a bit: immediately
following Dijkstra, there was a minimalist structured programming
movement, and a horrible debate about the Case statement, and
similar constructs. The minimalists didn't want it, the people who
wanted to get the job done did.

> (Oh, I adored that wonderful piece of programming madness, an
"article"
> where somebody proposed a new super-Fortran where GO TO was replaced
by
> CAME FROM statement, and the language had an inspiring
non-deterministic
> conditional:   If <<condition>> THEN WHAT?.
> Any references, anybody?)

It's been around for quite a while - I remember it from GUIDE meetings
in the '70s. There's a COBOL variant, and I'm certain there are others.
Once Dijkstra focused attention on the GOTO, someone came up
with COME FROM. It's actually very logical in a very warped way.
If the problem is that you can GOTO anywhere, leading to perplexity
as to how you got there, then why not have a construct that allows you
to say where you came from, thus removing the perplexity? I knew
people back then whose eyes glazed over considering it!

You might try the history of computing people -
although they might want this one to escape!

>
> Thank you for this interesting exchange.
>
> Jerzy Karczmarczuk
> Caen, France




From kaci_tizi_ouzou2000 at yahoo.ca  Mon Jun  3 09:14:08 2002
From: kaci_tizi_ouzou2000 at yahoo.ca (Kaci Tizi Ouzou)
Date: 3 Jun 2002 06:14:08 -0700
Subject: WIN32 Decimal point/coma
Message-ID: <68599ccc.0206030514.4364e899@posting.google.com>

Greetings all,

I am having a problem running my application on Win32
for if in the regional settings the user changes the
decimal dot to say ',' then floating point operations
do not work anymore!!!

After few experiments, I came to the conclusion that, it is
at 'compile' time that the dot/coma is decided. This means
if one compiles py files with dot as the decimal 'separator'
and then changes the settings so that coma is the new separator
then the code won't work anymore!

My question is: 

Is there a call to make so that to ignore the regional settings ?

Is there a solution/suggestion for this problem?

Thanks all


From tebeka at cs.bgu.ac.il  Thu Jun 20 04:45:04 2002
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 20 Jun 2002 01:45:04 -0700
Subject: Help: win32gui & threads
References: <33803989.0206170318.9d3a225@posting.google.com> <3D0E6849.3050909@skippinet.com.au> <33803989.0206190041.43b788ac@posting.google.com>
Message-ID: <33803989.0206200045.490e3d55@posting.google.com>

Hello Mark,

I've set the writer and tcp server threads to be deamonic and
this seems to solve the problem.

Thanks for you help.
Miki


From tack at cscs.ch  Mon Jun 24 07:25:31 2002
From: tack at cscs.ch (Davide Tacchella (CSCS))
Date: Mon, 24 Jun 2002 13:25:31 +0200
Subject: 64-bit Python?
References: <87elfl3yvu.fsf@mathdogs.com> <C0FL8.41941$n4.9496108@newsc.telia.net> <j41ybk69ac.fsf@informatik.hu-berlin.de> <83660wgvrm.fsf@panacea.canonical.org>
Message-ID: <3D1701AB.C862F82F@cscs.ch>

Kragen Sitaker wrote:
> 
> Yes, this is correct.  Even the buffer interface, which my
> arrayfrombuffer package[0] uses to provide access to mmapped files as
> Numeric arrays inexplicably uses int instead of size_t.
> 
> The problem is that all of the 64-bit platforms I have access to
> (various Linuxes on Alpha and IA-64, Tru64 on Alpha) have 4-byte ints
> and 8-byte longs, at least by default.  This is called "LP64": longs
> and pointers are 64, but ints are 32, so all the code that assumes
> ints are 32 will continue to work.
> 

Did somebody successfully built Python on AIX with 64 bit support ?

dynload_aix.c seems not to be 64 bit safe code ...
---- (function aix_getoldmodules)
offset = (unsigned int)ldiptr->ldinfo_next;
ldiptr = (struct ld_info *)((unsigned int)ldiptr + offset);
---

Davide


From david.bear at asu.edu  Wed Jun 12 12:31:07 2002
From: david.bear at asu.edu (David Bear)
Date: Wed, 12 Jun 2002 09:31:07 -0700
Subject: postscript processing in python
Message-ID: <ae7t0a$f4t$1@news.asu.edu>

I'm looking for tools to assist in postscript processing and pdf generation 
in python.  Hoping to find 'free' code -

I'm looking for something that will

1) validate and correct questionable postscript  (check document structure)
2) convert postscript to pdf
3) convert text to postscript

I know these are tall orders but I've written a filter in python that takes 
a postscript stream generated by an IBM mainframe and uses ghostscript to 
convert it to pdf.  The problem is ghostscript seems rather picky about 
some strange things that the mainframe puts in the postscript stream -- so 
I was hoping to do some cleanup on it and conversion without having to 
depend on ghostscript.  

Any pointers appreciated.

-- 
David Bear
College of Public Programs/ASU
480-965-8257
...the way is like water, going where nobody wants it to go


From gregor at hoffleit.de  Tue Jun  4 10:34:02 2002
From: gregor at hoffleit.de (Gregor Hoffleit)
Date: Tue, 4 Jun 2002 16:34:02 +0200
Subject: libpython as a shared library?
In-Reply-To: <20020604124540.E9162@phd.pp.ru>
References: <Pine.BSF.4.32.0206031620070.82931-100000@localhost> <20020604124540.E9162@phd.pp.ru>
Message-ID: <20020604143402.GB22993@hal.mediasupervision.de>

* Oleg Broytmann <phd at phd.pp.ru> [020604 10:51]:
> On Mon, Jun 03, 2002 at 08:24:28PM +0000, Gregory (Grisha) Trubetskoy wrote:
> > Has there been any progress in making an option of building libpython as a
> > shared lib (.so)? I saw some feature requests, but couldn't quite figure
> > out the ultimate status of this.
> 
>    Implemeneted in Python 2.3. Download CVS version or nightly tarball.

Debian has had libpython as a shared lib since a while. Recently, we got
a report that the performance of the shared version is dramatically
slower than the statically linked version (e.g. about 30%). I made some
pystone tests on different architectures of our GNU/Linux, and tried
different flags for building and linking the library (and different
Python versions), and in fact there seems a significant speed penalty
involved as soon as the interpreter is linked against the shared
library. The penalty has some correlation to the architecture, but it is
significant on all tested constellations.

I'll give a try at the 2.3 tree and see if the problem can be confirmed
there as well.

I would be interested if people could try to confirm this with other
operating systems and other variants of GNU/Linux as well.

    Gregor




From claird at starbase.neosoft.com  Tue Jun 25 08:38:42 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 25 Jun 2002 07:38:42 -0500
Subject: Python daemon
References: <mailman.1024999953.15765.python-list@python.org>
Message-ID: <F733C874CD014E16.148E57261FFD5E7A.AFFA4BD17B2F8D8B@lp.airnews.net>

In article <mailman.1024999953.15765.python-list at python.org>,
Roman Suzi  <rnd at onego.ru> wrote:
			.
			.
			.
>> 3. restarting of the program if it fails, or the watchdog doesn't trigg=
>er
>
>Run the program from the /etc/inittab - if it terminates,=20
>init will restart it for you.
			.
			.
			.
This is one of the great FMMs under Unix--the under-utilization
of init(1).  From everything I know, Roman's exactly right, that
/etc/inittab is the right way to set up a process that you truly
want to keep running.  However, many, MANY Unix hosts have all
kinds of ugly home-brewed hacks to duplicate this functionality.
I don't have an explanation, beyond the rather tautologous ob-
servation that init(1) simply isn't as widely understood as it
deserves to be.
-- 

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


From markus.vonehr at ipm.fhg.de  Mon Jun  3 06:11:10 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Mon, 03 Jun 2002 12:11:10 +0200
Subject: Python program problem involving reading a web site.
References: <ad6nvd$4mr$1@peabody.colorado.edu>
Message-ID: <3CFB40BE.8B822E69@ipm.fhg.de>

You can use win32com to access IE or netscape if you are on
a win32 system.

Look at the testNetcape.py example.

Mike Varney schrieb:

> Hello.
>
> I wrote a quick little program that is supposed to read the html from a web
> site and to update the web counter to show another hit.
> I am able to read the html code fine, but the counter on the web site is not
> updating.
> Here is the code.
>
> Any suggestions?
> Thanks.
>
> import urllib # imports url library
> from time import sleep # imports sleep timer
>
> # Initilize variables
> a = 0
> t = 0
>
> # User Input
> x = input("How many hits do you want to add to counter? ")
> t = input("How long do you want to pause between each hit? (Recommend > 1
> min) ")
>
> # while loop
>
> while a < x:
>  # open and reads page
>  sock = urllib.geturl(http://foo.bar)
>  hampage = sock.read()
>  sock.close()
>  print a
>  sleep(t)
>  a = a+1
> # print goodbye
> print "you have increased the page count by "
> print a
> print "hits"



From zayats at blue.seas.upenn.edu  Mon Jun  3 10:39:41 2002
From: zayats at blue.seas.upenn.edu (Salim Zayat)
Date: 3 Jun 2002 14:39:41 GMT
Subject: Changing TK
Message-ID: <adfv3d$jsp$1@netnews.upenn.edu>

Hey all.  I have a quick question about something.  If I were to install 
a newer version of Tcl/Tk on my computer, would Tkinter.py recoginze it 
automatically, or do I need to tweak the Tkinter.py code itself?

Salim


From fperez528 at yahoo.com  Wed Jun 19 18:24:18 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 19 Jun 2002 16:24:18 -0600
Subject: OS X debugger?
References: <aeqotj$dm$1@panix1.panix.com> <aeqp4m$c09$2@peabody.colorado.edu> <aeqs9j$cfn$1@panix1.panix.com>
Message-ID: <aer0aq$f3d$1@peabody.colorado.edu>

Aahz wrote:

> From some quick Googling, it appears that a bigger issue would be lack
> of support for Python 2.2.  Have you used DDD with 2.2?

No, sorry. I only used ddd/python back in 2001, with 2.1. So I have no idea 
about the issues with 2.2 (didn't even know there were issues to begin with).

On the other hand, if you can make it work it's quite nice. It integrates 
withh gnuplot for array display, which is nifty.

good luck,

f.


From bl at netgroup.dk  Mon Jun  3 04:01:56 2002
From: bl at netgroup.dk (Bo Lorentsen)
Date: 03 Jun 2002 10:01:56 +0200
Subject: Embedding and threads ?
In-Reply-To: <j4d6vac9b7.fsf@informatik.hu-berlin.de>
References: <mailman.1022667286.17372.python-list@python.org>
	<j44rgrnkfm.fsf@informatik.hu-berlin.de>
	<mailman.1022740196.32586.python-list@python.org>
	<j4d6vdwvyk.fsf@informatik.hu-berlin.de>
	<1022771667.689.136.camel@netgroup>
	<j4znyhzguh.fsf@informatik.hu-berlin.de> <1022830475.893.17.camel@netgroup>
	<j4znygd7iv.fsf@informatik.hu-berlin.de> <1022839266.823.66.camel@netgroup>
	 <j4d6vac9b7.fsf@informatik.hu-berlin.de>
Message-ID: <1023091316.685.40.camel@netgroup>

On Sun, 2002-06-02 at 11:35, Martin v. L?wis wrote:

> I see. This is necessary indeed. The builtins are normally inherited
> from some context when doing "exec" or "eval" at the Python level; on
> the C API, no such inheritance happens automatically.
Ok, I see. Its just that I assumed that the interpreter state was more
isolated, regarding real threads, because of this.
For me it looked more like the library had i thread problem than the
core interpreter.

> Nearly everything in Python
> - changing and accessing dictionaries
> - changing and accessing lists
> - writing to files
> - allocating memory (atleast with pymalloc)
Ok, so the word "everything" is really more appropriate :-)

> etc. None of these operations is thread-safe per se; it is only the
> GIL that prevents breakage.
And only one CPU thread can execute it at a time. As You may have more
threads but only one can execute Python code at a time, because the GIL
will serialize the execution !

> > Arhh, so Python can execute in a MT environment, but only one
> > interpreter at a time, the rest is pseudo MT ?
> 
> I don't think I understand this question. As I said before, restrict
> yourself to "one interpreter" (in the sense of PyInterpreterState*).
> This is the normal case even in the presence of threads.
What I meant is that, it is only possible to make threads that are
software scheduled (Python (and its lib functions) deside when to
release the GIL), and if I try to use real threads the result will still
be limitet by this software threading.

> I don't know what pseudo MT is (in what sense is it pseudo?). It
> certainly has multiple concurrent threads.
I'm not sure it a proper term, but this is how I se it : software thread
= pseudo thread

Pseudo, is because it can't really take advantage of SMP (or the system
process scheduler), and its the software that controlles the
"scheduling".

> For most people, it is "real" MT today, since they have only one
> processor in their computers, anyway.
I know, and as for now I don't know how important it is for me. But as I
looked into the Python interface, I was in the beginning hoping that the
threads was more "naturly" implimentet in the language engine, and I was
hoping to take advantage of this in the future :-) 

> To make it "more real", you would have to introduce what is known as
> "free threading". People have tried this (ask Google) and found that
> it slows down the interpreter significantly.
Hmm, this sounds like they had not done it the proper way :-) But I will
take a look in Deja/Google !

But ok, I'll use what Python provide, and I'm sure it will work out just
fine, and then I'll see hov things in Python progress regarding threads
in the future :-)

/BL





From bokr at oz.net  Tue Jun 25 21:35:44 2002
From: bokr at oz.net (Bengt Richter)
Date: 26 Jun 2002 01:35:44 GMT
Subject: SOLVED (Re: Python hits the spot)
References: <mailman.1025043213.28349.python-list@python.org> <afb48j$vl2$0@216.39.172.122>
Message-ID: <afb5pg$86f$0@216.39.172.122>

On 26 Jun 2002 01:09:39 GMT, bokr at oz.net (Bengt Richter) wrote:

>On Tue, 25 Jun 2002 15:12:47 -0700, "Jimmy Retzlaff" <jimmy at retzlaff.com> wrote:
>
>>Bengt Richter wrote:
>>>>> Well, lucky guess on the heat. That _probably_ accounts for laptop
>>>>> going to 87% and then shutting down, but how about the other huge
>>>>> slowdowns in your big loop of 8? Did you get that diagnosed too?
>>>>> Just curious ;-)
>>>>
>>>>On my stationary machine definitely not. The 8 steps remain constant.
>>I=20
>>>
>>>Still curious ;-)
>>
>>I believe that newer mobile CPUs will throttle back their clock speed in
>>response to heat. This could explain the slow down. Apparently the
>                                      ^^^-- or "some of the"
>>lowest clock speed wasn't low enough to keep the CPU happy...
>>
>Yeah, I understood that (having introduced the concept into this thread, I think ;-)
                                                             ^^^^ oops, original, not this one
Siegfried certainly advertised the many shutdowns/slowdowns/"freezes". I just hypothesized:
<pats self on back>
  "I wonder how/whether the percentage calculated is affected by power management
  that slows down the CPU clock to keep it cool? Wonder if that could be happening.
  I've never looked into power management stuff. I guess there is a temperature
  where it just shuts down too, to protect itself?"
</pats self on back><Need it today ;-p /> RPM1 was apparently also onto it when he
cryptically said, "Just curious, is your laptop a Hewlitt Packard?"

>But it was not clear to me whether that explained _everything_ (see other post re two
>readings of "On my stationary machine definitely not. The 8 steps remain constant." ;-)
>

Regards,
Bengt Richter


From R.Barrett at ftel.co.uk  Mon Jun 24 06:23:55 2002
From: R.Barrett at ftel.co.uk (Richard Barrett)
Date: Mon, 24 Jun 2002 11:23:55 +0100
Subject: Python and remote sensing
In-Reply-To: <3D16E528.5040207@kfunigraz.ac.at>
Message-ID: <5.1.0.14.2.20020624104306.0341dd48@pop.ftel.co.uk>

At 11:23 24/06/2002 +0200, Siegfried Gonzi wrote:
>Hi,
>
>
>a technician of us asked me whether it is possible to do the following:
>
>
>Starting a script on Unix at lets say 2 am and calling a remote site 
>(http://www.whatever...) and that site should become passively processed 
>(the site holds 41 observing stations). Is it possible to start a remote 
>script (automatically) connect to a http page and retrieve information 
>from that html-page?
>I answerd Perl or Python can do the job. But I am not sure whether that is 
>correct.
>
>I am sure the above topic has been asked 1000 times. Or is there a tool or 
>program availabe which can do the job better?; I mean our technican has 
>never heard of Python and I am not interested to make web based (put in 
>the right term here) programming or the like.
>
>If Python can do it - okay; if not please say so (if Python is dead - 
>okay; if Python lives - okay)!
>
>
>Thanks,
>S. Gonzi

You can write a script in Python or Perl to be executed by the UNIX cron 
daemon at the time you specify (see the UNIX crontab command man page). 
There is nothing out of the ordinary about such scripts beyond their being 
listed in an appropriate user's installed crontab and you will not want it 
to prompt for input parameters.

In the case of Python you can simply use the urllib, urllib2 or httplib to 
retrieve data using HTTP; these modules being listed in order of reducing 
simplicity of use.

Try using urllib from the python command line and you'll see how easy it is.

Probably the biggest issue will be if you have to use any form of 
authentication in order to get a response from the HTTP server. But that 
isn't a problem

Perl almost certainly has equivalent modules if you prefer it to Python.

With Python 2.2.1 a basic script like the one that follows will do it if 
you have to use HTTP basic authentication. You'll need to add some extras 
to handle problems with opening the URLs, supplying parameters if the URLs 
need to incorporate them etc. But basically its really quite easy to do.


#! /usr/bin/env python

import urllib

# Hardwired HTTP basic authentication parameters
basic_auth_username = '....'
basic_auth_passwd = '....'
# URL we want data from
url = 'http://.......'
# Destination filename for data retrieved
destination_filename = '.....'

class myURLopener(urllib.FancyURLopener):

     def prompt_user_passwd(self, host, realm):
         return basic_auth_username, basic_auth_passwd

urllib._urlopener = myURLopener()

stuff = urllib.urlretrieve(url)

open(destination_filename, 'w').write(open(f[0], 'r').read())






From python at tykebsd.net  Thu Jun  6 07:56:02 2002
From: python at tykebsd.net (Dave Moor)
Date: 6 Jun 2002 04:56:02 -0700
Subject: Interfacing to 1-Wire LAN, especially Dallas temperature sensors?
References: <a7iefuk650noamrqcmpvroqj60mtgbpf9v@4ax.com>
Message-ID: <1bf96e7f.0206060356.417a7394@posting.google.com>

Hi

I have just started producing a PyOneWire module it currently can
search the one wire network for families of devices and read the
temperatures from DS1820 devices.  What platform are you wanting to
run it on, I could send you a prerelease versiob to try.

Dave Moor

Christian Aastorp <christian.aastorp.killspam at siemens.no> wrote in message news:<a7iefuk650noamrqcmpvroqj60mtgbpf9v at 4ax.com>...
> I'm considering doing some temperature control and logging, both
> enviromental and inside computers. Did a google search, and found
> references to the range of temperature sensors from Dallas
> Semiconductors. 
> 
> I also found schematics for interfacing with the serial port, and
> timing diagrams for implementing 1-Wire Lan using direct control of
> the port.
> 
> As I'm using  Python for my programming, these days, I would like some
> pointers to modules or techniques to use Python to read the
> temperature sensors.
> 
> Christian Aastorp


From tmagna at online.no  Sun Jun  2 17:59:20 2002
From: tmagna at online.no (Tobias Brandvik)
Date: Sun, 02 Jun 2002 21:59:20 GMT
Subject: Python Turing machine
References: <d9akfu0c7ro0sebio1vd9c6ik2i0fg59le@4ax.com> <Xns9221CF0CC998Dcliechtigmxnet@62.2.16.82>
Message-ID: <b45lfukgk0uaigv02pucaf2790rknf8453@4ax.com>

On 2 Jun 2002 20:19:05 +0200, Chris Liechti <cliechti at gmx.net> wrote:

>hint: post some URLs of those implementations to the list. that way we 
>don't need to look up what it is and people here have fun trying to solve 
>things pythonic...

Here is a page describing the implementation of a UMT in Scheme
(together with the code):

http://web.mit.edu/manoli/turing/www/turing.html

--Tobias Brandvik


From carel.fellinger at iae.nl  Fri Jun 28 23:15:56 2002
From: carel.fellinger at iae.nl (Carel Fellinger)
Date: Sat, 29 Jun 2002 05:15:56 +0200
Subject: I'm an idiot
In-Reply-To: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16>
References: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16>
Message-ID: <20020629031556.GB2953@schaduw.felnet>

On Sat, Jun 29, 2002 at 01:00:22AM +0000, David wrote:
> OK, I am the first to admit it.  I am an idiot.  I have RTFM on this over 

Hm, that makes you the first idiot to admit he's an idiot, not likely so.

...
> And just to explain my stupidity, my reference langauge is BASIC.  I 

Hm, nothing wrong with having BASIC as your reference language, it just
indicates that you haven't had the time to familiarize yourself with
Python.  Don't worry, in a few days all will be swell.


> f=open('c:\\temp\\temp.txt', 'r')

Windows allows / in dir path, less error prone, and "r" is the default
mode and hence often left out.

> g=open('c:\\temp\\temp1.txt', 'w')
> while 1:
>     try:
>         s=f.readline

here you merely create an extra name for the file readline function.
The ( and ) are really needed to instruct the interpreter to perform
the function instead of creating a new name for it.

>         g.write(s.split())

the split method removes all spaces/newlines/tabs from a string, braeking
the string in those places and returning a list of al the leftovers.

  "a      bc d".split()  ==  ["a", "bc", "d"]

>     except IOError:
>         break

In python the file read(line) method(s) don't bark at the end of file,
they simply return an empty string. So you won't get an EndOfFile
Exception.  The easiest thing is to test here for the empty string.

> 
> g.close
> f.close

again, without the ()'s nothing happens here.

Fixing the above and replacing the non-working try-except you get:

   f = open('c:/temp/temp.txt')
   g = open('c:/temp/temp1.txt', 'w')
   while 1:
       line = f.readline()
       line = line.strip()
       if not line:   # line is empty
            break
       g.write(line)
   g.close()
   f.close()


But we can do better then this by replacing the while loop with a for loop:

   f = open('c:/temp/temp.txt')
   g = open('c:/temp/temp1.txt', 'w')
   for line in f.readlines():
       g.write(line.strip())
   g.close()
   f.close()


And in Python2.* the above for line can be further simplified as:

   f = open('c:/temp/temp.txt')
   g = open('c:/temp/temp1.txt', 'w')
   for line in f:
       g.write(line.strip())
   g.close()
   f.close()


Many people would rely on Python's automatic file closing behaviour
and write it like this:

   g = open('c:/temp/temp1.txt', 'w')
   for line in open('c:/temp/temp.txt'):
       g.write(line.strip())


-- 
groetjes, carel





From geoff at gerrietts.net  Fri Jun  7 20:14:32 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Fri, 7 Jun 2002 17:14:32 -0700
Subject: Medium to Large Scale Python Deployments
In-Reply-To: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com>
References: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com>
Message-ID: <20020608001432.GA18346@isis.gerrietts.net>

Quoting Sam Penrose (spenrose at intersight.com):
> [Domenic R. Merenda raises the topic of Python used in large contexts]
> 
> I'm also curious about large amounts of Python, period. Who knows of 
> ~50K Python lines-of-code (LOC) systems? 100K LOC? More?

I'm working on a large web project whose name I hafta jump thru
permission hoops to reveal (though a little perfunctory Googling
old newsgroup posts would probably give you the answers if you care).

I don't have a nice summary version printing out of the version of
pycount I found on starship, so mine may not be as good as others, but
I show:

lines:    173810
code:     127443
doc:        1387
comment:   17873
blank:     27107

I would put the number of python enthusiasts working on this code
somewhere around 50%; others appreciated its merits but were largely
indifferent to concerns of language, and still others were inductees
in the cult of Java.

Some of the code indexed above is autogenerated, for use with
ILU-based services. Some is boilerplate, for similar reasons. But most
has been seen by human eyes and shaped by human minds.

--G.

-- 
Geoff Gerrietts             "Punctuality is the virtue of the bored." 
<geoff at gerrietts net>                               --Evelyn Waugh




From ianb at colorstudy.com  Mon Jun  3 02:05:13 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 03 Jun 2002 01:05:13 -0500
Subject: What is a classmethod for?
In-Reply-To: <adebub0mdl@enews1.newsguy.com>
References: <addgv30f4d@enews4.newsguy.com> <addlm6$g44$1@panix1.panix.com>
	 <adebub0mdl@enews1.newsguy.com>
Message-ID: <1023084314.10680.12551.camel@localhost>

On Sun, 2002-06-02 at 19:06, Dennis Peterson wrote:
> > The point is that you can call class methods *without* creating an
> > instance:
> 
> >C.foo()
> 
> But you can do that with a static method:
> 
> class C:
>     def foo(x):
>         print x
>     foo = staticmethod(foo)
> 
> C.foo(1)

Maybe a better example would be per-class caching.  Like:

class C:
    cache = {} # Contains values of (expire_time, cache_value)
    def cleanCache(klass):
        for key, (expire_time, value) in klass.cache.items():
            if expire_time < time.time(): del[klass.cache[key]]

In general it only makes sense when you use class variables -- just like
if you don't use any instance variables it might make more sense to use
a plain function instead of a method.

  Ian






From loewis at informatik.hu-berlin.de  Sun Jun  2 05:35:24 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 02 Jun 2002 11:35:24 +0200
Subject: Embedding and threads ?
In-Reply-To: <1022839266.823.66.camel@netgroup>
References: <mailman.1022667286.17372.python-list@python.org>
	<j44rgrnkfm.fsf@informatik.hu-berlin.de>
	<mailman.1022740196.32586.python-list@python.org>
	<j4d6vdwvyk.fsf@informatik.hu-berlin.de>
	<1022771667.689.136.camel@netgroup>
	<j4znyhzguh.fsf@informatik.hu-berlin.de>
	<1022830475.893.17.camel@netgroup>
	<j4znygd7iv.fsf@informatik.hu-berlin.de>
	<1022839266.823.66.camel@netgroup>
Message-ID: <j4d6vac9b7.fsf@informatik.hu-berlin.de>

Bo Lorentsen <bl at netgroup.dk> writes:

> > I lost track. How do you set the builtins, and why do you have to?
> Before calling "PyEval_EvalCode", I have to make this call, or I was not
> even able to use the build in functions like "range" !
> 
> PyDict_SetItemString( pDict, "__builtins__", PyEval_GetBuiltins());
> 	
> pDict is both the local and the global dict. for the PyEval_EvalCode
> function.

I see. This is necessary indeed. The builtins are normally inherited
from some context when doing "exec" or "eval" at the Python level; on
the C API, no such inheritance happens automatically.

> One day I may understand this, but until then I hope you bear with me
> :-) So, besides the ThreatStates, what else does GIL protect ? 

Nearly everything in Python
- changing and accessing dictionaries
- changing and accessing lists
- writing to files
- allocating memory (atleast with pymalloc)

etc. None of these operations is thread-safe per se; it is only the
GIL that prevents breakage.

> Arhh, so Python can execute in a MT environment, but only one
> interpreter at a time, the rest is pseudo MT ?

I don't think I understand this question. As I said before, restrict
yourself to "one interpreter" (in the sense of PyInterpreterState*).
This is the normal case even in the presence of threads.

I don't know what pseudo MT is (in what sense is it pseudo?). It
certainly has multiple concurrent threads.

> Now, this makes more sense ! Do you know if there is any plans to make
> Python more flexible for (real) MT processing in the future ? When
> looking into the code, it looks like someone have tried to modulize it
> so that it could be real MT in the future, at least as far as the core
> language goes.

For most people, it is "real" MT today, since they have only one
processor in their computers, anyway.

To make it "more real", you would have to introduce what is known as
"free threading". People have tried this (ask Google) and found that
it slows down the interpreter significantly.

Regards,
Martin




From akuchlin at ute.mems-exchange.org  Mon Jun 10 14:22:06 2002
From: akuchlin at ute.mems-exchange.org (A.M. Kuchling)
Date: 10 Jun 2002 18:22:06 GMT
Subject: M2Crypto: select() behaves weird on SSL sockets
References: <3D04ECED.2030202@NOSPAMREMOVETHISxs4all.nl>
Message-ID: <slrnag9rie.enn.akuchlin@ute.mems-exchange.org>

In article <3D04ECED.2030202 at NOSPAMREMOVETHISxs4all.nl>,
	Irmen de Jong wrote:
> Basically, it seems that a SSL socket is no longer in ready
> state when entering select() again, after reading *a part*
> of the available data on the socket.

Perhaps there really is no more data on the socket because OpenSSL has
read it all, but the decrypted data is now sitting in a buffer
somewhere, probably inside an OpenSSL data structure.  

--amk





From aahz at pythoncraft.com  Tue Jun 25 17:49:38 2002
From: aahz at pythoncraft.com (Aahz)
Date: 25 Jun 2002 17:49:38 -0400
Subject: Suggestions for good programming practices?
References: <mailman.1024951532.5616.python-list@python.org> <af8n9i$3at$1@panix1.panix.com> <afafee$fc$0@216.39.172.122>
Message-ID: <afaohi$682$1@panix1.panix.com>

In article <afafee$fc$0 at 216.39.172.122>, Bengt Richter <bokr at oz.net> wrote:
>
>Do you recommend using def foo(*args) and testing len(args) instead?

Only if you really want a list of unnamed parameters.  Personally, I
usually prefer to pass a sequence directly in such cases, to distinguish
between None (i.e. no list) and the empty list.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From oren-py-l at hishome.net  Sun Jun 30 16:38:56 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Sun, 30 Jun 2002 16:38:56 -0400
Subject: yield equivalent in C/JavaScript?
In-Reply-To: <AS+BzLARS2H9EwUy@jessikat.fsnet.co.uk>
References: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk> <20020630180139.GA38821@hishome.net> <AS+BzLARS2H9EwUy@jessikat.fsnet.co.uk>
Message-ID: <20020630203856.GA59010@hishome.net>

On Sun, Jun 30, 2002 at 09:05:37PM +0100, Robin Becker wrote:
> In message <20020630180139.GA38821 at hishome.net>, Oren Tirosh <oren-py-
> l at hishome.net> writes
> 
> very smart! That Duff did it Fortran as I recall.

I don't think there is switch with case fallthrough in fortran.  Duff's 
device is most definitely a unique abuse of the C language.

http://www.lysator.liu.se/c/duffs-device.html

	Oren




From christophe.delord at free.fr  Wed Jun 12 18:27:37 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Thu, 13 Jun 2002 00:27:37 +0200
Subject: map -> class instance
References: <m2r8jcp3pv.fsf@video.paradise.lost>
Message-ID: <20020613002737.09044c6c.christophe.delord@free.fr>

self.x=y is a call to self.__setattr__, so if you write self.x=y in __setattr__ it is an infinite recursion.
(see http://www.python.org/doc/current/ref/attribute-access.html#l2h-112)

in __setattr__, self.m[key]=val may be written self.__dict__['m'][key]=val
in __init__, self.m=m may be written self.__dict__['m']=m


$ cat map2class.py
class Map2Class:
    def __init__(self,m):
        self.__dict__['m'] = m

    def __getattr__(self, key):
        return self.m[key]

    def __setattr__(self, key, val):
        self.__dict__['m'][key] = val

    def __delattr__(self, key):
        if self.m.has_key(key):
            del self.m[key]

m = {'first' : 'John',
     'last' : 'Hunter',
     'age' : 34}

c = Map2Class(m)
print c.first, c.last, c.age
$ python map2class.py
John Hunter 34
$


Christophe.


On Wed, 12 Jun 2002 16:49:32 -0500
John Hunter <jdhunter at nitace.bsd.uchicago.edu> wrote:

> 
> I want to convert dictionary instances to instances of a class that
> has the keys of the map as attributes.  I naively tried this:
> 
> class Map2Class:
>     def __init__(self,m):
>         self.m = m
> 
>     def __getattr__(self, key):
>         return self.m[key]
> 
>     def __setattr__(self, key, val):
>         self.m[key] = val
> 
>     def __delattr__(self, key):
>         if self.m.has_key(key):
>             del self.m[key]
> 
> m = {'first' : 'John',
>      'last' : 'Hunter',
>      'age' : 34}
> 
> c = Map2Class(m)
> print c.first, c.last, c.age
> 
> But got an infinite recursion:
> 
> ~/python/test $ python map_to_class.py 
> Traceback (most recent call last):
>   File "map_to_class.py", line 20, in ?
>     c = Map2Class(m)
>   File "map_to_class.py", line 3, in __init__
>     self.m = m
>   File "map_to_class.py", line 8, in __setattr__
>     self.m[key] = val
>   File "map_to_class.py", line 5, in __getattr__
>     return self.m[key]
>   File "map_to_class.py", line 5, in __getattr__
>     return self.m[key]
>         [snip ... lots more line fivers ]
> 
> Is there a better/right way to do what I want?
> 
> Thanks,
> John Hunter


-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From gmcm at hypernet.com  Fri Jun  7 08:53:32 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 07 Jun 2002 12:53:32 GMT
Subject: problem with isinstance and relative/absolute import
References: <87adq8jsq1.fsf@tux.ntw23.fr>
Message-ID: <Xns92265AA1E2FABgmcmhypernetcom@199.171.54.216>

Sylvain Thenault wrote:

> does someone can explain to me the following test case, or should it be
> considered as a bug ?
> 
> [syt at tux syt]$ ls tests/
> __init__.py  module2.py  module.py

[snip]

> [syt at tux syt]$ cd tests

Bad move. This makes the modules in the tests 
directory accessable in 2 ways - as a direct
import (sys.path[0]), or as a package. The
way you do your imports, you end up getting
module twice, once as module, once as tests.module.

Just look at the repr's of a, b and A to see.

And stay out of package directories.

-- Gordon
http://www.mcmillan-inc.com/


From whisper at oz.net  Sun Jun  9 18:16:22 2002
From: whisper at oz.net (David LeBlanc)
Date: Sun, 9 Jun 2002 15:16:22 -0700
Subject: Socket error problem
In-Reply-To: <m3elfg9ncp.fsf@mira.informatik.hu-berlin.de>
Message-ID: <GCEDKONBLEFPPADDJCOECELJDHAA.whisper@oz.net>

Ok, that solved the problem. It was caused by the app using it's own version
of sockets; overriding the default implementation to create something called
a "TimeoutSocket", which don't work on Windows for some reason (something to
worry about later if necessary). Fortunately, all that was required to go
back to using standard sockets was to remove the override. After chasing
down a few additional bugs (it seems to ignore it's own config options or
the doc isn't good enough to describe how to set them properly), it now
actually slurps mail!! :)

The author of the package said it would never work on Windows - hah! :)
(Actually, I'm not even going to bother with the QT GUI interface, although
I'm sure that's doable as well.)

Thanks for the help!

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: martin at v.loewis.de [mailto:martin at v.loewis.de]
> Sent: Sunday, June 09, 2002 14:03
> To: David LeBlanc
> Cc: python-list at python.org
> Subject: Re: Socket error problem
>
>
> "David LeBlanc" <whisper at oz.net> writes:
>
> > I've put print statements around all this before I first tryed
> the ml, but
> > here they are again:
>
> This is still not the right place.
>
> >         print 'poplib._init__: self.host = %s ' % self.host,
> 'self.port =
> > %s' % self.port
>
> Notice that host and port are not passed directly to socket. Instead,
> getaddrinfo is used
>
> >         for res in socket.getaddrinfo(self.host, self.port, 0,
> > socket.SOCK_STREAM):
> >             af, socktype, proto, canonname, sa = res
> >             try:
>
> Please add a line printing res inside of this try statement. Also add
> a print statement between the socket call and the connect call, so
> that we know which call it is that fails.
>
> Regards,
> Martin





From Blobbybirdman at hotmail.com  Mon Jun 17 06:03:33 2002
From: Blobbybirdman at hotmail.com (Mark Robinson)
Date: Mon, 17 Jun 2002 11:03:33 +0100
Subject: unhashable type when using pickle.load()
Message-ID: <aekdei$n7e$1@helios.herts.ac.uk>

I wonder if anyone can shed some light on a troublesome intermittent
problem I am having.

Firstly the necessary:
I am running on windows XP, using jython 2.1 and java 1.3.1_03.

A pickled file is created as follows:
    pickle.dump((self.seq.getMotifs(), self.s[self.seqIdx]),
                    newFile)

self.seq.getMotifs() simply returns a dictionary containing various complex
python objects (which will have been created and used sucessfully)

self.s[self.seqIdx] is a python sequence object.


Intermittently, when I try and reload from this pickle file I get the
following exception:


Exception occurred during event dispatching:
Traceback (innermost last):
  File "pars.py", line 129, in OpenFile
  File "D:\program files\jython-2.1\Lib\pickle.py", line 947, in load
  File "D:\program files\jython-2.1\Lib\pickle.py", line 567, in load
  File "D:\program files\jython-2.1\Lib\pickle.py", line 885, in
load_setitem
TypeError: unhashable type

I have read that this error typically can occur when a pyDictionary or
pyList is inserted into a hashtable. Is this a problem that anyone else is
familiar with, is it a limitation of the jython implementation of pickle?

Any help will be appreciated

Blobby






From deltapigz at telocity.com  Thu Jun 13 23:45:34 2002
From: deltapigz at telocity.com (Adonis)
Date: Thu, 13 Jun 2002 23:45:34 -0400
Subject: More Tkinter Help please...
References: <B92EC762.70EC%sarmstrong13@mac.com>
Message-ID: <3d096af6_1@nopics.sjc>

i ran the code, and i had no errors on loading it, you might want to try
placing all your functions above the __init__ call, i.e.:

class PyShell:
    def expyth
        ...
    def ...
    def __init__
        ...

hope thie helps

Adonis

"SA" <sarmstrong13 at mac.com> wrote in message
news:B92EC762.70EC%sarmstrong13 at mac.com...
> Hi Everyone-
>
>     I have another Tkinter question for you. I'm still trying to translate
> "that previously mentioned" Tcl/Tk program into Python/Tkinter. The
> following is all of the code I've managed so far:
>
> from Tkinter import *
> import sys
> import os
>
> class PyShell:
>     def __init__(self, top):
>
>         f = Frame(top)
>         f.pack()
>
>         self.t1 = Text(f, height="12", width="84", font="Courier 12")
>         self.t1.pack(side=TOP, pady=2)
>
>         self.t2 = Text(f, height="12", width="84", bg="lightblue",
> font="Courier 12")
>         self.t2.pack(side=TOP, pady=2)
>
>         self.b1 = Button(f, text="Execute", command=self.expyth)
>         self.b1.pack(side=LEFT)
>
>         self.b2 = Button(f, text="Clear Input", command=self.clearin)
>         self.b2.pack(side=LEFT)
>
>         self.b3 = Button(f, text="Clear Output", command=self.clearout)
>         self.b3.pack(side=LEFT)
>
>         self.b4 = Button(f, text="Save Input", command=self.savin)
>         self.b4.pack(side=LEFT)
>
>     def clearin(self):
>         self.t1.delete(0,END)
>
>     def clearout(self):
>         self.t2.delete(0,END)
>
>     def expyth(self):
>         output = os.popen("python -c").t1()
>         self.t2.delete(0,END)
>         sys.stdout.write.t2(output)
>
>     def savin(self):
>         pass
>
> root = Tk()
>
> app = PyShell(root)
>
> root.mainloop()
>
> When I run this I get the following error:
>
> Traceback (most recent call last):
>   File "PyShell.py", line 45, in ?
>     app = PyShell(root)
>   File "PyShell.py", line 17, in __init__
>     self.b1 = Button(f, text="Execute", command=self.expyth)
> AttributeError: PyShell instance has no attribute 'expyth'
>
>
> Any ideas on what I'm doing wrong here?
>
> Thanks.
> SA
>




From fgeiger at datec.at  Thu Jun  6 14:50:11 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Thu, 6 Jun 2002 20:50:11 +0200
Subject: Emulating classmethod in Python 2.1
References: <915a998f.0206060833.12ad76a8@posting.google.com>
Message-ID: <adoatv$2bk$1@newsreaderg1.core.theplanet.net>

There's a receipe about this in the ASPN, by Alex Martelli. Concise and
effective. Strongly recommended.

Cheers
Franz

"Hamish Lawson" <hamish_lawson at yahoo.co.uk> schrieb im Newsbeitrag
news:915a998f.0206060833.12ad76a8 at posting.google.com...
> Is it possible to define a 'classmethod' class or function for Python
> 2.1 that would effectively provide the functionality of Python 2.2's
> built-in classmethod function and have the same usage pattern? I'd
> like to be able run code like that below against both 2.2 (using its
> builtin classmethod function) and 2.1 (using some added-in wrapper).
>
>     class SomeClass:
>         def hello(cls, name):
>             print "Hi there", cls, name
>         hello = classmethod(hello)
>
>     SomeClass.hello("Peter")
>
> I've come across Thomas Heller's recipe in the online Python Cookbook,
> but this would require SomeClass to be modified, and so fails my
> criteria.
>
> I've also come across Alex Martelli's recipe for emulating
> staticmethod, but of course that doesn't deal with passing a class to
> the wrapped method.
>
>
> Hamish Lawson




From gerhard.haering at gmx.de  Wed Jun 19 23:27:10 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Thu, 20 Jun 2002 05:27:10 +0200
Subject: working with secure shell
In-Reply-To: <ILbQ8.1749$CZ4.555208@news1.news.adelphia.net>
References: <ILbQ8.1749$CZ4.555208@news1.news.adelphia.net>
Message-ID: <20020620032709.GA13875@lilith.my-fqdn.de>

* Jon J. Morin <rotundo52 at hotmail.com> [2002-06-20 03:13 +0000]:
> One question: Is there a python library or module for working with the 
> secure shell (ssh)?

Yes.


http://pyssh.sourceforge.net/

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From gyromagnetic at excite.com  Tue Jun 11 08:29:28 2002
From: gyromagnetic at excite.com (gyromagnetic)
Date: 11 Jun 2002 05:29:28 -0700
Subject: using m2crypto to encrypt credit card numbers
References: <mailman.1023748154.2706.python-list@python.org>
Message-ID: <4620daca.0206110429.1417b389@posting.google.com>

Hi,
Managing keys is a fundamental problem with encryption. I would highly
recommend that you not store the credit card numbers at all. Instead,
store a (SHA, MD5) hash of the number, and then validate against the
hash.

-g


"Mark McEahern" <marklists at mceahern.com> wrote in message news:<mailman.1023748154.2706.python-list at python.org>...
> I'm considering using M2Crypto (http://www.post1.com/home/ngps/m2/) to
> encrypt credit card numbers.  My part of the problem starts when the credit
> card arrives at the web server.  I plan to encrypt it with a public key and
> send it to a database that the web server only has write access to.
>


From Jeff_Donahoo at Baylor.edu  Wed Jun 12 09:03:35 2002
From: Jeff_Donahoo at Baylor.edu (Jeff Donahoo)
Date: Wed, 12 Jun 2002 08:03:35 -0500
Subject: Wanted:  Potential Python Authors
Message-ID: <ae7gke$mhb$1@ccis12.baylor.edu>

I'm looking for somebody to author a short book on Python in my programming
technology series.  Who do you think would make a great author that hasn't
already written a book on Python?  Candidates include:

- Competent and frequent posters to this newsgroup
- Author of articles in the various trade magazines
- Expert programmer




From opengeometry at NOSPAM.yahoo.ca  Thu Jun 27 14:30:15 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 27 Jun 2002 18:30:15 GMT
Subject: How to get rid the new line
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org>
Message-ID: <affljn$e94us$2@ID-99293.news.dfncis.de>

SiverFish <occeanlinux at linuxmail.org> wrote:
> Anyone tell me how to skip the empty line when you read through the file
> and copy the information in to the list(no empty line )

A bit off topic... but this question just cries out for 
    sed 's/^$/d'

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin


From gherron at islandtraining.com  Thu Jun  6 16:27:46 2002
From: gherron at islandtraining.com (Gary Herron)
Date: Thu, 6 Jun 2002 13:27:46 -0700
Subject: numeric limits?
In-Reply-To: <adofuv$6v$1@bob.news.rcn.net>
References: <adofuv$6v$1@bob.news.rcn.net>
Message-ID: <200206061327.46814.gherron@islandtraining.com>

On Thursday 06 June 2002 01:15 pm, David Abrahams wrote:
> Is there some straightforward way to find out the range of a Python Int (in
> code)?

Use sys.maxint :

>>> import sys
>>> sys.maxint
2147483647

Gary Herron





From Chris.Barker at noaa.gov  Thu Jun 20 15:56:46 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Thu, 20 Jun 2002 12:56:46 -0700
Subject: How to represent the infinite ?
References: <3d11ee9b$1@news.swissonline.ch> <20020620193230.53d58f8f.christophe.delord@free.fr> <3D121A29.A733609F@noaa.gov>
Message-ID: <3D12334C.59EB9911@noaa.gov>


Chris Barker wrote:
> whether
> this works is a function of the machine, compiler and library used to
> compile Python. I'm not sure how common it is, but if you want your code
> portable, it may not be reliabel on all machines, which is too bad,
> because IEEE 754 is pretty darn good standard.

After posting that, I went and read some more:

http://cch.loria.fr/documentation/IEEE754/wkahan/754story.html

A quote from that article:

"Programming languages new ( Java ) and old ( Fortran ), and their
compilers, still lack competent support for features of IEEE 754 so
painstakingly provided by practically all hardware nowadays. S.A.N.E.,
the Standard Apple Numerical Environment on old MC680x0-based Macs is
the main exception. Programmers seem unaware that IEEE 754 is a standard
for their programming environment, not just for hardware." 

Too bad. Python can't do anything that isn't supported by the compilers
it is built with.

Full disclosure: I took a Numerical Analysis course with Kahan when I
was at Berkeley. I understood about 1/10 of what he was trying to teach
us, but I did learn this: Floating point is hard!


-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From missive at frontiernet.net  Mon Jun 10 18:31:55 2002
From: missive at frontiernet.net (Lee Harr)
Date: Mon, 10 Jun 2002 22:31:55 -0000
Subject: ZODB and non-python languages
References: <b02e30f6.0206101009.156c6dda@posting.google.com>
Message-ID: <ugaa6ra0uvoudc@corp.supernews.com>

> Hi. I am working with a group of people who have an extensive Zope
> system with lots of data in their ZODB database. I am obliged to
> implement my parts in JSP. Can anybody point me in the right direction
> for using ZODB from Java?
> 


How about xml-rpc?



From johnsogg at cs.colorado.edu  Mon Jun 10 14:09:57 2002
From: johnsogg at cs.colorado.edu (Gabe Johnson)
Date: 10 Jun 2002 11:09:57 -0700
Subject: ZODB and non-python languages
Message-ID: <b02e30f6.0206101009.156c6dda@posting.google.com>

Hi. I am working with a group of people who have an extensive Zope
system with lots of data in their ZODB database. I am obliged to
implement my parts in JSP. Can anybody point me in the right direction
for using ZODB from Java?

Thanks

-gabe


From whisper at oz.net  Wed Jun  5 05:39:52 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 5 Jun 2002 02:39:52 -0700
Subject: writing a text in hyphenless justification
In-Reply-To: <32d0f89.0206042221.1910798c@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEMELIDGAA.whisper@oz.net>

Justifying text, with or without hyphens, is SO not a trivial job to do
right.

I think the tkinter edit control has a justification setting (I'm almost
positive) which does not do hyphenation and so pouring your text into a Tk
edit control and then reading it out line by line would be the simpliest
solution imo.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Backflip
> Sent: Tuesday, June 04, 2002 23:22
> To: python-list at python.org
> Subject: writing a text in hyphenless justification
>
>
> Hi everybody,
> i?ve got a little problem again......
> i want to take a raw_input and makes it to a hyphenless
> justification output.....
> how is this possible?????
> The most beautiful solution would be the one, that i can define
> the width....
> has got smb a idee??? it would be very nice, if smb can help me
> with my problem.....
> THankfull
> Backflip
> --
> http://mail.python.org/mailman/listinfo/python-list





From peter at engcorp.com  Tue Jun 11 00:12:55 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 11 Jun 2002 00:12:55 -0400
Subject: Correct code/unit testing (was Re: Efficient python programming...)
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com> <834rgd5isv.fsf@panacea.canonical.org> <3D02FB3B.CF75E925@engcorp.com> <83u1oaq0jt.fsf@panacea.canonical.org>
Message-ID: <3D0578C7.ACDBF41A@engcorp.com>

Kragen Sitaker wrote:
> 
> Peter Hansen <peter at engcorp.com> writes:
> > Have you never spent a large amount of time carefully
> > designing and coding a wonder little program, and checked it over
> > very carefully, even with a peer, and later found a bug in it?
> 
> Of course!  People are fallible, it's much easier to do a bad code
> review than a good one.

I have to confess I have never seen a successful code review, nor
an environment in which they were used regularly and which had a
high level of code quality.  Clearly as a result I'm fairly biased
against them -- I should keep a more open mind on the issue perhaps.

I'm swayed also by the thought that programmers generally rebel
against them, and that therefore they are likely not the best
approach in most shops.  Learning XP has taught me that a process
which works in a way programmers find natural is much more likely
to succeed than one which goes against the grain.

> I think now we have each committed the same arrogant gaffe in this
> discussion: "It is only possible to disagree with me if you do not
> know what I am talking about."  I was wrong, and so are you.

Point taken; my apologies.  (I'm about to go on vacation to NYC for
the rest of the week, so if I suddenly go silent, it's not personal. :)

> I think inspection can find bugs that are very difficult to find by
> testing, and vice versa.  That is the reason for my second statement;
> a balance of careful reasoning and testing can produce better code
> than the same amount of time spent on either alone.

For code not written test first, I would strongly concur.  For code
written test first, I believe the resulting modularity and cohesion
in the code mostly negates the value of inspection.  Again, other
practices from XP are probably polluting my mind against such things --
in this case pair programming.  Assuming the formal 'inspection'
advocated by some heavy-weight processes, I think it's of little
use on code pair-programmed.  (We could also be more general and
just say that pair programming is on par with inspection and that
either is effective in an appropriate context.)

Cheers,
-Peter


From emile at fenx.com  Tue Jun 11 09:01:30 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 11 Jun 2002 13:01:30 GMT
Subject: Defining a method final
References: <6glN8.35376$86.884025@twister1.libero.it> <mailman.1023796095.1427.python-list@python.org> <phmN8.36164$K4.893500@twister2.libero.it>
Message-ID: <KwmN8.153269$352.7754@sccrnsc02>

"Fabio Corneti" <fcorneti at libero.it> wrote in message
news:phmN8.36164$K4.893500 at twister2.libero.it...
> I would like to define a method for a class which has
> not to be overriden. Of course, since the developer of
> this project are two, this could be established by
> convention, but if the project will grow consistently
> another developer could override the method by mistake,
> causing unpleasant side effects.
>

Accepted practice is to use a single underscore for objects to be
considered private.  This does not prevent their use, but strongly
signals that re-use ought not to be done.

>>> class Test:
...     def _private(self): pass
...

HTH,

--

Emile van Sebille
emile at fenx.com

---------



From shalehperry at attbi.com  Thu Jun  6 20:29:36 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 06 Jun 2002 17:29:36 -0700 (PDT)
Subject: Efficient python programming...
In-Reply-To: <adoa9u$vqed$1@ID-146704.news.dfncis.de>
Message-ID: <XFMail.20020606172936.shalehperry@attbi.com>

On 06-Jun-2002 Shagshag13 wrote:
> 
>  I'm looking for any state of art, optimizing techniques, tips, that a
> python beginner couldn't know by itself, but should learn and
> would be aware to write "efficient and good looking"python's code...
> 
>  (not the kind of thing which made you an obfuscated python nerd,
> but that "map" should be preferred to "for in loop" if possible, that
> dictionary are very well optimized, etc.)
> 

the consensus on this list over the last year seems to be that optimized code is
usually the simplest implementation.  If you feel you are going through
hoops getting something working you probably need to take a step back, ponder
and maybe ask a question.

BTW tutor at python.org is a great place for new python coders to hang out, read,
and learn.




From tom at huno.net  Fri Jun 28 15:52:13 2002
From: tom at huno.net (thomas)
Date: Fri, 28 Jun 2002 21:52:13 +0200
Subject: generate all combinations of a list (with variable output length)
Message-ID: <3D1CBE6D.30002@huno.net>

hi,

i'm looking for a way to generate all possible combinations of all items 
in a list. list = ["A","D","$","5","R"]

but i also want it to use one item several times or not at all and in no 
particular order. the length of the string is 1-10. so a possible output 
would be:

DD$
A5RD$D555
A
5RR$$A
DRRR55$5$A
...


i find it particualy difficult to tell python to use one item several 
times or not at all, and the variable length thing. i don't want to 
reinvent the wheel so someone probably has already a soultion for that 
:) and if not someone can probably push me in some direction.

thx,
thomas



From garrett at bgb.cc  Mon Jun 10 21:04:50 2002
From: garrett at bgb.cc (Don Garrett)
Date: Tue, 11 Jun 2002 01:04:50 GMT
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
References: <Xns9229EE722467Fgustafl@195.100.94.182> <mailman.1023746539.13373.python-list@python.org> <Xns922AD3C74280gustafl@195.100.94.182>
Message-ID: <3D054C64.50406@bgb.cc>

Gustaf Liljegren wrote:
> Cliff Wells <logiplexsoftware at earthlink.net> wrote: 
> 
> Thanks both of you. I tried both solutions. Since some intervals are very 
> high (1000000+), it takes too long to add them to a list. Cliff's solution 
> was really fast, but it took a while to understand. :)
> 
> Gustaf

   In all truth, I didn't pay attention to the subject line, and slightly 
misunderstood the question. I was thus considering it to be a mostly newbie 
question..... yes, his answer is much better for what you are doing.

-- 
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc



From russblau at hotmail.com  Fri Jun 28 15:00:43 2002
From: russblau at hotmail.com (Russell Blau)
Date: Fri, 28 Jun 2002 15:00:43 -0400
Subject: testing type of an object
References: <pan.2002.06.28.14.49.33.858327.1787@presidency.com>
Message-ID: <afibp0$ekn25$1@ID-76829.news.dfncis.de>

"Rajarshi Guha" <rajarshi at presidency.com> wrote in message
news:pan.2002.06.28.14.49.33.858327.1787 at presidency.com...
> Hi,
>   I'm writing some code that depends on the type of an argument
> passed to it.
>
> Is there a standard trick to do this? My versions don't seem to work:
>
> if type(s) == 'str':
>   do something
>
> or
>
> if string.find(type(s), 'str') != -1:
>   do something
>
> dont seem to work

They don't work because the type() function returns a type _object_, not
a string containing the type name.

You can use

if type(s) == type('abc'):
    do something

or, even better, import the types module (which you can look up in the
docs).

--
I don't actually have a hotmail account; but I do have one on excite.com
if you really want to get in touch with me.





From fredrik at pythonware.com  Sat Jun 29 08:32:02 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 29 Jun 2002 12:32:02 GMT
Subject: question od default args
References: <o3vjhuguen6gfm8vnq7rbk5volmi80shl5@4ax.com> <YynS8.13571$p56.4313990@newsb.telia.net> <aff4ap$aqm$2@news.wrc.xerox.com>
Message-ID: <6NhT8.45584$n4.10824585@newsc.telia.net>

Mark Jackson wrote:

> More quickly:
>
>     def myfunc(arg=[]):
>         if arg is myfunc.func_defaults[0]:
>             print "no argument"
>         else:
>             ...

while the func_ attributes are documented, a quick peek in
the standard library should tell you that it's best to treat them
as an implementation detail.

the less such stuff you depend on in your production code,
the more portable it gets...

</F>




From a.schmolck at gmx.net  Fri Jun 21 13:21:03 2002
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 21 Jun 2002 18:21:03 +0100
Subject: What language - platform use to data acquisiton/numerical work?
References: <1c849ea9.0206130130.6cdaa36f@posting.google.com> <1c849ea9.0206210457.36d494db@posting.google.com>
Message-ID: <yfsk7osfszk.fsf@black132.ex.ac.uk>

mdtorre at freemail.it (Matteo) writes:

> Thank you all for the contributions.
> 
> I have understood that Python is good for GUIs, for serial port data
> acquisition and for numercal computation.
> 
> My only problem is that MatLab has some huge numerical libraries that
> Phyton does not have. I have to do experiments with pattern recognitin
> algoritmhs (pca, neural networks etc.) and these are already
> efficently coded in MatLab. I can't afford the effort to code a radial
> basis function network from scratch (I'm able to do it, but it's
> tiring and very time consuming).
> So I'd like to use pyhon for all except numerical computation. I've
> seen Pymat to conetct MatLab to Python, but it seem a little outdated
> and unsupported. Is there any other way to embed MatLab into Python?
> 

I've hacked up a module based on pymat, so that you now can use matlab pretty
much like it were just some python library.

Examples: 

>>> mlab.plot([1,2,3])
[plots a line]

or 

>>> mlab.sort([[3.,2.,1.]])
array([       [ 1.,  2.,  3.]])

You can even do things like

>>> help(mlab.sort)
[prints the docu of sort]

and pickle objects in matlab. Speaking of objects, unlike pymat you can not
only work with arrays and strings but also with other types of matlab objects.

I have been using it for my work for some time, but I haven't yet got around
to release it (as open source). If you are interested I could send you
the code, but you'll need python2.2.1 and it might still have rough edges.

> Thank you in advance,
> 
> Matteo Della Torre

alex


From bokr at oz.net  Thu Jun 27 12:36:23 2002
From: bokr at oz.net (Bengt Richter)
Date: 27 Jun 2002 16:36:23 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com> <3D1A8CD9.C8B8EC28@engcorp.com> <3d1b1bd7.294171125@netnews.attbi.com>
Message-ID: <affeu7$qoa$0@216.39.172.122>

On Thu, 27 Jun 2002 14:09:09 GMT, candiazoo at attbi.com wrote:

>I am a newbie... this is my first Python "project" so I am probably writing
>horribly inefficient code...  I have never used the profiler but I'll try it!
>
>I am not opening/closing the file each time.  I am extracting 700000 rows from a
>mysql database, extracting additional data from our primary, Oracle database per
Are you by chance opening and closing your mysql connection for each row? Easy
to do if you just wrote a quick get-me-one-row wrapper function and forgot to revise it.

>row, then stuffing each piece of data into a class which preformats the data (I
>need to output the data into a fixed format string/record for another
>application which reads them) and returns a single string... which I write out
>to the file.
>
If it's working, probably the best thing is to measure actual timing before optimizing.

Regards,
Bengt Richter


From ark at research.att.com  Fri Jun  7 13:56:47 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 7 Jun 2002 13:56:47 -0400 (EDT)
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <15616.62131.521803.383860@12-248-41-177.client.attbi.com>
	(message from Skip Montanaro on Fri, 7 Jun 2002 12:51:47 -0500)
References: <yu99vg8v6orb.fsf@europa.research.att.com>
        <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
        <200206071649.g57GnmC00708@europa.research.att.com> <15616.62131.521803.383860@12-248-41-177.client.attbi.com>
Message-ID: <200206071756.g57HulT01147@europa.research.att.com>

Andrew> That could be it, but I have been setting LD_RUN_PATH explicitly
Andrew> when building Python, so I don't understand why it would have
Andrew> changed.

Skip> It's been ages since I used a Solaris system, but shouldn't that be
Skip> LD_LIBRARY_PATH?  Perhaps there is no ldconfig command (anymore) on Solaris.

No -- LD_LIBRARY_PATH tells the runtime system which libraries to
use for this particular execution; LD_RUN_PATH tells the linker to
generate .so files that use particular libraries whenever they are run,
regardless of the value of LD_LIBRARY_PATH.

I want to build python in such a way that its users do not have to
set LD_LIBRARY_PATH in order for it to work; setting LD_RUN_PATH
does just that.  Or at least it did until I installed gcc 3.1 :-(







From markus.vonehr at ipm.fhg.de  Mon Jun 17 05:40:26 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Mon, 17 Jun 2002 11:40:26 +0200
Subject: bug in Tkinter?
Message-ID: <3D0DAE8A.E7637F07@ipm.fhg.de>

Hi everybody,

I am marking a region in a photo (photoimage in canvas) with a
rectangle:

self.canvas.create_rectangle(..., outline="white")

Then I bind the functions to mouse events:

self.canvas.bind("<ButtonPress-1>",self.Down)
self.canvas.bind("<B1-Motion>",self.Drag)
self.canvas.bind("<ButtonRelease-1>",self.Release)

and I change the coordinates using:

def Drag(self, ev):
       self.canvas.coords(item, ...new_coords...)

there still remain wrong white pixels over my photo, this seems to
happen only in
the lower left corner. How can I eliminate them???

Thanks for any comments,

Markus


-------------- next part --------------
A non-text attachment was scrubbed...
Name: markus.vonehr.vcf
Type: text/x-vcard
Size: 245 bytes
Desc: Visitenkarte f?r Markus von Ehr
URL: <http://mail.python.org/pipermail/python-list/attachments/20020617/54e4e64e/attachment.vcf>

From fperez528 at yahoo.com  Thu Jun 20 12:29:30 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 20 Jun 2002 10:29:30 -0600
Subject: changing the viewing area with gnuplot in python
References: <3D1200B6.639CEBE5@lanl.gov>
Message-ID: <aesvta$g1k$1@peabody.colorado.edu>

Nathan Given wrote:

> right now gnuplot auto scales everything (I don't have control over the
> viewing area)...
> 
> 
> How do I change it?  (I want to view from -4 to 2 on the x axis, and 0
> to 10 on the y axis....  right now gnuplot automatically chooses -4 as
> xmin, 2 as xmax, 4 as ymin, and 10 as ymax)

Assuming 'g' is your gnuplot instance:

g('set xrange [-4:2]')
g('set yrange [0:10]')

Open a gnuplot session in a separate terminal and use its builtin help. Once 
you figure out which commands you need interactively, you can just pass them 
to the python-controlled gnuplot by calling it as a function with the command 
you want in a string.

cheers,

f.


From dfackrell at DELETETHIS.linuxmail.org  Thu Jun 13 13:07:03 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Thu, 13 Jun 2002 11:07:03 -0600
Subject: why not "'in' in 'in'"?
References: <mailman.1023978378.11084.python-list@python.org> <aeadtt06rq@drn.newsguy.com>
Message-ID: <3d08d139$1_2@hpb10302.boi.hp.com>

When all else fails, lets try throwing in some newbie exploration:

Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help
>>> print type("a")
<type 'str'>
>>> print type("ab")
<type 'str'>
>>> print type([1])
<type 'list'>
>>> print type([1,2])
<type 'list'>
>>> print type(1)
<type 'int'>
>>> if "a" in "ab":
    print "yes"
else:
    print "no"

yes
>>> if [1] in [1, 2]:
    print "yes"
else:
    print "no"

no
>>> if 1 in [1, 2]:
    print "yes"
else:
    print "no"

yes
>>>

I can see the rationale here, as there isn't really a "character" type in
Python.  So what the "in" operator does in the case of a string as the
second operand is check the first operand to see if it can be interpreted as
a element of a string, which would be a character.

It does in fact appear to be inconsistent, but I'll have to leave it up to
people who know more about what they are doing than myself to figure out
what to do about it, if anything.  I'm not sure that the ideas that come to
mind (creating an actual "character" type or modifying the behavior as
proposed since we really have an exception here already) are any good.

--
Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.


"Grant Griffin" <not.this at seebelow.org> wrote in message
news:aeadtt06rq at drn.newsguy.com...
> That's not a bad suggestion, Mark, but if I did that I guess I would be
forever
> wondering if I was testing whether the first one was in the second or the
second
> one was in the first.  Again, it doesn't read nearly as well as:
>
>    if x in subject:
>        ...
>
> which leaves no doubt.
>
> Ironically, that's already perfectly legal--so long as x has no more than
one
> character:
>
> >>> subject = "i like spam, i do."
> >>> x = "s"
> >>> y = "h"
> >>> if x in subject:
> ...    print "x is in subject and x has no more than one character"
> ...
> x is in subject and x has no more than one character
> >>> if y in subject:
> ...    print "same goes for y"
> ... else:
> ...    print "can't say the same for y"
> ...
> can't say the same for y
> >>>
>
> reluctant-to-discard-my-hard-won-near-mastery-of-"find"-ly y'rs,




From peter at engcorp.com  Mon Jun 10 12:52:00 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 10 Jun 2002 12:52:00 -0400
Subject: Is python suitable for embedded programming?
References: <aa104658.0206100707.75889ccf@posting.google.com>
Message-ID: <3D04D930.DFC50E26@engcorp.com>

Dennis Nash wrote:
> 
> I am considering using Python to control embedded hardware, and would
> like to know whether other people have done this, what their experiences
> were, and whether they recommend it.

We've been doing similar things for the last two and half years with
a great deal of success.  We use serial ports, CAN interfaces (connected
via USB), and GPIB interfaces to control various RF equipment and
embedded systems.

> The system in question is in three parts: Industrial hardware talks to a
> control system (a Linux or Windows2000 box). The control system outputs
> the data from the hardware onto a web server. A 3rd computer contains a
> web browser which will interact with the web server, allowing an operator
> to view what's happening with the hardware and change parameters.
> 
> I am thinking of using Python for the middle part, i.e. the control
> system which polls the hardware continually to get its state, and updates
> the web pages accordingly.
> 
> Some specific questions:
> 
> * how good is serial port access in Python?

It was poor when we started, but we hacked together our own wrapper
on win32comm (at the time), then win32file in the more recent Windows
extensions to Python.  There are now at least two, maybe three released
serial port packages for Python, running on Linux and Windows.  (When
we get a chance, we're going to either integrate their best bits into
ours and release it, or vice versa.)  A quick google search would turn
up the various options.

> * the program will be running forever, in a big loop. Is it liable to
> gradually use all the memory up and then fall over?

No, never in my experience.  We have applications running for months.
Python is an excellent choice for long-running applications.

> * output to web pages: how good is Python Server Pages? is this the best
> way to do it?

"Best" is entirely dependent on your own needs of course, not all of
which you've presented. :)  You might, for example, output the data
as simple XML and just dump it into folders somewhere with a web 
server in front serving them up.  No need for any Python (or code)
at that point, if you're willing to write a little XSLT.

You could also use something like Zope and make the data available
through that.  Anything in between these two would work okay too,
depending on what you really care about.

-Peter


From robin at jessikat.fsnet.co.uk  Sun Jun 30 16:05:37 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 30 Jun 2002 21:05:37 +0100
Subject: yield equivalent in C/JavaScript?
In-Reply-To: <20020630180139.GA38821@hishome.net>
References: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk>
 <20020630180139.GA38821@hishome.net>
Message-ID: <AS+BzLARS2H9EwUy@jessikat.fsnet.co.uk>

In message <20020630180139.GA38821 at hishome.net>, Oren Tirosh <oren-py-
l at hishome.net> writes

very smart! That Duff did it Fortran as I recall.

>On Sat, Jun 29, 2002 at 07:12:46PM +0100, Robin Becker wrote:
>> Is there any way to do yielding easily in C? I have a bunch of programs
>> which are normally connected together using unix pipes. I would like to
>> merge them and the problem is then how to get them to communicate nicely
>> using a single process.
>
>It's possible to implement this in C using Duff's Device:
>
>#define BEGIN_GENERATOR static int _GenState=0; \
>  switch(_GenState) { \
>    case 0:
>
>#define END_GENERATOR } \
>  _GenState = 0;
>
>#define YIELD(x) _GenState=__line__; return (x); case __line__:
>
>
>int generator(arguments) {
>
>  static int foo, bar, etc;    /* Make all variables static */
>
>  BEGIN_GENERATOR 
>  for(...) {
>      if(...) {
>        ...
>        YIELD(x)
>        ...
>      } else {
>        ...
>      }
>      ...
>      YIELD(y)
>      ...
>   }
>   END_GENERATOR
>}
>
>If you want this to be reentrant use a struct passed as an argument
>to the function instead of static variables.
>
>       Oren
>
>

-- 
Robin Becker




From peter at engcorp.com  Mon Jun 17 20:33:04 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 20:33:04 -0400
Subject: a question on SOAP
References: <pan.2002.06.13.14.57.08.774939.27301@presidency.com>
Message-ID: <3D0E7FC0.8C35FCF8@engcorp.com>

Rajarshi Guha wrote:
> 
> Hi,
>   I'm, using SOAP.py (0.9.7) and Python 2.2. I have a small python script
> acting as a 'server' and another script calls functions on the server.
> Currently both client and server are on the same machine. However when
> the client calls the server I get the following errors:
> '
> Traceback (most recent call last):
[...]
>   File "/usr/local//lib/python2.2/httplib.py", line 379, in connect
>     raise socket.error, msg
> socket.error: (111, 'Connection refused')
> 
> My firewall is off, nmap shows that the port 8080 is open, so why would
> the server be refusing connections?

Presumably you've already checked the basic things like that
your code is really connecting on port 8080 and not the standard
http port?  Have you also tried connecting manually with telnet?
Something like "telnet localhost 8080" should work, and if you type
"get / HTTP/1.0" and hit enter a couple of times you ought to get
some reply if an HTTP server is listening.

You don't give enough for anyone to do more than suggest that
"you're right, it doesn't look like it's working". :-)

-Peter


From duncan at NOSPAMrcp.co.uk  Wed Jun 12 06:09:38 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Wed, 12 Jun 2002 10:09:38 +0000 (UTC)
Subject: Python Virtual File System: A spike test
References: <mailman.1023872902.2540.python-list@python.org>
Message-ID: <Xns922B70C12302duncanrcpcouk@127.0.0.1>

"Mike C. Fletcher" <mcfletch at rogers.com> wrote in
news:mailman.1023872902.2540.python-list at python.org: 

> An initial implementation (spike test) exploring the viability of the 
> approach.  Implementation provides:
> 
>      equality testing
>      parent/child/ancestor/descendent retrieval/testing
>      filesystem-role testing (file/dir/root)
>      root retrieval and root comparisons
>      "walking" (callback-based recursive iteration)
>      listing (returning Paths which include their parent directory)
>      fragment determination (i.e. c:\\temp\\test -> "c:\\", "temp",
>      "test" recursive directory-tree removal (including files)
>      open( ) method for path specifiers
> 
> The current implementation is a sub-class of str which provides a
> number of new methods (and overrides some built-in methods) to provide
> the above functionality.
> 

My initial thoughts:
What is your use-case for equality testing? Especially given that it is 
virtually impossible to implement this reliably.
For example the normcase implementation in the standard library lowercases 
all pathnames under windows. While this is useful most of the time, not all 
filesystems are case insensitive, even on windows.

Your class should support iterators: probably three flavours, one for 
contained files, one for contained directories and one for everything. The 
list, walk, and remove methods would probably then benefit from reusing the 
iterators.

The FileSystemPath seems a bit windows oriented with drive() and unc() 
methods.

The root method returns a drive such as 'c:'. This means that:
 >>> temp = filepath.path('c:\\temp')
 >>> temp
 FileSystemPath('c:\\temp')
 >>> temp.root()
 FileSystemPath('c:')
 >>> temp.root()+'fred'
 FileSystemPath('c:fred')

I would have expected temp.root()+'fred' to give a reference to 'fred' in 
the root directory, not a relative reference.

There seems to be a problem with the implementation of canonical() when 
working with some UNC pathnames (watch w2):
 >>> from filepath import path
 >>> w = path(r'\\osprey\work')
 >>> w1 = path(r'd:\work')
 >>> w2 = path('\\.\d:\work')
 >>> w==w1
 0
 >>> w1==w2
 0
 >>> w.root()
 FileSystemPath('\\\\osprey\\work')
 >>> w1.root()
 FileSystemPath('d:')
 >>> w2.root()
 FileSystemPath('d:')
 >>> w2.canonical()
 'd:\\d:\\work'
 >>> w.canonical()
 '\\\\osprey\\work'

I suspect however, this is a problem with the underlying os.path functions.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From marc at informatik.uni-bremen.de  Sat Jun 15 09:39:11 2002
From: marc at informatik.uni-bremen.de (Marc Recht)
Date: Sat, 15 Jun 2002 15:39:11 +0200
Subject: IDE recommendations??
References: <b9veea.l85.ln@majestix.gallien.de>
Message-ID: <aefg1v$6fm$04$1@news.t-online.com>

> but this lacks some comfort. I would appreciate any recommendation for a
> suitable IDE

IMHO WingIDE is the best IDE around.


From not.this at seebelow.org  Thu Jun 13 18:34:22 2002
From: not.this at seebelow.org (Grant Griffin)
Date: 13 Jun 2002 15:34:22 -0700
Subject: Type subclassing: bug or feature
References: <aeau48$111$1@panix1.panix.com>
Message-ID: <aeb6le024nk@drn.newsguy.com>

In article <aeau48$111$1 at panix1.panix.com>, aahz at pythoncraft.com says...
>
>Consider the following code:
>
>class MyStr(str):
>    def contains(self, value):
>        return self.find(value) >= 0
>
>s = MyStr("hello, world!")
>s = s.capitalize()
>if s.contains('Hello'):
>    print "Found it!"
>
>It fails with an AttributeError when it calls s.contains(), because
>s.capitalize() returned a str instead of a MyStr.  Anyone want to take a
>whack at defending this as the correct behavior?

Sounds like a bug to me, but perhaps fixing it at some point was part of the
plan from the beginning.

Editorially speaking, although I hadn't thought of this problem specificially, I
guess it might be another reason to use subclassing to augment behavior, not
change it.

I haven't figured out whether this sort of thing could happen with other, more
usefully subclassed types like list and dict, but I guess it could occur
whenever a built-in type returns its own (er, "parent") type.  I bet somebody
can think of an example.

straight-man-ly y'rs,

=g2

_________________________________________________________________________

Grant R. Griffin                                           g2 at dspguru.com
Publisher of dspGuru                               http://www.dspguru.com
Iowegian International Corporation                http://www.iowegian.com



From pinard at iro.umontreal.ca  Tue Jun 25 20:01:49 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 25 Jun 2002 20:01:49 -0400
Subject: RELEASED: Pymacs 0.17
Message-ID: <oqadpioqky.fsf@titan.progiciels-bpi.ca>

Hi!  A new release of Pymacs is available as:

    http://www.iro.umontreal.ca/~pinard/pymacs/Pymacs.tar.gz

Pymacs allows Emacs users to extend Emacs using Python, where they might have
traditionally used Emacs LISP.  Pymacs runs on systems having sub-processes.

The oldish `import pymacs; from pymacs import lisp' interface is gone.
The usual interface is definitely `from Pymacs import lisp'.  The `setup'
script has been reduced even more, it now only takes care of installing the
Emacs Lisp side of Pymacs.  The Distutils `setup.py' script does the rest.
So, the `-P', `-b', `-p' and `-x' options of `setup' are gone.  Syver also
gave me a few hints about how to modify `setup' for Win32 systems.

Some tuning was needed for Python 2.2 hexadecimal escape sequences, which
Emacs does not grok in some cases.  This is the reason for this release.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From cben at tx.technion.ac.il  Mon Jun 17 08:47:51 2002
From: cben at tx.technion.ac.il (Beni Cherniavksy)
Date: Mon, 17 Jun 2002 15:47:51 +0300
Subject: '__builtin__' vs. __builtins__
Message-ID: <Pine.GSO.4.33.0206171543060.8570-100000@techst02.technion.ac.il>

Why is the '__builtin__' module available under the __builtins__ name?
By what's the logic one is supposed to remember it?  (If I do remember
that it's different I would still confuse which is what...)

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>




From estama at dblab.ntua.gr  Wed Jun 19 16:20:39 2002
From: estama at dblab.ntua.gr (Freeman Stopjohn)
Date: 19 Jun 2002 13:20:39 -0700
Subject: Extending long
Message-ID: <69310ed7.0206191220.3bbf68d8@posting.google.com>

I'm trying to extend the long basic type with some bit manipulation
methods (using Python 2.2).
Below is the code fragment.

-----
class bitnum(long):
    def getbit(self,bitn):
        if self & (1 << bitn) >0:
            return 1
        else:
            return 0
    def setbit(self,bitn):
        self=self ^ ( 1 << bitn )
-----

 The getbit method works ok.
 The setbit doesn't (it doesn't change the number).

 What am i doing wrong?

 thanks.


From geoff at elj.com  Wed Jun  5 10:03:00 2002
From: geoff at elj.com (Geoff Eldridge)
Date: 5 Jun 2002 07:03:00 -0700
Subject: new SmallEiffel elj-win32 release (-0.74)
Message-ID: <8cd6cc47.0206050602.1d69a237@posting.google.com>

A new version of elj-win32 has been prepared and is now available from ..

 * http://www.elj.com/elj-win32/
 * http://www.elj.com/elj-win32/download/

It is based on ..

 * SmallEiffel -0.74 (07 May 2002) [1,2,3]
 * lcc-win32 v3.8 (02 Jun 2002) [4]
 * GOBO library [5]
 * GOBO's gexace [6] and geant [7] tools

This version of SmallEiffel brings ace file support and agents (see small
example below).

This release of elj-win32 is that which is needed for the elj project ..

 * http://elj.sourceforge.net/

which provides interfaces to:

 * ewxw - a wxWindows interface ..

   * http://elj.sourceforge.net/projects/gui/ewxw/

 * win32-api ..

   * http://elj.sourceforge.net/projects/os/win32api/

 * databases: FireBird, mysql, Postgres, SleepyCat (ie Berkeley DB) ...

   * http://elj.sourceforge.net/projects/db/firebird/
   * http://elj.sourceforge.net/projects/db/mysql/
   * http://elj.sourceforge.net/projects/db/postgres/
   * http://elj.sourceforge.net/projects/db/sleepycat/

 * the Pervasive btrieve ..

   * http://elj.sourceforge.net/projects/db/btrieve/

We hope this distribution is useful to you.

Any comments and suggestions appreciated.

Geoff Eldridge

-- geoff at elj.com
-- http://elj.sourceforge.net/

References:

[1] http://smalleiffel.loria.fr/
[2] http://smalleiffel.loria.fr/misc/HISTORY.html#-0.74
[3] http://groups.google.com/groups?selm=e416d3df.0205100712.18847189%
40posting.google.com
[4] http://www.cs.virginia.edu/~lcc-win32/
[5] http://elj.sourceforge.net/projects/other/gobo/
[6] http://elj.sourceforge.net/phpwiki/index.php/GeXace
[7] http://elj.sourceforge.net/phpwiki/index.php/GeAnt

SmallEiffel Agent Example:

--8<-- from a cle post by Greg Compestine --
Here's a simple program that takes an array of strings, converts
all the strings to uppercase and then prints them out.

class UPPER
creation
   make
feature
   make is
      local
         lines: ARRAY[STRING]
      do
         create lines.from_collection (<<"These ","are ","some strings.">>)
         lines.do_all (agent {STRING}.to_upper)
         lines.do_all (agent print)
      end
end

Compiled with SmallEiffel -0.74, the output is:

c:\test\eiffel\agent\upper>compile upper

c:\test\eiffel\agent\upper>upper
  THESE ARE SOME STRINGS.
-->8--


From mkc+dated+1024627496.6928dd at mathdogs.com  Wed Jun  5 22:49:09 2002
From: mkc+dated+1024627496.6928dd at mathdogs.com (Mike Coleman)
Date: 05 Jun 2002 21:49:09 -0500
Subject: 64-bit Python?
Message-ID: <87elfl3yvu.fsf@mathdogs.com>

Has anyone looked at making Python 64-bit clean?

I was looking at trying to use the new mmap module to map in a large (>2G)
file, and the comments in the source note that the mmap regions are limited to
int-size (31 bits), primarily because Python objects are also limited to this
size.

Does anyone know if there's an inherent reason why it would be impossible or
difficult to make Python 64-bit clean (say as a compilation option)?

Mike


From dfaken at brandeis.edu  Wed Jun  5 09:49:18 2002
From: dfaken at brandeis.edu (Daniel Faken)
Date: 5 Jun 2002 06:49:18 -0700
Subject: Simple implementation of tail-call-elimination
Message-ID: <541988bb.0206050549.48d5ee28@posting.google.com>

Hello,

  Just thought I'd share this.
  For more info. about tail-call-elimination, continuations, etc. see
the Scheme language or articles about functional programming in
Python.

  Anyone have a suggestion for a better way to do this?
  (more efficient, elegant, etc..)
  This should be easy to implement in C, for speed.

cheers,
Daniel

dfaken at brandeis.edu

---------------
import types

# Tail-Recursion-Elimination tools
#
# Usage: include a keyword parameter with initial value 'TRE' in the
function(s)
#        from which you wish to eliminate tail-recursion - e.g.
'cont=TRE'.
#        Then, where you would normally use "return Fn(a,b,c..)",
instead
#        call the keyword with the function and its arguments as
parameters
#        - e.g. 'cont(Fn,a,b,c,...)'
#        This allows TRE to control the program flow, and eliminate
deep recursion.
#
# Example:
#   def factorial(n,acc=1,cont=TRE):
#	if n <= 1: return acc
#	else:      return cont(factorial,n-1,acc*n)
#
# N.B. this is not yet set up for use with functions
#      which accept keywords or return multiple values

TRE_magic = "TRE magic string"

def TREcont(fn, *rest): return (TRE_magic, fn, rest)

def TRE(fn, *rest):
	x = fn(cont=TREcont, *rest)
	while isinstance(x, types.TupleType) and (x[0] is TRE_magic):
	  x = x[1](*x[2])
	return x


From robin at jessikat.fsnet.co.uk  Sun Jun 30 12:20:07 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 30 Jun 2002 17:20:07 +0100
Subject: yield equivalent in C/JavaScript?
References: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk>
 <7xy9cxhlvx.fsf@ruckus.brouhaha.com> <w3xdKUAuljH9EwgH@jessikat.fsnet.co.uk>
 <7xit41i7i9.fsf@ruckus.brouhaha.com> <14ms5VAygsH9Eww2@jessikat.fsnet.co.uk>
 <7xelepnk84.fsf@ruckus.brouhaha.com> <WrXUtJA3EvH9EwER@jessikat.fsnet.co.uk>
 <5GFT8.369062$cQ3.24122@sccrnsc01>
Message-ID: <r2HLqVA3+yH9EwSz@jessikat.fsnet.co.uk>

In article <5GFT8.369062$cQ3.24122 at sccrnsc01>, Emile van Sebille
<emile at fenx.com> writes
>Robin Becker
>> The IE stuff seems to rely too much on the URL extension so provided
>it
>> thinks it's htm or html IE 5.x seems OK.
>>
>
>Cool... last time I looked into this it didn't work on IE, and I found
>this on msdn
>
>If the server-provided MIME type is either known or ambiguous, the
>buffer is scanned in an attempt to verify or obtain a MIME type from the
>actual content.
>
>... that I read to mean 'if you tell it, it will make its own check
>anyway and ignore you' so I gave it up.
>
>Can you post an example url that causes IE to properly interpret
>Content-Encoding: x-gzip?
>
>Thanks,
>
I'll try and do that tomorrow, due to the latest apache worm scare I'm
rebuilding the reportlab.co.uk server at the moment. According to my
latest info this won't work for IE 4/5 on Macs, but I may be mistaken on
that.

The URL that I tried looked something like
http://www.reportlab.co.uk/cgi-bin/tgzenc.cgi?ttt.html

ie it had an extension '.html' when I tried with the extension
.html.gz IE only wanted to save the file.
-- 
Robin Becker


From cliechti at gmx.net  Sat Jun 29 12:52:55 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 29 Jun 2002 18:52:55 +0200
Subject: exception handing
References: <pan.2002.06.29.17.41.28.926443.440@myself.com>
Message-ID: <Xns923CC087D1C6Acliechtigmxnet@62.2.16.82>

Rhymes <rhymes at myself.com> wrote in 
news:pan.2002.06.29.17.41.28.926443.440 at myself.com:
...
>>>> ex1 = "spam"
>>>> try:
> ...    raise ex1
> ... except ex1:
> ...    print 'got it'
> ...
> got it
...

exceptions are instances of classes nowdays... don't use strings. and avoid 
troubles...

>>> class Ex1(Exception): pass	#inherit from the root of all exceptios
... 
>>> try: raise Ex1("hello")     #raise instances
... except Ex1: print "got it"  #catch by specifying classes
... 
got it
>>> try: raise Ex1("hello")
... except Exception: print "got it"    	#baseclass works too
... 
got it

as you can see, you have much finer control with classes than strings as 
exceptions. e.g. you can catch the baseclass etc...

chris
-- 
Chris <cliechti at gmx.net>



From emile at fenx.com  Wed Jun  5 09:50:28 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 05 Jun 2002 13:50:28 GMT
Subject: python cookbook at oreilly's web site
References: <dCnL8.5$566.1250@news.ecrc.de> <adl2nq$3hrm$1@ID-41653.news.dfncis.de> <adl3ii$ofq$1@newsreaderg1.core.theplanet.net>
Message-ID: <EGoL8.116405$cQ3.3284@sccrnsc01>

F. GEIGER
> Yes, and I'd really like to know the difference.
> 

I guess we'll have to wait for the book.  ;-)


--

Emile van Sebille
emile at fenx.com

---------



From cliechti at gmx.net  Wed Jun 12 18:24:05 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 13 Jun 2002 00:24:05 +0200
Subject: map -> class instance
References: <m2r8jcp3pv.fsf@video.paradise.lost>
Message-ID: <Xns922C47C58FCcliechtigmxnet@62.2.16.82>

John Hunter <jdhunter at nitace.bsd.uchicago.edu> wrote in 
news:m2r8jcp3pv.fsf at video.paradise.lost:
> I want to convert dictionary instances to instances of a class that
> has the keys of the map as attributes.  I naively tried this:
> 
> class Map2Class:
>     def __init__(self,m):

          self.__dict__['m'] = m
 
>     def __getattr__(self, key):
>         return self.m[key]
> 
>     def __setattr__(self, key, val):
>         self.m[key] = val
> 
>     def __delattr__(self, key):
>         if self.m.has_key(key):
>             del self.m[key]

see above... you have to avoid __setattr__ for your attributes and __dict__ 
is found witouth calling __gettattr__ thus the above works.

chris

-- 
Chris <cliechti at gmx.net>



From -$P-W$- at verence.demon.co.uk  Wed Jun 26 05:49:11 2002
From: -$P-W$- at verence.demon.co.uk (Paul Wright)
Date: 26 Jun 2002 10:49:11 +0100
Subject: Reminder of a python Project.
References: <2218.941T781T12224051threeseas@earthlink.net>
Message-ID: <afc2mn$2kn$1@verence.demon.co.uk>

In article <2218.941T781T12224051threeseas at earthlink.net>,
Timothy Rue <threeseas at earthlink.net> wrote:
>First off, it's ok that I got such a high level of resistance to my
>initial posting here about this project. Thanks to some research that /.
>had an article on, regarding the size of GPL project teams.
>
>But anyway.... Remember the autocoding project I mentioned?


                      ______
                     /      \
                   .' PLEASE `.
                   |  DO NOT  |      _____
                   | FEED THE |    ,'.....`.
                   `. TROLLS ,'  ,'........ )
                     \_    _/   |........ ,'
                       |  |     `. .... _/
                       |  |      ,'.,'-'
                       |  |     /../
                       |  |   ,'.,'
                       |  |  /../
                     . |  | /..'
                   .\_\|  |/_/,
                   ___ |  | ___
                     . `--' .
                      .    .

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


From catunda at pobox.com  Tue Jun 25 10:59:14 2002
From: catunda at pobox.com (Marco Catunda)
Date: Tue, 25 Jun 2002 11:59:14 -0300
Subject: Thread Memory Leak
References: <pan.2002.06.25.12.37.12.535288.17223@pobox.com> <mailman.1025010754.9180.python-list@python.org> <3D18745D.554A7C44@engcorp.com>
Message-ID: <pan.2002.06.25.14.59.10.965526.18707@pobox.com>

Hummmm. The sounds like good. Perhaps, python 2.0 doesn't have
this problem. I will test it in Python 2.0.

I have RedHat 7.3, Python 2.2.1

-- 
Marco Catunda


On Tue, 25 Jun 2002 10:47:09 -0300, Peter Hansen wrote:

> Skip Montanaro wrote:
> Actually, the list is rebound to a new empty list each pass through the
> while loop...
> 
> I can't duplicate the symptoms on my machine (RedHat 7.1, Python 2.0).
> 
> -Peter


From sameer_ at email.com  Mon Jun 10 19:23:58 2002
From: sameer_ at email.com (sameer)
Date: 10 Jun 2002 16:23:58 -0700
Subject: getting files from a windows network
Message-ID: <6bd9f01b.0206101523.7af66fc8@posting.google.com>

is there a way to read files in python that are found on a windows
network, i.e. files that users choose to share on their computers on a
windows network.


thanks ahead for the info.


From shagshag13 at yahoo.fr  Wed Jun  5 04:58:48 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Wed, 5 Jun 2002 10:58:48 +0200
Subject: How could i get execution times ?
Message-ID: <adkjro$12183k$1@ID-146704.news.dfncis.de>

I'm looking for a way to get execution time at different points of a program, by now i use the
code below, but i've seen in some post that time.clock() could be a better way to measure this,

So my question, is how you pythoners do it ??? and to understand _why_ ?

Thanks

s13.

------
import time

class Chrono:
 def __init__(self):

  self._tags = []
  self._time = []
  self._t0 = time.time()

 def add(self, tag = ''):

  self._time.append(time.time())
  self._tags.append(tag)


 def __str__(self):
  maxL = 0
  for tag in self._tags:
   maxL = max(maxL, len(tag))
  tp = self._t0
  s = ' '
  for i in range(maxL + 13):
   s += '-'
  for i in range(len(self._time)):
   s += "\n " + '['
   for j in range(maxL - len(self._tags[i])):
    s += ' '
   s += self._tags[i]
   s += '] : ' + time.strftime("%H:%M:%S", time.gmtime(self._time[i] - tp))
   tp = self._time[i]
  s += "\n " + '['
  for j in range(maxL):
   s += '='
  try:
   s += '] : ' + time.strftime("%H:%M:%S", time.gmtime(self._time[-1] - self._t0))
  except IndexError:
   pass
  return s




From BPettersen at NAREX.com  Tue Jun 11 12:27:48 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 11 Jun 2002 10:27:48 -0600
Subject: Changing the path on Win2k/XP
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019221519B@admin56.narex.com>

> From: David LeBlanc [mailto:whisper at oz.net] 
>
> > From: Bjorn Pettersen 
> > Subject: Changing the path on Win2k/XP
> >
> > Is there a way to change the path on Win2k/XP through Python? (I'm 
> > trying to write a .dll deployment script for our production 
> > department, and I'd like to give them a "one button" install...)
>
> 
> You should be able to modify the path on Win2k/XP by 
> modifying the associated registry key(s). There is one for 
> the "system" environment one for the "per user" environment. 
> This won't take effect for the current process though, but 
> only for new ones (at least comd.exe processes).
> 
> Take a look in (changes depending on user logged in):
> 	HKEY_CURRENT_USER\Environment
> and:
> 	HKEY_LOCAL_MACHINE\SYSTEM\Current Control Set\Control\Session
> Manager\Environment	
> HKEY_LOCAL_MACHINE\SYSTEM\ControlSet001\Control\Session
> Manager\Environment	
> HKEY_LOCAL_MACHINE\SYSTEM\ControlSet002\Control\Session
> Manager\Environment
> 
> Of the 3 in HKLM\SYSTEM, I think only "Current Control Set" 
> is relevent.
> 
> You can modify these with the _winreg package ("import 
> _winreg"). BE VERY CAUTIOUS! Incorrect modifications of your 
> registry can render your machine unusable and possibly 
> requiring an OS reinstall. There is a procedure for creating 
> a backup copy of your registry and a backup disk and I 
> suggest you do both until you are comfortable with what you're doing.

Thanks! Hers's what I came up with (use at your own risk etc. etc.)

-- bjorn

import _winreg

def setpath(entry):
    assert(type(entry) == unicode)
    assert(entry[-1] == u';')
    SYSTEM = _winreg.OpenKey(_winreg.HKEY_LOCAL_MACHINE, 'SYSTEM')
    CurrentControlSet = _winreg.OpenKey(SYSTEM, 'CurrentControlSet')
    Control = _winreg.OpenKey(CurrentControlSet, 'Control')
    SessionManager = _winreg.OpenKey(Control, 'Session Manager')
    Environment = _winreg.OpenKey(SessionManager, 'Environment')
    EnvironmentWrite = _winreg.OpenKey(SessionManager, 'Environment', 0,
_winreg.KEY_SET_VALUE)
    path, type = _winreg.QueryValueEx(Environment, 'Path')
    path = entry + path
    assert(len(path) < 2048)
    _winreg.SetValueEx(EnvironmentWrite, 'Path', 0,
_winreg.REG_EXPAND_SZ, path)
    _winreg.CloseKey(EnvironmentWrite)
    _winreg.CloseKey(Environment)
    _winreg.CloseKey(SessionManager)
    _winreg.CloseKey(Control)
    _winreg.CloseKey(CurrentControlSet)
    _winreg.CloseKey(SYSTEM)





From greg at cosc.canterbury.ac.nz  Thu Jun  6 23:42:48 2002
From: greg at cosc.canterbury.ac.nz (Greg Ewing)
Date: Fri, 07 Jun 2002 15:42:48 +1200
Subject: ANN: Pyrex 0.3 -- Extension Types
Message-ID: <mailman.1023502562.5896.clpa-moderators@python.org>

Pyrex 0.3 is now available:

   http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/

Major new feature: EXTENSION TYPES

You can now define new built-in types in Pyrex, and
use them to wrap arbitrary C data structures. See
the Language Overview for details.

For an example, see Demos/spam.pyx in the distribution.

What is Pyrex?
--------------

Pyrex is a new language for writing Python extension modules.
It lets you freely mix operations on Python and C data, with
all Python reference counting and error checking handled
automatically.

-- 
Greg Ewing, Computer Science Dept, 
University of Canterbury,	  
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg




From rotundo52 at hotmail.com  Wed Jun 12 01:18:57 2002
From: rotundo52 at hotmail.com (Jon J. Morin)
Date: Wed, 12 Jun 2002 05:18:57 GMT
Subject: newbie: copying dictionaries
References: <gdAN8.8432$nn1.3195949@news1.news.adelphia.net>
Message-ID: <5RAN8.8438$nn1.3203473@news1.news.adelphia.net>

Jon J. Morin wrote:

> Hi.  I'm a newbie to python and have a question about copying dictionary
> objects.  I'm working out of Learning Python by Mark Lutz & David Ascher
> (O'Reilly).  The following bit of code was an exercise in the book and it
> works, but what I'm asking is why?  I see where the dictionary keys get
> copied from one object to the other, but how do the values get copied over
> as well?
> 
> #!/usr/bin/python
> # copyDict.py - copy a dictionary argument and return it
> 
> def copyDict(aDict):
>     newDict = {}
>     for key in aDict.keys():
>         newDict[key] = aDict[key]
>     return newDict
> 
> D = {'jon':1, 'allison':2, 'zachary':3}
> print copyDict(D)
> 
> Output:
> [jmorin at cisx1 python]$ ./copyDict.py
> {'zachary': 3, 'jon': 1, 'allison': 2}
> 
> Any help with this is appreciated.  Thanks.
> 
> Jon J. Morin
Thanks, guys.  I think that I understand this now.  Thanks for the help.

Jon


From kurlberg at math.chalmers.se  Sun Jun 23 16:19:51 2002
From: kurlberg at math.chalmers.se (=?iso-8859-1?q?P=E4r?= Kurlberg)
Date: 23 Jun 2002 22:19:51 +0200
Subject: Python hits the spot
References: <mailman.1024772248.9399.python-list@python.org>
Message-ID: <yppfsn3dlpco.fsf@hotlips.cs.chalmers.se>

> It's no wonder that 1000 lines of uncommented code behave like that, taking
> into account that you were trying to integrate several (Tcl/Tk, Fortran, C,
> Python) systems in a hurry, and, I presume, not knowing most of them.

Funny you should mention this.  On my laptop (IBM Thinkpad with 128
megs of ram) the following two-liner gets me a lockup (i.e. keyboard
and mouse completely unresponsive, powertoggle needed):

for i in range(50*1000*1000):
   pass

(I am serious.)  Happened some time ago, but I never got around to
filing a bug report (this thread was a good reminder.)  Seems like the
python code exercises the linux VM/swap to the breaking point.  It is
rather annoying that a user space program can do this...

Can someone else reproduce this behaviour?  (Make sure to quit all
important applications before you try!)

Some trivia:
Python 2.2.1 (#1, Jun 12 2002, 17:57:17)
[GCC 2.96 20000731 (Mandrake Linux 8.1 2.96-0.62mdk)] on linux2






From neal at metaslash.com  Tue Jun  4 18:26:14 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Tue, 04 Jun 2002 22:26:14 GMT
Subject: Fatal Python error: GC object already in linked list
References: <cf43c01d.0206032005.647e3355@posting.google.com>
Message-ID: <pan.2002.06.04.22.26.09.393459.10123@metaslash.com>

On Tue, 04 Jun 2002 00:05:00 -0400, Amit Mongia wrote:

> Hi,
>   I am running a program using pycurl, to upload files on a server. I
> am using Python 2.2.1. I am using threads and lists in my program. the
> application just keeps waiting for new files to arrive and uploads them.
> There are times my program errors out and crashes with this error
> message.
> 
> Fatal Python error: GC object already in linked list
> 
> I am not using the Garbage Collector anywhere in my program. Any idea
> what can cause this error? Any solutions. I desperately need one. There
> is nothing given anywhere in any of the books. I am not even sure if its
> an error with the Garbage Collector. Any comments?

The problem is within the interpreter and could be a bug in Python
or within curl (or any other module).  Have you talked to the curl people?
What version of curl are you using?

If you believe this is a bug in Python, post a bug in SF:

	https://sourceforge.net/tracker/?func=add&group_id=5470&atid=105470

Neal


From fperez528 at yahoo.com  Wed Jun 19 18:44:34 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 19 Jun 2002 16:44:34 -0600
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <um7Q8.44210$n4.10292249@newsc.telia.net>
Message-ID: <aer1gr$fj1$1@peabody.colorado.edu>

Fredrik Lundh wrote:

> 
> you can find more Python tutorials here:
> 
>     http://www.python.org/doc/Newbies.html
> 

I've always been curious. Why isn't the standard tutorial linked to here? It 
seems like the obvious first link to have, no?

Cheers,

f


From bernie at 3captus.com  Tue Jun 11 12:38:55 2002
From: bernie at 3captus.com (Bernard Yue)
Date: Tue, 11 Jun 2002 10:38:55 -0600
Subject: getopt
References: <3D061A9F.7CDC7132@3captus.com> <15622.9979.675543.351236@12-248-41-177.client.attbi.com>
Message-ID: <3D06279E.2452DA86@3captus.com>

Skip Montanaro wrote:
> 
>     Bernard> I was expecting an exception from the following getopt.getopt()
>     Bernard> called, but didn't.  Is it normal?
> 
> Yes.  There's nothing invalid about "-d" as an argument to the "-c" flag, so
> getopt gobbles it up.
> 

Fair enough.  Thx.

> --
> Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
> Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html

Bernie




From scholeyd at logica.com  Tue Jun 18 03:56:14 2002
From: scholeyd at logica.com (Dan Scholey)
Date: Tue, 18 Jun 2002 08:56:14 +0100
Subject: Newbie - Process Management
Message-ID: <1024387201.371093@ernani.logica.co.uk>

Hi,

I am hoping you can help me, I am fairly new to python (and Unix) but am
trying to solve the following problem...

I want to be able to check if a parent process has died.

At the moment I can check that a child process has died using...

        status = (os.waitpid(child, os.WNOHANG))[0]
        if status!=0:

I've looked at several manuals and can't find a similar function for parent
processes.
Obviously I know both the parent and child process id ( get_pid() /
get_ppid() ).

I have a workaround of....

        ps_line=string.split(commands.getoutput("ps -lp %s"%my_pid),"\012")
        if string.atoi(string.split(ps_line[1])[4])!=daddy:

... but I would rather not use ps.

Any ideas?

Cheers
Dan




From fperez528 at yahoo.com  Fri Jun 28 15:52:36 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Fri, 28 Jun 2002 13:52:36 -0600
Subject: C API for Python's RNG?
References: <3D1CBB3A.E1C43D8A@astro.cornell.edu> <3D1CBBED.58EAC61E@astro.cornell.edu>
Message-ID: <afieq8$o17$1@peabody.colorado.edu>

Tom Loredo wrote:

> So I guess it's Numeric or some third-party RNG if
> I want to do this quickly.  Any suggestions?

The Numeric RNG is reasonably fast and versatile. Otherwise just code your own 
from one of the bazillion RNGs out there. Something that simple can be 
wrapped up in a minute using weave or even SWIG.

cheers,

f


From amuys at shortech.com.au  Mon Jun 24 00:25:52 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 23 Jun 2002 21:25:52 -0700
Subject: Iterating through two lists
References: <mailman.1022234790.26965.python-list@python.org> <7934d084.0206031700.377c691e@posting.google.com> <c.tee-FF2C9D.10574923062002@news.bellatlantic.net>
Message-ID: <7934d084.0206232025.3d7d9ce5@posting.google.com>

Curtis Taylor <c.tee at verizon.net> wrote in message news:<c.tee-FF2C9D.10574923062002 at news.bellatlantic.net>...
> In article <7934d084.0206031700.377c691e at posting.google.com>,
>  amuys at shortech.com.au (Andrae Muys) wrote:
> 
> > "Denis S. Otkidach" <ods at fep.ru> wrote in message 
> > news:<mailman.1022234790.26965.python-list at python.org>...
> > > On Fri, 24 May 2002, jb wrote:
> > > 
> > > j> I have two lists, x and y with the property len(x) = len(y).
> > > j>
> > > j> I should like to achive this (x is a list of class instances)
> > > j>
> > > j>   for (a,b) in (x,y): a.f(b)
> > > j>
> > > j> Is there a fancy way of doing this or have I to introduce an
> > > j> auxillary
> > > j> counter (that is very easy but maybe not very "lispy", that
> > > j> is
> > > j> "python-like").
> > > 
> > > If lists are not very lange try this:
> > > 
> > > for a, b in zip(x, y): a.f(b)
> > 
> > If the lists are large then you probably need something along the lines of 
> > xzip.
> > 
> > def xzip(*args):
> >     iters = [iter(a) for a in args]
> >     while 1:
> >         yield tuple([i.next() for i in iters])
> > 
> > Andrae Muys
> 
> 
> Hi Andrae,
> 
> I've got a similar situation I'm dealing with, in that I have 2 lists of 
> dictionaries, each of which is of equal length. What's different 
> however, is that I need to do an equality check for the values of 2 of 
> the keys (both of which exist in each dictionary), then set the value of 
> the 2nd dictionary's key to the value of that of the first's, e.g.:
> 
> if dict2[key1] == dict1[key1]:
>    dict2[key2] = dict1[key2]
> 
> I'm concerned with overhead, as the 2 lists could potentially be very 
> large. Can your xzip function be extended to do such a thing?
> 

1. I don't understand where you're getting key1 or key2 from?  

2. xzip() is reasonably well defined semantically, so I doubt you
really want to extend it, rather I suggest you consider a custom
generator (or 2).

3. I suspect you should probably read the section on iterators and
generators in the What's New in Python 2.1 bulletin.

4. One useful thing to note is that you can always collapse a
generator/iterator into a tuple/list by using the tuple()/list()
builtin functions.

Andrae


From whisper at oz.net  Thu Jun  6 15:40:05 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 6 Jun 2002 12:40:05 -0700
Subject: Using AT on XP to run python script - MARK HAMMOND PLEASE READ
In-Reply-To: <fa7108b0.0206061106.339d547b@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEEEBMDHAA.whisper@oz.net>

I did my tests on Win 2000 pro - what OS are you using?

When it comes to using sockets etc. you might be right that some permissions
are involved. Would be interesting to know what user account 'at' runs
under.

Since this isn't strictly a Python issue, maybe comp.os.windows (or
something like that) might better be able to answer the question. You might
also check the discussion boards on http://www.codeguru.com/index.shtml or
http://www.codeproject.com/

If you discover the answer I would be interested in knowing too!

Regards,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Chris
> Sent: Thursday, June 06, 2002 12:07
> To: python-list at python.org
> Subject: Re: Using AT on XP to run python script - MARK HAMMOND PLEASE
> READ
>
>
> Interesting findings.  I know mine didn't work bcs the script I want
> AT to execute sends me an email.  The script sends an email, then I
> want it to run another AT command to set itself up to run again at a
> later time.  But I can't get it to work.  If I run the script
> manually, it runs, sends email and executes the proper AT command.
> But when AT fires off the task the script set up, nothing happens and
> no more AT commands are set up.
>
> "David LeBlanc" <whisper at oz.net> wrote in message
> news:<mailman.1023320948.13063.python-list at python.org>...
> > This is most curious, but I believe the commands _are_ getting executed!
> >
> > I tried the following (assume time is 16:30) "at 16:32 pythonw
> > j:\python22\tools\idle\idle.py" (without the qoutes of course). At the
> > appointed time, there was disk activity, but no idle window
> popped up (idle
> > was chosen just to get something to pop up!). However, there
> _was_ a pythonw
> > process running in the task manager! I think there might be a
> way to attach
> > a console or get access to the GUI, but i'm not sure - maybe
> Mark Hammond
> > might have a clue (he being the Python Win guru :) ).
> >
> > I tried the following:
> > #at_test.py
> > import sys
> > import time
> >
> > tst = open("testfile.txt", 'a+')
> >
> > tst.write(time.asctime() + "\n")
> >
> > tst.close()
> >
> > put into an 'at' job as: at xx:xx python j:\python22\at_test.py
> >
> > I expected the output from this to be in j:\python22, but to my
> surprise, it
> > was in c:\winnt\system32 where cmd.exe resides! testfile.txt
> contained the
> > exact time the 'at' job was set to run at too!
> >
> > I suspect your last example that produced no visible output did
> in fact work
> > :)
> >
> > In any case, it works - now if one could just get visible output!
> >
> > David LeBlanc
> > Seattle, WA USA
> >
> > > -----Original Message-----
> > > From: python-list-admin at python.org
> > > [mailto:python-list-admin at python.org]On Behalf Of Chris Stromberger
> > > Sent: Tuesday, June 04, 2002 17:35
> > > To: python-list at python.org
> > > Subject: Using AT on XP to run python script
> > >
> > >
> > > I'm having trouble getting AT on XP to run a python script for me.
> > > Have tried:
> > >
> > > at xx:xx "c:\python21\python.exe c:\scripts\script.py"
> > > at xx:xx "start c:\scripts\script.py"
> > > at xx:xx "c:\scripts\script.py"
> > >
> > > and get "the system cannot find the file specified" for the first two.
> > > The last version does nothing apparently--no error indication or
> > > anything.
> > >
> > > I imagine it is related to user accounts and such.  Anyone have any
> > > pointers?
> > > --
> > > http://mail.python.org/mailman/listinfo/python-list
> --
> http://mail.python.org/mailman/listinfo/python-list





From ak at silmarill.org  Sat Jun  1 12:25:51 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Sat, 01 Jun 2002 16:25:51 GMT
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org> <3CF8E382.3A44F0DA@engcorp.com>
Message-ID: <slrnafhtdv.407.ak@ak.silmarill.org>

In article <3CF8E382.3A44F0DA at engcorp.com>, Peter Hansen wrote:
> Andrei Kulakov wrote:
>> 
>> Hello,
>> 
>>     "hello world" type program from wxPython tutorial takes 6
>> seconds to start on my cel 366 128mb system (running Debian). Is
>> this typical? Why is it so slow? I'm looking for a graphic
>> toolkit that'd be fairly quick. I don't want to use Tk because of
>> the looks.
> 
> How fast do you need it to be?  (That's a serious question.
> Are you planning to open and close the application so frequently
> that six seconds is unacceptable?)
>
For me it's not a big deal, but I plan to write apps for
other people, and this 6 seconds just feels wrong. I guess it's
more of a psychological thing than anything else.. Anyway, it's
still not that big a deal, I was just hoping someone would say
"of course, you have to use this flag or option:
--start-in-one-second" or "pygtk will start in 2 sec". If
everythign is this slow, I'll just use wxPython and won't mind it
too much :-).

> 
> On my blazing 233 MHz Windows 98 machine, with Python 2.2.1
> and WxPython 2.3.2.1, a hello world program I just wrote
> runs in about two seconds, for what it's worth.
> 
> -Peter

Odd.. if not too much trouble, could you please try this one 
(from wxpython tutorial)?

from wxPython.wx import *

class MyApp(wxApp):
    def OnInit(self):
        frame = wxFrame(NULL, -1, "Hello from wxPython")
        frame.Show(true)
        self.SetTopWindow(frame)
        return true

app = MyApp(0)
app.MainLoop()





-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From aoeu at go.com  Wed Jun 19 20:15:19 2002
From: aoeu at go.com (Thinkit)
Date: 19 Jun 2002 17:15:19 -0700
Subject: Unicode hexadecimal characters
Message-ID: <3eeda89d.0206191615.71b0629d@posting.google.com>

It's a bit OT but does Unicode support the digits ten through fifteen
(for hexadecimal)?


From maxm at mxm.dk  Mon Jun 17 13:52:15 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 17 Jun 2002 19:52:15 +0200
Subject: Long pathes on windows
References: <mailman.1024334302.2943.python-list@python.org>
Message-ID: <3D0E21CF.7010405@mxm.dk>

Roman Yakovenko wrote:
> Good evening. I have big problem Windows 2000 and small one.
> The small one is long pathes on windwos. How can I convert long path with ' '
> into short path? 


come agan?


Max M



From johnroth at ameritech.net  Sat Jun 29 18:35:37 2002
From: johnroth at ameritech.net (John Roth)
Date: Sat, 29 Jun 2002 18:35:37 -0400
Subject: private
References: <pan.2002.06.29.17.38.51.789527.440@myself.com> <3D1DE132.4A890D9D@engcorp.com> <mailman.1025382159.23425.python-list@python.org>
Message-ID: <uhsdf5irpg9n25@news.supernews.com>

"Dave Reed" <dreed at capital.edu> wrote in message
news:mailman.1025382159.23425.python-list at python.org...
> > From: Peter Hansen <peter at engcorp.com>
> > X-Accept-Language: en
> > Newsgroups: comp.lang.python
> > Xref: news.baymountain.net comp.lang.python:170296
> > Sender: python-list-admin at python.org
> > Date: Sat, 29 Jun 2002 12:32:50 -0400
> >
> > Rhymes wrote:
> > >
> > > Is there any possibility to build up a _real_ private attribute?
> >
> > Please define what "_real_" means to you.  It's not apparent.
> >
> > > Such as C++ private data members...
> >
> > You realize that it's quite possible to get at private data members
> > in C++, don't you?
>
> Kind of off-topic, but this got me curious. The only way I could think
> of is to assume (which I think is always true) that the private data
> members are in order from the starting address so based on the offset
> from the address of the object you could access each one.
>
> Is there another way?

I believe he's talking about the preprocessor macro trick
of redefining the word 'private' as 'public.'

John Roth
>
> Dave
>
>




From peter at engcorp.com  Sat Jun  8 09:43:37 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 08 Jun 2002 09:43:37 -0400
Subject: Medium to Large Scale Python Deployments
References: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com> <mailman.1023409393.3361.python-list@python.org> <3D003EC7.C72D27BE@engcorp.com> <3D006320.5040805@web.de> <3D00A209.EC24F671@engcorp.com> <adr2ao$bei$0@216.39.172.122> <3D0140BF.A239E39E@engcorp.com> <mailman.1023522673.25519.python-list@python.org>
Message-ID: <3D020A09.7BF6F985@engcorp.com>

Nicola Larosa wrote:
> 
> > (we rarely use doc strings, having avoided that aspect of
> > Python when writing on an embedded system where all the .pyc
> > files were stored in our limited 1MB of flash along with the
> > interpreter...).
> 
> python -OO ?

I don't recall whether we actually didn't know about that at
the time (I think that was it), or were not willing to accept
the /other/ changes -OO made to the output.

-Peter


From djc at object-craft.com.au  Sun Jun 23 06:11:07 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 23 Jun 2002 20:11:07 +1000
Subject: Web templating/db tool with best designer/coder separation?
References: <af2gh6$2c2$0@216.39.172.122> <slrnah9mj3.rdk.jon+usenet@snowy.squish.net> <3D159696.9010209@mxm.dk>
Message-ID: <m3r8iye24k.fsf@ferret.object-craft.com.au>

>>>>> "Max" == Max M <maxm at mxm.dk> writes:

Max> Jon Ribbens wrote:
>> In article <af2gh6$2c2$0 at 216.39.172.122>, Bengt Richter wrote: Take
>> a look at http://jonpy.sourceforge.net/ . It almost completely
>> separates code and HTML so that the designer and the coder do not
>> interfere with each other.

Max> I do understand that argument, but I almost certainly don't agree
Max> that it is of much use.

I agree, I am not sure it is such a good argument myself.  If you are
working in an environment where HTML editors do not understand the
templating system then you are going to need at least one person who
can post process the HTML and integrate it with the templating logic.
No magic templating system is going to solve that problem.

Max> In smaller one-of systems it makes sense, but in an app-server
Max> like Zope where the idea is that different components can be
Max> joined together to create a bigger application, there has to be
Max> stronger rules for layout and design of the components so that
Max> they will automatically fit into the bigger whole.

Whenever you have a medium to large size system you need to establish
rules to ensure that components fit together.  Zope is certainly not
unique in this regard.

Max> Page template languages is the wrong approach in this regards.

Are you able to explain why?

Max> The development of the Cmf and Plone are good examples of this
Max> wrongfull approach. They make it easy to create one single type
Max> of application, but makes it very hard to re-use their components
Max> to build other types of applications.

Whenever you build any toolkit you have to decide on the set of
problems that you wish to solve with the toolkit.  The smaller the
problem domain the more useful you can make the toolkit.

If your problem falls within the domain of the toolkit then you will
like the toolkit.  If your problem is not addressed by the toolkit
then you are likely to question the usefulness of the approach.

Max> And I believe that the reason is that they are using pagetemplate
Max> languages to bind it together, which leads to this kind of
Max> design.

I suggest that the problem that they are trying to solve has lead to
the implementation.  I could be misunderstanding you, but could you
explain an alternative approach to solving the problem which would
deliver a more general solution?

Max> I my experience, tools which separates logic from view in the
Max> "pagetemplate" approach is doing it wrong.

Again, can you explain why?

Max> The layout may be separated, but reuse of code/componentes
Max> greatly suffers from this approach.

Not sure I agree here.  Maybe I am missing something because I think
that a clear separation of presentation and implementation increases
the cohesiveness of code and reduces the incidence of pathological
coupling.

- Dave

-- 
http://www.object-craft.com.au


From guy at lightwork.co.uk  Thu Jun 27 09:57:29 2002
From: guy at lightwork.co.uk (Guy)
Date: 27 Jun 2002 06:57:29 -0700
Subject: Var substitution
Message-ID: <c07d0edc.0206270557.28180893@posting.google.com>

Hi

What i'm wanting to know, is there any way of looking at a variables
value rather than its name.

Example

Unix : ${MyVar}
Dos  : %MyVar%

Both of the above will be subtituted with the values.

Example python script (below causes error):

>>> guy="name"
>>> import os
>>> os.guy

This is what should happen :

>>> import os
>>> os.name
'nt'


From mcfletch at rogers.com  Sun Jun 16 15:16:17 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Sun, 16 Jun 2002 15:16:17 -0400
Subject: [Extending] Py_BuildValue( "((OiiO))",...) creates empty string?
References: <3D0CDEEB.2070208@rogers.com>
Message-ID: <3D0CE401.8030605@rogers.com>

Argh.  Sorry about that, was operator error (the test was wrong, not the 
C code it was testing).  I was receiving the test value and forgetting 
to store it before checking it.

Sorry for the wasted bandwidth,
Mike

Mike C. Fletcher wrote:
> I'm working on a C module, and have come across something I don't 
> understand.  Basically, I have a method definition/specification which 
> says I will pass my clients a 4-item tuple to their append method.

<dumb details omitted>





From bokr at oz.net  Thu Jun 27 10:44:21 2002
From: bokr at oz.net (Bengt Richter)
Date: 27 Jun 2002 14:44:21 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com> <afdrlv$r83$0@216.39.172.122> <nd0mhu8omept0dkl1deeac7pp4u2c17ufh@4ax.com>
Message-ID: <aff8c5$stl$0@216.39.172.122>

On Thu, 27 Jun 2002 15:16:46 +0300, Christos "TZOTZIOY" Georgiou <DLNXPEGFQVEB at spammotel.com> wrote:

>On 27 Jun 2002 02:01:35 GMT, rumours say that bokr at oz.net (Bengt
>Richter) might have written:
>
>>Try accumulating your chunks in a list,
>>like sl=[] followed by sl.append('chunk') instead, and then do f.write(''.join(sl))
>>instead of the f.write(s) you would have done.
>
>A minor correction, if I may add:
>
>f.writelines(sl)
>might be quite more efficient than
>f.write(''.join(sl))
>
>AFAIK the output is the same, as long as one remembers to insert the
>necessary '\n' for line breaks where needed :)

Thank you for pointing that out. (That should teach me to throw phrases around
like "typical newbie performance killer." ;-) I have not been in the habit of
using writelines, and did not remember that since readlines preserves newlines,
writelines would presumably, for symmetry, not add them.

I expect to be using writelines more in the future ;-)

Regards,
Bengt Richter


From BPettersen at NAREX.com  Tue Jun 25 13:39:34 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 25 Jun 2002 11:39:34 -0600
Subject: How do i COPY files from 1 directory to another..
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E201922151AD@admin56.narex.com>

> From: Dirgesh Patel [mailto:Kemp100 at attbi.com] 
> 
> How do i Copy multiple files from 1 directory to 
> another..this is the code I have so far..but it gives me a 
> PERMISSION DENIED ERROR..
> 
> 
> import shutil
> import os
> 
> shutil.copy2(r"C:\test",r"C:\test1")
> 
> I wanna copy everythign from C:\test to C:\test1....

I'm guessing you really want shutil.copytree()?

-- bjorn





From cliechti at gmx.net  Sun Jun 23 15:17:39 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 23 Jun 2002 21:17:39 +0200
Subject: sys.exc_info() into string
References: <KZoR8.44805$n4.10504231@newsc.telia.net>
Message-ID: <Xns9236D90D4E2FAcliechtigmxnet@62.2.16.82>

Magnus <spammers.do.not.bother at void.bogus> wrote in
news:KZoR8.44805$n4.10504231 at newsc.telia.net: 
> in my program I am catching exceptions that I might have missed by
> using something like:
> 
> try:
>         doMyStuff()
> except:
>         type = sys.exc_info()[0]
>         syslog.syslog(syslog.LOG_EMERG, "My program: unexpected error"
>         + type) 
> 
> The thing is that I can not figure out how to make "type" into a
> string above so I can use it for e.g. syslog(). Is there an easy way
> of doing this?

yes, the traceback module has functions to format an exception as string or 
to print.

format_exc is a candidate for you. it return a list of lines from which you 
can select to mosr informative or just join it the one big string.

chris

-- 
Chris <cliechti at gmx.net>



From kragen at pobox.com  Sat Jun  8 20:31:42 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 08 Jun 2002 20:31:42 -0400
Subject: check on numerical value?
References: <3cff56b4.511834453@news.skynet.be> <l5JL8.123663$cQ3.3519@sccrnsc01> <adnqh7$ri6f$1@ID-146704.news.dfncis.de>
Message-ID: <83lm9p5m35.fsf@panacea.canonical.org>

"Shagshag13" <shagshag13 at yahoo.fr> writes:
> can't you use (after importing types) type(x) == types.FloatType (or type.IntType)

Don't do that.  See "isinstance considered harmful",
http://lists.canonical.org/pipermail/kragen-tol/2002-January/000659.html.



From usenet at thinkspot.net  Sat Jun  1 12:23:27 2002
From: usenet at thinkspot.net (Sheila King)
Date: Sat, 01 Jun 2002 09:23:27 -0700
Subject: How to open a HTML file when the python cgi program is executing?
References: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de> <ad9uv4$g2n$2@agate.berkeley.edu> <ad9vo8$utftp$1@ID-49758.news.dfncis.de>
Message-ID: <ada3qg.lo.1@kserver.org>

On Sat, 1 Jun 2002 18:16:28 +1000, "Ken" <ken at hotmail.com> wrote in
comp.lang.python in article <ad9vo8$utftp$1 at ID-49758.news.dfncis.de>:
 
> Yes, I want to display the content of another html file (main.html in this
> example) onto the browser. How do I do that? And also, how do I set target
> to which frame to display?

If you are trying to open some file and output it to the browser, then:

f = open('main.html', 'r')
myhtml = f.read()
f.close()
print myhtml

Depending on what else the script has already done. For example, if you are
printing the HTML output, instead of redirecting, you do need to print the
Content-type before printing the contents of main.html:

print "Content-type: text/html\n"

But, if you are redirecting to another URL instead, then omit the line show
above and print this line:

print "Location: http://replacement-url\n"


--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/





From rjones at ekit-inc.com  Sun Jun 23 21:38:19 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Mon, 24 Jun 2002 11:38:19 +1000
Subject: PIL, Python and CGI
In-Reply-To: <Pine.GSO.4.33.0206231929060.26987-100000@infinity.infineon.net>
References: <Pine.GSO.4.33.0206231929060.26987-100000@infinity.infineon.net>
Message-ID: <200206241138.19665.rjones@ekit-inc.com>

On Mon, 24 Jun 2002 11:30, Sascha Ferley wrote:
> Thanks for replying back to me, this thing has been driving me nuts for a
> while now. Anyways, i removed the try/catch block from the Image.py lib,
> and checked the error output:
>
> Traceback (most recent call last):
>   File "/export/server/web/CPSC461/cgi-bin/interfce.py", line 15, in ?
>     print Image.open(urlretrieve(sourcePath, '/tmp/img1')[0]).histogram()
>   File "/usr/local/lib/python2.2/site-packages/PIL/Image.py", line 577, in
> histogram
>     self.load()
>   File "/opt/sfw/lib/python2.2/site-packages/PIL/ImageFile.py", line 126,
> in load
>     self.load_prepare()
>   File "/opt/sfw/lib/python2.2/site-packages/PIL/ImageFile.py", line 180,
> in load_prepare
>     self.im = Image.core.new(self.mode, self.size)
> AttributeError: 'module' object has no attribute 'core'

"core" is assigned just after the "import _imaging" line in Image.py. Are you 
sure there isn't another traceback we're missing? What happens if you just 
have a simple CGI script that just does "import _imaging"?


   Richard





From krissepu at vip.fi  Wed Jun  5 03:28:21 2002
From: krissepu at vip.fi (Pekka Niiranen)
Date: Wed, 05 Jun 2002 10:28:21 +0300
Subject: Fastest way to read / procsess / write text?
References: <3CFD5229.2060803@excite.com>
Message-ID: <3CFDBD95.D6DFE028@vip.fi>

Use map() instead of  for -loop:

pdata = []
pdata = map(re.sub(whatever), data)

I managed to get 10x speed increase myself with map()

Try alsol earning mxtexttools for further speedups (so I can ask help from
you ;))

-pekka-

Brent Miller wrote:

> I'm somewhat new to python and programing in general and I trying to
> write a simple script that I can use to format large text files so that
> they are suitable for importing into MySQL. So far, I have came up with
> this:
>
> #!/usr/bin/python2
>
> import sys, re
>
> infile = sys.stdin
> data = infile.read()
> infile.close()
>
> data = re.sub('[.](?=\d\d\d\d\d\d)|[.](?=\w+[ ][<|>])|[.](?=\w+[:])|[
> ](?!0x)', '\t', data)
>
> outfile = open("/mnt/storage/output.txt", "w")
> outfile.write(data)
> outfile.close()
>
> This works beautifully most of the time (it's super fast), except when I
> pipe large files (>50megs) to it and then it usually dies half way
> through complaining of memory errors because it ran out of ram.
>
> Short of buying more ram, is there a way I can make this more effecient
> without taking a big peformance hit? I tried the above where I used a
> "for line in data:" loop so it would only process one line at a time but
> this seemed to take forever! I'm imagining there's a way to process the
> data in something like 1 meg chunks, but I'm not too sure how to do that.
>
> I'm using python 2.2.1 on a RH linux 7.2 system with 256 ram and a
> Athlon XP1700 if it makes any diffrence.
>
> Any suggestions?
>
> TIA,
> Brent



From lmbrtmarc at aol.comlaissermoi  Wed Jun 26 13:41:20 2002
From: lmbrtmarc at aol.comlaissermoi (Marc)
Date: 26 Jun 2002 17:41:20 GMT
Subject: Python and Windows XP
References: <3d193d85$0$20999$afc38c87@news.optusnet.com.au>
Message-ID: <20020626134120.09399.00000674@mb-cp.aol.com>

Dans l'article <3d193d85$0$20999$afc38c87 at news.optusnet.com.au>, "Steven Adams"
<sadams123 at optushome.com.au>a ?crit:

>
>I just open up python.exe in my Python directory and it brings up the Python
>interpreter, the standard 2.2 install setup shortcuts to it as

   Well I do that, too.
>
>Start->Programs->Python -->Python (command line)
>
>There is a command prompt that behaves like a DOS prompt - though I can't
>recall where it sits in the default menu.
>
>I'm using XP Pro so YMMV

  Me too, and my problem has be solved, thank you.
-- 
Marc
if we can't benefit from an error by making jokes, what's the  point of being
here?


From smarsh at hotmail.com  Sat Jun 29 00:53:23 2002
From: smarsh at hotmail.com (scott)
Date: Sat, 29 Jun 2002 04:53:23 GMT
Subject: Is Python growing?
References: <mailman.1025267736.21565.python-list@python.org> <slrnahoucn.140.gerhard@lilith.my-fqdn.de>
Message-ID: <3D1D3D34.6080001@hotmail.com>

Gerhard H?ring wrote:
> Mark McEahern wrote:
> 
>>[...] What's so
>>beautiful about Python is that all I had to do was fire up the interactive
>>interpreter and do this to remind myself of what map does:
>>
>> >>> print map.__doc__
> 
> 
> help(map) in 2.2.
> 
> Gerhard

Or 2.1.1.



From gerhard.haering at gmx.de  Wed Jun  5 13:55:54 2002
From: gerhard.haering at gmx.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Wed, 5 Jun 2002 19:55:54 +0200
Subject: [Pypgsql-users] Re: [ANN] pyPgSQL 2.1 released
In-Reply-To: <20020605170847.2631.qmail@www4.nameplanet.com>
References: <20020605170847.2631.qmail@www4.nameplanet.com>
Message-ID: <20020605175554.GB33711@gargamel.hqd-internal>

* paul at boddie.net <paul at boddie.net> [2002-06-05 17:08 -0000]:
> On Wed, 5 Jun 2002 18:40:11 +0200 Gerhard H?ring wrote:
> >CC-ing to pypgsql-users list, maybe other users have comments on this issue,
> >too.
> 
> [Arrays vs. 'IN']
> 
> >QUERY: DECLARE PgSQL_08120A6C CURSOR FOR select * from cages where name
> >       in '{"mammals","apes"}'
> >
> >... which obviously cannot work. The reason is that if pyPgSQL encounters a
> >tuple or a list in the statement parameters, it will quote them suitable for
> >use as an ARRAY. Unfortunately, the ARRAY syntax is different from what you
> >would need for the IN operator in SQL.
> 
> Right, and there's no way of knowing what the database engine expects for
> that parameter until it has parsed the statement.

Unless you also parse at the client side, which I'd rather not ;-)

There might be a solution with controlling the behaviour using a flag in the
cursor object, for example.

> I suppose there isn't a way of getting type information from PostgreSQL about
> parameters - indeed, does PostgreSQL actually support parameters in the way
> that other database systems do?

You mean prepared statements? No, PostgreSQL doesn't currently support these.
All it gets is a query string. From what I hear on IRC and on
postgresql-hackers, people are working on prepared statements.

Unfortunately, we'll have a problem with them and pyPgSQL, as as far as I see,
prepared statements don't play nice with the DB-API 2.0 'format' and 'pyformat'
parameter quoting styles. I believe that the 'qmark' or 'numeric' styles are
designed exactly for prepared statement use. It might be possible to rewrite
the query and continue using our 'pyformat' style, but I haven't tried, yet.

> > [Unicode patch] 
> That would be great. Where XML data and PostgreSQL come together, it quickly
> becomes a requirement to handle the data gracefully or transparently,
> although I suspect that many database systems still struggle with Unicode.

Hopefully, pyPgSQL with PostgreSQL works fine. PostgreSQL can even use the
UTF-8 encoding at server side. With the Unicode patch and appropriate
parameters for the connect() call, you should be able to use full Unicode:

>>> from pyPgSQL import PgSQL
>>> conn = PgSQL.connect(client_encoding="utf-8", unicode_results=1)
>>> cursor = conn.cursor()
>>> cursor.execute("insert into test(name) values (%s)", unicode("?sterreich", "latin1"))
>>> cursor.execute("select * from test")
>>> res = cursor.fetchone()
>>> print repr(res.name)
u'\xd6sterreich'

Btw. the Unicode support that PySQLite recently gained is based on the same
patch, and it includes an example for mass-importing Freshmeat project data
from XML into a database. As you said, XML is Unicode-based, so this is a
relatively practical example, I hope.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From tim at vegeta.ath.cx  Wed Jun  5 16:33:45 2002
From: tim at vegeta.ath.cx (Tim Hammerquist)
Date: Wed, 05 Jun 2002 20:33:45 GMT
Subject: Detecting OS and Window Manager
References: <ufsnh8pmmde6cf@corp.supernews.com>
Message-ID: <slrnafstki.3ne.tim@vegeta.ath.cx>

Andrew Wilkinson graced us by uttering:
> I'm writing a program that integrates with a number of different OSes
> and Window Managers (under Linux), but I need to check what the user
> is running.  I know that...
> 
> import os
> if os.name == 'nt':
> 
> will tell me they're running NT. But under Linux I'd like to know if
> they are running KDE or Gnome, does anyone know a (preferably) simple
> test for this?

Well, windowmanagers aren't as static as operating systems are.  It's
actually possible to run more than one wm at a time.

In any case, I went and experimented with the various wms on my system
and in all cases the $WINDOWMANAGER shell var was set to the full path
of the running wm executable.  This appears to be the only consistent
and reliable var in my environment; for instance, here's a dump of my
wm-related vars:

QTDIR=/usr/lib/qt2
GNOME_PATH=:/opt/gnome:/usr:/opt/gnome:/usr
KDEDIR=/opt/kde2
GNOMEDIR=/opt/gnome
EROOT=/usr/X11R6/share/enlightenment
WINDOWMANAGER=/usr/X11R6/bin/enlightenment

The qt lib on my box is used only by KDE apps.  I am indeed running KDE2
apps.  Gnome is in fact located under the /opt/ dir.  The Enlightenment
tree does start at $EROOT.

There are, of course, more KDE-related vars in a KDE session, and more
Gnome-related vars in a Gnome session, but $WINDOWMANAGER confirms that
I'm actually running Enlightenment (with the Gnome panel and one KDE
app).

However, I'm not sure how this works in instances (which I've had) when
the window manager and desktop environment are different. eg: KDE env
with sawfish wm.

KDE apps run fine under Gnome, Gnome apps run under KDE, and both KDE
and Gnome run under E.  As long as you have the correct libs available
(gtk, qt, etc) you shouldn't need to know what windowmanager is running
anyway.

Thoughts? Ideas?

Tim Hammerquist
-- 
Ford, you're turning into a penguin. Stop it.
    -- Arthur Dent, "The Hitchhiker's Guide to the Galaxy"


From lt6789remove at hotmail.com  Thu Jun 13 14:28:08 2002
From: lt6789remove at hotmail.com (Larry Taitt)
Date: Thu, 13 Jun 2002 11:28:08 -0700
Subject: Post installation problem with Python
References: <9dafguci28vdiv5lr3qbhset2jko78t6lm@4ax.com> <3d07b91c$0$21420$91cee783@newsreader02.highway.telekom.at>
Message-ID: <nuohgug051t6rgpg456h0jo6hjj4u0v29s@4ax.com>

That's exactly what it was.  

Thanks.


On Wed, 12 Jun 2002 23:12:37 +0200, "Guenter Szolderits"
<g.szolderits at gmx.at> wrote:

>"Larry Taitt" <lt6789remove at hotmail.com> wrote:
>> Installed ActivePython 2.2.1 build 222 on my NT4.0 (sp6a) workstation.
>> After the installation I'm unable to execute any .exe, .com and .bat.
>> program (e.g. notepad).  It tells me that it can't find the files.
>> However, if I add the proper extension (e.g., notepad.exe), the
>> programs execute without a problem.  I checked the associations and
>> filetypes, and everything appears normal.  Any ideas?
>
>Maybe your PATHEXT environment variable is damaged. It should look similar
>as mine:
>
>PATHEXT=.COM;.EXE;.BAT;.CMD;.py;.pyw;.pyc;.pys;.pyo
>
>It should at least include .COM;.EXE;.BAT
>
>G?nter
>



-----------== Posted via Newsfeeds.Com - Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Ulimited Fast Downloads - 19 Servers =-----


From skip at pobox.com  Thu Jun 20 18:20:10 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 20 Jun 2002 17:20:10 -0500
Subject: Bug in Python 2.3 ?
In-Reply-To: <20020620205335.0c1d31b4.christophe.delord@free.fr>
References: <20020620205335.0c1d31b4.christophe.delord@free.fr>
Message-ID: <15634.21786.425480.991507@beluga.mojam.com>

    Christophe> It seems there's a bug in python 2.3 (available on CVS).

I won't comment on your note except to say that if you think there is a bug
(getting a segfault sure sounds like it), you should submit a bug report on
SourceForge: 

    http://sourceforge.net/projects/python/

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html




From cliechti at gmx.net  Thu Jun 20 15:23:28 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 20 Jun 2002 21:23:28 +0200
Subject: Inter Process Communication with Numeric arrays
References: <wx1yb2uvt3.fsf@velocita.nasoftware>
Message-ID: <Xns9233DA07E6417cliechtigmxnet@62.2.16.82>

Ian McConnell <ian at nasoftware.co.uk> wrote in
news:wx1yb2uvt3.fsf at velocita.nasoftware: 

> I want to send the results of some calculation on a Numeric array to a
> second process. These arrays will be fairly large (about 1Mb) so can
> anyone recommend a package for simple IPC?

you could use pickle and send it trough a pipe or socket.
pyro.sf.net solves remote objects problems in python. it uses pickle 
over sockets and the other app can be on any machine on th net.

> I'm writing both processes (though one might be in C) so ORB/CORBA
> seems to be overkill and I can't get through to the memory mapped
> files at 
>         http://pobox.com/~kragen/sw/arrayfrombuffer/
> (Internal server error)

the link works for me. but it seems that you would still need some 
locking/communication mechanism.

> Suggestions?



-- 
Chris <cliechti at gmx.net>



From rayed at saudi.net.sa  Sat Jun  1 05:10:51 2002
From: rayed at saudi.net.sa (Rayed Al-Rashed)
Date: Sat, 01 Jun 2002 12:10:51 +0300
Subject: Interactive socket connection
Message-ID: <3CF88F9B.1080308@saudi.net.sa>

Hello,

I am trying to build POP3 proxy, I want it to run from "inetd" so it should 
communicate with the user using regular stdin, stdout.
I built a small code that should run as general proxy for any protocol, but 
I faced the following problem, it doesn't show any output from the server 
until I pass it a command, for example:

-----------------------------------------
user rayed
+OK POP3 pop.mydomain.com server ready
pass mypass
+OK User name accepted, password please
rert 1
-ERR Bad login
quit
-ERR Unknown AUTHORIZATION state command

+OK Sayonara
-----------------------------------------

Any ideas what the problem is?
Thanks in advance,

- rayed




This the proxy code
--------------------------
#!/usr/bin/python
# POP3 proxy

import sys
import string
import socket
import asyncore
import select

class proxy (asyncore.dispatcher):
         def __init__(self, server, port):
                 asyncore.dispatcher.__init__(self)
                 self.create_socket(socket.AF_INET,socket.SOCK_STREAM)
                 self.connect ((server, port))

         def handle_connect (self):
                 pass

         def handle_read (self):
                 buf = self.recv(8012)
                 sys.stdout.write( buf )

         def handle_write (self):
                 buf = sys.stdin.readline()
                 self.send(buf)

         def handle_close (self):
                 print "Closed"

if __name__ == '__main__':
         p = proxy("pop.mydomain.com", 110)
         asyncore.loop()



From cotemark at globetrotter.net  Sun Jun  9 19:58:59 2002
From: cotemark at globetrotter.net (Marc Cote)
Date: Sun, 09 Jun 2002 23:58:59 GMT
Subject: python interacting with other programs
References: <64eb9145.0206041454.4f70e86e@posting.google.com>
Message-ID: <7ZRM8.9358$bt2.116439@charlie.risq.qc.ca>

"Jakob" <da_woym at hotmail.com> wrote in message
news:64eb9145.0206041454.4f70e86e at posting.google.com...
> now I know python can interact with programs like Microsoft wor and
> access and all but has anyone ever tried to interact with Lotus Notes?
> Is it possible?

Jakob,

Google Search: "Lotus Notes Python" gives:

    Using Python to create a command line interpreter for Notes

URL: http://www.dominopower.com/issues/issue200008/command001.html






From maxm at mxm.dk  Mon Jun 17 03:24:47 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 17 Jun 2002 09:24:47 +0200
Subject: I feel stoopid ... this code is to verbose
References: <3D0A1791.3060307@mxm.dk> <mailman.1024074512.2890.python-list@python.org>
Message-ID: <3D0D8EBF.3050804@mxm.dk>

Joseph Wilhelm wrote:
> On Fri, 2002-06-14 at 09:19, Max M wrote:
> <snip> 
> 
>>def stringSlicer(string, chunkSize=3):
>>     chunkList = []
>>     reverseString = list(string)
>>     reverseString.reverse()
>>     for i in range(0, len(string), chunkSize):
>>         chunk = reverseString[i:i+chunkSize]
>>         chunk.reverse()
>>         chunk = ''.join(chunk)
>>         chunkList.append(chunk)
>>     chunkList.reverse()
>>     return '.'.join(chunkList)

> import string
> 
> def stringSlicer( str, chunkSize=3 ):
> 	parts = list( str )
> 	for x in range( ( len( parts ) - chunkSize ), 0, -chunkSize ):
> 		parts.insert( x, "." )
> 	return string.join( parts, "" )
> 
> print stringSlicer( '42000000' )

Yeah ... this was the one I was looking for. Why the h... didn't I do 
this one myself? Well I know why! I didn't take the time to do a 
pseudo-run on paper before coding. Let this be a lesson to me... once 
more :-)

I gotta stop eating all those simple Carbo hydrates. It makes the brain 
implode into uselessness.

Well I made a few stylechanges and it fit me perfectly:

def stringSlicer(str, chunkSize=3):
     parts = list(str)
     for x in range((len(parts) - chunkSize), 0, -chunkSize):
         parts.insert(x, ".")
     return ''.join(parts)

Easy on the eyes and the mind.

I find it really amusing that even in Python there is such a wide array 
of approaches to such a relatively simple and well defined problem.

Thanks to all for the suggestions.


regards Max M



From shalehperry at attbi.com  Wed Jun  5 23:18:04 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 05 Jun 2002 20:18:04 -0700 (PDT)
Subject: simple pattern matching question
In-Reply-To: <20020606022819.GA72435@attbi.com>
Message-ID: <XFMail.20020605201804.shalehperry@attbi.com>

On 06-Jun-2002 Corey G. wrote:
> Gurus,
> 
> How would I match with an item from a list?  I would imagine something
> this simple would not require a regular expression.
> 
> web_servers=('www01','www02','www03')
> 
>         for server_match in web_servers:
>            if (trap == "Service:"server_match "State:down"):
>             print "found match"
> 
> The comparsion should be: if trap == Service:www01 State:down
> 
> server_match prints "as is, but not from the list".  I tried many 
> combinations but decided to leave the most obviously wrong one.
> 

if trap == ("Service:%s State:down" % server_match):
  print "found match"




From phd at phd.pp.ru  Mon Jun 24 05:49:51 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Mon, 24 Jun 2002 13:49:51 +0400
Subject: mimedecode.py version 1.1.4
Message-ID: <20020624134951.D24065@phd.pp.ru>

Hello!

                                mimedecode.py

WHAT IS IT

   Mail users, especially in non-English countries, often find that mail
messages arrived in different formats, with different content types, in
different encodings and charsets. Usually this is good because it allows us to
use apropriate format/encoding/whatever. Sometimes, though, some unification is
desireable. For example, one may want to put mail messages into an archive,
make HTML indicies, run search indexer, etc. In such situations converting
messages to text in one character set and skipping some binary atachmetnts is
much desireable.

   Here is the solution - mimedecode.py.

   This is a program to decode MIME messages. The program expects one input
file (either on command line or on stdin) which is treated as an RFC822 mesage,
and decoded to stdout. If the file is not an RFC822 message it is just piped to
stdout one-to-one. If the file is a simple RFC822 message it is just decoded as
one part. If it is a MIME message with multiple parts ("attachments") all parts
are decoded. Decoding can be controlled by command-line options.


WHAT'S NEW in version 1.1.4
   Fixed a minor bug with decoding of empty headers.


WHERE TO GET
   Master site: http://phd.pp.ru/Software/Python/#mimedecode

   Faster mirrors: http://phd.by.ru/Software/Python/#mimedecode
      http://phd2.chat.ru/Software/Python/#mimedecode

   Requires: Python 2.0+ (actually tested with 2.1 and 2.2),
      configured mailcap database.

   Documentation (also included in the package):
      http://phd.pp.ru/Software/Python/mimedecode.txt
      http://phd.by.ru/Software/Python/mimedecode.txt
      http://phd2.chat.ru/Software/Python/mimedecode.txt


AUTHOR
   Oleg Broytmann <phd at phd.pp.ru>

COPYRIGHT
   Copyright (C) 2001-2002 PhiloSoft Design

LICENSE
   GPL

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From erict at millfilm.co.uk  Thu Jun 27 10:06:31 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Thu, 27 Jun 2002 15:06:31 +0100
Subject: How to get rid the new line
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org>
Message-ID: <3D1B1BE7.D482AE@millfilm.co.uk>

## remove all the leading and tail white space
line = string.strip(aline)
if line:
    # this is an emptyLine


or better

    def   isNotEmpTxt(text):
        empty = re.compile('^\s*$')
        if empty.match( text):
                return 0
        else:
                return 1




SiverFish wrote:

> Anyone tell me how to skip the empty line when you read through the file
> and copy the information in to the list(no empty line )



From aahz at pythoncraft.com  Fri Jun 14 02:31:56 2002
From: aahz at pythoncraft.com (Aahz)
Date: 14 Jun 2002 02:31:56 -0400
Subject: newbie: from .. import *
References: <56e1eff0.0206131437.4ca8ed9e@posting.google.com>
Message-ID: <aec2ks$t5$1@panix1.panix.com>

In article <56e1eff0.0206131437.4ca8ed9e at posting.google.com>,
=?ISO-8859-1?Q?Jean-S=E9bastien_Bolduc?= <jseb at cs.mcgill.ca> wrote:
>
>I've been working with Python for a while, but never really felt
>confortable with the way namespaces are imported.

What you're dealing with isn't, strictly speaking, a namespace issue but
a scope issue.  Python uses lexical scope to determine visibility of
module globals and locals, so that static analysis of source can
determine visibility of names.  Therefore, when you define a function,
its scope stays with the module where you defined the function.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From gerhard.haering at gmx.de  Fri Jun 21 03:47:46 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 21 Jun 2002 09:47:46 +0200
Subject: compile python witout tkinter
In-Reply-To: <Pine.LNX.4.44.0206210124060.585-100000@aesaeion.com>
References: <Pine.LNX.4.44.0206210124060.585-100000@aesaeion.com>
Message-ID: <20020621074746.GA765@lilith.my-fqdn.de>

* kosh at aesaeion.com <kosh at aesaeion.com> [2002-06-21 01:26 -0600]:
> On a machine I need to build python on I need to build it without tkinter.

Why? Even if Tkinter gets compiled, it doesn't change anything for the
rest. You could even delete the tkinter modules afterwards, if you're
short on disk (quota).

> How can I do that. I didn't notice any configure options to do that
> and I also looked at Modules/Setup and that seems to have _tkinter
> commented out.

That's all there is to it.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From mdk at mdk.com  Thu Jun 20 14:05:50 2002
From: mdk at mdk.com (mdk)
Date: 20 Jun 2002 18:05:50 GMT
Subject: Proper number alignment
Message-ID: <aet5hu$9leln$1@ID-98166.news.dfncis.de>

I am using \t\t (tabs) to align these numbers (disregard the math):

Total RAM                2,141,056
Available RAM            12,934,032
Used RAM                 1,024

But they are left justified and I need them right justified.

Any ideas?

Thanks.


From mhuening at zedat.fu-berlin.de  Wed Jun 26 13:14:10 2002
From: mhuening at zedat.fu-berlin.de (Matthias Huening)
Date: 26 Jun 2002 17:14:10 GMT
Subject: how to remove HTML  attributes from web pages ( HTML parseing)
References: <63170f57.0206260345.2b61b7ec@posting.google.com>
Message-ID: <Xns9239C3A3F43E9mhueningzedatfuberli@130.133.1.4>

sanjay2kind at yahoo.com (sanjay) wrote in 
news:63170f57.0206260345.2b61b7ec at posting.google.com:

> Hi,
> 
>  New to python and doing one application. i would like to  web page
> content after removing specific html tag, attribute etc. 

Below you find some sample code that uses htmllib.
Hope this helps. 
- Matthias

-----------

import htmllib, formatter, StringIO, urllib

class HTMLStripper(htmllib.HTMLParser):
    def __init__(self):
        self.bodytext = StringIO.StringIO()
        writer = formatter.DumbWriter(self.bodytext)
        htmllib.HTMLParser.__init__(self,
            formatter.AbstractFormatter(writer))

    def anchor_end(self):
        if self.anchor:
            self.handle_data('')
            self.anchor = None

    def gettext(self):
        return self.bodytext.getvalue()

f = urllib.urlopen('http://www.python.org/').read()       
st = HTMLStripper()
st.feed(f)
st.close()
print st.gettext()
print st.anchorlist

-----------


From maxm at mxm.dk  Fri Jun 14 12:19:29 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 14 Jun 2002 18:19:29 +0200
Subject: I feel stoopid ... this code is to verbose
Message-ID: <3D0A1791.3060307@mxm.dk>

Hmm ... I am working on a problem. In Danish we have a number format 
that looks like:

42.000.000,00 which is 42 millions

So I need to insert dot's at every three character from end of the 
string for the integer value of the number.

Let's discard decimal points and just focus on the meat. I have written 
a funcion which works::

def stringSlicer(string, chunkSize=3):
     chunkList = []
     reverseString = list(string)
     reverseString.reverse()
     for i in range(0, len(string), chunkSize):
         chunk = reverseString[i:i+chunkSize]
         chunk.reverse()
         chunk = ''.join(chunk)
         chunkList.append(chunk)
     chunkList.reverse()
     return '.'.join(chunkList)


print stringSlicer('42000000')

 >>> 40.000.000

I just find that it's a lot of code for such a little function an it 
annoys my sense of aestetics. I have tried a lot of different approaches 
including using zip on a list like ['','','.'], and other weird stuff :-)

I just cannot seem to find the nice 3-liner I expected from the 
beginning. Has anybody got a better solution ? I thought somebody might 
find it a fun exercise. Well I have...


regards Max M



From listuser at nicetomeetyou.to  Fri Jun 28 01:52:38 2002
From: listuser at nicetomeetyou.to (Markus)
Date: Fri, 28 Jun 2002 07:52:38 +0200
Subject: crypt for windows?
References: <mailman.1025241336.1209.python-list@python.org>
Message-ID: <afgtj8$us1$07$1@news.t-online.com>

Klaus Reinhardt wrote:

> Am 27.06.02 19:29:52, schrieb Gerhard H?ring <gerhard.haering at gmx.de>:
> 
>>> Is there a crypt for windows-python, so
>>> I can manage some password-files?
>>
>>Probably not. I believe Cygwin's Python has this module. Is that an
>>option for you?

I use a module named fcrypt. It is a pure Python implementation and works 
fine for Unix *and* Windows.

I'm sure a search on Parnassus or Google will find it.

--Markus



From tim.one at comcast.net  Thu Jun 27 23:47:01 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 27 Jun 2002 23:47:01 -0400
Subject: Python 2.2 __slots__ Bug?
In-Reply-To: <B940C17E.D101%jonathan@onegoodidea.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEKGAAAB.tim.one@comcast.net>

[Jonathan Hogg]
> It appears that GjR is on the case ;-)

Na, I assigned the bug to GvR just to irritate him <0.7 wink>:  he wrote the
__slot__ deallocation code that's blowing up, yet has also called the funky
trashcan mechanism (which you described well later) "evil" more than once.
The irritating part is that the evil trashcan mechanism is the only hope of
avoiding stack overflow in absurd cases like this.

> ...
> I can't see an easy way to fix this without changing the way that the
> "delete later" list is constructed.

Sometimes we just agree to pretend that we never saw the bug report <wink>.

more-than-one-way-to-close-a-bug-ly y'rs  - tim





From osuchw at ecn.ab.ca  Fri Jun 14 02:03:40 2002
From: osuchw at ecn.ab.ca (waldekO)
Date: 13 Jun 2002 23:03:40 -0700
Subject: Help with ADO and dynamic insert statement...
References: <3d0826ec.240888125@netnews.attbi.com>
Message-ID: <c5b3dbee.0206132203.793dc5a3@posting.google.com>

You should at least post a traceback if you expect a resonable answer
to your question.
 
candiazoo at attbi.com wrote in message news:<3d0826ec.240888125 at netnews.attbi.com>...
> I have a program written in VB that creates an ado COMMAND object, sets the text
> to a dynamic sql string (insert into table (?, ?, ?, etc))...  it works in VB...
> this is my first foray into doing something similar with Python and it isn't
> working.  Could someone take a gander at this and tell me what I might be doing
> incorrectly?
> 
> Thanks!
> 
> Mike
> 
> ============================================
> segment of code below
> ============================================
> 
>     if status == SUCCESS:
>         #   Generate our dynamic insert statement, create new
>         #   ADODB connection and command objects which we'll
>         #   use to connect to our access database (via the
>         #   CatalogMDB DSN object)...
> 
>         sql = "insert into works (title, wrk_inst, trs_inst, aas_inst) " + \
>               "values (?, ?, ?, ?)"
>     
>         cnnTempDB = win32com.client.Dispatch('ADODB.Connection')
>         cmdInsert = win32com.client.Dispatch('ADODB.Command')
>         
>         cnnTempDB.Open(TEMP_DB)
>         cmdInsert.ActiveConnection = cnnTempDB
>         cmdInsert.CommandType = 1
>         cmdInsert.CommandText = sql
>         
>         if not rstOracle.EOF:
>             #   Let's bind up our dynamic sql statement...
>             
>             cmdInsert.Parameters.Append(cmdInsert.CreateParameter("title", 200,
> 1, len(rstOracle.Fields("title").Value), rstOracle.Fields("title").Value))
>             cmdInsert.Parameters.Append(cmdInsert.CreateParameter("wrk_inst",
> 131, 1, 4, rstOracle.Fields("wrk_inst").Value))
>             cmdInsert.Parameters.Append(cmdInsert.CreateParameter("trs_inst",
> 131, 1, 4, rstOracle.Fields("trs_inst").Value))
>             cmdInsert.Parameters.Append(cmdInsert.CreateParameter("aas_inst",
> 131, 1, 4, rstOracle.Fields("aas_inst").Value))
>             cmdInsert.Execute()
>             rstOracle.MoveNext()


From mcfletch at rogers.com  Tue Jun  4 12:03:57 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 04 Jun 2002 12:03:57 -0400
Subject: Looking for a list subclassing example...
References: <adi4dv$118tko$1@ID-146704.news.dfncis.de>
Message-ID: <3CFCE4ED.5000105@rogers.com>

Well, here's a generic list-subclass example...

class rlist( list ):
     """Sub-class of list which calls a method before any addition is 
allowed"""
     def __setslice__( self, start, stop, value ):
         """__setslice__ with value checking"""
         value = self.beforeMultipleAdd([ self.beforeAdd(item) for item 
in value ])
         return list.__setslice__( self, start, stop, value )
     def extend( self, value ):
         """extend with value checking"""
         value = self.beforeMultipleAdd([ self.beforeAdd(item) for item 
in value ])
         return list.extend( self, value )
     __iadd__ = extend
     def append( self, value ):
         """append with value checking"""
         value = self.beforeAdd( value )
         return list.append( self, value )
     def insert( self, index, value ):
         """insert with value checking"""
         value = self.beforeAdd( value )
         return list.insert( self, index, value )
     def __setitem__( self, index, value ):
         """__setitem__ with value checking"""
         value = self.beforeAdd( value )
         return list.__setitem__( self, index, value )
     def beforeAdd( self, value ):
         """Called before all attempts to add an item"""
         return value
     def beforeMultipleAdd( self, value ):
         """Called before attempts to add more than one item (beforeAdd 
has already be called for each item)"""
         return value

HTH,
Mike


Shagshag13 wrote:
> Hello,
> 
> I'm looking for a list subclassing example, and i can't find it...
> 
> Do you have one to show ?
> 
> Thanks,
> 
> s13.
> 
> 
> 


-- 
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From dan at eevolved.com  Mon Jun  3 14:37:34 2002
From: dan at eevolved.com (Daniel Parisien)
Date: Mon, 03 Jun 2002 14:37:34 -0400
Subject: encoding and decoding repr'd version of strings
References: <mailman.1023127086.24340.python-list@python.org>
Message-ID: <hKOK8.15289$MH4.610679@weber.videotron.net>

Gustavo Cordova wrote:

>> 
>> Hi
>> is there an codec that bundles with python so you can
>> transform strings into the repr'd (escaped) version and
>> back? If not, how easy would it be to code? I don't want
>> to use eval(...) because I don't trust the source of the
>> string and I might mistakenly eval an expression like "'
>> '*(2**30)" which would allocate about 1 GB of data (eek!)
>> 
>> Thanks,
>> Dan
>> 
> 
> I might be going out on a limb here, but how about using pickle
> or cpickle?
> 
> -gustavo

Valid suggestion, but it's a lot of overhead for naught (I just want to 
store escaped strings, not data of all types). It's also a little too 
python-specific.

Dan


From peter at engcorp.com  Thu Jun 20 21:22:40 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 20 Jun 2002 21:22:40 -0400
Subject: Dispatching a Win32 COM in the background?
References: <mailman.1024531044.10120.python-list@python.org> <9OpQ8.90552$ks6.8015@nwrddc02.gnilink.net>
Message-ID: <3D127FE0.74D47795@engcorp.com>

"Steven M. Castellotti" wrote:
> 
> Thanks a lot guys, your suggestions have worked perfectly!
> 
>         Just for the record, here's what the full implementation looks like.

Beautiful... thanks for the followup Steven.  I especially like
your use of deferred imports to speed up initial program loading.

[...]
>       if not(self.text_queue.empty()):
>          text = self.text_queue.get()
>          speech.Speak(text)
>          time.sleep(.001)

Note that the time.sleep() as shown above is doing nothing
useful for you.  I suspect you had it unindented but your
posting probably used tabs and spaces which makes it hard
to see properly on some readers.

The reason it does nothing useful is that Python will 
automatically switch between different threads every so
often anyway, so the concept of giving up the CPU for 
what amounts to a tiny fraction of a second is somewhat
unnecessary.

-Peter


From dbasch at yahoo.com  Sat Jun 29 23:14:23 2002
From: dbasch at yahoo.com (Derek Basch)
Date: Sat, 29 Jun 2002 20:14:23 -0700 (PDT)
Subject: Win32/Python serial port event character problem
Message-ID: <20020630031423.93113.qmail@web20806.mail.yahoo.com>

Thanks for the ideas Peter and Chris. I would use
pySerial but I have spent the last few months building
my own serial module that is very similiar to pySerial
but with some extra features. It was probably silly of
me to build my own from scratch but it was a good
challange and I certainly understand serial ports much
better now. So you think I should ditch the event
character and just read a preset number of characters
from the buffer on each loop? That sounds like a good
idea for OS compatability. I will use something like
this (from pySerial):
http://www.geocities.com/dbasch/tempers.txt

I dont understand what Chris is doing in the middle
though. Maybe you can answer that Chris?

I hadn't had a chance to write any code for detecting
the termination character and breaking up the data but
your pseudocode was just what I needed Peter.
Thanks!,
Derek Basch

P.S. Next time I will be more patient. Need to drink
less tea ;)  

Derek Basch wrote:
> 
> If anyone reads this and can show me a better way to
> work with serial event characters in win32 please
feel
> free to :)

I second Chris' comments about using something like
PySerial
and just handling this in a thread. Event characters
are
probably non-portable, while you could easily write
code
with PySerial that runs under Linux or Windows.

But give it time... you definitely didn't give it
enough time
before you reposted. :)

By the way, here's some typical (?) code for handling
a termination 
sequence on incoming data for something like this
(while inside of
some kind of serial port object):

def read(self, timeout=None):
buf = self.leftover
self.leftover = ''

while true:
data = FUNCTION_TO_READ_A_RAW_BLOB_OF_DATA()

# if new data received, add to the buffer and check
for
# a contained input termination sequence
if data:

buf = buf + data

termIndex = string.find(buf, self.inputTerminator)
if termIndex >= 0:
# move past the terminator
termIndex = termIndex + len(termIndex)

# preserve anything lying after input termination 
# sequence in the leftover buffer for next time
buf, self.leftover = buf[:termIndex], buf[termIndex:]

break

return buf

Each time you call it, it reads raw data and adds it
to a buffer
which it then rescans for the termination sequence. If
the buffer
contains the sequence, it splits it on that and
returns everything
up to and including the sequence, preserving
everything after it 
for the next call.

-Peter

__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com




From db3l at fitlinxx.com  Thu Jun  6 21:58:42 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 06 Jun 2002 21:58:42 -0400
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet> <3CFE23DB.994F8A6C@engcorp.com> <3CFFC67E.3070609@bgb.cc>
Message-ID: <uhekfu9wt.fsf@ctwd0143.fitlinxx.com>

Don Garrett <garrett at bgb.cc> writes:

>     It would be even more useful if there was an easy way to produce a
> standalone RPM or MSI file (depending on the plateform). The MSI file is
> probably more useful. The unix people are more capable of dealing with the
> extra steps of getting a python program to work.

While not MSI, there are free setup programs (I'm a fan of InnoSetup)
that will produce a perfectly standard looking Windows installation
executable for your application.  You can couple that with the
installer/py2exe output to look like a standard Windows setup.

If you're already doing this and just specifically want an MSI format,
then ignore the point. :-)

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From -$P-W$- at verence.demon.co.uk  Sun Jun  9 13:15:17 2002
From: -$P-W$- at verence.demon.co.uk (Paul Wright)
Date: 9 Jun 2002 18:15:17 +0100
Subject: Python MTA
References: <mailman.1023489912.30571.python-list@python.org>
Message-ID: <ae02f5$5fn$1@verence.demon.co.uk>

In article <mailman.1023489912.30571.python-list at python.org>,
David LeBlanc <whisper at oz.net> wrote:
>Has anyone developed a simple/stupid Mail Transfer Agent using Python
>that does't rely on any external mail agents (and thus would be
>platform independent)?
>
>What I have in mind is an app that sits between my mail client (outlook
>(dev/null=flames)) and, at most, a few mail hosts. It would act like a
>client to those mail hosts and as a host to outlook. The idea is to be
>able to insert a spam filter like ASK (or something better?).

You don't want an SMTP MTA then, you want a filtering POP3 proxy.

I have the beginnings of such a thing in Python but my motivation to
finish it kind of went away when SpamPal was released:
http://www.twinlobber.org.uk/spampal/

It is not written in Python and the source is not open, but I'm told it
works well.

Should I ever finish my proxy I'll let this group know, I should think.

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


From schaffer.SPAM at optonline.net  Thu Jun 20 16:41:23 2002
From: schaffer.SPAM at optonline.net (Les Schaffer)
Date: Thu, 20 Jun 2002 16:41:23 -0400
Subject: Problems with mayavi, pyvtk
References: <7396d2b2.0206191025.3351a888@posting.google.com> <aeqme8$avd$1@peabody.colorado.edu> <7396d2b2.0206201046.4f0530fa@posting.google.com> <aet89k$jfh$1@peabody.colorado.edu>
Message-ID: <aeteln$9n9tu$1@ID-147588.news.dfncis.de>

Fernando P?rez wrote:

> Keep in mind that Prabhu doesn't use windows either (at all). Maybe
> someone on the mayavi list does use windows and can lend you a hand
> though

I am in the process of updating the Python 2.2 DLLs for VTK on my page in 
the next several days. so if there are problems getting it to work, write 
me directly and we'll get them fixed.

Les Schaffer



From peter at engcorp.com  Sat Jun  1 14:03:45 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 01 Jun 2002 14:03:45 -0400
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org> <3CF8E382.3A44F0DA@engcorp.com> <slrnafhtdv.407.ak@ak.silmarill.org>
Message-ID: <3CF90C81.5A141A97@engcorp.com>

Andrei Kulakov wrote:
> 
> > On my blazing 233 MHz Windows 98 machine, with Python 2.2.1
> > and WxPython 2.3.2.1, a hello world program I just wrote
> > runs in about two seconds, for what it's worth.
> >
> > -Peter
> 
> Odd.. if not too much trouble, could you please try this one
> (from wxpython tutorial)?
> 
[snip short hello world]

First time running: 4 seconds.

Subsequent invocations: 2 seconds.

-Peter


From kragen at pobox.com  Tue Jun  4 18:37:48 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 04 Jun 2002 18:37:48 -0400
Subject: Problems with exec scoping (solved)
References: <adep6r01ntg@enews1.newsguy.com> <aderf60vsg@enews2.newsguy.com>
Message-ID: <83it4y4qmb.fsf@panacea.canonical.org>

Erik Walthinsen <omega at temple-baptist.com> writes:
> Well, it turns out that I was being an idiot and not checking the source
> of the returned HTML page.  When repr() is run on an object, it returns a
> string such as:
> <captiveclient.CaptiveClient instance at 0x8189264>
> ...
> I will now go and see about writing a HTML-textizer for use in such
> situations.

Take a look at
http://lists.canonical.org/pipermail/kragen-hacks/2002-January/000291.html
--- it's an HTML version of repr() called htmlstr().  It generates
CSS-skinnable HTML that renders builtin types much more nicely than
the builtin repr() or str() do, using tables to make things wrap
nicely.

There's also HTMLRepr from pydoc.



From hesterloli at hotmail.com  Sat Jun 15 18:58:36 2002
From: hesterloli at hotmail.com (George Hester)
Date: Sat, 15 Jun 2002 22:58:36 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de>
Message-ID: <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com>

Python ActiveX Scripting Engine error '80020009'

Traceback (innermost last): File "<Script Block >", line 5, in ?
Response.Write(sys.version) NameError: name 'sys' is not defined

/python.asp, line 14

Response.Write(sys.version)

I think I need a script to do this.

--
George Hester
_________________________________
"Jeremy Yallop" <jeremy at jdyallop.freeserve.co.uk> wrote in message
news:aegfck$6tkre$1 at ID-114079.news.dfncis.de...
> * George Hester
> | What is a simple script that will tell me the version of python I am
currently
> | running?  IIS 5.0 Windows 2000.
>
> Look up sys.version and sys.version_info.
>
> Jeremy.




From mwh at python.net  Tue Jun 18 07:35:38 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 18 Jun 2002 11:35:38 GMT
Subject: Listing all available (not just loaded) modules
References: <aem9om$888fb$1@ID-140151.news.dfncis.de>
Message-ID: <lkd6uolt0q.fsf@pc150.maths.bris.ac.uk>

"Michael Steuer" <mike at radiomag.com> writes:

> I've been searching the newsgroup for information on this but so far
> have been unsuccessful:
> 
> Is there a way to dynamically list all available modules (not just those that
> are already loaded) in python? 

In general, no, because of __import__ hooks and the like.

> By available I mean those modules, which are
> installed on the machine executing python (in the lib path(s)).

pyrepl contains some nasty code to do this, if you get stuck with the
other packages you were pointed to.

Really must do some work on pyrepl soon.

Cheers,
M.

-- 
  MGM will not get your whites whiter or your colors brighter.
  It will, however, sit there and look spiffy while sucking down
  a major honking wad of RAM.              -- http://www.xiph.org/mgm/


From hwcowan at hotmail.com  Fri Jun  7 10:43:12 2002
From: hwcowan at hotmail.com (Hugh Cowan)
Date: 7 Jun 2002 07:43:12 -0700
Subject: Creating Dynamic Web Pages (ZOPE)
References: <46ca81a0.0206062016.789d3af6@posting.google.com> <mailman.1023424390.6172.python-list@python.org>
Message-ID: <46ca81a0.0206070643.3d1a0856@posting.google.com>

Thanks Michael for the reply and information.  It's good to know that
I am not "barking" up the wrong tree.  Now that I know I am headed in
the right direction I will do more reading and research and check-out
their web-site.  I just didn't want to spend a lot of time up front
just to find out in the end that it wouldn't work for me.

Hugh,



Michael Gilfix <mgilfix at eecs.tufts.edu> wrote in message news:<mailman.1023424390.6172.python-list at python.org>...
> Hiya!
> 
> On Thu, Jun 06 @ 21:16, Hugh Cowan wrote:
> > I don't know if ZOPE would fit my situation or not.  I know that it is
> > / can be used to create complex E-commerce sites, but I am wondering
> > if it is overkill for what I need done.  Would I be better off trying
> > to find some sort of RAD Web Tool instead (anyone have any good
> > suggestions) ?
> 
>   I think Zope is definitely the way to go as it is designed to
> address exactly what you need and easily. Zope allows you to add
> 'drop-in' components to your web service (such as an SQL query,
> or called a 'ZSQL query' in Zope-speak), which tends to be a big
> time saver, and then reference that query to dynamically produce a
> webpage. More so, it allows for easy remote management with delegation
> of roles and authority to multiple parties; seems like a good way to
> structure your intranet.
> 
>                      -- Mike


From noone at here.invalid  Sun Jun 23 20:30:16 2002
From: noone at here.invalid (JohnJacob)
Date: Sun, 23 Jun 2002 19:30:16 -0500
Subject: ConfigParser & converting strings to lists
References: <3D152A28.2177B054@tds.net> <3D159F06.2154D3F3@tds.net> <Xns9236975A47F67cliechtigmxnet@62.2.16.82>
Message-ID: <af5p6l0pf2@enews4.newsguy.com>

"Chris Liechti" <cliechti at gmx.net> wrote in message
news:Xns9236975A47F67cliechtigmxnet at 62.2.16.82...
> "Edward K. Ream" <edream at tds.net> wrote in news:3D159F06.2154D3F3 at tds.net:
>
> > To complete this topic, let me restate the problem and JohnJacob's
> > elegant solution.
> >
> > Suppose files is a list of strings, and we have written a configuration
> > section like this:
> >
> >   config = ConfigParser.ConfigParser()
> >   config.set("recent files", "recentFiles, files)
> >
> > Then the elegant way of retrieving files _as a list_ is:
> >
> >   files = eval(config.get("recent files", "recentFiles"))
>
> it isn't that elegant i tend to call that sloppy ;-) the problem is that
> you execute arbitrary python expressions and you hope to get a list. but
> that is not sure, the user could have edited the config file.
> exec and eval are always a risk in your code. e.g they could rise any
> exception because you don't know what they realy execute.
>
> the split version is much more secure. you could also try a pickle. witch
> pickle.dumps you can get a string and reterive the old object using
loads()
> but that isn't as secure as split.
>
> if thats all too complicated, use your version with individual names
> "file1", "file2", ... it's much saver for a progrom that you release in
the
> wild.
>
> chris
>
> PS you mentioned you see double backslashes. maybe you looked at the repr
> of a string:
> >>> s = "he\tllo"
> >>> repr(s)
> "'he\\tllo'"
> >>> str(s)
> 'he\tllo'
> >>> print s
> he llo
> >>>
>
> as you can see there is not realy a double backslash...
>
> --
> Chris <cliechti at gmx.net>


Ugh! Not elegant? Seriously, though, Chris is right. I wouldn't distriute
code like that. It is dangerous in the security sense and in the stability
sense. For scripting your own system I wouldn't think that would be too big
of a problem.

If it's going to be distributed I would go with the split() method others
have suggested.


greg

P.S. Do you have an Erdos number? I've never talked to anyone who actually
has one (that I know of).




From djmitchell at NOSPAM.optushome.com.au  Fri Jun 21 05:27:15 2002
From: djmitchell at NOSPAM.optushome.com.au (David Mitchell)
Date: Fri, 21 Jun 2002 19:27:15 +1000
Subject: I'd give up Perl tomorrow if only...
Message-ID: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au>

.. there was something equivalent in scope and breadth to CPAN.

More and more I find myself in positions where I can use pretty well
whatever language I like, and I'm increasingly drawn to Python.  It's
elegant, it's easy to support, development time is really fast, ... (you
guys know the rest!)

However, I'm probably equally proficient in Perl, and Perl has this
incredible resource in CPAN.  For just about any library module I've ever
wanted, I can download one or more modules from CPAN that gets me 90% of the way
there.  Perl modules like DBI/DBD, SOAP::Lite, GD::Graph, MQSeries::*,
Tie::*, Convert::EBCDIC, and so on (and on) make me exceptionally
productive in the eyes of the people I work for, and tend to just destroy
any reservations about using Perl in those organizations.

Don't get me wrong: Python also has some outstanding modules that are
reliable, powerful and make life much easier as well.  I also enjoy using
Python much more than Perl these days, for most of the well-documented
reasons.  There just isn't any Python resource with the breadth
of coverage of what's at CPAN.

If there was only some way to subclass Perl/CPAN modules as Python
modules (i.e. similar to how you can subclass Java modules and use them
from Jython), I'd jump on it in a heartbeat.  For the vast majority of what I
do, performance isn't that big a deal, so I'd be happy to live with some
cross-language delay.  I've been hoping something along these lines might
arise out of the Parrot initiative, but so far it hasn't.

Is ANYONE working on this, or thinking about working on it, or knows
someone who's thinking about working on it, or dreaming about it, or ...?

Regards

David Mitchell


From dgallion1 at yahoo.com  Fri Jun 21 08:28:06 2002
From: dgallion1 at yahoo.com (Darrell)
Date: 21 Jun 2002 05:28:06 -0700
Subject: Looking for Python tools for extracting information from PDF files.
References: <mailman.1024567408.12098.python-list@python.org>
Message-ID: <b0c5da3a.0206210428.7648bd28@posting.google.com>

Yes Pyrex is cool also. 
But I Plex is the tool I used to parse PDF.

Pyrex + distutils is a must show friends and family.
Ok most people aren't too impressed to see Python and C in the same
file converted to a .pyd with minimal effort. But it sure tickles me
:)

--Darrell


From hancock at anansispaceworks.com  Wed Jun 19 13:53:25 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Wed, 19 Jun 2002 10:53:25 -0700
Subject: Unicode hexadecimal characters
References: <20020620015059.11518.52464.Mailman@mail.python.org>
Message-ID: <3D10C515.463EF9D7@anansispaceworks.com>

Take a look at http://www.unicode.org/ for more information,
I was able to download the v3.2.0 UCD and scan it (I was
curious too).  I just searched it for the strings "DIGIT",
"HEX", and "NUMBER". 

The answer appears to be "No", though, which I think is
kind of a funny oversight.  Surely, HEXADECIMAL DIGIT A
would be at least as conceptually different from, say,
LATIN CAPITAL LETTER A as that is from CYRILLIC CAPITAL LETTER A.
Yet there are separate characters for the latter two, but it looks like 
LATIN CAPITAL LETTER A has to do double duty as HEXADECIMAL DIGIT A.

You could of course use one of the CIRCLED, DOUBLE CIRCLED,
FULL STOP or other "NUMBER" characters this way, but it
would probably be considered pretty nonstandard.

Makes me wonder what the procedure is for suggesting a new
code block or if it's already been proposed and rejected
for some reason.  I suppose a search on the site above would
probably reveal this.

> "Peter Hansen" <peter at engcorp.com> wrote in message
> news:3D112C60.9C3D2F9F at engcorp.com...
> > Thinkit wrote:
> > > It's a bit OT but does Unicode support the digits ten through fifteen
> > > (for hexadecimal)?
> >
> > Do you mean "does Unicode have unique glyphs representing the
> > hexadecimal digits ten through fifteen other than the standard
> > A, B, C... ?"
> 
> Yes, that's what I mean.  I looked at the spec and saw a lot of different
> numerals, including 1-20 for circled numbers.  Will have to examine it
> again.
> 
> > Unicode of course supports various glyphs for A through F, but
> > it has nothing to do with other representations of bases other
> > than ten and under, as far as I know.
> >
> > Not sure if that's exactly what you meant...

BTW...
Just in the spirit of being really pedantic: this post
confuses the term "glyph" (which should mean the graphical
representation of a character) with "character" which
should be used for the conceptual atomic element of
information. Unicode doesn't support *any* glyphs, just
characters -- the *font* supports glyphs.  What Peter Hansen
probably meant would be better phrased as "the glyph A may
be represented by various unicode characters".  Technically,
of course, this is only strictly correct when the character
A that I'm sending you is rendered into a glyph by your email
client and window manager (which actually might render
"LATIN CAPITAL LETTER A" and "CYRILLIC CAPITAL LETTER A"
as different glyphs, even though mine uses the same one).

Of course, this only matters if you want to speak Unicode
jargon -- the English language is more forgiving. ;-)

But I guess it is off topic after all.

Cheers,
Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
Anansi Spaceworks                 
http://www.anansispaceworks.com 
P.O. Box 60583                     
Pasadena, CA 91116-6583
------------------------------------------------------




From tim.one at comcast.net  Sat Jun 29 00:22:08 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 29 Jun 2002 00:22:08 -0400
Subject: Python needs better error reporting
In-Reply-To: <GCEDKONBLEFPPADDJCOEEEOMDMAA.whisper@oz.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEODAAAB.tim.one@comcast.net>

>>> "Syntax Error: invalid syntax" isn't very informative, yet it's
>>> tossed out frequently. HOW is the syntax invalid?  For example:
>>>  if s == ' '

>> That is supposed to start a new block so it should be
>> if s == ' ':

> That's not the point - I know that.

Or, IOW, SyntaxError is all the clue you actually needed.  If the parser
were to tell you what *it* thinks is wrong, it would be along the lines of
"the only arcs leading out of this state are 23, 167, and 528".  Then you'd
really have a point <wink>.





From jepler at unpythonic.net  Mon Jun 10 21:37:34 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Mon, 10 Jun 2002 20:37:34 -0500
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
In-Reply-To: <Xns922AD3C74280gustafl@195.100.94.182>
References: <Xns9229EE722467Fgustafl@195.100.94.182> <mailman.1023746539.13373.python-list@python.org> <Xns922AD3C74280gustafl@195.100.94.182>
Message-ID: <20020610203728.A8589@unpythonic.net>

On Mon, Jun 10, 2002 at 11:16:46PM +0000, Gustaf Liljegren wrote:
> Cliff Wells <logiplexsoftware at earthlink.net> wrote: 
> 
> Thanks both of you. I tried both solutions. Since some intervals are very 
> high (1000000+), it takes too long to add them to a list. Cliff's solution 
> was really fast, but it took a while to understand. :)
> 
> Gustaf

You can do one better by keeping the list sorted.  I'd personally express
the ranges in the half-open way that range() does it, and always store
tuples, so that
    ((1,3), (5,100), (200,201))
means exactly 1, 2, 5, ..., 99, 200.

This means that you can use a Python implementation of a priority queue to
manage the lists (since all items are comparable to each other with the
normal Python comparison operator) and you can test for membership in O(lg
n) where n is the number of intervals.

As an added bonus, something like
    reduce(operator.add, map(range, intervals))
would give you an expanded Python list of all the numbers contained.

The following code does something similar to what you want .. I just wrote
_find, so it's probably wrong.  The rest has been slightly tested, though.
This is not a proper priority queue implementation, because insertions
would appear to be O(n), not O(lg n) as they should be (list.insert to the
middle of the list has to move len()/2 pointers to complete the move).
Similarly, the insertion procedure doesn't coalesce neighboring or
overlapping ranges, you have to call clean() to do that.

def _find(l, n):
    lo = 0
    hi = len(l)
    while lo < hi:
        mid = (lo+hi)/2
	a, b = l[mid]
	if a <= n < b: return 1
        elif n < a: hi = mid
        else: lo = mid+1
    return 0

def _insort(l, i):
    lo = 0
    hi = len(l)
    while lo < hi:
        mid = (lo+hi)/2
        if i < l[mid]: hi = mid
        else: lo = mid+1
    l.insert(lo, i)

class RangeList:
    def __init__(self):
        self.l = []

    def addrange(self, low, hi):
        _insort(self.l, (low, hi))

    def clean(self):
        l = []
        if not self.l: return
        ol, oh = self.l[0]
        for (nl, nh) in self.l[1:]:
            if nl >= oh+1:
                l.append((ol, oh))
                ol, oh = nl, nh
            else:
                oh = nh
        l.append((ol, oh))
        self.l = l

    def get(self):
        return self.l

    def __contains__(self, item):
	return _find(self.l, item)

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




From bokr at oz.net  Mon Jun  3 04:45:15 2002
From: bokr at oz.net (Bengt Richter)
Date: 3 Jun 2002 08:45:15 GMT
Subject: create a tmp file for system execution
References: <3CF4E711.8A175798@millfilm.co.uk> <ad3bkt$83b$0@216.39.172.122> <ad3e34$223s$1@nntp6.u.washington.edu>
Message-ID: <adfaar$6vr$0@216.39.172.122>

On 29 May 2002 20:35:48 GMT, "Donn Cave" <donn at u.washington.edu> wrote:

>Quoth bokr at oz.net (Bengt Richter):
>...
>| wouldn't it be faster to do a single ls F.* command piped to a grep for
>| the decimal extensions, then to a python script for sorting and final formatting
>| or whatever you want to do with the file names? Or am I missing something?
>
>You might have missed Jurgen Hermann's observation that something
>like this might be done most easily directly in Python.  If the
>question really is only "how can I list these files", then I'd go
>with os.listdir() and fnmatch.fnmatch().
>
Yes. Looks like our mail crossed too.

Regards,
Bengt Richter


From ark at research.att.com  Fri Jun  7 12:49:48 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 7 Jun 2002 12:49:48 -0400 (EDT)
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
	(message from Skip Montanaro on Fri, 7 Jun 2002 11:42:19 -0500)
References: <yu99vg8v6orb.fsf@europa.research.att.com> <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
Message-ID: <200206071649.g57GnmC00708@europa.research.att.com>

Skip> My guess is you have a new libgcc.so file that needs to be
Skip> introduced to the system using the ldconfig command.  Try
Skip> running ldd on the struct.so file and see if it finds everything
Skip> the file references.

That could be it, but I have been setting LD_RUN_PATH explicitly when
building Python, so I don't understand why it would have changed.




From peter at engcorp.com  Wed Jun 26 00:23:22 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 26 Jun 2002 00:23:22 -0400
Subject: OT: autocoder redux (was Re: Reminder of a python Project.)
References: <2218.941T781T12224051threeseas@earthlink.net> <3D193728.663CA774@engcorp.com> <1708.941T2398T14045676threeseas@earthlink.net>
Message-ID: <3D1941BA.7AFF5C92@engcorp.com>

Timothy Rue wrote:
> 
> On 25-Jun-02 22:38:16 Peter Hansen <peter at engcorp.com> wrote:
> >I'm afraid there is no longer any need for an autocoder.
> 
> Geee, maybe you should tell that to NIST

No need.  They called me the other day and I let them know, but
it was too late to get it into the Computerworld article.

Seriously, there's really no issue with software development any
more.  The problems are solved.  I guess if you've invented the
world's first autocoder, it kinda sux to be you then, eh?

-Peter


From david at dataovation.com  Fri Jun 28 08:53:00 2002
From: david at dataovation.com (David McInnis)
Date: Fri, 28 Jun 2002 05:53:00 -0700
Subject: Dialing out on MODEM
In-Reply-To: <3D1C5B65.6CBA96A3@engcorp.com>
Message-ID: <007501c21ea2$bc298c90$221d7041@bigdaddy>

This is my object

>>> import serial
>>> modem = serial.Serial
>>> modem = serial.Serial(2)
>>> modem.write("atdt3120992\n")

I have also tried with 

>>> import serial
>>> modem = serial.Serial
>>> modem = serial.Serial(2)
>>> modem.write("atdt3120992")

David McInnis

-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On Behalf Of Peter Hansen
Sent: Friday, June 28, 2002 5:50 AM
To: python-list at python.org
Subject: Re: Dialing out on MODEM

David McInnis wrote:
> 
> I need to write a small script that opens com3 and writes an ATDT to
> dial out, answer the line and send a numeric message.  Basically, I
> paging utility.
> 
> I am using pyserial and cannot seem to get my modem to dial with
> 
> Modem.write("ATDT5555555")

Is Modem an object from pyserial, or one of yours?  And does it
automatically append the \n or CR or do you have to do that yourself?

-Peter
-- 
http://mail.python.org/mailman/listinfo/python-list





From phlip_cpp at yahoo.com  Sun Jun  2 19:07:56 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 02 Jun 2002 23:07:56 GMT
Subject: Reading coords from a scrolled Tkinter Canvas widget?
References: <addeiv$df1$1@news6.svr.pol.co.uk> <addjad$3cp@dispatch.concentric.net> <ade0em$1u9$1@newsg1.svr.pol.co.uk>
Message-ID: <ade8gc$3cr@dispatch.concentric.net>

G. Willoughby wrote:

>>Your code appears not to grok canvasx and canvasy. They translate event.x
>>style coordinates to canvas model coordinates.
> 
> Hmmm i know, have you any code to grab the canvasx and canvasy?

x = self.thumbnail.canvasx(event.x)

>>BTW you could be using a Python Megawidgets Pmw.ScrolledCanvas here. More
>>features less pain.
> 
> I could do but im trying to keep imports to a minimum and keep its
> portability.

But everyone else is doing it!

-- 
  Phlip
            http://www.greencheese.org/SonseOne
  --  Creationism: The belief that God is too stupid to be able
      to invent a Universe where life can spontaneously arise  --



From deltapigz at telocity.com  Sat Jun 29 21:44:00 2002
From: deltapigz at telocity.com (Adonis)
Date: Sat, 29 Jun 2002 21:44:00 -0400
Subject: possibly a dumb question
References: <3d1e5eff$1_7@nopics.sjc> <3D1E6415.8128F3F8@engcorp.com>
Message-ID: <3d1e6251$1_3@nopics.sjc>

sure:

class foo:
    def __init__(self, value):
        return value
x = foo(0)
print x ;yeilds 0

i know the code provided is wrong, but its the general idea.

Adonis


"Peter Hansen" <peter at engcorp.com> wrote in message
news:3D1E6415.8128F3F8 at engcorp.com...
> Adonis wrote:
> >
> > can a class return a value? i have treid to google this, but have found
> > myself short of an answer, i know i could place it in a member and
retrieve
> > it or through a function, but im trying to simplify my life.. just a
> > little..
>
> You probably don't actually mean a class.  I think you mean an object,
> which is an instance of a class.  (If I've got that wrong, I think
> we need clarification of the question.)
>
> If that's the case, there's still one more problem left. :-)
> "Returning a value" is a concept related to calling a method
> or function.  It doesn't have any particular meaning if you
> ask "can an object return a value?" unless you mean something
> like "can a method of an object return a value?".  That is
> certainly possible, although I'm not this is what you wanted,
> either:
>
> >>> class MyClass:
> ...    def myMethod(self):
> ...        return 'spam'
> ...
> >>> c = MyClass()
> >>> print 'Green eggs and %s.' % c.myMethod()
> Green eggs and spam.
>
> Can you ask the question in a different way, or show an example
> of what you were hoping to do?
>
> -Peter




From fgeiger at datec.at  Fri Jun 21 15:51:35 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Fri, 21 Jun 2002 21:51:35 +0200
Subject: [wxPython] How can I have a scrolled window on a notebook page, that autom. fills the whole pane?
Message-ID: <af005n$mve$1@newsreaderm1.core.theplanet.net>

Hi everyone,

I'm currently writing an app, that has a Notebook front end. The 1st page
(there's yet only one) should contain a wxScrolledWindow, which in turn
should contain a VirtualListCtrl.

In an earlier app I had text controls on the notebook pages, which
automatically could adapt their size to the notebook page automatically,
which in turn could adapt its size to the main frame.

Now I wanted this for my scrolled window, but no chance.

I tried to add a wxScrolledWindow to the wxNotebook, but the list control on
it was so tiny, you even could not see the column headers.

I tried to add a wxPanel to the wxNotebook. The wxPanel in turn contained
the wxScrolledWindow, which in turn contained the list control. Same result.

So I ended up with setting the size for the list control explicitly. Now I
see the list control's contents, but no scroll bars, no auto sizing.

Any hint, what I'm missing?

Cheers
Franz GEIGER

P.S.: To have the controls auto size themselves I tried size=wxDefaultSize,
size=(-1, -1), style=wxMAXIMIZE_BOX.

ASPython 2.1.3, wxPython 2.3.2.1.





From chris.gonnerman at newcenturycomputers.net  Wed Jun 26 08:02:57 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Wed, 26 Jun 2002 07:02:57 -0500
Subject: Python and Windows XP
References: <20020625233842.29966.00000386@mb-ma.aol.com>
Message-ID: <005101c21d09$6dcf2b80$0101010a@local>

----- Original Message -----
From: "Marc" <lmbrtmarc at aol.comlaissermoi>


> Hello all,
>    I apologise for this newbie question, but I use Windows XP and wonder
how I
> can open a shell session to use Python?  There used to be the DOS command
> option, but now...

Start, Run, then type CMD and press ENTER.  The Windows XP command prompt
even
included BASH-like filename completion (type part of the name and press TAB
to
see the first match).  Pretty cool considering it's M$.

You can do the same thing typing COMMAND instead of CMD but the COMMAND.COM
interpreter is brain-damaged.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net





From jepler at unpythonic.net  Thu Jun  6 08:59:03 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Thu, 6 Jun 2002 07:59:03 -0500
Subject: check on numerical value?
In-Reply-To: <3cff56b4.511834453@news.skynet.be>
References: <3cff56b4.511834453@news.skynet.be>
Message-ID: <20020606075858.A30258@unpythonic.net>

You could see if eg float() will handle it:
    def isfloat(s):
        try:
            float(s)
        except ValueError:
            return False
        else:
            return True
This method will work on any type that defines a __float__ method as
well (eg an instance of a rational number class)

Or you could check it against a regular expression (only good for
strings, obviously):
    import re, tokenize
    _isfloat_re = re.compile(tokenize.Number + "$")
    def isfloat(s):
        if _isfloat_re.match(s): return True
        else: return False

(tokenize.Number recognizes hex, octal, and decimal integer constants;
floating point constants with and without exponents; and imaginary
numbers with the "J" or "j" suffix.  You might prefer to use
    DecFloat = tokenize.group(tokenize.PointFloat, tokenize.ExpFloat,
                              tokenize.DecNumber)
    _isfloat_re = re.compile(DecFloat + "$")
)

(True and False may not appear in your version of Python.  Substitute 1
and 0 if they're not there)

Jeff




From whisper at oz.net  Wed Jun 19 20:09:18 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 19 Jun 2002 17:09:18 -0700
Subject: Windows versions of Python---pros and cons?
In-Reply-To: <20020619154654.C619@ActiveState.com>
Message-ID: <GCEDKONBLEFPPADDJCOEEEFADKAA.whisper@oz.net>

<snip>
> > In particular, I write many C extensions; can one use distutils and
> > the Borland or gcc free compilers to build C extensions in both
> > versions?
>
> My understanding is that for C extensions to work they must be built
> with the same compiler as that used to build the Python with which they
> will run. (I may be wrong. Someone please correct me if I am.) Visual
> C++ 6.0 is used to build ActivePython and a python.org's Python for
> Windows. This means that you would have to build Python yourself from
> source with your free compiler to be able to run your extensions built
> with the free compiler.
>
> Cheers,
> Trent
>
> --
> Trent Mick
> TrentM at ActiveState.com
>

It is always true (or has been historically) that you can't mix the output
of C++ compilers (due to name mangling not being standardized). It was, at
one time, true that Borland and Microsoft C compilers could use each other's
output, but that may no longer be so - MS may have changed it's .obj format.
If you have Borland C or GCC, I don't think it's all that hard to build
Python and friends (dunno about Mark Hammonds' COM extensions on Borland
though).

Dave LeBlanc
Seattle, WA USA





From eppstein at ics.uci.edu  Mon Jun  3 13:12:24 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Mon, 03 Jun 2002 10:12:24 -0700
Subject: Graph distances, how can I speed it up?
References: <33803989.0206030235.4134d33e@posting.google.com> <d571ae53.0206030815.2e52aac@posting.google.com> <B92114F6.B79E%pereira@cis.upenn.edu>
Message-ID: <eppstein-347A24.10122403062002@news.service.uci.edu>

In article <B92114F6.B79E%pereira at cis.upenn.edu>,
 Fernando Pereira <pereira at cis.upenn.edu> wrote:

> Dijkstra's algorithm is not a good choice here, because distances from all
> sources appear to be needed, not from a single source.

Dijkstra is not a bad choice for all pairs shortest paths -- just repeat 
it for each source in the graph.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From julian.gollop at ntlworld.com  Mon Jun  3 08:20:59 2002
From: julian.gollop at ntlworld.com (Julian Gollop)
Date: Mon, 3 Jun 2002 13:20:59 +0100
Subject: Interrupting a continuous script
Message-ID: <_6JK8.1535$376.74360@newsfep1-win.server.ntli.net>

I have a few programs continually processing text files on a Linux machine.
Occasionally I need to stop them, but I don't want to use a keyboard
interrupt because it may cause the script to stop in the middle of doing
something.
Does anyone know a simple way to allow me to stop the process by pressing a
key on the keyboard - without stopping inside something crucial?

Julian.





From shagshag13 at yahoo.fr  Tue Jun  4 07:01:14 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Tue, 4 Jun 2002 13:01:14 +0200
Subject: Looking for a list subclassing example...
References: <adi4dv$118tko$1@ID-146704.news.dfncis.de> <3cfc994e$0$3851$e4fe514c@dreader4.news.xs4all.nl> <adi5hc$10ulbe$1@ID-146704.news.dfncis.de>
Message-ID: <adi6lb$107klr$1@ID-146704.news.dfncis.de>

Is is possible to write a method in an object subclassing from list and which "frees" it ?

reset doesn't work

> class Shag(list):
>  def __init__(self, emptyValue = None):
>   self._emptyValue = emptyValue
>
>  def reset(self):
>   del self
>   self = []

thanks,

s13.




From kragen at pobox.com  Sat Jun  8 17:27:28 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 08 Jun 2002 17:27:28 -0400
Subject: int, float and string...
References: <adhut1$10rtkl$1@ID-146704.news.dfncis.de> <mailman.1023206585.1785.python-list@python.org> <adirg9$11hv44$1@ID-146704.news.dfncis.de> <mailman.1023212772.9349.python-list@python.org>
Message-ID: <83y9dp5um7.fsf@panacea.canonical.org>

"Mike C. Fletcher" <mcfletch at rogers.com> writes:
> See object.h in the Python source code.  Basically, Python objects 
> (PyObjects) are a C structure with book-keeping information (what is 
> your class/type, and your reference count) + the actual data of the 
> object.    Depending on the object type, that can be anywhere from 
> (minimum) 8 bytes to 8+x for every object.

PyObjects are generally allocated on the heap, which adds additional
overhead, typically about another 8 bytes per object.

> BTW, I'm not a C person, so I may be mis-calculating some of those 
> values, basic principle still applies: don't worry about differentiating 
> between int and float based on storage concerns save when using huge 
> numbers of them (preferably in Numeric arrays).

Even the built-in 'array' module might be an improvement.

> not mistaken).  Still, this is probably premature optimisation unless 
> you actually find yourself running out of memory.

Dunno if you've been watching, but Shagshag is building a text-search
engine that is running out of (gigabytes of) memory.



From tebeka at cs.bgu.ac.il  Mon Jun  3 06:35:03 2002
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 3 Jun 2002 03:35:03 -0700
Subject: Graph distances, how can I speed it up?
Message-ID: <33803989.0206030235.4134d33e@posting.google.com>

Hello All,

I'm trying to calculate distances in a graph. (The ultimate goal is 
to calculate the distances between verbs in WordNet)

The graph is built edge after edge and each time I recalculate the 
distances matrix. However this takes ages to complete. (will finish in ~50H).

The code:
(Note: If you try to use the following code as is mail me for a patch to pywordnet.)

----------------
#!/usr/bin/env python
'''Find distance between all verbs in wordnet
'''

__author__ = 'Miki Tebeka <tebeka at cs.bgu.ac.il>'
__version__ = '$Revision: 1.1.1.1 $'

# $Source: /home/miki/.cvsroot/verbdist/verbdist.py,v $

import sys
import wordnet
import time

class GraphDist:
    '''Graph distances class
    
    Distances between vertexs are initially infinite, when an edge
    is added/updated between vertexes a new distance is calculated.
    
    Distances are held in hash table of pairs (x,y) -> distance.
    For each vertex a hash of all vertexes conncted to it is saved
    to improve distance update.
    '''
    def __init__(self, directed = 0):
        '''Constructor. Defaults to undirected graph'''
        self.directed = directed # Directed graph flag
        self.inf = long(sys.maxint) # Infinite
        self.clear()

    def clear(self):
        '''Clear graph'''
        self.distances = {} # Distances hash
        self.backwards = {} # All vertexes that has path to x
        self.forward = {} # All vertexes that x has path to
        
    def distance(self, x, y):
        '''Distance between x and y
        
        If (x,y) not in graph then distance is inf
        '''
        if (x == y):
            return 0
        return self.distances.get((x,y), self.inf)

    def add_edge(self, x, y, distance = 1):
        '''Add edge between x and y with distance distance'''
        if (x == y): # Ignore edge to self
            return
        olddist = self.distance(x, y)
        newdist = min(olddist, distance)
        if (newdist == olddist): # No change
            return
        
        self.distance_change(x, y, newdist)

    def add_path(self, x, y):
        '''Add path from x to y'''
        # Add backwardss path
        h = self.backwards.get(y, {})
        h[x] = 1
        self.backwards[y] = h
        # Add forward path
        h = self.forward.get(x, {})
        h[y] = 1
        self.forward[x] = h
        
    def path_to(self, x):
        '''All vertexes that have path to x'''
        return self.backwards.get(x, {}).keys()

    def path_from(self, x):
        '''All vertexes that x hash path to'''
        return self.forward.get(x, {}).keys()
        
    def directed_distance_change(self, x, y, newdist):
        '''Change distance for x -> y to newdist'''
        print 'directed_distance_change %s->%s' % (x, y)
        # Update distance
        self.distances[(x, y)] = newdist

        # Update distances in all vertexes connected to x
        for target in [y] + self.path_from(y):
            for source in [x] + self.path_to(x):
                dist = self.distance(source, target)
                nd = min(dist, newdist + \
                               self.distance(source, x) +\
                               self.distance(y, target))
                if (nd == dist):
                    continue # No change
                self.distances[(source, target)] = nd
                self.add_path(source, target)
        
    def distance_change(self, x, y, newdist):
        '''Distance between x and y was changed to newdist'''
        self.add_path(x,y) # Update connected hashes
        self.directed_distance_change(x, y, newdist)
        if not self.directed:
            self.add_path(y, x)
            self.directed_distance_change(y, x, newdist)
        

def _test(directed):
    dg = GraphDist(directed)
    print 'Adding c->d'
    dg.add_edge('c', 'd')
    print 'Adding b->c'
    dg.add_edge('b', 'c')
    return dg

def _get_keys(dict):
    '''Hack around wordnet's dict.keys() bug'''
    keys = []
    for i in xrange(len(dict)):
        keys.append(dict[i])
    return keys

def prompt(msg):
    '''Prompt a string and flush stdout'''
    print msg,
    sys.stdout.flush()

def promptln(msg):
    '''Prompot a string + newline and flush stdout'''
    print msg
    sys.stdout.flush()
    
def vd():
    '''All verbs distances in wordnet'''
    st = time.time() # Start time
    gd = GraphDist() # Graph distances
    promptln('Getting all verbs...')
    verbs = _get_keys(wordnet.V) # Get all verbs
    promptln('Got %d verbs' % len(verbs))
    i = 0
    promptln('Calculating distances...')
    for verb in verbs:
        i += 1
        if not (i % 100):
            prompt('.')
        vstr = verb.form # Verb form
        for sense in verb:
            for pointer in sense.getPointers(wordnet.HYPERNYM): # Find hypernyms
                for s in pointer.target():
                    gd.add_edge(vstr, s.form) # Add edge to hypernym
    et = time.time()
    promptln('* Done. [Took %d seconds]' % int(et - st))
    return gd                   
                 
if __name__ == '__main__':
    # Main program, run and store results in text file
    gd = vd()
    outfname = 'verbdist.txt'
    promptln('Writing to file ...')
    fp = open(outfname, 'w')
    for src, dst in vd.distances.keys():
        print >> fp, '%s\t%s\t%s' % (src, dst, vd.distance((src, dst)))
    print '* Done. %s created' % fname
        
    
Any ideas how to speed it up?

Thanks.
Miki


From johnroth at ameritech.net  Thu Jun 13 17:31:22 2002
From: johnroth at ameritech.net (John Roth)
Date: Thu, 13 Jun 2002 17:31:22 -0400
Subject: why not "'in' in 'in'"?
References: <ae8dg70217g@drn.newsguy.com>
Message-ID: <ugi3nqirt970d3@news.supernews.com>

"Grant Griffin" <not.this at seebelow.org> wrote in message
news:ae8dg70217g at drn.newsguy.com...
> Hi Gang,
>
> One of the things I like about Python is that it works as expected.
However, I
> sometimes accidently expect it to allow one to use "in" to determine
if one
> string is "in" another, e.g.:
>
> >>> 'in' in 'in'
>
> which yields a traceback:
>
>    Traceback (most recent call last):
>      File "<interactive input>", line 1, in ?
>    TypeError: 'in <string>' requires character as left operand

That doesn't seem very intuitive, does it? However, limiting
it as defined actually makes sense, because otherwise you
get ambiguities and boundary conditions.

> Evidently one is invited to use "count" or "find" instead:
>
> >>> 'in'.count('in') > 0
> 1
> >>> 'in'.find('in') >= 0
> 1
>
> I prefer "find" over "count" because it (presumably) punts after the
first
> occurrence.  However, neither one reads nearly as well as the
(illegal) "in"
> form.

Actually, the .count form is probably preferable, since it goes nicely
into an if statement - 0 is false and anything else is true. The .find
function is not so nice.

> So why is the "in" operator limited to single characters when used as
a
> membership operator?  This peculiar lack of generality seems a bit
(dare I say
> it?) "non-Pythonic".  (That said, "Pythonic" seems to be a bit like
> "enlightenment": those who claim to have mastered it wisely avoid
spelling out
> any concrete criteria by which others can judge their mastery <wink>.)

Actually, as far as I'm concerned, it's an anomaly, although I'm not
complaining.
When 'in' is used against any other kind of object, it takes whatever
type the
actual member of the sequence happens to be. Strings are a special case,
since there is no character type in Python.

> Perhaps its somehow due to the overloading of "in", which is used both
as a
> membership operator and as a separator gizmo inside "for" loops.
("It's a floor
> wax/No, it's a desert topping/No it's both!")  But is that any good
reason to
> exclude the handy, sensible, and downright intuitive use of "in" as a
substring
> detector?

No, that isn't the reason. 'in' is defined on sequences and mappings,
and has
the same semantics on either: does the particular member exist or not
within
the sequence or mapping. The usage in the 'for' statement is entirely
other.

Whether it's intuitive or not depends entirely on the speaker's
intuition. I,
for one, don't find it intuitive.

> Or maybe support for multiple characters would penalize the
performance of the
> single-character "in".

Python is not concerned with performance over clarity of expression. If
your suggestion would help clarity, it would be considered. However, it
does
not help clarity.

> Or maybe Tim just doesn't want to burden us with a third way to do it
<wink>.
>
> gotten-spoiled-to-Python's-habits-of-working-as-expected
>    -and-reading-as-pseudo-code-ly y'rs,
>
> =g2
>
>
________________________________________________________________________
_
>
> Grant R. Griffin
g2 at dspguru.com
> Publisher of dspGuru
http://www.dspguru.com
> Iowegian International Corporation
http://www.iowegian.com
>




From chris.gonnerman at newcenturycomputers.net  Thu Jun 13 23:43:52 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Thu, 13 Jun 2002 22:43:52 -0500
Subject: system command
References: <aeajuh$5b5eo$1@ID-108351.news.dfncis.de> <3d08dba0_2@hpb10302.boi.hp.com> <aeamjt$5eflk$1@ID-108351.news.dfncis.de>
Message-ID: <008f01c21355$f1ff8d40$0101010a@local>

----- Original Message -----
From: "Thor" <thor__00 at yahoo.com>


> Daniel Fackrell wrote:
>
> > Assumption: You are referring to the "source" command as provided by the
> > bash shell on some *nix.
> >
> > The man page for bash suggests that you probably want:
> >
> > print "source "+root+".in"
> > os.system("bash "+root+".in")
>
> Changing to bash does the same.. i mean, nothing, no error, just finishes
> the script

Short answer is, no matter how you do it, you can't do it.

The "source" command executes the given file in the *current* shell.  A
Python script, or a Perl script, or for that matter a shell script runs
in a *subshell* and *can not change the parent shell's environment*.
This is a FAQ for shell users somewhere...

BTW this also means that a Python script on Unix/Linux/BSD can't change
the current shell's environment variables either.

Just what, precisely, are you trying to accomplish?  Likely there is some
sort of workaround...

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net





From sholden at holdenweb.com  Thu Jun  6 10:18:47 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 6 Jun 2002 10:18:47 -0400
Subject: array, list, performance...
References: <3CFF5F98.2060406@destiny.com> <mailman.1023373030.31871.python-list@python.org>
Message-ID: <HfKL8.314106$%u2.179672@atlpnn01.usenetserver.com>

"Christopher Armstrong" <radix at twistedmatrix.com> wrote in message
news:mailman.1023373030.31871.python-list at python.org...
>
> > Michael Chermside <mcherm at destiny.com> writes::
> >   li[n] = x      assignment to the middle is O(len(li))
>
> This doesn't seem right. Shouldn't that be O(1), just like
> access-time?
>

If you can do it in O(1) you should go work for Zope corporation :-)

The problem is the need to make a new entry in the middle of the existing
structure. This can involve memory allocation and value copying, etc.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From emile at fenx.com  Thu Jun 27 11:35:24 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 27 Jun 2002 15:35:24 GMT
Subject: sys.argv[0] - Returns path and filename.
References: <c07d0edc.0206270432.faf1fd3@posting.google.com>
Message-ID: <0hGS8.326386$352.35079@sccrnsc02>

Guy
> The only major problem that I have come across at the moment is the
> argv[0], on my win32 box it retruns the path and a filename, on the
> mac osx and the linux only the filename are returned.
>

What you get is the almost the same on win2k and linux when similarly
invoked.

F:\Python22>python test.py
['test.py']
F:\Python22>test.py
['F:\\Python22\\test.py']


[emile at lnxfc emile]$ python test.py
['test.py']
[emile at lnxfc emile]$ test.py
['./test.py']
[emile at lnxfc emile]$ /home/emile/test.py
['/home/emile/test.py']

--

Emile van Sebille
emile at fenx.com

---------




From tdelaney at avaya.com  Tue Jun  4 01:54:23 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 4 Jun 2002 15:54:23 +1000 
Subject: prototyping good OOdesign in Python?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A34D@natasha.auslabs.avaya.com>

> From: Greg Ewing [mailto:look at replyto.address.invalid.avaya.com]
> 
> "Delaney, Timothy" wrote:
> > 
> > Unfortunately, this will only work for 1 or 2 projects ... 
> after that they
> > will *expect* you to bring everything in under time and budget ...
> 
> So, don't tell them it's finished until just before the
> deadline... and find something plausible-sounding to
> spend the extra money on... :-)

Hmm ... so *that's* how I get the raise that was being talked about ... ;)

Sometimes, there's such a thing as being too honest ...

Tim Delaney




From dougz at cs.washington.edu  Sun Jun 16 21:21:14 2002
From: dougz at cs.washington.edu (Douglas Zongker)
Date: Mon, 17 Jun 2002 01:21:14 GMT
Subject: what *is* a class?
References: <MPG.17774343cc5a589098968a@news.rz.uni-karlsruhe.de>
Message-ID: <Gxtt3F.upA@beaver.cs.washington.edu>

Uwe Mayer <merkosh at hadiko.de> wrote:
: i've got an object 'Group'. i instanciate it by calling, f.e.
: g = Group()

: what actually does Group() return and is there any way to get hold of 
: that again?

"Group()" creates a new object which is an instance of the class
Group.  It calls the object's initializer ("__init__" method), then
returns the newly created object.

: i want to write a method which returns the instance, so that f.e. the 
: following is possible:
:   g = Group()
:   g2 = g.foobar()
:   g2 is g
: so that g2 is identical to g.
: any idea?

    class Group:
        def __init__( self ):
            print 'creating instance'

        def foo( self ):
            print 'in the foo method'
            return self

    >>> g = Group()
    creating instance
    >>> g2 = g.foo()
    in the foo method
    >>> g2 is g
    1
    >>>		

Note that there's usually no reason for a method (such as "foo") to
return the self object, because the caller should already have the
object to call the method in the first place.  I could leave "return
self" out of the definition of "foo", and just write:

    g = Group()
    g.foo()
    g2 = g

The result would be the same.

dz


 


From djmitchell at NOSPAM.optushome.com.au  Fri Jun 21 06:48:43 2002
From: djmitchell at NOSPAM.optushome.com.au (David Mitchell)
Date: Fri, 21 Jun 2002 20:48:43 +1000
Subject: I'd give up Perl tomorrow if only...
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au> <mailman.1024653868.4542.python-list@python.org>
Message-ID: <pan.2002.06.21.10.48.42.651127.4527@NOSPAM.optushome.com.au>

On Fri, 21 Jun 2002 20:03:16 +1000, Gerhard H?ring wrote:

> I think the main point of CPAN is the automagical installation, that
> pythonsiphon tries to bring to Python.

That's right.  
> 
> As for a searchable database, there's the Vaults of Parnassus, and the
> Python modules category at dmoz.org/directory.google.com/you.name.it.

I hadn't heard of pythonsiphon, but I'm pretty familiar
with the Vaults of Parnassus.  My chief concern isn't with the *quality* of
what's availabile in Python (and some of what's at CPAN is pretty grim in
this respect); it's the sheer *breadth* of what's at CPAN that makes it
so useful.

> 
>> For just about any library module I've ever wanted, I can download one
>> or more modules from CPAN that gets me 90% of the way there.  Perl
>> modules like DBI/DBD, SOAP::Lite, GD::Graph, MQSeries::*, Tie::*,
>> Convert::EBCDIC,
> 
> No idea what Tie::* is, but EBCDIC sounds easy to implement, should it
> not exist already, I haven't heard of an MQSeries binding for Python,
> but I've personally used Python libraries for the first three items, and
> they worked well.

ASCII<->EBCDIC conversion is pretty easy to implement in just about any
language, but it's also easy to make a typo somewhere when you're typing in
lots and lots of ASCII codes.  I know; I've done it ;->

The advantage of the pre-existing Perl module is that someone's already
written it, and 500 others have probably beaten out the bugs; there's a
degree of confidence there that I wouldn't have for some time if I was
typing in my own solution from scratch.  Not that I couldn't do it; just
that it'd take some time for all my typos to be sorted out.

> 
>> If there was only some way to subclass Perl/CPAN modules as Python
>> modules
> 
> PyPerl exists. That's all I can say to that (never used it myself), as
> my brain is incompatible with Perl.

PyPerl looks very interesting - hadn't heard of that either!  Glad I
posted the original message, and thanks to everyone who's responded for
some really useful ideas.  Reading through various newsgroup postings
on PyPerl courtesy of Google, it looks pretty robust as well, once you
get used to how it works.

For me, Perl has one big advantage over *every* other language.  If I have
to write a one-off piece of code to do some relatively simple task, and I
know that I'll truly never look at that code again, I can rip out a solution
in Perl almost as fast as I can think.  The problem is that that code
will be completely illegible to anyone else, including myself a few weeks
later, but as long as the job gets done and I don't have to worry about
it later, who cares what the code looks like?  I suspect lots of
experienced Perl coders will nod their heads as they read this...

Using Python, that same task will take me a bit longer (although
nothing like it'd take using, say, C or Java), but the code will be
relatively easy to go back to at some later date.  It'll be better
structured; the variables will have better names, and so on.

However, when speed of development is absolutely crucial, Perl can't be
beat.  (Thinks to himself: "Please don't let this turn into yet another
Perl vs. Python advocacy thread")

Dave Mitchell


From olc at ninti.com  Tue Jun  4 13:41:24 2002
From: olc at ninti.com (Michael Hall)
Date: Wed, 5 Jun 2002 03:11:24 +0930 (CST)
Subject: CGI questions
Message-ID: <Pine.LNX.4.21.0206041508270.902-100000@localhost.localdomain>

I've been using PHP for a couple of years now, but have begun to look at
Python as a CGI language. So far so good, I really like writing straight
OO code and keeping all the HTML spaghetti out of the way in templates.
Anyway, a couple of questions:

1.
Does Python have the same or a similar problem that PHP had until recently 
with regard to automatic global registration of GET, POST and COOKIE
variables within a script's namespace? My guess is that it doesn't, but
I'd like to know for sure.
In other words, is all user input in the form of query string or POST
variables 'filtered' through cgi.Fieldstorage? Is user input ever directly
available to the script? 
Let's say I have a variable 'loggedon' in my script. If a user passes
'loggedon=yes' in a query string to my script, will it have any effect or
would it have to be accessed via something like this first:

form = cgi.Fieldstorage
loggedon = form["loggedon"].value

2.
I'm trying to use the md5 module to create a session id string. It
produces some output, but it is pretty weird stuff, not the 32 character
alphanumeric string I'd expect from PHP. Most of it is definitely not
[a-zA-Z0-9], and it is around 10 characters long. How can I get a PHP
style md5 result?

3.
What is the decimal bit on the end of time.time()'s output? Hundredths of
a second?

4.
Is the following Python the (best) equivalent to the PHP?

PHP				Python
ereg("bit","bigbits")		re.compile("bit").search("bigbits",1)
				

TIA 


--------------------------------
Michael Hall     ninti at ninti.com
--------------------------------






From opengeometry at NOSPAM.yahoo.ca  Wed Jun 12 19:09:29 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 12 Jun 2002 19:09:29 -0400
Subject: Wanted:  Potential Python Authors
References: <ae7gke$mhb$1@ccis12.baylor.edu> <db3fgu40sbnt3pno5v1i2h9lnmpcoti2bk@4ax.com>
Message-ID: <3d07d4a8@news.eol.ca>

Lothar Scholz <llothar at web.de> wrote:
> On Wed, 12 Jun 2002 08:03:35 -0500, "Jeff Donahoo"
> <Jeff_Donahoo at Baylor.edu> wrote:
> 
>>I'm looking for somebody to author a short book on Python in my programming
>>technology series.  Who do you think would make a great author that hasn't
>>already written a book on Python?  Candidates include:
> 
>>- Competent and frequent posters to this newsgroup
> This can be an indicator. 
> 
>>- Author of articles in the various trade magazines
> You really mean "trade" magazines ?
> Normally they are written by marketing peoples who don't
> have the competence to write more then a 4 page overview.
> Hands off !!!!
> 
>>- Expert programmer
> No, definitely not. Most programmers are not trained on writing
> tutorials or books - and normally they are really bad.
> Look more for someone trained as technical writer.

Bull.  I agree about trade magazines above, though there isn't any magazine
covering Python.  But, whether you want "techical writer" or "real
programmer" would depend on your target audience.

If your audience is people who needs some hand-holding and will never
re-read the book again, then you may want "technical writer".  But, if your
audience is people who learned C from "K&R book" over the weekend and
learned Awk/Bash from manpages, then you may want "real programmer" who can
write.

There is no such thing as "trained as writer", whether it's for computer
books or fiction/nonfiction novels.  You either can write or you can't.

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin


From admin at NO-SPAM-at-2host.com  Sat Jun  1 13:34:02 2002
From: admin at NO-SPAM-at-2host.com (2Host.com)
Date: Sat, 01 Jun 2002 10:34:02 -0700
Subject: Migrating from PHP to Python...
References: <xWyk7.4824$v%4.492814@news20.bellglobal.com> <l4ihfu0nhp1a9p53vgch3n3kcc25s8c00p@4ax.com> <slrnafhj0d.dp4.gerhard@lilith.my-fqdn.de> <m3znyf3x2o.fsf@mumonkan.sunstarsys.com> <slrnafhr3p.e0j.gerhard@lilith.my-fqdn.de> <m3ptzb3q5q.fsf@mumonkan.sunstarsys.com>
Message-ID: <3CF9058A.8624ECFC@NO-SPAM-at-2host.com>


Joe Schaefer wrote:
> 
> Gerhard =?iso-8859-15?Q?H=E4ring?= <gerhard at bigfoot.de> writes:
> 
> > Sorry. I should have just PLONKed him and been done with it.
> 
> No harm, no foul.  IMO the clp.misc denizens have recently
> elevated troll management to near artform, and it has been
> quite a pleasure to witness the collective triage effort.
> For instance, check out the subthread in clp.misc that
> revolves around this post
> 
> Message-ID: <3CF686D1.608DE457 at NO-SPAM-at-2host.com>
> 
> That 2Host.com guy can draw blood from a stone.
> 
> Cheers.
> --
> Joe Schaefer

I hope I didn't miss something and question if you were saying I was a
troll, or if you maybe meant I had dealt a blow to a troll? :-/

Regards,
Robert
-- 
Email: admin at -NO-SPAM-2host.com | Phone: 530-941-0690
Server administrator/support - 2Host.com web hosting.
Secure, reliable hosting - What you expect & deserve!
How are we different? Find out: http://www.2host.com/


From mcalla at insightbb.com  Tue Jun 18 18:55:46 2002
From: mcalla at insightbb.com (Mike Callahan)
Date: Tue, 18 Jun 2002 22:55:46 GMT
Subject: Question about kinfxdb.py
Message-ID: <STOP8.247602$cQ3.10370@sccrnsc01>

I am running kinfxdb on an HPUX system. I am new to interfacing Python to
Informix and I have some problems:

1) The 'unload to' commands don't work. "select * from location" is fine but
"unload to 'test' select * from location" gives a 201 error.

2) The info command doesn't work.

3) Dates before 1900 are returned as 12/31 1969.

If I switch to informixdb will these problems go away? Or are these
well-known limitations of the Informix interface. Thanks.

Mike Callahan





From marklists at mceahern.com  Fri Jun 14 12:50:45 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 14 Jun 2002 11:50:45 -0500
Subject: NEWbie asks "is Python a good tool for automated file/folder generation?"
In-Reply-To: <6k6kgukhuk8pfgos97a38h0d4r5jr0u3sr@4ax.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBGELFLEAA.marklists@mceahern.com>

> I've never used Python, and I'm wondering if you, collectively, think
> that Python is a good language to do this.

Sure.  I'd recommend starting with simple things:

Fire up Python and do this:

>>> import os
>>> path = "c:\type your path here"
>>> print os.listdir(path)

listdir returns a list of the contents of path.  You can use os.path.isdir()
to distinguish files from directories.

Also, note that strings are objects with methods that you can use to do
string manipulation:

>>> s = "foobar"
>>> dir(s)

dir(object) tells you the the attributes of the object.  This encourages
interactive exploration of Python, which is one of the things I think makes
Python eminently easy to (re)learn.  ;-)

I'd recommend reading any one of the fine Python tutorials:

  http://www.python.org/doc/Newbies.html

If you have questions, you can always post them here or:

  http://www.python.org/psa/MailingLists.html#tutor

Cheers,

// mark

p.s.  If you want people to cc you, you can make it easier by using that
email address when sending messages.

-





From david_griswold1 at yahoo.com  Fri Jun 28 21:34:24 2002
From: david_griswold1 at yahoo.com (David Griswold)
Date: Sat, 29 Jun 2002 01:34:24 GMT
Subject: I'm an idiot
References: <mailman.1025313281.20462.python-list@python.org>
Message-ID: <Xns923BBCF03699Adavidgriswold1netsca@204.127.202.16>

Thanks.  I actually figured out the missing parentheses, and the typo on 
the split/strip..  I will blame Active Python for being TOO helpful on 
the former and a brainfart on the latter.

Now one more thing is happening I can't figure out.  When readline gets 
to the end of the file (the file is only 34k) it just keeps on reading 
blank lines.  Am I missing something?  I even changed my error I am 
checking for to EOFError.

Thanks again.

David

"Sean 'Shaleh' Perry" <shalehperry at attbi.com> wrote in
news:mailman.1025313281.20462.python-list at python.org: 

>> 
>> f=open('c:\\temp\\temp.txt', 'r')
>> g=open('c:\\temp\\temp1.txt', 'w')
>> while 1:
>>     try:
>>         s=f.readline
> 
> should be 's = f.readline() # I am a method, not a variable'
> 
>>         g.write(s.split())
> 
> split() takes a string and breaks it into pieces, this is not exactly
> what you want, for instance 'this is a line' will beceome ['this',
> 'is', 'a', 'line']. 
> 
>>     except IOError:
>>         break
>> 
>> g.close
>> f.close
> 
> these too should have () on the end.
> 
> 
> 



From guillermo.fernandez at epfl.ch  Tue Jun 25 01:19:01 2002
From: guillermo.fernandez at epfl.ch (Guillermo Fernandez)
Date: Tue, 25 Jun 2002 14:49:01 +0930
Subject: File "in3sun34.py", line 33, in AttributeError: HTTPConnection
Message-ID: <3D17FD45.EDE1EC36@epfl.ch>

Hello!

Please, if you answer me, include my e-mail in the answer as I am not
abonned to this mailing list.

I've written a little program to test python (I'm a beginner). Here it
is:

import httplib

conn = httplib.HTTPConnection("www.python.org")
print "CONNECTED"
conn.request("GET", "/index.html
response=conn.getresponse()
answer=response.read()
conn.close()
print answer

When I execute it, I obtain an error:
[112]host-cfernand% python test.py
Traceback (innermost last):
  File "test.py", line 3, in ?
    conn = httplib.HTTPConnection("www.python.org")
AttributeError: HTTPConnection

Line number 3 is:
conn = httplib.HTTPConnection("www.python.org")

I have searched in the internet and in the documentation and I have not
been able to solve it.

On the other hand, this program it works, so it's not a network problem:
import urllib
sock = urllib.urlopen("http://www.python.org/index.html")
htmlSource = sock.read()
sock.close()
print htmlSource

Thank you for your help!

Guille




From gerhard at bigfoot.de  Tue Jun 11 17:05:26 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Tue, 11 Jun 2002 23:05:26 +0200
Subject: Embedding Python 2.2 in non-reentrant application
In-Reply-To: <ae5o37$37q$03$1@news.t-online.com>
References: <ae5n1f$qtm$07$1@news.t-online.com> <m3k7p5wnt1.fsf@mira.informatik.hu-berlin.de> <ae5o37$37q$03$1@news.t-online.com>
Message-ID: <20020611210526.GA28153@lilith.my-fqdn.de>

* Stefan Werner <stefan1werner at hotmail.com> [2002-06-11 22:52 +0200]:
> Excuse my ignorance, but what does GIL mean?

Global Interpreter Lock.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From peter at engcorp.com  Thu Jun 20 23:21:27 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 20 Jun 2002 23:21:27 -0400
Subject: newbie string question
References: <aeu40s$9jg2s$1@ID-145503.news.dfncis.de>
Message-ID: <3D129BB7.C7EA15D2@engcorp.com>

Don Low wrote:
> 
> There's probably a very simple explanation, but I can't see it for now, so
> bear with me.
> 
> >>>pystr = 'Python'
> >>>pystr = [5]
> 'n'
> >>>pystr [2:5]
> 'tho'
> 
> The question is, why isn't pystr [2:5] 'thon' instead of 'tho'. Could
> somebody explain?

The numbers don't represent the position of the items in the sequence,
but the places *between* items where you would "slice" the sequence
to get the subsequence.  That's why it's called a slice operation.

Picture it this way:

   0   1   2   3   4   5    <-- subscript index
 +---+---+---+---+---+---+
"| P | y | t | h | o | n |"
 +---+---+---+---+---+---+
 0   1   2   3   4   5   6  <-- slice index

Your slice went from 2 to 5, so it got only the "tho" part of the string.
Here's what you got:

 +---+---+---+
 | t | h | o |
 +---+---+---+
 2   3   4   5

Other languages confuse the indices used when slicing things (as in
the common substring() method) and when subscripting things (when you 
take just a single item, such as pystr[4] which is "o" in the above).

Python avoids up this confusion by realizing they are not the same
thing.  (Of course, this can confuse newcomers who don't realize 
what happened, but at least the pain happens for you only once. :-)

-Peter


From jjl at pobox.com  Sun Jun  2 15:24:09 2002
From: jjl at pobox.com (John J. Lee)
Date: Sun, 2 Jun 2002 20:24:09 +0100
Subject: HTML and python
In-Reply-To: <gstK8.148299$%u2.116279@atlpnn01.usenetserver.com>
References: <ad94md$viphf$1@ID-49758.news.dfncis.de> <gstK8.148299$%u2.116279@atlpnn01.usenetserver.com>
Message-ID: <Pine.SOL.4.30.0206022022220.23311-100000@mimosa.csv.warwick.ac.uk>

On Sun, 2 Jun 2002, Steve Holden wrote:
> "Ken" <ken at hotmail.com> wrote in message
> news:ad94md$viphf$1 at ID-49758.news.dfncis.de...
> > How do you call to a static HTML page from the script?
> >
> > <FORM METHOD="POST" ACTION="frame2a.html" target=_parent>
> > This doesn't seem to be working..... Can someone help?
> >
>
> It really makes no sense to use method POST to access a static page. Using
> method GET will probably owrk, though you will see the for contents in the
> URL.

Why doesn't it make sense?

(following from memory, may be wrong)

from urllib2 import urlopen
from urllib import urlencode
result = urlopen(url, urlencode(data))


Sorry if this is no help -- I didn't see the original message.


John



From kragen at pobox.com  Wed Jun 26 15:52:17 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 26 Jun 2002 15:52:17 -0400
Subject: Correct code/unit testing (was Re: Efficient python programming...)
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com> <834rgd5isv.fsf@panacea.canonical.org> <3D02FB3B.CF75E925@engcorp.com> <83u1oaq0jt.fsf@panacea.canonical.org> <3D0578C7.ACDBF41A@engcorp.com>
Message-ID: <837kkllswe.fsf@panacea.canonical.org>

Peter Hansen <peter at engcorp.com> writes:
> Kragen Sitaker wrote:
> > Of course!  People are fallible, it's much easier to do a bad code
> > review than a good one.
> 
> I have to confess I have never seen a successful code review, nor
> an environment in which they were used regularly and which had a
> high level of code quality.  Clearly as a result I'm fairly biased
> against them -- I should keep a more open mind on the issue perhaps.

You are experiencing code reviewing whenever you are writing
production code.  Do you find that it improves the code quality?

I've found that both pair programming and less-intense kinds of
reviewing are usually extremely helpful.

> I'm swayed also by the thought that programmers generally rebel
> against them, and that therefore they are likely not the best
> approach in most shops.

Yeah, that can be a problem.  That problem has made it hard to
introduce pair programming in some places I've worked; it largely
depends on the social dynamics of the place.

> Assuming the formal 'inspection' advocated by some heavy-weight
> processes, I think it's of little use on code pair-programmed.  (We
> could also be more general and just say that pair programming is on
> par with inspection and that either is effective in an appropriate
> context.)

I mostly agree that inspection during pair programming finds the bugs
any inspection is likely to find.


From claird at starbase.neosoft.com  Mon Jun 24 20:02:49 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 24 Jun 2002 19:02:49 -0500
Subject: Embed terminal into a Tkinter application
References: <3D1758B4.8FA63EBE@club-internet.fr> <Xns9237F055C1CABcliechtigmxnet@62.2.16.82>
Message-ID: <653AB281E1346AA6.7E81367537EEC2BE.CF6BC8C39D83D26D@lp.airnews.net>

In article <Xns9237F055C1CABcliechtigmxnet at 62.2.16.82>,
Chris Liechti  <cliechti at gmx.net> wrote:
>Fabien H?non <ffjhenon at club-internet.fr> wrote in 
>news:3D1758B4.8FA63EBE at club-internet.fr:
>
>> The title says it all.
>> Has someone already managed to include a Linux terminal into a Tkinter
>> GUI ?
>> 
>> Redirecting the output of a console into the application does not do
>> what I want.
>
>well, what do you want? let us help you!
>
>there is also popen[2,3,4] - you can also redirect stdin.
>
>chris
>
>-- 
>Chris <cliechti at gmx.net>
>

My guess is he wants an xterm(-like thing) in a Tkinter frame.

Information on the subject appears in <URL: http://wiki.tcl.tk/xterm >.
I haven't made the time to translate this into idiomatic Python.  Speak
up if you need help doing so.
-- 

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


From siegfried.gonzi at kfunigraz.ac.at  Sun Jun 23 05:05:02 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Sun, 23 Jun 2002 11:05:02 +0200
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122>
Message-ID: <3D158F3E.1010907@kfunigraz.ac.at>

Bengt Richter wrote:

> Unless you are doing a lot of i/o from the program, sounds like maybe that
> 13% (100-87) may be waiting for swapping to get more virtual space? Or it could be
> coincidental background stuff? (Do you trigger any other processes at some point?)

To cut it here: I tested the Mie code again and it works! The images 
what I get are exactly the same as the one in a specific peer reviewed 
paper. After that calculation I wait 20 minutes in order to watch 
whether the system will freeze or not: it freezes not. THE calculation 
does not use any swap! I have got 256MB RAM but the calculation consumes 
about 20 MB on Linux; there is no need for swapping (side note: my swap 
is 756MB).

In turn I tried the following combinations:

1. The complete calculations without the Mie code (note: the Mie code 
works) and dummy values instead.
Result: The first 10 minutes the system consumes 99.8% of the processor 
(memory consumption is up to 5%) but after 10 minutes it sucks down to 
87% and after 15 minutes the system freezes (I mean there is only the 
power off option then!).

2. I did the same without the Mie and Fortran code. Same situation as in 
point 1.

3. I did the same without Fortran/C/Satellite binary reading -> same 
situation as in point 1 or 2.

4. I turned off every power management for my laptop, but that won't 
cure the situation.


Summarize: The system frezzes after 15 minutes and sucks down to 87% 
system performance, without any notice why (I cannot see why under "top" 
; sometimes I notice that many "kdeinits" appear right after the python 
process)? Exactly the same situation appears on Windows: system goes 
down from 99% to 87% after some time, but on Windows I do not get a 
frozen system.

I then wrote a small Python script in order to test the swap. The Python 
script does constantly claim memory and my swap memory consumption goes 
up to 500MB (system perfromance is only some 5%). After 30 minutes the 
calculation is completed and I can work on without any problems.


That means, if I have got a calculation of the form: processor 
consumption 99% and memory consumption 5% the system fails; if I have 
got a calculation in the reverse order: 90% memory and 5% processor the 
system remains stable.


Maybe this all can be explained that the ActivePython version has a bug 
or something like that. Or that the combination Linux/Python/laptop is 
not optimal. But I am not willing to install the normal Python version 
again (I did it until I noticed that idle was not installed; I  invested 
2 days in order to install idle but was not successfull; ActiveState was 
my isle in the following).

The offer remains valid: if someone is interested in my source code he 
can get it.

I am not any longer willing to try the calculation on my Linux box and 
Python 2.2 (20 Linux power off  shut downs are enough in a day; should I 
destroy my laptop?). In the afternoon I will decide whether I will wait 
6 hours for the calculation on Windows or not. And I will also decide 
whether it will be a good idea to rewrite it in Fortran 90. Python can 
then be used in combination with DISLIN for the plots.


S. Gonzi







From logiplexsoftware at earthlink.net  Tue Jun 18 17:18:23 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 18 Jun 2002 14:18:23 -0700
Subject: python version?
In-Reply-To: <D7NP8.46301$GY.14487785@twister.nyroc.rr.com>
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com>
	<D7NP8.46301$GY.14487785@twister.nyroc.rr.com>
Message-ID: <20020618141823.7b6c36d3.logiplexsoftware@earthlink.net>

On Tue, 18 Jun 2002 20:56:03 GMT
George Hester wrote:

> I just cannot believe what I ran into in this newsgroup.  If you people want
to
> be paid for your suggestions then that is fine.  But this newsgroup does not
> have the facility set up for that to happen.  Then since it don't it is
> courteous not to give advice if that is what you really want. It almost seems
> that I was given half an answer so that the other half could be provided by
> someone that would try to make a buck off me.
> 
> Thank you Duncan for finally seeing the error of thier ways.
> 
> Don't worry I won't return here.  This is not a group to get help but a group
of
> individuals that think the world owes them something when they participate
here.

That'll be $12.95 



-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From phd at phd.pp.ru  Thu Jun  6 08:22:03 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 6 Jun 2002 16:22:03 +0400
Subject: Urllib thread safe?
In-Reply-To: <j43cw07qpz.fsf@informatik.hu-berlin.de>; from loewis@informatik.hu-berlin.de on Thu, Jun 06, 2002 at 10:31:52AM +0200
References: <mailman.1023280808.19305.python-list@python.org> <m3it4xobkb.fsf@mira.informatik.hu-berlin.de> <mailman.1023304328.9010.python-list@python.org> <just-9B7B3B.21511805062002@news1.xs4all.nl> <mailman.1023308888.6405.python-list@python.org> <j43cw07qpz.fsf@informatik.hu-berlin.de>
Message-ID: <20020606162203.D6199@phd.pp.ru>

On Thu, Jun 06, 2002 at 10:31:52AM +0200, Martin v. L?wis wrote:
> >    I am afraid you would have problems mixing calls to the global
> > _urlopener from different threads.
> 
> Can you be more specific as to how precisely such problems could
> arise?

   I said "I am afraid" - that is, I am not sure if this could lead to
probles. But I learned hard that global variables are enemies of
multithreaded programs. And in this particular case the global _urlopener
is an instance of class with long and complex methods. I would be very
worried to share the instance between threads.

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From SBrunning at trisystems.co.uk  Fri Jun  7 11:52:53 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Fri, 7 Jun 2002 16:52:53 +0100 
Subject: Medium to Large Scale Python Deployments
Message-ID: <31575A892FF6D1118F5800600846864DCBD2F6@intrepid>

> From:	Domenic R. Merenda [SMTP:dom at edgereport.put_a_c_o_m_here]
>    Sure, Pedro.  The database server is homegrown C and RPG II code
> running
> on the AS/400 for the majority of our "archived" data.  The live database
> is
> mySQL on a PC.  User interface is PC/400 by IBM, and communcation is
> handled
> over IPX through twisted pair cabling.  Kludgy, but it works so far. :-)
> OS/400 is a wacky operating system.
 
Wacky? Hmm, well, it's pretty different to anything else that I've come
across. Coming from an OS/400 background, Windows and Unix look like
slightly different flavours of the same thing to me. (What do you mean a
program is a file? In OS/400, a file is a file and a program is a program,
dammit, not the same thing at all.)

RPG II? Now that *is* wacky! No structured op-codes at all - you had to do
everything with IFs and GOTOs. Yuck. I hope that anything new is done in
RPGIII at least! (We use RPGIV exclusively for new stuff - the free
formatting rocks. Well, compared to fixed formatting, that is - I still hate
semi-colons and braces everywhere.)

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.




From amckay at merlintechnologies.com  Mon Jun 10 16:07:48 2002
From: amckay at merlintechnologies.com (Andy McKay)
Date: Mon, 10 Jun 2002 13:07:48 -0700
Subject: No such directory?
In-Reply-To: <886c5e4b.0206101158.680c6278@posting.google.com>
References: <886c5e4b.0206101158.680c6278@posting.google.com>
Message-ID: <20020610200748.GE3098@mckay.merlintechnologies.com>

On 10/06/02 12:58 -0700, Robin Siebler wrote:
> When I try to change into a certain directory, I get an errorm message
> that states the directory doesn't exist.  The directory does exist,
> however, and the listdir function shows that it exists.  What gives?
> 
> >>> os.chdir('c:\\18_510GMC1\\18_510GMC1')

Im assuming that you the typo is a cut and paste one? How about:

os.chdir('c:\\18_510GMC1\\18_510GMC')

(No trailing "1")

> >>> os.listdir(os.getcwd())
> ['18_510GMC']

-- 
  Andy McKay
  Merlin Technologies




From paul at boddie.net  Mon Jun  3 08:06:19 2002
From: paul at boddie.net (Paul Boddie)
Date: 3 Jun 2002 05:06:19 -0700
Subject: How to install Python without root priviledges ?
References: <acvmin$svjfb$1@ID-146704.news.dfncis.de> <3cf866fa@news.poly.edu>
Message-ID: <23891c90.0206030406.74e9f423@posting.google.com>

"Chandan Mathur" <chandan_mathur at hotmail.com> wrote in message news:<3cf866fa at news.poly.edu>...
> you have done the right thing.., only without SU priviliges, I think you can
> run python only on the directory Python-2.2.1 as follows,  ./python
> I have the same problem... and it really sucks :-(

Just to clear up this misinformation, inside the 'Python-2.2.1'
directory you should issue the following commands:

  ./configure --prefix=/home/myuser/local
  make
  make install

You get to choose the prefix directory, but let's go off the example
given before - you've set up a 'local' directory in your own home
directory.

Now set up your path. This does require some knowledge of your UNIX
environment, but guessing that you're using 'bash', I would suggest
adding the following line to '/home/myuser/.bashrc':

  export PATH=/home/myuser/local/bin:${PATH}

This won't take effect straight away and I've never had too much luck
with re-reading the shell configuration, so just issue that command at
the shell prompt or restart your shell. You should now be able to
start 'python'.

Paul

P.S. This is undoubtedly covered in the documentation. Indeed, it's
probably in the "instuctions for the impatient" section at the very
start of the 'INSTALL' file...


From pyth at devel.trillke.net  Thu Jun 13 17:28:11 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 13 Jun 2002 23:28:11 +0200
Subject: Type subclassing: bug or feature
In-Reply-To: <aeau48$111$1@panix1.panix.com>; from aahz@pythoncraft.com on Thu, Jun 13, 2002 at 04:08:40PM -0400
References: <aeau48$111$1@panix1.panix.com>
Message-ID: <20020613232811.I6609@prim.han.de>

Aahz wrote:
> Consider the following code:
> 
> class MyStr(str):
>     def contains(self, value):
>         return self.find(value) >= 0
> 
> s = MyStr("hello, world!")
> s = s.capitalize()
> if s.contains('Hello'):
>     print "Found it!"
> 
> It fails with an AttributeError when it calls s.contains(), because
> s.capitalize() returned a str instead of a MyStr.  Anyone want to take a
> whack at defending this as the correct behavior?

class Names(str):
    def __init__(self, namelist):
        str.__init__(self, ":".join(namelist))

what would you expect 

    Names([1,2,3]).capitalize()

to return? An exception? I think it's dangerous to assume anything
about the subtyped classes' constructor. You *do* need a constructor
because strings are immutable so 'str.capitalize' can't modify
anything in place.

if-you-subtype-cake-you-gotta-eat-it-all-ly y'rs, holger




From mcfletch at rogers.com  Thu Jun  6 03:23:41 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Thu, 06 Jun 2002 03:23:41 -0400
Subject: Out-of-process Execution (was:Re: constructive critisism on pythonwin
 IDE)
References: <6bd9f01b.0206050736.f469082@posting.google.com> <admuke$gu8$1@newsreaderg1.core.theplanet.net>
Message-ID: <3CFF0DFD.3030705@rogers.com>

More verbosely, the "stale modules" problem is an issue related to the 
"easy way" to implement IDEs for Python, that is, providing an 
interpreter window.  The problem is that when a Python modules changes 
you need to reload each and every loaded module and figure out all sorts 
of tricky corner cases with inter-module dependencies and imports. 
PythonWin uses the interpreter-based approach, and punts on the reload 
requirement (likely because it's a messy problem).

Reload Requirement:
	In an interpreter window, you generally _want_ to retain previous results 
of calculations.  Some issues:

	You want all objects to update their class pointers to the newly loaded 
classes in the new modules (requires finding every object in the 
system).  You similarly need each import in the system to be re-done, in 
the same order they originally occured (and watching out for weird 
corner cases where the same names in different packages confuse the 
import mechanisms).

	There is code lying around in various usenet posts to address this issue 
by walking the object hierarchy and updating each encountered object. 
It's pretty ugly, but it's often possible to get it working right 80% 
(or whatever) of the time.  The rest of the time, something or other 
hasn't been updated (worst being exceptions and dynamically-generated 
constants).  You spend a lot of time tracking down subtle errors (the 
exception is thrown, you catch it, but it doesn't get caught, 
etceteras).  This approach needs a considerable amount of work to make 
it standard, stable and reliable.

	One approach to simplifying the problems above would be to re-run the 
interactive session on reload/change of a module.  But interpreters are 
often used for one-off calculations.  You don't want to have to re-run a 
30 minute calculation just to reload a particular module, but maybe you 
want all the other objects to be updated save your particular objects. 
If you're going to do this, you need to provide some way of marking a 
particular item static across a reload boundary.  Doable (store a 
reference in some dictionary for later usage), ugly though.  Arguably, 
you could even store all the objects, load the modules "clean" and if 
the user asks for a variable from the pre-reload session, scan it with 
algo above before giving it to them.


But, to my mind, though the approaches above might work, they miss the 
best solution, and the one you see in most (all?) of the newer IDEs for 
Python, namely out-of-process execution IN ADDITION to an interpreter 
(which is still used for "playing around" with what-ifs and interactive 
calculation sessions).  What we could really use is a robust 
(BSD-licensed) library for the following:

	running a python script in a seperate process (cross-platform, with 
ability to specify particular executable, and command-line flags 
(including -i for dropping into gui mode)) in either "standard" or 
"debug" mode

	support stdout, stderr stdin suitable for using as hooks with GUI 
libraries (i.e. ability to drop into an "interpreter" window when the 
script completes, type responses to console-mode programmes, and 
generally interact with the script as if it were being run from the 
console) GUI will be able to keep this window around after execution if 
it just wants to show user results from completed script.

	allow debugging a client script (with inspection support (i.e. be able to 
run a GUI inspector across the pipe/process barrier)
		[ you could do this with a wrapped set of stdout, stderr, stdin pipes 
(differentiating between system and "data" messages) or an out-of-band 
client-server comm mechanism (COM, CORBA, RPC) ]

	kill child process

	stop-at-breakpoint


And, add the following while you're at it ;) :

	provide for "interupt-to-debug" support (press a key, move the client 
process into debug mode) for debugging situations where code falls into 
weird states unpredictably

	"slow execution" modes for use with watch variables

	watch-variables (same interface as inspector above), basically the value 
is passed back on every single step of the programme being debugged.
	

I can't see adding this to PythonWin myself (though I use PythonWin as 
my IDE and find it the most comfortable by far, I can't see spending the 
time figuring out the raw win32api calls to create the various dialogs 
and support you'd need in the application when I just use the console 
for 98% of this kind of work), but if sameer (the original poster) feels 
like putting together a robust, reliable and complete module for 
out-of-process execution, he'd probably find lots of people willing to 
work on it with him, and one of them might be willing to do the Win32API 
coding.  Such a library would also make writing new Python IDEs 
considerably easier.

Just a thought,
Mike


F. GEIGER wrote:
> This a Python issue, not a PythonWin IDE issue.
> 
> Cheers
> Franz
> 
> "sameer" <sameer_ at email.com> schrieb im Newsbeitrag
> news:6bd9f01b.0206050736.f469082 at posting.google.com...
> 
>>First of all, let me start of by saying that Pythonwin is a great IDE.
>> Now for the constructive critisism.  The thing I hate about PythonWin
>>is, that despite the fact that it's a development environment, there
>>is a lot of caching done of modules, and the cached copy is not
>>checked against a potentialy modified copy.  Whenever I change the
>>contents of a file and try to run a module that depends on that file,
>>I always end up with the old copy.  I have to then import and then
>>reload the changed module in the interactive window for it to refresh.
>> It would greatly increase my productivity, if I can specify which
>>modules can be cached.  I understand that there is a speed and
>>development time issue, but I would like a little more fine grained
>>control on how caching works on PythonWin, as the way it's currently
>>implemented now is not acceptable.
> 
> 
> 


-- 
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From eppstein at ics.uci.edu  Sun Jun  2 11:07:31 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Sun, 02 Jun 2002 08:07:31 -0700
Subject: Pop a list from beginning ? and memory saving...
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de> <3CF9EB89.7CA7204B@doc.ic.ac.uk>
Message-ID: <eppstein-564DF8.08073102062002@news.service.uci.edu>

In article <3CF9EB89.7CA7204B at doc.ic.ac.uk>,
 Benjamin Tai <bt98 at doc.ic.ac.uk> wrote:

> You could try a.pop(0), which pop the first instance of the list, like a FIFO.

But he said the list was huge, so this will be very slow.
See e.g. <http://groups.google.com/groups?th=e79e99edc138cb02> for prior 
discussion of this issue and comparison of implementation strategies.

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From zopestoller at thomas-guettler.de  Tue Jun 11 06:03:46 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Tue, 11 Jun 2002 12:03:46 +0200
Subject: Changing the path on Win2k/XP
References: <mailman.1023768255.16926.python-list@python.org>
Message-ID: <3D05CB02.2090406@thomas-guettler.de>

Bjorn Pettersen wrote:

> Is there a way to change the path on Win2k/XP through Python? (I'm
> trying to write a .dll deployment script for our production department,
> and I'd like to give them a "one button" install...)


Do you want to change the current working directory or the PATH
environment variable?

The first can be done with os.chdir()

The second is answered in the other posts


  thomas



From shalehperry at attbi.com  Tue Jun 18 21:38:10 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 18 Jun 2002 18:38:10 -0700 (PDT)
Subject: tomorrow in yyyymmdd format
In-Reply-To: <20020619004409.GA2583@lilith.my-fqdn.de>
Message-ID: <XFMail.20020618183810.shalehperry@attbi.com>

>> 
>> Sure there is :-) time.mktime() can take a "malformed" time tuple and
>> do something sensible with it. You can therefore get a time tuple from
>> gmtime, add one to the day position and call time.mktime() on the
>> result:
>> 
>> >>> x = (2002, 2, 29, 0, 31, 42, 2, 170, 0)
>> >>> time.gmtime(time.mktime(x))
>> (2002, 3, 1, 7, 31, 42, 4, 60, 0)
>> >>>
> 
> Pure luck that this works. Don't complain if it stops workin in Python
> 2.7 and your code breaks ;-)
> 

actually that is a function of the mktime() function in the C library which the
python module wraps.  Based on my reading of a man page or two this behaviour
seems guaranteed by POSIX.  Now, the assumption made is the mktime() function
on YOUR system follows POSIX.




From pyth at devel.trillke.net  Thu Jun 13 20:18:43 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 14 Jun 2002 02:18:43 +0200
Subject: Why does Python mix OO concepts and non OO concepts for opera tion s  on basic types?
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8BF0A387@natasha.auslabs.avaya.com>; from tdelaney@avaya.com on Fri, Jun 14, 2002 at 09:57:12AM +1000
References: <B43D149A9AB2D411971300B0D03D7E8BF0A387@natasha.auslabs.avaya.com>
Message-ID: <20020614021843.R6609@prim.han.de>

Delaney, Timothy wrote:
> > From: holger krekel [mailto:pyth at devel.trillke.net]
> > 
> > probably. With python 1.6 many methods of the string *module* were
> > included included directly with the *type* string ('str'). Most 
> > python developers think that the string module should be deprecated 
> 
> I think "most" is a little bit strong. The general feeling that I get seems
> to be pretty much split down the middle.

maybe. IIRC GvR is in favor of deprecating it (but not any time soon).
He does have not-to-be-underestimated influence on such issues :-)
 
> I personally *don't* favour getting rid of the string module - there are
> many things which lend themselves better to a non-object-oriented style -
> and many of the string functions are included.
> 
> Now, if the str type included *all* of this functionality, *and* was aliased
> to string, then you may have a case 

which i never made. 

> ... although 'import string' would still
> need to work (for backwards compatibility) until 3.0.

yes. My personal favorite is to *revive* the string module and just
deprecate superflous functions for 3.0. But i don't feel this currently is
a majority oppinion over on python-dev.

    holger




From travis at vaught.net  Thu Jun 20 00:08:09 2002
From: travis at vaught.net (Travis N. Vaught)
Date: 19 Jun 2002 21:08:09 -0700
Subject: Pedagogic advice needed
References: <3D104D35.E0932297@info.unicaen.fr> <aeqfuh$88n$1@peabody.colorado.edu>
Message-ID: <ec4802fb.0206192008.318f0e45@posting.google.com>

Fernando P?rez <fperez528 at yahoo.com> wrote in message news:<aeqfuh$88n$1 at peabody.colorado.edu>...
<snip>
> If someone is there to do the install for them, I'd say scipy could satisfy 
> your needs. Installing it is a major chore right now, but once up and running 
> it works very well. You need to run off of CVS though, all releaseed code is 
> outdated now.
> 
<snip>

Actually, nightly builds (with a test scoreboard) are creating
downloadable binaries (from the cvs) that make installation _much_
nicer.  You still need Numeric and wxPython, but windows installs
(;)), at least, are quite simple now.

Check it out:

http://www.scipy.org/site_content/download_list
http://www.scipy.org/site_content/scoreboard


From cgale1 at cox.net  Wed Jun 19 04:25:30 2002
From: cgale1 at cox.net (Van Gale)
Date: Wed, 19 Jun 2002 08:25:30 GMT
Subject: Caching query resuls
References: <aentug$m8d$1@news.onet.pl> <mailman.1024470864.21143.python-list@python.org>
Message-ID: <_dXP8.49532$3R2.1721528@news1.west.cox.net>

"Gerhard H?ring" <gerhard at bigfoot.de> wrote in message
news:mailman.1024470864.21143.python-list at python.org...
> * Pawel Lewicki <lewy0lewy at poczta.onet.pl> [2002-06-18 20:25 +0200]:
> > Hallo,
> > Is there any good solution to cache the results of SQL queries?
>
> Your database has one. It should save at least the query execution plan.
>

Check out SQL Relay at http://www.firstworks.com/sqlrelay.html

It might do what you are looking for and has Python bindings, but I'm in the
same skeptical camp as Gerhard :)

Van





From david.abrahams at rcn.com  Sun Jun 16 10:07:34 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Sun, 16 Jun 2002 10:07:34 -0400
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de> <aeftmj$m29$1@bob.news.rcn.net> <3d0c6d67@nntp.server.uni-frankfurt.de>
Message-ID: <aei645$o66$1@bob.news.rcn.net>

"Michael 'Mickey' Lauer" <mickey at tm.informatik.uni-frankfurt.de> wrote in
message news:3d0c6d67 at nntp.server.uni-frankfurt.de...
> David Abrahams <david.abrahams at rcn.com> wrote:
> > With Boost.Python, what you write looks like a kind of IDL (interface
> > description language), which just happens to be built out of C++ source
> > code. Lots of jobs (function argument type checking, conversion, arity
> > checking) are done automatically, but the library doesn't decide which
> > elements to wrap. Unlike SWIG and a few others, there's no attempt to
parse
> > your source code. Instead, the compile-time introspection capabilities
of
> > C++ are exploited to automatically build wrappers from function and
member
> > function pointers.
>
> I see. So it seems in order to get a nice object oriented Python
interface,
> I can 1.) use SWIG to parse the header file and automatically produce a
wrapper
>       1.1.) and then use this low-level wrapper to write some high-level
python classes around it or
>       2.) use Boost.Python and do the two steps in one by writing a
"handcrafted" interface?

I have no deep familiarity with SWIG, so I really don't know it takes to
produce high-level classes on the Python side with that tool. If you use
Boost.Python, you do get true classes on the Python side which are wrapped
around your C++ classes.

> How would sip come into this picture? Is it more like SWIG for C++ or more
like Boost.Python?
> (Apart from not beeing concerned about the total lack of documentation ;)

I wrote this page more than a year ago; it's surely out-of-date:

http://www.boost.org/libs/python/doc/comparisons.html

Someone needs to freshen it up for the upcoming release of Boost.Python v2;
Since you seem to be doing a survey, would you care to volunteer?

-Dave





From max at alcyone.com  Wed Jun 19 14:20:54 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 19 Jun 2002 11:20:54 -0700
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com> <3D1010C3.C321F8B0@engcorp.com> <raXP8.47926$GY.15260255@twister.nyroc.rr.com> <3D107948.B3FEE4BB@engcorp.com> <CT%P8.33113$uk2.13710586@twister.nyroc.rr.com>
Message-ID: <3D10CB86.3D3D5558@alcyone.com>

George Hester wrote:

> Well I know quite a lot about limits.  Nothing about Maple. But the
> reason I
> mentioned is that I hoped that my students would take an interest in
> the subject
> and learn the concept.

Good for you.  Now did you understand the purpose of his analogy?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Who'd ever think it / Such a squalid little ending
\__/ The American and Florence, _Chess_
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From usenet at thinkspot.net  Sat Jun  1 03:48:33 2002
From: usenet at thinkspot.net (Sheila King)
Date: Sat, 01 Jun 2002 00:48:33 -0700
Subject: How to open a HTML file when the python cgi program is executing?
References: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de>
Message-ID: <ad95l1.i0.1@kserver.org>

On Sat, 1 Jun 2002 16:43:01 +1000, "Ken" <ken at hotmail.com> wrote in
comp.lang.python in article <ad9q91$vlrsb$1 at ID-49758.news.dfncis.de>:

> How do I open a HTML file when the cgi program is executing?
> 
> The only method I can think of, which doesn't work is this:
> <meta HTTP-EQUIV="REFRESH" CONTENT="0; URL=main.html">

What do you mean by "open an HTML file when the cgi program is executing"?

Are you talking about displaying the resulting HTML output.

HTML output from a cgi script is simply printed to standard output.

Here is a very simply cgi script (basically "Hello World"):

=============(begin hello script)===============
#!/usr/bin/env python

print "Content-type: text/html\n"
print "<HTML><HEAD><TITLE>Hello, World!</TITLE></HEAD>"
print "<BODY><H1>HELLO, WORLD!!!</H1></BODY></HTML>"

=============(end hello script)===============

Change the interpreter path

#!/usr/bin/env python

as appropriate for your server.

Save the above as something like hello.py or hello.cgi
upload it to the cgi-bin on your server. Make sure file permissions are set
for executable (this will probably be 777 or 755 depending on your server
configuation). Make sure you upload it in ASCII/TEXT mode.

Then point your browser to

http://www.example.com/cgi-bin/hello.py

where you replace example.com with your actual domain name.

This should display an HTML page with the H1 header saying "Hello, World!!"

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/





From znu at znu.dhs.org  Fri Jun  7 01:31:10 2002
From: znu at znu.dhs.org (ZnU)
Date: Fri, 07 Jun 2002 01:31:10 -0400
Subject: Method call from string
Message-ID: <znu-39AB22.01311007062002@news.fu-berlin.de>

Say I've got a variable with the value 'do_something', and I have an an 
object that happens to have a do_something method. How can I call that 
method based on the value of the variable? Preferably without using 
eval() or similar.

I can't turn up anything about this on Google or in the documentation 
I've searched, but maybe Python just doesn't use the terminology I know 
from Objective-C.

-- 
Right now the Internet gives an awfully good imitation of providing superhuman
intelligence capability, both in terms of the total hardware that is out there
and the fact that the 'net has all these human-equivalent peripheral devices
called "users" that can be appropriated in a distributed way to attack problems.
                    --Vernor Vinge


From gerhard at bigfoot.de  Fri Jun  7 06:22:56 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 7 Jun 2002 10:22:56 GMT
Subject: problem with distutil....
References: <adq006$e29eb$4@hades.rz.uni-sb.de>
Message-ID: <slrnag12c4.9rv.gerhard@lilith.my-fqdn.de>

Uwe Schmitt wrote in comp.lang.python:
> Hi, I use the following setup.py for generating a distribution:
> 
>     from distutils.core import setup
>     setup(name="graph",
> 	  version="1.0", 
> 	  author="Uwe Schmitt",
> 	  author_email="uwe.schmitt at procoders.net",
> 	  url="http://www.procoders.net",
> 	  py_modules=["graph"])
> 
> The module to distribute is graph.py and is located in the same
> directory as setup.py.
> When I do "python setup.py sdist" a dist-folder and a .tar.gz-file
> are generated. But when I unpack this tar.gz, I cant find the module
> graph.py...

Works for me. I tested with _only_ setup.py and graph.py, no other
fancy stuff like MANIFEST* lying around, and it worked.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From SSchukat at dspace.de  Tue Jun  4 04:05:50 2002
From: SSchukat at dspace.de (Stefan Schukat)
Date: Tue, 4 Jun 2002 09:05:50 +0100 
Subject: getting windows processes list
Message-ID: <84257D042AA7D411B3F700D0B7DB9B7C13B17B@PDC-DSPACE>

The idea came from the from the 1.5.2 script from Mark Hammond
[PythonRoot]\win32\scripts\killProcName.py 

A little bit changed : 

#########################################################################
import win32api, win32pdhutil, win32con, sys

def CreateTaskManager():
	try:
		import win32pdh
		import win32pdhutil
	except ImportError:
		print "No performance data helper installed"
		return None

	processinfo,processes
=win32pdh.EnumObjectItems(None,None,"Process",-1)
	TaskManager = {}
	for process in processes:
		TaskManager[process] = {}
		for info in processinfo:
			TaskManager[process][info] =
win32pdhutil.FindPerformanceAttributesByName(process, counter=info)
	return TaskManager

if __name__ == '__main__':
	TaskManager = CreateTaskManager()
	# Show info about process	
	if TaskManager.has_key("pythonwin"):
		for Key in TaskManager["pythonwin"]:
			print "%-40s : %s " % (Key,
TaskManager["pythonwin"][Key][0])



	Stefan


-----Original Message-----
From: Grant Griffin [mailto:Grant_member at newsguy.com]
Sent: Monday, June 03, 2002 6:53 PM
To: python-list at python.org
Subject: getting windows processes list


Hi Gang,

Does anybody know offhand how to use Python to get the list of processes
that
are currently running on Windows?  (I'm essentially trying to get the
information that appears in Window's Task Manager's "Processes" tab.)

not-gifted-at-finding-needles-in-haystacks-ly y'rs,

=g2

_________________________________________________________________________

Grant R. Griffin                                           g2 at dspguru.com
Publisher of dspGuru                               http://www.dspguru.com
Iowegian International Corporation                http://www.iowegian.com

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




From pyth at devel.trillke.net  Mon Jun 17 04:47:37 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 17 Jun 2002 10:47:37 +0200
Subject: a possibly foolish question about slices
In-Reply-To: <4c877253.0206162149.bff3ae3@posting.google.com>; from garth@deadlybloodyserious.com on Sun, Jun 16, 2002 at 10:49:32PM -0700
References: <yu99sn3prbqm.fsf@europa.research.att.com> <eppstein-FCC2A4.12082514062002@news.service.uci.edu> <4c877253.0206162149.bff3ae3@posting.google.com>
Message-ID: <20020617104737.M15079@prim.han.de>

Garth T Kidd wrote:
> > x[p:][:n]
> 
> Does the parser take the appropriate short-cut, here, or do we end up
> with a temporary list?

the parser looks at syntax rather than semantics. You probably mean the 
compiler or bytecode-generator. This is the roughly the bytecode:

>>> f=lambda x: x[3:][:7]
>>> import dis
>>> dis.dis(f)
          0 SET_LINENO               1
          3 LOAD_FAST                0 (x)
          6 LOAD_CONST               1 (3)
          9 SLICE+1
         10 LOAD_CONST               2 (7)
         13 SLICE+2
         14 RETURN_VALUE

The SLICE+1 outcome/sideeffects cannot be predicted by the compiler,
so he simply does what you say.

regards,

    holger




From bokr at oz.net  Sun Jun 23 22:25:52 2002
From: bokr at oz.net (Bengt Richter)
Date: 24 Jun 2002 02:25:52 GMT
Subject: Pythonic clean Ctrl-C loop termination?
Message-ID: <af5vvg$1lg$0@216.39.172.122>

Wondering if this is safe, and/or if there's a more Pythonic(tm) way to do it:
(also wondering how signal.signal etc. plays in embedded interpeter contexts)

--
import signal
class SigHandler:
    def __init__(self):
        self.signaled = 0
    def __call__(self, sn, sf):
        self.signaled += 1

sh = SigHandler()
oldHandler = signal.signal(signal.SIGINT,sh)

# the idea is loop until Ctrl-C
while not sh.signaled:
    # whatever ...
# may one assume loop code 'whatever' has completed as if undisturbed here?

signal.signal(signal.SIGINT,oldHandler)
--

>From the point of view of the handler, what is the granularity of the interrupt? A byte code?
A line? A statement? A code block? Or what is "atomic"?
.../Python22/Doc/lib/module-signal.html says

"Although Python signal handlers are called asynchronously as far as the Python user is concerned,
they can only occur between the ``atomic'' instructions of the Python interpreter. This means that
signals arriving during long calculations implemented purely in C (such as regular expression matches
on large bodies of text) may be delayed for an arbitrary amount of time. "

E.g., what happens in an extension C function that is ignoring interrupt issues? Does it get suspended
while an interpreter instance or thread goes to execute the handler? From the above, apparently not.

But what happens if an extension is running python code via embedding? How would a handler in the
parent python code get control? Does any single handler globally prevent an interrupt from being
turned into a KeyboardInterrupt exception? Or will the embedded python code see an exception and
the handler in the outer context never see anything? If a single handler from any context causes
queing of the handling until the particular context is re-entered, then I guess embedded Python code
would not see anything, and there would just be a long delay until it got done, and the first
context was returned to. Probably the way it should work? Also, if the embedded code called signal.signal()
to set its own handler, IWT that should take precedence. But it would have to restore the other handler,
unless there's some magic cleanup when leaving embedded interpreting?

I guess I could experiment, but I guess someone's been there and/or there's another writeup somewhere?

Regards,
Bengt Richter


From shagshag13 at yahoo.fr  Tue Jun  4 04:48:48 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Tue, 4 Jun 2002 10:48:48 +0200
Subject: int, float and string...
Message-ID: <adhut1$10rtkl$1@ID-146704.news.dfncis.de>

Still wondering questions about pythoner's way...

Why does i have :

>>> int('0.0123456789')
Traceback (most recent call last):
  File "<pyshell#157>", line 1, in ?
    int('0.0123456789')
ValueError: invalid literal for int(): 0.0123456789

The correct way to avoid is it :

>>> int(float('0.0123456789'))
0

???

And to check if a number should be use as an int rather than a float, can i use :

>>> int(float('0.0123456789')) == float('0.0123456789')
0
>>> int(float('5.00000000000009')) == float('5.00000000000009')
0
>>> int(float('5.0000000000000')) == float('5.0000000000000')
1

or i miss a better way  ???

By doing this kind of stuff do i save memory ?
(having a huge list of int and float, casting them to float only for math operations)

thanks,

s13.




From fredrik at pythonware.com  Thu Jun 27 19:14:22 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 27 Jun 2002 23:14:22 GMT
Subject: Newbie: GUI
References: <3D1B9687.8060606@linuxfreemail.com>
Message-ID: <i%MS8.13738$p56.4384836@newsb.telia.net>

"coh" <leaf00 at linuxfreemail.com>:

> Is there a native gui, widget set, for python, or any in the works?

http://www.python.org/doc/FAQ.html#4.13

according to many, tkinter is still the de-facto standard:
    http://www.python.org/topics/tkinter/

if you want to write tkinter-compatible widgets in pure
python, use the Tkinter 3K WCK:
    http://tkinter.effbot.org

search google (or google groups) for more information,
and endless discussions of why toolkit A is better than
toolkit B, or not.

</F>




From contactNOSPAM at id-cool.jasm.org  Tue Jun  4 06:18:38 2002
From: contactNOSPAM at id-cool.jasm.org (id-cool)
Date: Tue, 04 Jun 2002 12:18:38 +0200
Subject: Python and openoffice.com
References: <0n%K8.578$86.11871@twister1.libero.it>
Message-ID: <adi43o$10g1$1@norfair.nerim.net>

You might want to have a look at pyUNO (http://polysorbate.org/?work/pyuno)
UNO is a Corba-like interface that OpenOffice seems to have choosen to 
access its internals.

/mig

CARbomboniere spa wrote:

> Hi,
> 
> Is there a way to access to Openoffice.org 1.0 's spreadsheet (Calc) in
> Python ? Maybe through COM ?
> 
> Any examples (even very basic ones; e.g. writing into a cell) ?
> 
> Thanks in advance,
> 
> Fabrizio



From wurmy at earthlink.net  Thu Jun  6 22:35:48 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Fri, 07 Jun 2002 02:35:48 GMT
Subject: MS SQL connectivity
References: <pan.2002.06.06.16.47.25.714491.1989@abargon.com>
Message-ID: <3D001B84.3030306@earthlink.net>

Joakim Ziegler wrote:
> I'm currently faced with the... interesting task of talking to an MS SQL
> Server 8.0 database running on W2k, from Python running on GNU/Linux. 

MS SQL 8.0, that's the same as 2000, right? If so, you might consider using 
SQLXML. You talk to the server through HTTP, sending either a GET statement 
with an "URL" containing SQL, or a POST statement with an updategram containing 
SQLXML. The result set also comes back as XML.

This way of working is very portable, and you need no drivers, but it's not 
perfect. NULL values are by default excludes from the result set (although you 
can change this), and posting binary data would be a bit of a problem. Also, 
you'll need to write your own basic set of functions/methods to talk to it with 
Python, interpret the results, etc.

If you can't get any of the other methods to work, this may be something to 
consider.

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/



From loewis at informatik.hu-berlin.de  Thu Jun 13 03:40:56 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 13 Jun 2002 09:40:56 +0200
Subject: Extending distutils?
References: <mailman.1023894093.18616.python-list@python.org>
Message-ID: <j4wut37hiv.fsf@informatik.hu-berlin.de>

Martin Sj?gren <martin at strakt.com> writes:

> Does anybody have any hints on what I should do?

To extend a command, you subclass the existing command (or write a new
one from scratch), then pass an argument cmdclass to install, like
this:

       # Override certain command classes with our own ones
       cmdclass = {'install_data':install_Data_Files,
                   'bdist_wininst':wininst_request_delete
                   },

HTH,
Martin


From dave at pythonapocrypha.com  Tue Jun  4 16:04:23 2002
From: dave at pythonapocrypha.com (Dave Brueck)
Date: 4 Jun 2002 13:04:23 -0700
Subject: urllib, urllib2, httplib -- Begging for consolidation?
References: <mailman.1020962597.19539.python-list@python.org> <b0f083db.0205091346.397e35db@posting.google.com>
Message-ID: <477762c2.0206041204.6993e7e2@posting.google.com>

jeremy at alum.mit.edu (Jeremy Hylton) wrote in message news:<b0f083db.0205091346.397e35db at posting.google.com>...
> brueckd at tbye.com wrote in message news:<mailman.1020962597.19539.python-list at python.org>...
> > On 9 May 2002, Paul Boddie wrote:
> > 
> > > For me, what I've mostly been doing with urllib is to connect to
> > > locations and to download files. Indeed, having this functionality in
> > > the standard library is incredibly useful when a server
> > 
> > Right, and I'm not in any way advocating the elimination of this, only
> > that more of the meat of the functionality be built in the corresponding
> > protocol libraries (and not urllib), on top of which is built a thin,
> > general-purpose API that does simple "protocol triage" if you will.
> 
> It's not at all clear that you want all possible logic associated with
> processing HTTP inside httplib.  Rather, specific http-based
> applications may want to use some set of the features.  Thus, I think
> it's helpful to have a separation between the base protocol module and
> a higher level application toolkit.

(this thread was dead, but I ran into a problem today that made me
think of it)

That's fine, but why would so much useful http-specific knowledge live
outside of an http-specific module? If there's some division of the
feature set (base protocol and higher level) it would make sense for
the two modules to be httpcore and httplib (or something) with urllib
built on top.

> That's the rationale for having urllib and urllib2 separate from
> httplib, I believe.  The rationale for having urllib2 in addition to
> urllib is flexibility. urllib requires applications to get exactly one
> set of features; urllib2 lets an application pick the features it
> needs.

Ok, maybe I'm just not understanding how httplib/urllib/urllib2 work
together. For example, what is the correct way to do a HTTP HEAD
request that follows redirects? It's not hard, but it's silly to have
to code it yourself if somebody already did the work for you. Well,
httplib doesn't know how to follow an HTTP redirect, but lo and
behold, urllib does. Unfortunately, there's no way to use it because
in this case because urllib is for opening URLs (the GET is hardcoded
and not easily overridable).

I don't disagree that a layering or division of functionality can be
helpful, I just disagree that the current layering/division makes
sense. I think it's great that opening a URL via urllib will
automatically follow redirects, it's just a shame that you can't take
advantage of that functionality if what you're doing falls outside the
narrow design domain of urllib transactions. :(

-Dave


From bit_bucket5 at hotmail.com  Thu Jun  6 18:59:11 2002
From: bit_bucket5 at hotmail.com (Chris Stromberger)
Date: Thu, 06 Jun 2002 22:59:11 GMT
Subject: Using AT on XP to run python script - MARK HAMMOND PLEASE READ
References: <fa7108b0.0206061106.339d547b@posting.google.com> <mailman.1023392498.12869.python-list@python.org>
Message-ID: <tjqvfucuejl2pbtg1eits0qiuhon7uv4qc@4ax.com>

Thanks for all the suggestions.  Please see the original thread (same
subject, minus the MARK HAMMOND part) for the answer that worked for
me...or read the pertinent snippet below:

    Change "c:\python21\python.exe c:\scripts\script.py" to
"c:\python21\python.exe" "c:\scripts\script.py" otherwise it's trying
to turn
that whole thing into a single file name. 

On Thu, 6 Jun 2002 12:40:05 -0700, "David LeBlanc" <whisper at oz.net>
wrote:

>I did my tests on Win 2000 pro - what OS are you using?
>
>When it comes to using sockets etc. you might be right that some permissions
>are involved. Would be interesting to know what user account 'at' runs
>under.
>
>Since this isn't strictly a Python issue, maybe comp.os.windows (or
>something like that) might better be able to answer the question. You might
>also check the discussion boards on http://www.codeguru.com/index.shtml or
>http://www.codeproject.com/
>
>If you discover the answer I would be interested in knowing too!
>
>Regards,
>
>David LeBlanc
>Seattle, WA USA
>
>> -----Original Message-----
>> From: python-list-admin at python.org
>> [mailto:python-list-admin at python.org]On Behalf Of Chris
>> Sent: Thursday, June 06, 2002 12:07
>> To: python-list at python.org
>> Subject: Re: Using AT on XP to run python script - MARK HAMMOND PLEASE
>> READ
>>
>>
>> Interesting findings.  I know mine didn't work bcs the script I want
>> AT to execute sends me an email.  The script sends an email, then I
>> want it to run another AT command to set itself up to run again at a
>> later time.  But I can't get it to work.  If I run the script
>> manually, it runs, sends email and executes the proper AT command.
>> But when AT fires off the task the script set up, nothing happens and
>> no more AT commands are set up.
>>
>> "David LeBlanc" <whisper at oz.net> wrote in message
>> news:<mailman.1023320948.13063.python-list at python.org>...
>> > This is most curious, but I believe the commands _are_ getting executed!
>> >
>> > I tried the following (assume time is 16:30) "at 16:32 pythonw
>> > j:\python22\tools\idle\idle.py" (without the qoutes of course). At the
>> > appointed time, there was disk activity, but no idle window
>> popped up (idle
>> > was chosen just to get something to pop up!). However, there
>> _was_ a pythonw
>> > process running in the task manager! I think there might be a
>> way to attach
>> > a console or get access to the GUI, but i'm not sure - maybe
>> Mark Hammond
>> > might have a clue (he being the Python Win guru :) ).
>> >
>> > I tried the following:
>> > #at_test.py
>> > import sys
>> > import time
>> >
>> > tst = open("testfile.txt", 'a+')
>> >
>> > tst.write(time.asctime() + "\n")
>> >
>> > tst.close()
>> >
>> > put into an 'at' job as: at xx:xx python j:\python22\at_test.py
>> >
>> > I expected the output from this to be in j:\python22, but to my
>> surprise, it
>> > was in c:\winnt\system32 where cmd.exe resides! testfile.txt
>> contained the
>> > exact time the 'at' job was set to run at too!
>> >
>> > I suspect your last example that produced no visible output did
>> in fact work
>> > :)
>> >
>> > In any case, it works - now if one could just get visible output!
>> >
>> > David LeBlanc
>> > Seattle, WA USA
>> >
>> > > -----Original Message-----
>> > > From: python-list-admin at python.org
>> > > [mailto:python-list-admin at python.org]On Behalf Of Chris Stromberger
>> > > Sent: Tuesday, June 04, 2002 17:35
>> > > To: python-list at python.org
>> > > Subject: Using AT on XP to run python script
>> > >
>> > >
>> > > I'm having trouble getting AT on XP to run a python script for me.
>> > > Have tried:
>> > >
>> > > at xx:xx "c:\python21\python.exe c:\scripts\script.py"
>> > > at xx:xx "start c:\scripts\script.py"
>> > > at xx:xx "c:\scripts\script.py"
>> > >
>> > > and get "the system cannot find the file specified" for the first two.
>> > > The last version does nothing apparently--no error indication or
>> > > anything.
>> > >
>> > > I imagine it is related to user accounts and such.  Anyone have any
>> > > pointers?
>> > > --
>> > > http://mail.python.org/mailman/listinfo/python-list
>> --
>> http://mail.python.org/mailman/listinfo/python-list
>
>



From db3l at fitlinxx.com  Wed Jun  5 20:22:36 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 05 Jun 2002 20:22:36 -0400
Subject: win32: popen works, popen3 causes hang
References: <930dfcf0.0205301058.6917c599@posting.google.com> <adm8u2$8cu$1@news7.svr.pol.co.uk>
Message-ID: <uvg8x6ysz.fsf@ctwd0143.fitlinxx.com>

"G. Willoughby" <never at mind.info> writes:

> i think os.popen3 returns a tuple of file objects, so to use this command
> correctly it should be coded like this:
> 
> childstdin, childstdout, childstderr = os.popen3("spectrum.exe -file
> script.tcl")

How you store the results really doesn't matter.  In your example,
you're doing tuple unpacking, which I agree is generally convenient
and provides a simple way to name elements of the tuple, but there's
nothing "wrong" with just binding a name to the tuple itself.  E.g.,

    output = os.popen3("spectrum.exe -file script.tcl")

is just fine - output is just a tuple in that case.

> to read the error stream use childstderr.read() etc..

The equivalent to that would then be output[2].read().

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From peter at engcorp.com  Sun Jun  2 10:36:18 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 02 Jun 2002 10:36:18 -0400
Subject: Question on time module
References: <3cfa2287_1@news.iprimus.com.au>
Message-ID: <3CFA2D62.6831D1F2@engcorp.com>

Gold Fish wrote:
> 
> Can anyone tell me how to sort the file in the directory according to their
> modification time. Then if i want to look up a find say python.txt Mar 04
> 2002 in the directory can i use glob.glob to do it. I using
> time.localtime(time.time()) to get the current time in, how can i display
> only the file in the directory 5 days agos from the current time. Please
> help me to solve this problem. I tried to use os.stat to seperate the file
> time but it couldn't work.

C:\> rem Create file that appears five days old
C:\> touch -d 05-28-2002 target.txt

>>> import os,time,stat

>>> # Retrieve all files in current directory
>>> files = [x for x in os.listdir('.') if os.path.isfile(x)]

>>> # make each entry into tuple with modification time first
>>> files = [(time.localtime(os.stat(x)[stat.ST_MTIME]), x) for x in files]
>>> files.sort()
>>> print files
[((1999, 4, 23, 22, 22, 0, 4, 113, 1), 'COMMAND.COM'), ((2002, 1, 4, 20, 35, 32,
4, 4, 0), 'CONFIG.DOS'), ((2002, 1, 4, 20, 42, 14, 4, 4, 0), 'AUTOEXEC.DOS'), ((
2002, 1, 5, 0, 52, 58, 5, 5, 0), 'MSDOS.SYS'), ((2002, 1, 5, 1, 57, 10, 5, 5, 0)
, 'SETUPLOG.TXT'), ((2002, 1, 23, 21, 12, 24, 2, 23, 0), 'AUTOEXEC.BAT'), ((2002
, 1, 23, 21, 12, 24, 2, 23, 0), 'CONFIG.SYS'), ((2002, 5, 28, 10, 31, 46, 1, 148
, 1), 'TARGET.TXT'), ((2002, 6, 2, 1, 53, 42, 6, 153, 1), 'SCANDISK.LOG')]

>>> # create tuple for current date (only year,month,day needed)
>>> oldTime = time.localtime(time.time() - 5*24*60*60)[:3]
>>> # make list of files with matching dates
>>> fiveDayOldFiles = [x[1] for x in files if x[0][:3] == oldTime]
>>> print fiveDayOldFiles
['TARGET.TXT']

Note: I wouldn't use exactly the above code for this, since it's not highly
readable, but it should give you the basics.

-Peter


From dave5774 at hotmail.com  Mon Jun 17 23:59:25 2002
From: dave5774 at hotmail.com (Dave)
Date: Mon, 17 Jun 2002 23:59:25 -0400
Subject: Please help with Python Script/MS-Access (DAO)
References: <aell87$3uv$1@netnews.upenn.edu> <pan.2002.06.18.00.09.54.672123.11168@qwest.net>
Message-ID: <aemb70$8mt$1@netnews.upenn.edu>

Thank you Kirby,
I apologize, but I'm not that advanced in Python to know sgml.  I would like
to use more standard commands.  I can use SQL, and I know Python - Access
commands like OpenRecordSet(), TableDefs.count(), Execute(), while not
rs.EOF, Fields().....
Is there any other way to write the program without SGML?  To give you more
of an idea of the way I'd like the program, I wrote another program for
doing a similar thing in Excel, not Access, for a file named sales.txt and
the same 5 columns, but put from a file into 5 columns on excel, and certain
parameters to enter at the beginning... :

ProductID = float(raw_input("Please enter the ProductID: "))
OrderQty = float(raw_input("Please enter the cutoff order quantity: "))
CSV = open('c:/sales.txt','r')
import win32com.client
x1=win32com.client.Dispatch('Excel.Application')
x1.visible=1
x1.Workbooks.Add()
x1.Range("A1").Select()
x1.Range("A1:B1").Value = ("ProductID:",ProductID)
x1.Range("A2:B2").Value = ("OrderQty:",OrderQty)
x1.ActiveCell.Offset(4).Activate()
x1.ActiveCell.Range("A1:E1").Value=("OrderID","OrderDate","RegionID","Produc
tID","OrderQty")
x1.ActiveCell.Offset(2).Activate()
for line in CSV.readlines():
    line=line.rstrip()
inOrderID,inOrderDate,inRegionID,inProductID,inOrderQty = line.split(",")
if float(inProductID) == ProductID:
       if float(inOrderQty) >= OrderQty:
           xl.ActiveCell.Range("A1:E1").Value =
(inOrderID,inOrderDate,inRegionID,inProductID,inOrderQty)
xl.ActiveCell.Offset(2).Activate()

If you could help me out I would highly appreciate it...if you would rather
do this over e-mail I could do that, you can e-mail me at
hobiewon at hotmail.com

Thanks again,
Dave







From akineko at pacbell.net  Sun Jun 30 15:33:07 2002
From: akineko at pacbell.net (Aki Niimura)
Date: 30 Jun 2002 12:33:07 -0700
Subject: A problem with urllib
Message-ID: <b714b8de.0206301133.31c48f98@posting.google.com>

Hi,

I have noticed this problem for a while. 

Using FancyURLopener, you can provide user, passwd to authenticate a page
in your program. (You need to override prompt_user_passwd() function)

This feature works fine for most cases.

However, when I tried to read my DSL router's Status page which 
required an authentication, it failed.

The DSL router's default authentication is:
user   "" <<<-- empty string
passwd "admin"

urllib doesn't like user to be an empty string. Thus, it keeps asking
a new user, passwd input (another call to prompt_user_passwd() ).
And eventually it went crazy (recursive error?).

Although providing an empty string to user input is a rare situation,
I would like to read the DSL router's Status page to get the WAN
address. BTW, you can change the passwd but can not specify the user 
in my DSL router.

>From my Web browser, the following works just fine.
http://:admin at 192.168.1.1/Status.htm

Any workaround?

Thanks in advance.

Best regards,
Aki Niimura


From peter at engcorp.com  Mon Jun 17 20:23:15 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 20:23:15 -0400
Subject: What language - platform use to data acquisiton/numerical work?
References: <1c849ea9.0206130130.6cdaa36f@posting.google.com>
Message-ID: <3D0E7D73.241F8D5D@engcorp.com>

Matteo wrote:
> 
> The work involve:
> - a system of data acquisition (and control) from (of) a PLC via
> serial port. The data acquisition is in real time with low frequency
> (about 1 - 2 hz)
> - a big system of data analisys, pattern recognition, data
> visualization and heavy user interaction mostly for experiment with
> algorithms and for diagnostics
> - a end user system, mostly a wrapping of siplified GUI to the
> previous two subsystems.
> 
> I think I will handle serial port comm vith small C libraries, I don't
> know if Python is suited to the problem.
> Having the base drivers should not be a problem ro realize the data
> acquisition system in Python Or not?

Python, with appropriate serial port modules (as Chris suggested)
will do just fine.  No need for a small C library, and 1-2 Hz
is easily achievable (assuming you do not mean this is a mission
critical hard real-time system) on any PC with Python.

Others have already commented on the data analysis and GUI side.
I just wanted to mention that we use Python for roughly similar
kinds of work (haven't done much on the numerical processing side
yet, for lack of time) and it has never failed us.  Definitely 
a big time saver, and very robust, maintainable code seems to 
result from using it.

-Peter


From hesterloli at hotmail.com  Wed Jun 19 11:07:56 2002
From: hesterloli at hotmail.com (George Hester)
Date: Wed, 19 Jun 2002 15:07:56 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com>
Message-ID: <g71Q8.33179$uk2.13779733@twister.nyroc.rr.com>

Welp:

<%@LANGUAGE=Python%>
<HTML>
<head></head>
<body>
<h1>Python Test</h1>

<%
#do some python stuff here
import sys
Response.Write('Python Test<br>')
Response.write('<h3>Smaller heading</hr>')
Response.Write(sys.version)
%>

</body>
</html>

Thank you Mark McEaheren.  He didn't give me this but pointed me in the right
direction.  I suppose others tried to do the same thing but he gave a nudge
using the command prompt and that is all it took.  Since my knowledge was <0 in
this language all that I read looked Greek to me.

--
George Hester
_________________________________
"George Hester" <hesterloli at hotmail.com> wrote in message
news:K7PO8.16057$uk2.8406537 at twister.nyroc.rr.com...
> What is a simple script that will tell me the version of python I am currently
> running?  IIS 5.0 Windows 2000.
>
> --
> George Hester
> _________________________________
>
>




From tdelaney at avaya.com  Wed Jun  5 20:03:45 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 6 Jun 2002 10:03:45 +1000 
Subject: self
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A358@natasha.auslabs.avaya.com>

> From: Vojin Jovanovic [mailto:vjovanov at stevens-tech.edu]
> 
> Sorry, but it is not that my problem is hard.  It is that the 
> concept of
> having
> to reference instance variables always through self. is 
> making it hard and
> inconvenient which was my point in my original post.

Actually, I think it's more that the problem is not specified correctly.

How would you do this in C? You haven't got any member variables there to do
it with.

How would you do it in C++? You can't execute arbitrary code.

How would you do it in Java? You can't execute arbitrary code.

You are trying to take advantage of a feature in Python (extreme dynamism)
and are then bemoaning that it doesn't work exactly as you would like?

What has happened is that you are no longer concentrating on the problem,
but on how to work around Python to make it work the way you think it should
work for this one problem domain.

The standard way of dealing with this is to create a parse tree for the
expressions and evaluate them, storing the actual values separately.

I think that in this case, the standard method is also the correct one.

Tim Delaney




From ajw126 at york.ac.uk  Wed Jun  5 19:11:40 2002
From: ajw126 at york.ac.uk (Andrew Wilkinson)
Date: Thu, 6 Jun 2002 00:11:40 +0100
Subject: Detecting OS and Window Manager
References: <mailman.1023309760.22815.python-list@python.org>
Message-ID: <uft6liipd6m60e@corp.supernews.com>

Thanks for all the replies guys.

Basically I want to have a program that hides itself in the taskbar of
whatever the user is running. Unfortunately there isn't (or at least I don't
know of) a portable way of doing this. I guess I'll just have to mess around
with looking for the environment variables and then code it up for
KDE/Gnome/whatever.

Thanks again
Andrew

"Sean 'Shaleh' Perry" <shalehperry at attbi.com> wrote in message
news:mailman.1023309760.22815.python-list at python.org...
>
> On 05-Jun-2002 Tim Hammerquist wrote:
> > Andrew Wilkinson graced us by uttering:
> >> I'm writing a program that integrates with a number of different OSes
> >> and Window Managers (under Linux), but I need to check what the user
> >> is running.  I know that...
> >>
> >> import os
> >> if os.name == 'nt':
> >>
> >> will tell me they're running NT. But under Linux I'd like to know if
> >> they are running KDE or Gnome, does anyone know a (preferably) simple
> >> test for this?
> >
> > Well, windowmanagers aren't as static as operating systems are.  It's
> > actually possible to run more than one wm at a time.
> >
> > In any case, I went and experimented with the various wms on my system
> > and in all cases the $WINDOWMANAGER shell var was set to the full path
> > of the running wm executable.  This appears to be the only consistent
> > and reliable var in my environment; for instance, here's a dump of my
> > wm-related vars:
> >
>
> WINDOWMANAGER is not one I am familiar with -- is KDM or GDM setting it?
None
> of the machines I looked at had it set.
>
>




From i.linkweiler at gmx.de  Sat Jun  8 20:29:59 2002
From: i.linkweiler at gmx.de (Ingo Linkweiler)
Date: Sun, 09 Jun 2002 02:29:59 +0200
Subject: Comparison: HTTP with Java, Perl or Python
References: <3D009911.9000904@gmx.de> <3D0127D2.2060604@gmx.de> <B9273737.4982%js_nntp@nospam_blackrocksoftware.com> <3D022634.9050304@gmx.de> <6E63F7FCD191848B.20A99585505D1A67.96A3FE18E66D1AF7@lp.airnews.net>
Message-ID: <3D02A187.5070505@gmx.de>

>
>
>Please don't generalize or conclude much from this.
>Different languages manage "run-time libraries" in
>different ways.
>
Yes, I agree. This is only one EXAMPLE, showing script languages and 
especially python CAN have shorter and better readable source.
This example is very short and to simple for a full review of Python.
A better comparision of languages can be found here: 
http://www.ipd.uka.de/~prechelt/Biblio/jccpprtTR.pdf

If you are interested, we can make a second comparison with a greater 
example and a more specific test condition.

Script languages are often used to solve very simple problems.
Perhaps this can be a better test, it can be solved within a half hour 
with all languages:

Input: An URL, a local dirctory
Task1: recursive download of all jpg-files to local directory from the 
given url, including all sub-URLs (but not links to other sites)
File name of first image in local directory: pic00001.jpg and so on.

or this one:
Task2: Create a "map" of all links: Make a list of all found local links 
and links to other sites (bot do not follow links to other URLs).
Output should be written in a HTML-file "map.htm".


For comparison:
- Time for coding
- Time for testing and debugging
- Number of changes during debugging
- Lines of Code / Packed bytes (without comments)
- Is code readable by a programmer who not has used the language before?

Are you interested? Any suggestions?
We need at least 2 programmers for each language (Python, Perl, Java, TCL)

Ingo



From bcw at rave.ch  Sat Jun 15 09:10:45 2002
From: bcw at rave.ch (Bill C. Wong)
Date: Sat, 15 Jun 2002 13:10:45 GMT
Subject: Integer conversion
References: <3D0B3257.E8572AE5@sympatico.ca>
Message-ID: <p1HO8.303$ua4.290@nwrddc03.gnilink.net>

"Colin J. Williams" <cjw at sympatico.ca> wrote in message
news:3D0B3257.E8572AE5 at sympatico.ca...
> int(string, base) => integer
>
> For int('127',  8), I had expected this function to deliver 87.
>
> The example below correctly gives errors at the top and bottom of the
> range and
> for int('127', 10) delivers 127.  The other cases are not what I
> expected.

Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> int( '127', 8 )
87
>>>


> I would appreciate any suggestions.
>
> Colin W.
>
> for b in range(2,38):
>   try:
>     print b, '[', int('127', b), ']'
>   except:
>     pass
>     print 'error'
>




From emile at fenx.com  Wed Jun 12 14:36:40 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 12 Jun 2002 18:36:40 GMT
Subject: methods to parse big foreign ascii-file?
References: <mailman.1023899956.13534.python-list@python.org>
Message-ID: <XwMN8.170062$cQ3.5000@sccrnsc01>

holger krekel
> tommorrow i want to convince an old skool C-programmer to
> use python. We are going to make a one-day project :-)
>
> It involves reading 50MB+ files of a strange ascii-format
> containing names, floats and flags all around. These items
> span multiple lines and should be grouped in lists or objects.
>
> How would you go about it? Options that come to my mind:
>
> - using the re module to parse line for line and
>   try to get it all right by hand.
>
> - come up with a grammar and let a parser do
>   the hard work (which one?)
>
> - using some module i don't yet know?
>
> thanks for pointers or comments,
>

I would just write it.  I'm assuming some kind of ascii data dump that's
got markers of some type indicating start/end of record/field.  I'd read
in the whole file, break out next record and parse that into
fields/objects/whatever.  I wouldn't try to use re or a parser: it
always makes me feel a step removed from the problem by having to define
it to something that I need to learn first.  You know the saying...
programmer says "... I know, I'll use regular expressions for this
problem" -- now he has two problems...  ;-)  I imagine that those who
are faced with this daily will be comfortable with other available
tools, but by the time I'd come up to speed I'd have moved on to another
project...

HTH,


--

Emile van Sebille
emile at fenx.com

---------



From aahz at pythoncraft.com  Sat Jun 29 21:36:21 2002
From: aahz at pythoncraft.com (Aahz)
Date: 29 Jun 2002 21:36:21 -0400
Subject: can't import generators
References: <m2hejlc060.fsf@mother.paradise.lost> <afl5je$4iu$1@panix1.panix.com> <m2d6u9by8e.fsf@mother.paradise.lost>
Message-ID: <aflnal$aak$1@panix1.panix.com>

In article <m2d6u9by8e.fsf at mother.paradise.lost>,
John Hunter  <jdhunter at nitace.bsd.uchicago.edu> wrote:
>
>> /usr/local/bin/python2.2 -V
>Python 2.2.1
>
>> /usr/local/bin/python2.2 /usr/local/lib/python2.2/test/test_generators.py
>Traceback (most recent call last):
>  File "/usr/local/lib/python2.2/test/test_generators.py", line 1, in ?
>    from __future__ import generators
>ImportError: cannot import name generators

Try re-downloading 2.2.1 and building in a completely new directory.
It's looking to me like you have cross-version contamination, and that's
probably the best way to test it.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From smithj at kryos.colorado.edu  Tue Jun 25 16:41:01 2002
From: smithj at kryos.colorado.edu (Jeff Smith)
Date: Tue, 25 Jun 2002 14:41:01 -0600
Subject: Reference Count Confusion
Message-ID: <3D18D55D.75CAC6EA@kryos.colorado.edu>

I'm a little puzzled about reference counts from Python objects created
in a C extension. In the following example, I return a Numerical Python
array from a C function (function 'add' ). Just before the return,
I printed the reference count, and is was 1, as expected. When I passed
the returned array into another C function (function 'clean'), it's
reference count was 3. This wasn't what I expected, and probably
explains why when I incorporate C functions into my Python programs
I chew up gobs of memory. I'm presumably doing something wrong. Am
I supposed to decrement the reference count somewhere?

Python code:

import Numeric
import foo

a = Numeric.zeros(1000000).astype('f')
b = Numeric.ones(1000000).astype('f')
for i in range(10):
    d = foo.add(a, b)
    foo.clean(d)

C code (module 'foo'):

static PyObject *add(PyObject *self, PyObject *args)
{

  PyArrayObject *a, *b, *c;
  float *cb, *cc, *ca;
  int i;
  long buffer_size;

  if (!PyArg_ParseTuple(args, "OO", &a, &b))
    {
      return NULL;
    }

  buffer_size = a->dimensions[0];

  ca = (float *)a->data;
  cb = (float *)b->data;
  cc = (float *)malloc(buffer_size*sizeof(float));
  for (i = 0; i < buffer_size; i++)
    {
      cc[i] = ca[i] + cb[i];
    }

  c = (PyArrayObject *)PyArray_FromDims(1, (int *)&buffer_size, PyArray_FLOAT);
  memcpy(c->data, cc, buffer_size*sizeof(float));
  free(cc);

  printf("add: %d\n", c->ob_refcnt);
  return Py_BuildValue("O", c);
}

static PyObject *clean(PyObject *self, PyObject *args)
{
  PyArrayObject *c;

  PyArg_ParseTuple(args, "O", &c);

  printf("clean: %d\n", c->ob_refcnt);
  if (c->data != NULL) free(c->data);
  if (c->dimensions != NULL) free(c->dimensions);
  if (c->strides == NULL) free(c->strides);
  Py_DECREF(c);
  printf("clean: %d\n", c->ob_refcnt);

  return Py_BuildValue("");
}

--







From frederic.michoulier at prisme.fr  Tue Jun 11 12:57:51 2002
From: frederic.michoulier at prisme.fr (=?iso-8859-1?Q?Fr=E9d=E9ric_Michoulier?=)
Date: Tue, 11 Jun 2002 18:57:51 +0200
Subject: This is a test.
Message-ID: <00e401c21169$1fa4af80$4064a8c0@pcfm>

This is a test.
Frederic
-------------- next part --------------
A non-text attachment was scrubbed...
Name: left.gif
Type: image/gif
Size: 172 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020611/b10ee4fb/attachment.gif>

From garrett at bgb.cc  Mon Jun 10 17:58:57 2002
From: garrett at bgb.cc (Don Garrett)
Date: Mon, 10 Jun 2002 21:58:57 GMT
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
References: <Xns9229EE722467Fgustafl@195.100.94.182>
Message-ID: <3D0520D4.8030403@bgb.cc>

Gustaf Liljegren wrote:
> I'm making a function that takes one integer and returns a boolean that 
> says if the integer matches a number in a certain group of numbers. This 
> group consists of single numbers aswell as ranges spanning thousands of 
> numbers. I'm looking for an elegant way to write this function.
> 
> Thanks in advance!
> 
> Gustaf

   This very much depends on the group of numbers.

   If the numbers you are testing against are in a list named group, you can 
just do "k in group" to see if the number you are testing is anywhere in the 
list. This is possibly the simplest way to do this.

   If you are going to do many tests against the same group of numbers, put 
the group in a dictionary (using the numbers as keys, and None or 1 or some 
other filler value for the values). Then you can use "k in group" (where the 
dictionary is named group) to see if k is in the group of numbers.

   This is just like using the list, except that the tests will use the hash 
table of the dictionary to speed up the testing. However, you have to spend 
the CPU/memory to build the dictionary, so this mechanism is slower if you are 
only doing a couple of lookups against the same group of numbers.

   Both mechanisms leverage the built in strengths of the language, require a 
minimum of work, and should be easy to understand.

-- 
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc



From tim.one at comcast.net  Sun Jun  9 17:25:32 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 09 Jun 2002 17:25:32 -0400
Subject: array, list, performance...
In-Reply-To: <79b2b0bb.0206091128.2b54271b@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEKDPLAA.tim.one@comcast.net>

[Kragen Sitaker]
> >>> import MetaPy.Speed
> >>> def mklist(n):
> ...     ar = []
> ...     for ii in xrange(n):
> ...             ar.append(ii)
> >>> MetaPy.Speed.seconds_per_call(lambda: mklist(0))
>  1.7825510264682151e-05
> >>> MetaPy.Speed.seconds_per_call(lambda: mklist(1))
>  2.862680041153042e-05
> >>> MetaPy.Speed.seconds_per_call(lambda: mklist(10))
>  7.8054009556034588e-05
> >>> MetaPy.Speed.seconds_per_call(lambda: mklist(100))
>  0.00057740376706547905
> >>> MetaPy.Speed.seconds_per_call(lambda: mklist(1000))
>  0.0055893541133707342
> >>> MetaPy.Speed.seconds_per_call(lambda: mklist(10000))
>  0.055918868174255905
> >>> MetaPy.Speed.seconds_per_call(lambda: mklist(100000))
> 0.59515897688580466
>
> Looks linear to me.

[Scott Gilbert]
> In a very practical sense your test is the one that matters, since
> it's usually time that we care about.  Your test could just as easily
> be measuring how linear the memory allocator is or whether cache
> effects are linear, since either of those could dwarf the cost of the
> list operations.  Who knows, you might actually have a realloc() that
> does something smart.

In the old days, Guido was running on a Solaris box whose realloc eventually
moved a growing list to the end of memory, so that subsequent growth only
resulted in an sbrk call to boost the VM highwater mark.  This was pretty
common, and gave linear time for such tests despite outraged theoreticians'
attempts to demonstrate it was quadratic time <wink>.  The sbrk gimmick
fails if you're growing *more* than one list repeatedly, though, and for
that reason simplified benchmarks may not have anything to do with real-life
performance.  More modern systems eventually move a growing thing into its
own mmap'ed memory segment, and then the true behavior depends on how some
other inscrutable OS subsystem behaves.

> Also, it's pretty tough to tell the difference between O(N) and
> O(N*log(N)) with such small numbers.

Yup!

> Here's some code that just counts the operations:
> [followed by code and some output]
> ...
> The powers of 10 wave up and down a bit, but the powers of 2 make a
> nice .08 .08 .05 pattern.  Still looks linear, but it has bouts of
> being non-linear before the next major step compensates for it.  It
> doesn't look to be O(N*N), and that's what really matters for large N.

It's linear.  Ponder the comment in listobject.c:

	/* Round up:
	 * If n <       256, to a multiple of        8.
	 * If n <      2048, to a multiple of       64.
	 * If n <     16384, to a multiple of      512.
	 * If n <    131072, to a multiple of     4096.
	 * If n <   1048576, to a multiple of    32768.
	 * If n <   8388608, to a multiple of   262144.
	 * If n <  67108864, to a multiple of  2097152.
	 * If n < 536870912, to a multiple of 16777216.
	 * ...
	 * If n < 2**(5+3*i), to a multiple of 2**(3*i).
	 ...

The last line implies that overallocation, when it occurs, is never by less
than n/32 "extra" elements [2**(3*i)/2**(5+3*i) == 1/2**5].  That puts a
lower bound of 3.125% on excess, and that alone is enough to guarantee
amortized linear time.  Let r = 1+1/32.  If the final size is N, then
(ignoring details <wink>) you had to copy N/r elements on the last resize,
N/r**2 on the penultimate resize, N/r**3 on the one before that, etc.  The
total number of element copies is thus bounded above by N/r + N/r**2 + ... =
N/r * (1 + 1/r + 1/r**2 + ...) = N/(r-1) = 32*N.  Add N more for the
individual appends to get an upper bound of 33*N operations.  Then your "N /
operations" statistic is bounded below by N/(33*N) = 1/33 ~= 0.03.  You
never saw a number near that low because overallocation actually varies
between factors of 1/32 and 1/4 (right after a boundary is crossed in the
table) "extra" elements.

This may seem like too-mild overallocation anyway, but "even on Win95" most
reallocs under this scheme don't actually copy anything.  So much for trying
to guess what happens <wink>.





From stuart at bmsi.com  Fri Jun  7 17:57:21 2002
From: stuart at bmsi.com (Stuart D. Gathman)
Date: Fri, 07 Jun 2002 17:57:21 -0400
Subject: Configuration recommendations
Message-ID: <adra81$qf9$1@nntp2-cm.news.eni.net>

I have an application - of the background daemon variety - written in
python.  It has some configuration data like the following:

scan_rfc822 = 1
scan_html = 1           # can be CPU intensive
block_chinese = 0
hide_path = ()          # domains that block visible private nodes
#hide_path = ( 'jcpenney.com' )
block_forward = {}      # people that hate forwards
#block_forward = { 'mycorp.com': ('egghead','busybee') }
log_headers = 0
blind_wiretap = 1
wiretap_users = {}
wiretap_dest = None
#wiretap_users = { 'bigcorp.com': ('disloyal', 'bigmouth') }
#wiretap_dest = '<spy at bigcorp.com>'
check_user = {}
#check_user = { 'mycorp.com': ('joe','mary','bill') }
reject_virus_from = ()
#reject_virus_from = ( 'mycorp.com' )

What is standard practice for incorporating such data?  I would like it
to remain executable python code for maximum flexibility (plus I don't
have to deal with parsing or XML). Should the application exec the config
script?  Can this be done so that the module namespace is updated?
Or should the config script be the entry point, setting up the data, then
calling the main application?


-- 
Stuart D. Gathman <stuart at bmsi.com>
Business Management Systems Inc.  Phone: 703 591-0911 Fax: 703 591-6154
"Confutatis maledictis, flamis acribus addictis" - Mozart background
song for the Microsoft "Where do you want to go from here?" commercial.


From jacob at boris.cd.chalmers.se.cd.chalmers.se  Mon Jun  3 12:36:15 2002
From: jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen)
Date: 3 Jun 2002 16:36:15 GMT
Subject: OT: Crazy Programming
References: <mailman.1021921632.6771.python-list@python.org>
Message-ID: <adg5tv$as2$1@nyheter.chalmers.se>

In article <mailman.1021921632.6771.python-list at python.org>,
David LeBlanc <whisper at oz.net> wrote:
> jacob at boris.cd.chalmers.se.cd.chalmers.se (Jacob Hallen) wrote in
> message news:<ac6lhu$j4n$1 at nyheter.chalmers.se>...
>>
>> The baroque went out of fashion over 200 years ago. Art is not merely
>> decorative and for amusement.
>
>For something so "out of fashion" why do examples cost a bazillion dollars?
>
>Art these days seems in large part to consist of anything that critics can
>be persuaded is art. There's a PBS show running currently called "Egg" that
>examines various facets of the American art scene. One recent artist they
>profiled has a work of art that consists of many strands of spagetti glued
>up into a shape that looks like a big blob of ... spagetti! In truth, some
>of the other stuff this guy did was imo artistic, but the spagetti piece was
>just old cold pasta! Another artist - Yoki Ono - has a piece that consists
>of a video tape of people cutting her clothes off of her; so called
>"performance art". If the people doing the cutting had worn jack boots and
>swastikas, we would have called it inhumane and criminal yet somehow the
>critics call it art.

Without the jack boots and swastikas it is called "lure for gullible idiots".
Witn the jack boots and swastikas it is called "beware of gullible idiots".

Jacob


-- 


From sarmstrong13 at mac.com  Sun Jun 16 10:24:44 2002
From: sarmstrong13 at mac.com (SA)
Date: Sun, 16 Jun 2002 09:24:44 -0500
Subject: Newbie Help Please.(OO and Tkinter)
Message-ID: <B93209DC.7281%sarmstrong13@mac.com>

Hi Everyone-

    I've had some problems finishing this code. I think it stems from my
lack of OOP understanding. Below is the code:

from Tkinter import *
import os
import commands
import tkSimpleDialog

class PyShell(tkSimpleDialog.Dialog):
    def clearin(self):
        self.t1.delete(0.0,END)
    def clearout(self):
        self.t2.delete(0.0,END)
    def expyth(self):
        self.t2.delete(0.0, END)
        self.output = commands.getoutput(self.t1.get(0.0, END))
        self.t2.insert(END, self.output)
    def __init__(self, top):
        self.t1 = Text(top, height="12", width="84", font="Courier 12")
        self.t1.pack(side=TOP, pady=2)
        f = Frame(top)
        f.pack()
        self.b1 = Button(f, text="Execute", command=self.expyth)
        self.b1.pack(side=LEFT)
        self.b2 = Button(f, text="Clear Input", command=self.clearin)
        self.b2.pack(side=LEFT)
        self.b3 = Button(f, text="Clear Output", command=self.clearout)
        self.b3.pack(side=LEFT)
        self.b4 = Button(f, text="Save Input", command=Saving.savin)
        self.b4.pack(side=LEFT)
        self.t2 = Text(top, height="12", width="84", bg="lightblue",
font="Courier 12")
        self.t2.pack(side=TOP, pady=2)
        
class Saving(PyShell):
    
    def savin(self, question):
        Label(question, text="Directory:").grid(row=0)
        Label(question, text="Filename:").grid(row=1)
        self.e1 = Entry(question)
        self.e2 = Entry(question)
        self.e1.grid(row=0, column=1)
        self.e2.grid(row=1, column=1)
        def apply(self):
            self.dir = self.e1.get()
            self.fn = self.e2.get()
            self.sav = self.dir + self.fn
            self.savfile = open(self.sav, "w")
            for line in self.t1.readlines():
                self.savefile.write(line)
            self.savefile.close()
            
root = Tk()
app = PyShell(root)
root.mainloop()



Everything runs until I try to save the text in t1 with the "Save Input"
button. I end up with the following error:

Exception in Tkinter callback
Traceback (most recent call last):
  File "/sw/src/root-python-2.2.1-6/sw/lib/python2.2/lib-tk/Tkinter.py",
line 1292, in __call__
    return apply(self.func, args)
TypeError: unbound method savin() must be called with Saving instance as
first argument (got nothing instead)

What am I doing wrong? Saving should inherit from PyShell correct? And in
order to work correctly saving needs tkSimpleDialog.Dialog, so PyShell
inherits this correct?

Or am I totally wrong here?

Thanks.
SA



From ngcontact at varco.ch  Tue Jun  4 05:02:07 2002
From: ngcontact at varco.ch (Steven Varco)
Date: Tue, 4 Jun 2002 11:02:07 +0200
Subject: Porting python to MVS Host
Message-ID: <adhvh4$v95q8$1@ID-28256.news.dfncis.de>

Hi python developers!

I know, I've got a not very common question, but I hope, to fnd some help
here to...

In our company we developed a python script thats helps us in a bit complex
program structure.
Our problem is now, that we have to run the script on our Host system, that
runs with MVS OS.

So we have to port it... Does anyone here have experiences with porting
python to MVS?

Would be a great help!

Tnx,
  /steven




From peter at engcorp.com  Sun Jun 23 16:55:14 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 23 Jun 2002 16:55:14 -0400
Subject: sys.exc_info() into string
References: <KZoR8.44805$n4.10504231@newsc.telia.net>
Message-ID: <3D1635B2.99BCD944@engcorp.com>

Magnus wrote:
> 
> in my program I am catching exceptions that I might have missed by using
> something like:
> 
> try:
>         doMyStuff()
> except:
>         type = sys.exc_info()[0]
>         syslog.syslog(syslog.LOG_EMERG, "My program: unexpected error" + type)
> 
> The thing is that I can not figure out how to make "type" into a string
> above so I can use it for e.g. syslog(). Is there an easy way of doing
> this?

try:
    1/0
except Exception, ex:
    print "Error: %s" % ex

Good enough??

-Peter


From fredrik at pythonware.com  Wed Jun 26 07:49:59 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 26 Jun 2002 11:49:59 GMT
Subject: Python file upload
References: <mailman.1025088634.8164.python-list@python.org>
Message-ID: <HThS8.13525$p56.4299073@newsb.telia.net>

Bjarne Christiansen wrote:

> #!c:\Python\python -d

do you really need optional debug output from the parser?
maybe you meant "-u"?

> I have changed the script to the following:
> ...
>   item = form["filename"]
>   if item.file:
>     data = item.file.read()
>     f = open("file1.jpg","wb")
>     f.write(data)
> ...
>
> But I still run into the same problem, any ideas?

how much data is missing?  if it's a couple of kilobytes missing
at the end, you might need to explicitly close the file:

    f.close()

</F>




From aahz at pythoncraft.com  Tue Jun 11 13:35:10 2002
From: aahz at pythoncraft.com (Aahz)
Date: 11 Jun 2002 13:35:10 -0400
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Jun 10)
References: <EE3CDB256E0A231C.577F9FCCC77B96EC.B451C214671092A5@lp.airnews.net>
Message-ID: <ae5cce$hpd$1@panix1.panix.com>

In article <EE3CDB256E0A231C.577F9FCCC77B96EC.B451C214671092A5 at lp.airnews.net>,
Kragen Sitaker <kragen at pobox.com> wrote:
>
>Python Development News
>
>    Oleg BroytMann posted glue that automatically moves vim users to

Broytman
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From zopestoller at thomas-guettler.de  Thu Jun 27 07:06:53 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 27 Jun 2002 13:06:53 +0200
Subject: Strings across lines
Message-ID: <3D1AF1CD.8010503@thomas-guettler.de>

Hi!

if I do:

print 'foo ................. %s ' + \
       'bar ................. %s ' % (foo, bar)

I get:

"Not all arguments converted"

What is the nicest way of writing strings across
line breaks?

  thomas



From duncan at NOSPAMrcp.co.uk  Thu Jun 13 04:40:05 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Thu, 13 Jun 2002 08:40:05 +0000 (UTC)
Subject: why not "'in' in 'in'"?
References: <ae8dg70217g@drn.newsguy.com> <3D080482.7A7FD514@alcyone.com> <3D082A62.CF800C75@seebelow.org>
Message-ID: <Xns922C61724D96Dduncanrcpcouk@127.0.0.1>

Grant Griffin <not.this at seebelow.org> wrote in 
news:3D082A62.CF800C75 at seebelow.org:

>> Grant Griffin wrote:
>>  This peculiar lack of generality seems a bit (dare I say
>>  it?) "non-Pythonic".
<lots of snips and then>
> But anyway, let's "just say no" to the tuple case rather than confuse
> ourselves; let's just make the string version of "in" a special case. 

I see a slight lack of consistency in your approach here. First you 
complain that the way 'in' works lacks generality, then when everyone 
points out that to make it work the way you would expect would make it 
break on lists or tuples you suggest special casing.

The Pythonic way is to maintain generality, not to introduce special cases.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From amartin at traza-si.com  Wed Jun 12 11:09:40 2002
From: amartin at traza-si.com (Aurelio Martin Massoni)
Date: Wed, 12 Jun 2002 17:09:40 +0200
Subject: parse tar file with python
References: <ae7498$4jven$1@ID-146704.news.dfncis.de>
Message-ID: <ae7o6s$4oh4v$1@ID-50400.news.dfncis.de>

"Shagshag13" <shagshag13 at yahoo.fr> escribi? en el mensaje
news:ae7498$4jven$1 at ID-146704.news.dfncis.de...
> hello,
>
> i 'm not sure this is the right place to ask, but maybe somebody had
already do some kind of python code on this :
>
> i had hundred thousand of small files, my program is very slow. I'm
thinking that i loose time waiting for opening/closing, and i
> think this could be best if i "tar" my files by blocks.
>
> But in this case i don't know how to check for the end of a file and how
to get its name (both in a tar archive) ...
> (i don't want to tar/untar with python, just to be able to do a sequential
reading of a tar file to get its content)
>
> do you have an idea ?
>
> thanks in advance,
>
> s13.
>
>

Maybe you can use the zipfile module, since it is available with the Python
Library. I have found tar.py in Zope, but only to write tar files, not to
read them. If you don't want to compress the files inside the archives you
create with zipfile, use the ZIP_STORED constant instead of ZIP_DEFLATED.

Hope this helps

    Aurelio





From donn at u.washington.edu  Thu Jun 20 13:26:16 2002
From: donn at u.washington.edu (Donn Cave)
Date: 20 Jun 2002 17:26:16 GMT
Subject: AF_UNIX + SOCK_DGRAM
References: <aes1he$gv2$1@news.idiom.com>
Message-ID: <aet37o$1l0o$1@nntp6.u.washington.edu>

Quoth "James T. Dennis" <jadestar at idiom.com>:

|  I was playing with the socket module and Unix domain sockets,
|  trying to make it work with SOCK_DGRAM.  I could get messages
|  recvfrom() but any attempt to send responses was failing, 
|  complaining about the return address.
|
|  Now I think I understand the problem.  The recvfrom() tuple
|  does contain two parts (which WOULD BE address and data in UDP)
|  but I'm guessing that the address is meaningless for Unix 
|  domain sockets.  It occurs to me that the "return" address 
|  must be the same socket (filesystem node of type 's') as the 
|  one that I did a .bind() to.  

Actually think you have run into a bug in Python's socket module.
If it's convenient for you to build your own socket module, change
it like this and see what happens (I'm looking at version 2.1 here.)

*** socketmodule.c.dist Sun Apr 15 17:21:33 2001
--- socketmodule.c      Thu Jun 20 10:10:12 2002
***************
*** 597,602 ****
--- 597,603 ----
        case AF_UNIX:
        {
                struct sockaddr_un *a = (struct sockaddr_un *) addr;
+               a->sun_path[addrlen - (sizeof(*a) - sizeof(a->sun_path))] = 0;
                return PyString_FromString(a->sun_path);
        }
  #endif


The address returned from recvfrom is (addr, len), but we've been
ignoring len on the theory that the string would be NUL-terminated.
It isn't, so there can be some garbage on the end.

|  This is academic curiosity, and I'll read through the C sources
|  to see what I can glean therefrom, eventually.  I'd like to know
|  how I can tell what process has opened a connection to my 
|  AF_UNIX socket and I'd love to know how to pass an open file
|  descriptor through an AF_UNIX socket (in Python).  I've read that
|  this is possible in C and that it is one of the few features of 
|  AF_UNIX sockets which is NOT available via other address/protocol 
|  families.

Yes, AF_UNIX supports that (SOCK_DGRAM or SOCK_STREAM.)  See man 2
sendmsg.  That would have to be added to the socket module.  Probably
easy enough to do, as long as you confine yourself to support for this
one particular thing (passing file descriptors.)

|  ... I've also heard that this can be used as a way for a
|  client to request read-only or append-only access to a file
|  --- since the UNIX APIs (in C) apparently offer ways for the server
|  to obtain credential information about the requesting process as
|  well as passing open file descriptors (which are then roughly akin
|  to "capabilities" since their access mode can't be changed)

Don't know, right off hand I can't think what API function you
might be thinking of there.

	Donn Cave, donn at u.washington.edu


From pyth at devel.trillke.net  Mon Jun 17 20:29:53 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 18 Jun 2002 02:29:53 +0200
Subject: Python Executable Code
In-Reply-To: <20020618000621.GA3119@lilith.my-fqdn.de>; from gerhard@bigfoot.de on Tue, Jun 18, 2002 at 02:06:21AM +0200
References: <1024357771.569c0ff8rchopra@myrealbox.com> <20020618000621.GA3119@lilith.my-fqdn.de>
Message-ID: <20020618022953.X15079@prim.han.de>

Gerhard H?ring wrote:
> Ger'URL-bot'hard

some day there will be a quick search engine where
people can just type in some keywords and get a good URL. 

oh wait ...




From xah at xahlee.org  Tue Jun 11 18:45:43 2002
From: xah at xahlee.org (Xah Lee)
Date: 11 Jun 2002 15:45:43 -0700
Subject: html parser etc help
Message-ID: <7fe97cc4.0206111445.43b196de@posting.google.com>

i'm learning python in hoping to do switch web programing from perl.
can anyone show me code snipets for the following problem to get
me started?

* snipet to traverse a directory and list files by typer or suffix.

* sample snipet that parse a given html file and print out all <a href> links.

* a sample snipet that fetch and print out the content of a given url.

Thanks. I've more or less read the python tutorial.

 Xah
 xah at xahlee.org
 http://xahlee.org/PageTwo_dir/more.html

perl sample code to show what i mean.

# perl samples
# traverse dir and print out files ending in .html
use File::Find;
find(\&wanted, '/home/xah');
sub wanted {if($_ =~ m/\.html&/){print "$File::Find::name\n";}}

--
# fetch a given webpage
use LWP::Simple;
$content = get('http://xahlee.org/');
print $content;
--
# i'm too sick of reading HTML::Parser
# to understand how it's used.


From jiri at baum.com.au  Sat Jun  1 23:43:50 2002
From: jiri at baum.com.au (Jiri Baum)
Date: Sun, 2 Jun 2002 13:43:50 +1000
Subject: How to open a HTML file when the python cgi program is executing?
References: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de> <ad95l1.i0.1@kserver.org> <ad9vsm$vhjnu$1@ID-49758.news.dfncis.de> <9m6ada.31b.ln@10.0.0.1> <adaeh8$uum7t$1@ID-49758.news.dfncis.de>
Message-ID: <m94cda.nbb.ln@10.0.0.1>

Please attribute stuff when you reply, I don't read everything and almost
missed your question.

>> If you realize this before you've output anything, you can use one of
>> the redirection / refresh headers, such as "Location".

>> print """Content-type: text/html
>> Location: http://replacement-url
>>
>> <p><a href="http://replacement-url">Click here</a></p>

Ken:
> Doesn't seem to work... :(

Does any CGI work for you at all?

If you do just a simple: print "Content-type: text/plain\n\nHello there!\n"
does it show up in the browser?

>> As for frames, there's the main document, which decides which frames get
>> shown; and then the server gets separate requests for each of the
>> frames.  You might want to browse a few pages with telnet to get a feel
>> for how things work.

> Where can I find this document?

Pick a page you like (preferably fairly small).
telnet www.site.com http
GET /rest-of-url.html

Study the output to see how it was done.


Alternatively, get a book on html. Any book on html should do.

Jiri
-- 
Jiri Baum <jiri at baum.com.au>           http://www.csse.monash.edu.au/~jirib
  MAT LinuxPLC project --- http://mat.sf.net --- Machine Automation Tools


From hamish_lawson at yahoo.co.uk  Fri Jun  7 10:11:09 2002
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: 7 Jun 2002 07:11:09 -0700
Subject: Emulating classmethod in Python 2.1
References: <915a998f.0206060833.12ad76a8@posting.google.com> <ado5v8$uih9$1@ID-59885.news.dfncis.de>
Message-ID: <915a998f.0206070611.295149d3@posting.google.com>

Thomas Heller wrote:

> Here's the closest I came up with (you must rewrite your code to
> derive SomeClass from object, the rest can remain unchanged). ...

While ideally I was hoping not to have change any client code, your
solution is the next best thing in requiring only a minimal change and
in working with both Python 2.2's builtin classmethod() and with an
emulating wrapper. Thanks!

Hamish


From rowen at cesmail.net  Thu Jun 27 19:57:19 2002
From: rowen at cesmail.net (Russell E. Owen)
Date: Thu, 27 Jun 2002 16:57:19 -0700
Subject: Help with an algorithm wanted: a solution (improvements gratefully accepted)
References: <rowen-CE3C0F.13020226062002@nntp1.u.washington.edu>
Message-ID: <rowen-A276BF.16571927062002@nntp2.u.washington.edu>

In article <rowen-CE3C0F.13020226062002 at nntp1.u.washington.edu>,
 "Russell E. Owen" <rowen at cesmail.net> wrote a long complicated message
about a generating what is basically a data converter object.

I really appreciate everybody's help on this. I apologize for not making 
the problem clearer in the first place.

I offer my solution below with enough comments to, I hope, make the 
problem clearar. Any critique would be appreciated.

"""Sample object that converts between different data
representations A-D, starting from a given value in one representation.

In reality there will be more representations and the conversion code
will be complicated (though much of it may reside in an external
library of subroutines). But the graph of data representations
connected by conversion functions will always be a constant
and will always be very simple (one hub with lines of varying
numbers of nodes radiating from it).

Figuring out which order to do the conversions (as a function of
what is a given) is solved with a rather simplistic algorithm
that tests too many cases but is short and easy to read.
Thanks to Bjorn Pettersen for his help (but the crudeness is mine).

The main thing I am worried about is: given the order in which to do
the conversions, how best to create an object that implements it.
The solution shown here works, but I suspect there's a better way.
It feels as if I'm creating some sort of meta-object
(by using a function dictionary) instead of making best use
of the existing optimized object mechanism.
"""
class CnvObj (object):
    def __init__(self, givenValue, whatGiven):
        assert whatGiven in ("A", "B", "C", "D")
        self.givenValue = givenValue
        self.funcDict = {whatGiven: self.getGiven}
        self.addFunctions(whatGiven)
    
    def addFunctions(self, fromWhat):
        for toWhat in ("A", "B", "C", "D"):
            if not self.funcDict.has_key(toWhat):
                funcName = "%sfrom%s" % (toWhat, fromWhat)
                if hasattr(self, funcName):
                    self.funcDict[toWhat] = getattr(self, funcName)
                    self.addFunctions(toWhat)
    
    def getItem(self, item):
        return self.funcDict[item]()

    def BfromC(self):
        return "BfromC(%s)" % self.getItem("C")
    
    def CfromA(self):
        return "CfromA(%s)" % self.getItem("A")
    
    def DfromC(self):
        return "DfromC(%s)" % self.getItem("C")
    
    def getGiven(self):
        return self.givenValue
    
    #...etc. only the functions needed for "A given" are shown

co = CnvObj("value for A", "A")
for rep in ("A", "B", "C", "D"):
    print co.getItem(rep)

"""this prints out:
value for A
BfromC(CfromA(value for A))
CfromA(value for A)
DfromC(CfromA(value for A))
"""


From vjovanov at stevens-tech.edu  Fri Jun  7 10:51:17 2002
From: vjovanov at stevens-tech.edu (Vojin Jovanovic)
Date: Fri, 7 Jun 2002 10:51:17 -0400
Subject: self
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu> <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com> <mailman.1023294728.2322.python-list@python.org> <sWsL8.82$v_7.4260962@newnews.cc.stevens-tech.edu> <mailman.1023323409.9676.python-list@python.org> <jpKL8.83$HV1.4324388@newnews.cc.stevens-tech.edu> <mailman.1023409885.8528.python-list@python.org>
Message-ID: <rL3M8.85$Cf4.4420736@newnews.cc.stevens-tech.edu>

> Why such a
> harsh reaction without bothering to correct the obvious 'typo'?

No, it is not the 'typo'.  I corrected that immediately.
Also you had in your message >>> e=Evaluator({'a':'3']), a wrong bracket.
Go back to your post and check it out.  That is why I thought that you
didn't test it.
However,  all that aside, the problems is that your code just din't work
after I corrected those 2 'typos'.  Take a look yourself ....

Python 2.1.2 (#31, Jan 15 2002, 17:28:11) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
IDLE 0.8 -- press F1 for help
>>> class Evaluator:
    def __init__(self, eqs={}):
        self.__dict__['equations']=eqs
    def __getattr__(self, name):
        value = self.equations[name]
        if type(value)!=str:
            return value   # if it's not a string, just give it back
        scope={}
        while 1:
            try:
                return eval(value, scope)
            except NameError, n:
                var=str(n).split("'")[1]
                scope[var]=getattr(self, var)
    def __setattr__(self, name, value):
        self.equations[name]=value


>>> class Berta:
    a=5


>>> e=Evaluator({'a':3})
>>> e.berta=Berta()
>>> e.g='berta.a*a'
>>> e.g
'berta.a*a'
>>>


V.J.




From brueckd at tbye.com  Wed Jun 19 11:55:40 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 19 Jun 2002 08:55:40 -0700 (PDT)
Subject: What If.....    Strong Types
In-Reply-To: <3D1014DC.7090600@bgb.cc>
Message-ID: <Pine.LNX.4.44.0206190800570.10661-100000@ns0.tbye.com>

On Wed, 19 Jun 2002, Don Garrett wrote:

> brueckd at tbye.com wrote:
> > I never understood all the hooplah over ints not being classes. I've only 
> > ever seen a handful of cases where it might be useful, and lots of cases 
> > where it'd get abused, but some people really seem to want it.
> 
>    My desire for it is simple desire for orthagonality. The curse of a CS 
> background, I guess. But I'm curious, how could you really abuse it?

Well, if you can subclass an int, for example, you can then override 
and/or extend its behavior, but since the behavior of ints is pretty well 
understood you're begging for bugs if your subclass does anything 
non-obvious (e.g. you override addition to coerce non-integers differently 
than Python normally would, or you extend subtraction to include some side 
effect). In general, too magical and subtle for me!

The reason I asked, BTW, is because from time to time I play with Ruby and 
nearly _every_ time somebody says they prefer Ruby over Python they 
include the 'ints are types, not classes!' argument on their list, even 
though they generally have no use for ints as classes, so whenever I hear 
that I'm curious to know if they really need it or not. ;-)

> > You might not consider it important, but it'd break a lot of _my_ code, so 
> > it'd be important to me! :) Python doesn't force you to use all its 
> > dynamic abilities, but I regard that as one of the tools of the Python 
> > toolbox, so yes, I'd be hurt by its removal.
> 
>    What kinds of code would it break? That's what I'm trying to understand. I 
> find that my code doesn't take advantage of these features at all.
> 
>    I use function pointers a lot, but I never change/add/delete methods
> at run time, and I always define all members in __init__, even if it's
> just to a default value (or None). Whenever I start to do anything else,
> it feels hacky and hard to follow. Obviously other people feel
> differently.

Dynamically adding and removing methods isn't something I do too often,
but when I do it's usually plugging in different callback methods for
others to use. For example, I've used it as a convenient way to manage
signal/event handling - if an object temporarily needs to be able to
ignore a particular type of event I'll plug in a do-nothing handler. This
isn't always the best approach, of course, but when it is it works well
because the event source doesn't need to know whether or not the object is
ignoring events, and the object itself doesn't need a separate flag or 
variable to track that state either - it's just a part of the object 
itself.

What I use much more often is dynamically adding/removing data members. A
common example is when an object is processed by different components in a
larger workflow, especially when the components in the workflow are not
decided until (or can change during) runtime. Rather than have each
component wrap the object up inside another object so it can add state
information, I can just add that to the object itself, and remove those
bits of info as processing by that component is complete and the object is
moving on to another component.

A simpler example is when it's convenient to use the presence of an object
attribute as an indicator that some work has been accomplished. Say you
have a Package object that represents a FedEx package to be shipped
somewhere, but you often have to weigh it first. 

One approach is to not even define a weight member variable until I've
weighed the package, and I can test to see if I've weighed it already by
checking for the presence of that variable. The alternative is to define a
weight variable up front and then initialize it to some dummy value, e.g.
-1. I believe that when I do this, though, it is more because of my
"upbringing" in other languages rather than it being a good idea. Does the
package really weigh -1? No.  Worse, to make it be "good code" I'll
probably define -1 to be NOT_WEIGHED or something, and check against it so
I'm not using magic numbers. 

If I can break away from my previous language habits, the former approach
often makes more sense than the latter, especially when not all modes of
processing the package would require the package to be weighed (customer
BigCorp just gets charged a flat rate, for example, so the weight member
variable would never be used at all perhaps).

The scenarios in which dynamic object modification works best are ones in
which it's convenient to think of the "type" of the object as one that
evolves over its lifetime, or one that reacts to its use. IOW, the object
may not even _have_ a type that is completely definable at compile time.

It's sometimes hard for me to model my objects this way because of having
had to always do it the other (completely statically defined) way, but
when this approach works the solutions are often quite elegant and
straightforward (btw - I believe that a good portion of bugs and program
cruftiness is related to forcing the programmer to clearly define, up
front, what the complete type of the object is. In many cases this forces
the programmer to make decisions without enough information, in others it
is too constricting, and in others it's just too tedious to make the
programmer worry about all those details, and in all cases general
hackiness ensues).

-Dave





From markus.vonehr at ipm.fhg.de  Mon Jun 10 06:05:46 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Mon, 10 Jun 2002 12:05:46 +0200
Subject: Numeric and PIL: 1D array from 2D-tuple
Message-ID: <3D0479FA.CEFFFAC2@ipm.fhg.de>

Hi,

I have a 2D tuple (an Image) and want to display it using ImageTk
in PIL.
Numeric.choose throws me an error:

Traceback (innermost last):
  File "<pyshell#14>", line 1, in ?
    res = Numeric.choose(MyImage, mask)
ValueError: invalid entry in choice array

Anybody knows how to get a 1D-array to construct the
ImageTk.PhotoImage?

Thanks for any comments,

Markus

code:

MyImage = self.cam.Image # MyImage is a 2D-tuple

mask = [1 for i in MyImage]

MyImage = Numeric.choose(MyImage, mask)
        
new_image = MyImage.tostring()
self.im.fromstring(new_image, "raw", "RGBX", 0, -1)

self.photoimage = ImageTk.PhotoImage(self.im)


From tdelaney at avaya.com  Sun Jun  2 21:37:24 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 3 Jun 2002 11:37:24 +1000 
Subject: Python Turing machine
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A33E@natasha.auslabs.avaya.com>

> From: Chris Tavares [mailto:christophertavares at earthlink.net]
> "G. Willoughby" <never at mind.info> wrote in message
>
> > I belive he means a chat bot that tries to fool humans into 
> beliving its a
> > real person.
> >
> > --G. Willoughby
> >
> 
> No, that's the Turing TEST. The universal Turing Machine 
> (both were created
> by Alan Turing) is the theoretically minimal general purpose 
> computer. It's
> an interesting bit of CS theory, but not particularly useful 
> in day-to-day
> life.

Yeah - the dependence on an infinitely-long tape can be a real bugger ...

:)

Tim Delaney




From dfackrell at DELETETHIS.linuxmail.org  Wed Jun 26 10:12:13 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Wed, 26 Jun 2002 08:12:13 -0600
Subject: getting python source
References: <afcgv0$aie$1@news.geneseo.net>
Message-ID: <3d19cbbf$1_2@hpb10302.boi.hp.com>

SteveK,

>From http://www.python.org you can choose Download, then your desired
version, and scroll down to the "All others should" paragraph which has the
link.  For 2.2.1, it leads to:

http://www.python.org/ftp/python/2.2.1/Python-2.2.1.tgz

--
Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.


"SteveK" <killes at geneseo.net> wrote in message
news:afcgv0$aie$1 at news.geneseo.net...
> Is the CVS repository the only way to get the source for Python 2.2?  Is
> there a .zip file somewhere I can just download?




From peter at engcorp.com  Fri Jun 28 08:49:41 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 28 Jun 2002 08:49:41 -0400
Subject: Dialing out on MODEM
References: <mailman.1025235517.13517.python-list@python.org>
Message-ID: <3D1C5B65.6CBA96A3@engcorp.com>

David McInnis wrote:
> 
> I need to write a small script that opens com3 and writes an ATDT to
> dial out, answer the line and send a numeric message.  Basically, I
> paging utility.
> 
> I am using pyserial and cannot seem to get my modem to dial with
> 
> Modem.write("ATDT5555555")

Is Modem an object from pyserial, or one of yours?  And does it
automatically append the \n or CR or do you have to do that yourself?

-Peter


From peter at engcorp.com  Wed Jun 19 19:46:32 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 19 Jun 2002 19:46:32 -0400
Subject: Dispatching a Win32 COM in the background?
References: <U_6Q8.39803$ks6.8500@nwrddc02.gnilink.net>
Message-ID: <3D1117D8.9AAEF2F4@engcorp.com>

"Steven M. Castellotti" wrote:
> 
> 2) Make the call inside a thread which produces the same effect.

This might work (I don't know anything about COM though).  Try putting
your own code inside the run method of a thread:

import threading
class Background(threading.Thread):
    def __init__(self):
        threading.Thread.__init__(self)

    def run(self):
        print 'Starting'
        pass   # your own code goes here
        print 'Done'

>>> b = Background()
>>> b.start()
Starting
<possibly a pause here?>
Done

If at this point you get a >>> prompt back right away, but the
speech generation is carrying on "in the background" then you're
all set.  Make sure you do this with code you've already tested
outside of the thread, so you'll know that any failure is not
the fault of your own code...

-Peter


From asencio at mitre.org  Tue Jun 18 20:59:17 2002
From: asencio at mitre.org (Angel Asencio)
Date: Tue, 18 Jun 2002 20:59:17 -0400
Subject: calling wxPython from the shell
Message-ID: <3D0FD764.4937BDB4@mitre.org>

Greetings:

  I have been requested to write a function that will take a list, open
a selection window, and then when the window is closed the function
would return a list with the selected items.

  Looks like wxPython is quite favored, and I decided to use the Boa
Constructor, finished the tutorial. Everything fine and dandy, until I
wanted to create the function that you can call from the shell.

  The scenario that I have tried is for the function to call the MainApp
(that has the
MainLoop) with the Frame that will do the selection. I understand about
the events, but what ever I do with the event, 'myModule.myFunction( [
"a","collection","to","select"])' return nothing on the combinations I
have tried.

  Bottom line is that I want a function call that I can call in the
shell, like as part of a  print, that I do the selection, and when I
press ok on the window, the window closes and the print completes.

  Thanks to any pointer,

--
-Angel Asencio
The MITRE corp.
asencio at mitre.org




From erict at millfilm.co.uk  Fri Jun 28 04:36:20 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Fri, 28 Jun 2002 09:36:20 +0100
Subject: Python 2.2 __slots__ Bug?
References: <mailman.1025089474.17780.python-list@python.org>
Message-ID: <3D1C2004.C845DB8B@millfilm.co.uk>

I am running on [GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
and your script DOESN'T   crash.
I am a  notSoNewPythonWriter/longTimeC++Writer and I am always
concern  about memory(I guess like anybody).
I just started writing a big app in python,  and  for
now I decided to not be to worry about memory and to rely as much as
possible on python.
My question is simple: Will it crash the same way if the range call what
in a function and I let python cleanning after itself:

def doStuff(x):
    o = None
    for i in xrange(x):
        o = foo(o)


# x = 43550
x = 3
doStuff(x)
print "I'm fine."
doStuff(x+1)
print "I'm dead."



Glyph Lefkowitz wrote:

> I'm sorry if this is a known issue.  I couldn't get through to the bugtracker
> at sourceforge.net; it's abysmally slow and I stopped waiting for the search to
> run after about 15 minutes.  Searches on google yielded nothing similar for
> 'python __slots__ bug'.
>
> I think I have discovered a weird bug in the Python interpreter.  Under
> Python2.2, when I run this script:
>
>     class foo(object):
>         def __init__(self, n):
>             self.n = n
>         __slots__ = ['n']
>
>     o = None
>     for i in xrange(43550):
>         o = foo(o)
>     del o
>     o = None
>     print "I'm fine."
>     for i in xrange(43551):
>         o = foo(o)
>     del o
>     print "I'm dead."
>



From derek at wedgetail.com  Thu Jun 27 07:13:50 2002
From: derek at wedgetail.com (Derek Thomson)
Date: Thu, 27 Jun 2002 11:13:50 GMT
Subject: Strings across lines
References: <3D1AF1CD.8010503@thomas-guettler.de>
Message-ID: <3D1AF36B.7000202@wedgetail.com>

Thomas Guettler wrote:
> Hi!
> 
> if I do:
> 
> print 'foo ................. %s ' + \
>       'bar ................. %s ' % (foo, bar)
> 
> I get:
> 
> "Not all arguments converted"
> 
> What is the nicest way of writing strings across
> line breaks?

Your problem is actually one of precedence. The % has higher precedence 
than the +.

This will work:

print ( 'foo ................. %s ' + \
         'bar ................. %s '     ) % (foo, bar)

But notice that now the '\' isn't necessary, as Python is smart enough 
to realize that the unclosed parantheses indicates that this line will 
continue. So this will work, too:

print ( 'foo ................. %s ' +
         'bar ................. %s '   ) % (foo, bar)

BTW I always use parentheses to split lines even if they aren't 
necessary (unlike this case). I just find it more readable and maintainable.

Regards,
Derek.



From deathtospam43423 at altavista.com  Sun Jun 30 09:10:04 2002
From: deathtospam43423 at altavista.com (netvegetable)
Date: 30 Jun 2002 13:10:04 GMT
Subject: Great OOP tutorial for newbies
Message-ID: <Xns923DE797FA160AusYourStandingInIt@130.133.1.4>

My fellow newby programmers. I have found a pretty good web tut on OOP. I 
found it pretty simple an straight forward. 

http://www.devshed.com/Server_Side/Python/OOPWithPython/OOPWithPython1/page
1.html

One tip I would offer, I found it was best to save the examples provided, 
to my python Lib directory as scripts, then import them into the command 
line, e.g...

from veryBigSnake import *

This worked much better for me actually typing the examples straight into 
the command line. (P'raps cuz I'm using win, and the examples were written 
on linux?)

Anyway, I really appreciate the help that people on this newsgroup have 
provided me and other newbies. I try to pitch in, by providing links for 
these other newbies. Hopefully, if the newbies find these links useful, the 
real pythonistas can waste a little less of their time answering stupid 
questions for people like you and me. (Well, it's something atleast isn't 
it?) 
-- 
netvegetable at excite.com


From gerhard.haering at gmx.de  Wed Jun 26 16:11:36 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 26 Jun 2002 22:11:36 +0200
Subject: py2exe erro message???
In-Reply-To: <3D19FB0A000002A8@www.zipmail.com.br>
References: <3D19FB0A000002A8@www.zipmail.com.br>
Message-ID: <20020626201136.GB18029@lilith.my-fqdn.de>

* jubafre at zipmail.com.br <jubafre at zipmail.com.br> [2002-06-26 16:41 -0300]:
> C:\Python22\Lib\SITE-P~1\py2exe>c:\python22\python setup.py py2exe
> usage: setup.py [global_opts] cmd1 [cmd1_opts] [cmd2 [cmd2_opts] ...]
>    or: setup.py --help [cmd1 cmd2 ...]
>    or: setup.py --help-commands
>    or: setup.py cmd --help
> 
> error: invalid command 'py2exe' (no module named 'distutils.command.py2exe')
> 
> what does mean this error??

Do you actually import py2exe in your setup.py script, like described in
http://py2exe.sourceforge.net/? Else the new command isn't registered.

Btw. what are you doing in the C:\Python22\Lib\SITE-P~1\py2exe
directory? Did you actually put your setup.py in this directory? Don't.
Any other directory outside c:\python22 will do fine.

It's also making ones life easier to put c:\python22 in the path. Dunno
why the Python installer doesn't do this itself. Also, at least on the
Windows' I use, you can start a Python script by just using its name,
without explicitely feeding it to python.exe, like this:

setup.py py2exe

This works because .py, .pyc, .pyo, .pyw are associated with python.exe.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From michael at damaru.com  Sun Jun  9 13:40:42 2002
From: michael at damaru.com (Michael Davis)
Date: Sun, 09 Jun 2002 13:40:42 -0400
Subject: How to capture stdout into a string?
References: <2YLM8.905$Yw3.119357@news20.bellglobal.com> <3D039285.72DFC71D@engcorp.com>
Message-ID: <moMM8.91$Vr2.130316@news20.bellglobal.com>

Peter Hansen wrote:

> Michael Davis wrote:
>> 
>> Hi,
>> 
>> I'm using libftp. The FTP.dir() function gets the remote directory
>> listing and prints it to standard out. How can I capture that in a
>> string?
>> 
>> If I were using an external process, I know I could open a pipe to it.
>> But it's not an external process.
>> 
>> I'm sure there's a tricky way of doing it by getting the file descriptor
>> for stdout and reading from it directly, but I was hoping to avoid going
>> that low-level. I will if I must though.
> 
> The best option by far (and really the easiest) to figure out
> stuff like this is to check the source.
> 
> In this case, you can see that dir() calls retrlines() with
> the LIST argument and no callback function.  As the notes
> say, this directs the output to stdout.
> 
> You must specify a callback function of some kind.  The best
> is probably a small class with a __call__ method, but this
> little baby seems to work too, although it may not be quite
> enough to work in the general case yet.  It grabs lines that
> are passed to it into a list, until it is called with an
> empty argument, at which point it returns the saved list
> and deletes it, ready for another bunch of lines:
> 
>>>> def grablines(line='', lines=[]):
> ...    if line:
> ...       lines.append(line)
> ...    else:
> ...       result = lines[:]
> ...       del lines[:]
> ...       return result
> ...
>>>> grablines()
> []
>>>> grablines('test')
>>>> grablines('test')
>>>> grablines('test')
>>>> grablines()
> ['test', 'test', 'test']
>>>> ftp.dir(grablines)
>>>> grablines()
> ['total 62', '-rw-r--r--  1 engcorp  netdial      0 Jan 11  2000
> [.addressbook',
> '-rw-------  1 engcorp  netdial   2285 Jan 11  2000 .addressbook.lu',
> '-rw------
> -  1 engcorp  netdial   6459 May 21 23:42 .bash_history', '-rw-r--r--  1
> engcorp
>   netdial     29 Jun 17  2001 .forward', '-rw-------  1 engcorp  netdial 
>   12221
> .............]
> 
> 
> -Peter

Thanks again, Peter. I would have accepted the answer RTFM too - I realized 
that I hadn't read the FTP.dir() documentation carefully enough!

You've saved me a lot of time today!

Michael

-- 
Michael Davis
Damaru
Custom Programming - Web Development - Database Design
http://www.damaru.com
416-540-1284


From mkelly2002 at adelphia.net  Sat Jun  8 13:23:30 2002
From: mkelly2002 at adelphia.net (Michael Kelly)
Date: Sat, 08 Jun 2002 13:23:30 -0400
Subject: Removing ^M
References: <mailman.1023549912.12643.python-list@python.org> <adteeb$250i1$4@ID-100555.news.dfncis.de>
Message-ID: <adteif$250i1$5@ID-100555.news.dfncis.de>

Michael Kelly wrote:

> If you're on a unix flavor this works pretty well
> 
> tr -d '\015' <infile >outfile

Whoops!  Should just be
tr -d '015' <infile >outfile

No need for the backslash.

-- 

"Only choice is an oxymoron."



From rnd at onego.ru  Tue Jun 25 00:33:34 2002
From: rnd at onego.ru (Roman Suzi)
Date: Tue, 25 Jun 2002 08:33:34 +0400 (MSD)
Subject: Suggestions for good programming practices?
In-Reply-To: <GCEDKONBLEFPPADDJCOEMEOODLAA.whisper@oz.net>
Message-ID: <Pine.LNX.4.44.0206250823070.21468-100000@rnd.onego.ru>

On Mon, 24 Jun 2002, David LeBlanc wrote:

>Grrrr!
>
>Look at it this way:
>
>	a,b,c = func()
>
>	abc = func()
>	a = abc[0]
>	b = abc[1]
>	c = abc[2]
>
>Which would you rather write? In C or C++ you have to use an analog of the
>second way since the first way isn't an option. I _like_ this behavior of
>Python.
>
>It's a piddling point as to how Python returns the info - they come to rest
>in discrete variables, not as elements of one aggregate variable. The
>original point was about the flexibility of Python not some technical
>hair-splitting about what Python might put on the stack.


The 'a,b,c = func()'-style is good as far as there will be 
no soon change to the number of variables.

I'd not advice using this style as I fight it now in my own 
old scripts...

Now I am using dicts or instances to hold values:
it gives context information:

       abc = func()
       ... use abc["first"]
       abc["second"] ...
       ... abc["third"]

or (in the less trivial cases):

       abc = func()
       ... use abc.first
       abc.second ...
       ... abc.third

Or I am using 'constants' to access return:

       FIRST,SECOND,THIRD = range(3)

       ...

       abc = func()
       ... use abc[FIRST]
       abc[SECOND] ...
       ... abc[THIRD]


- all these allow variable number of returned values thus
keeping compatibility. (for example, I can access old
and new logs by the same code).

Well, to be honest, I am also using this approach:


a,b,c,d,e = (func()+ (None,)*2)[:5]

when originally there were 3 vars and now I sometimes get
3, sometimes 5.
but it is not very nice...

- those my examples aren't probably good programming practices,
but they allow not to loose context when flexibility is needed.

>David LeBlanc
>Seattle, WA USA

Sincerely yours, Roman Suzi
-- 
rnd at onego.ru =\= My AI powered by Linux RedHat 7.2





From whisper at oz.net  Wed Jun 26 18:13:54 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 26 Jun 2002 15:13:54 -0700
Subject: Be the first on your block to register your computer.
Message-ID: <GCEDKONBLEFPPADDJCOEAEGKDMAA.whisper@oz.net>

Yes! exciting times are coming again as Intel, AMD and Microsoft (among
quite a few others) are banding together to once more put ID numbers (and
now basic encryption) onto next generation silicon. Microsoft will then
start offering it's "secure" (from M$? heh!) OS that will only permit
certified apps to run, not to mention certified music and video.

http://www.theregister.co.uk/content/3/25892.html - general story featuring
Intel's role
http://www.theregister.co.uk/content/4/25891.html - MS to trash GPL software
http://www.theregister.co.uk/content/4/25878.html - Ms' secure OS plans
(vaguely)
http://www.theregister.co.uk/content/54/25905.html  - AMD info on
"Palladium"

Looks like in the future, you'll use your money to buy their computer that
will run what they want you to.

David LeBlanc
Seattle, WA USA





From jepler at unpythonic.net  Mon Jun 17 07:53:52 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Mon, 17 Jun 2002 06:53:52 -0500
Subject: locale module & threads (was Re: I feel stoopid ... this code is to verbose)
In-Reply-To: <3D0D919C.1010101@mxm.dk>
References: <3D0A1791.3060307@mxm.dk> <871yb922el.fsf@bunty.ruud.org> <3D0D919C.1010101@mxm.dk>
Message-ID: <20020617065347.A987@unpythonic.net>

On Mon, Jun 17, 2002 at 09:37:00AM +0200, Max M wrote:
> "setlocale() is not thread safe on most systems. Applications typically 
> start with a call of::
[...]

Since the locale.format and ._group are written in Python, you could have
taken their code and rewritten them to not use the system locale, but
your values of grouping, thousands_sep, and decimal_point.

This is actually an interesting question.  I can certainly see the use of
having a web application that prepares a different locale for each user,
and where multiple simultaneous requests might run in different threads.
It seems that in such situations, it might be valuable for Python to
provide some sort of threadsafe locale support.

Is there something about this situation I'm missing?  It's bound to hurt
performance if every locale.format() / locale.gettext() call has to take a
lock.

Jeff




From Chris.Barker at noaa.gov  Wed Jun  5 15:35:53 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Wed, 05 Jun 2002 12:35:53 -0700
Subject: Fastest way to read / procsess / write text?
References: <3CFD5229.2060803@excite.com> <200206051234.g55CYsB06651@helios.gatwick.geco-prakla.slb.com> <mailman.1023280867.19639.python-list@python.org>
Message-ID: <3CFE67E7.EE0B28E8@noaa.gov>

Martin Franklin wrote:
> Havine re-read you original version this is how to do it;-)
> 
> while 1:
>     data=infile.read(100000)
>     if not data:
>         break ## EOF reached
>     ## process data.....

I think this could potentially split the file in the middle of one of
the expressions that is being searched for.

-CHB
-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From martin at strakt.com  Wed Jun 12 11:00:28 2002
From: martin at strakt.com (Martin =?iso-8859-1?Q?Sj=F6gren?=)
Date: Wed, 12 Jun 2002 17:00:28 +0200
Subject: Extending distutils?
Message-ID: <20020612150028.GA10539@strakt.com>

Hello.

There is so far no chapter in the distutils manual about extending
distutils. However, I think I need to do that. When talking about
bdist_rpm it says "a better way to do this is probably to override the
standard bdist_rpm command with one that writes whatever else you want to
the .spec file" so I suppose I want to do that.

Does anybody have any hints on what I should do?


Regards,
Martin

-- 
Martin Sj?gren
  martin at strakt.com              ICQ : 41245059
  Phone: +46 (0)31 7710870       Cell: +46 (0)739 169191
  GPG key: http://www.strakt.com/~martin/gpg.html




From observer at NOSPAM.space.pl  Thu Jun 13 04:39:40 2002
From: observer at NOSPAM.space.pl (Johann)
Date: Thu, 13 Jun 2002 10:39:40 +0200
Subject: How to check version of Python under CGI
Message-ID: <rbmggus7ln1uiutktf5libf4bbngu29ji7@4ax.com>

I would like to write a simple CGI script to get as much info about
installed Python as possible. But I even don't know how to check
version of Python. :( From shell command  it is easy, but how to do it
from CGI script? Eg.

#!/usr/bin/env python
print "Content-type: text/html\n\n"
print "Python version: %s" %s (IDoNotKnowWhatToChoose(),)

--
Johann



From menion at asylumwear.com  Mon Jun 24 10:25:24 2002
From: menion at asylumwear.com (Joshua Schmidlkofer)
Date: Mon, 24 Jun 2002 07:25:24 -0700
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at>	<3d149f19$0$53133$edfadb0f@dspool01.news.tele.dk>	<af28vp$j57$1@panix1.panix.com> <3D14AD7F.3D5D243F@engcorp.com> 	<af2ia8$a92$1@panix1.panix.com> <1024809043.14694.2.camel@localhost.localdomain>
Message-ID: <3D172BD4.3040001@asylumwear.com>

I have always said, 'Nut loose behind the keyboard'.

Cliff Wells wrote:

>On Sat, 2002-06-22 at 12:14, Aahz wrote:
>  
>
>>In article <3D14AD7F.3D5D243F at engcorp.com>,
>>Peter Hansen  <peter at engcorp.com> wrote:
>>    
>>
>>>Aahz wrote:
>>>      
>>>
>>>>Ya know, you could have saved yourself a lot of typing by simply saying
>>>>PBCAK.  ;-)
>>>>        
>>>>
>>>To save us from those who can't Google:  Problem Between Chair And Keyboard.
>>>(It was new to me, too.)
>>>      
>>>
>><grin>  Ya coulda ast me, too.
>>    
>>
>
>I'd always heard of it called "chair-keyboard interface problem".
>
>Cliff
>
>
>
>  
>







From phlip_cpp at yahoo.com  Thu Jun 13 23:26:18 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 14 Jun 2002 03:26:18 GMT
Subject: More Tkinter Help please...
References: <B92EC762.70EC%sarmstrong13@mac.com>
Message-ID: <aebnoq$ac0@dispatch.concentric.net>

SA wrote:

> class PyShell:
>     def __init__(self, top):
...
>         self.b1 = Button(f, text="Execute", command=self.expyth)
>         self.b1.pack(side=LEFT)

>     def expyth(self):
>         output = os.popen("python -c").t1()

Those two 'def' markers /look like/ they are at the same level. But if your 
editor expands tabs to 4 spaces, and if the first def had 4 spaces before 
it, and the second def has 1 tab, Python's wonderful indent-based blocking 
system will assume you like mixed indent modes, and then assume that a tab 
is 8 spaces.

Then it will perceive the expyth as /inside/ the def __init__, where it's 
not a member of PyShell. So the dot inside self.expyth will not be able to 
find such a member.

Turn off tabs in your editor, perform a Tab to Space conversion (some few 
editors support that), and then manually adjust the block indentions.

Style tip: Give everything longer, more pronouncable, and more intent 
revealing names after you get this working.

-- 
  Phlip
           http://www.greencheese.org/DontPlanDesigns
  --  Set phasers on illin'  --



From hesterloli at hotmail.com  Tue Jun 18 04:12:14 2002
From: hesterloli at hotmail.com (George Hester)
Date: Tue, 18 Jun 2002 08:12:14 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com>
Message-ID: <yXBP8.44417$GY.13865423@twister.nyroc.rr.com>

Listen you aren't my daddy.  If you do not want to help constructively then
please get off my boat.

--
George Hester
_________________________________
"Peter Hansen" <peter at engcorp.com> wrote in message
news:3D0E8AC8.6C5C96EE at engcorp.com...
> George Hester wrote:
> >
> > "Gerhard H?ring" <gerhard at bigfoot.de> wrote:
> > > You did not import the sys module.
>
> > Probably not.  Where do I import that from and how do it?  I am trying to
put a
> > script in ASP using Python as the script language to give me the version of
> > Python that I am running when I call the ASP.  The more the script works out
of
> > the box the better I'll be able to catch on to this language.  Thanks.
>
> It sounds to me like you should probably crawl before you walk.
> I suggest downloading the tutorial and installing Python on
> your own machine and using it interactively from the prompt
> before you advance to something like ASP scripting.  Your
> responses clearly show you aren't ready to write code in an
> environment which is doubtless going to mask from you many
> of the errors and messages which you will encounter as you learn.
>
> In the above case you would be able to experiment and learn about
> these things:
>
> C:\>python
> Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
> >>> sys.version
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> NameError: name 'sys' is not defined
>
> Here's where you would pause to learn a little more about Python....
>
> >>> import sys
> >>> sys.version
> '2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)]'
>
> -Peter




From peter at engcorp.com  Thu Jun 20 22:12:13 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 20 Jun 2002 22:12:13 -0400
Subject: RE strings (was: Variable Interpolation - status of PEP 215)
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com>
	        <4PfQ8.44245$n4.10307683@newsc.telia.net>
	        <KhiQ8.44269$n4.10312432@newsc.telia.net>
	        <29e28c51.0206201156.7e61a7f5@posting.google.com>
	        <3D127B89.EE1C17F4@engcorp.com> <mailman.1024622248.20011.python-list@python.org>
Message-ID: <3D128B7D.A87D3BE7@engcorp.com>

Skip Montanaro wrote:
> 
>     Peter> Given that the only real difference is you skip an import, and at
>     Peter> the cost of modifying the language definition in an unusual
>     Peter> manner, is there any point to this at all?
> 
> Yeah...  Makes it more like Perl (more magic under the covers instead of out
> where you can see it). ;-)

:-)  I guess I understood that to be the point.  What I should
have said was "is there any _benefit_ to this at all?"

-Peter


From peter at engcorp.com  Wed Jun 19 08:30:00 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 19 Jun 2002 08:30:00 -0400
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com> <3D1010C3.C321F8B0@engcorp.com> <raXP8.47926$GY.15260255@twister.nyroc.rr.com>
Message-ID: <3D107948.B3FEE4BB@engcorp.com>

George Hester wrote:
> 
> Yes my take on it was that you said "I needed to crawl before I walk."  Please
> don't hold my inference against me.

That's precisely what I said.  Nothing derogatory about it, except
in your own mind.  We /all/ need to crawl before we walk, in all
things, unless we're lucky.  I was just reminding you of this 
fact of life.

If I were attempting to learn limit calculus, and simultaneously
apply it to a difficult problem while using Maple, and I knew 
nothing about Maple and little about limit calculus, you would
be helping me if you said "crawl before you walk... just learn
the basics of limit calculus before you try to figure out how
to use Maple too".  I hope I would not be so sensitive as to 
take that the wrong way.

-Peter

P.S.: I know little about limit calculus, and nothing about 
Maple, so please forgive the above example if it combines
the two infeasibly.... :)


From russblau at hotmail.com  Tue Jun 11 12:05:43 2002
From: russblau at hotmail.com (Russell Blau)
Date: Tue, 11 Jun 2002 12:05:43 -0400
Subject: Sorting list (maybe stupid)
References: <mailman.1023789674.28761.python-list@python.org> <3D05F3FD.7020202@mxm.dk>
Message-ID: <ae574q$42cjd$1@ID-76829.news.dfncis.de>

"Max M" <maxm at mxm.dk> wrote in message news:3D05F3FD.7020202 at mxm.dk...
> afile = open('myfile.txt')
> tuplelist = []
> for aline in afile:
>      nummericValue = int(aline[:aline.find('.')])
>      tuplelist.append((nummericValue, aline))
> tuplelist.sort()
> sortedLines = [tup[1] for tup in tuplelist]
> for line in sortedLines:
>      print line

If the programmer is *certain* that each line will be in the format
given by the OP, which is a line number followed by a period and a
space, followed by arbitrary text, then the solution can be even
simpler:

linelist = []
for aline in afile:
    linelist.append(aline.split(".", 1))
# each entry in linelist is a list of two elements: [line_number,
line_body]
linelist.sort()
resultlist = []
for item in linelist:
    resultlist.append(".".join(item))
# used resultlist in case the OP wants to do something with these lines
besides just print them
# alternately, you could join the strings in-place in linelist
for line in resultlist:
    print line

--
I don't actually have a hotmail account; but I do have one on excite.com
if you really want to get in touch with me.





From liya.chernyakova at staarinc.com  Fri Jun 21 15:33:02 2002
From: liya.chernyakova at staarinc.com (liya)
Date: 21 Jun 2002 12:33:02 -0700
Subject: Showing MFC dialog from Python
Message-ID: <c51133b6.0206211133.1f8cf5f4@posting.google.com>

Please, help, I am trying to use the MFC dialog (in
"PyOmegaParamDlg.dll") in the Python script. I am not sure what is
missing, but the dialog is not shown and it's functions are not
accesible.
The Python code I am trying is following:
class TestDialogApp(dlgappcore.DialogApp):
	def CreateDialog(self):
		return PyWraperDlg()

class PyWraperDlg(dlgappcore.AppDialog):
	def __init__(self):
		self.edit = None
		print 'loading dll'
		mydll = win32ui.LoadLibrary("PyOmegaParamDlg.dll")
		print 'initializing dialog'
		mydll.AttachToMFC()
		IDD_CONFIG_DIALOG = 1000
		dlgappcore.AppDialog.__init__(self, IDD_CONFIG_DIALOG, mydll)
		print 'dialog initialized'
		

app.AppBuilder = TestDialogApp
print 'instantiating the application'

All the print statements are executed, but the dialog does not appear
on the screen.
IDD_CONFIG_DIALOG = 1000 is the resource ID for the dialog in my dll's
resource.h file.
I am sure I am missing something, but I am not sure what. Ideally I
want to acces the dialog functions, that are declared as public in my
dll.
Thank you
Liya


From emile at fenx.com  Sat Jun 22 14:36:37 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 22 Jun 2002 18:36:37 GMT
Subject: Numerical Methods in Python
References: <%32R8.853$xv1.36390126@newssvr15.news.prodigy.com>
Message-ID: <Vs3R8.290252$cQ3.14809@sccrnsc01>

"Russ Dilley" <rddlog at prodigy.net> wrote in message
news:%32R8.853$xv1.36390126 at newssvr15.news.prodigy.com...
> I am interested in the numerical capability of Python as well as it's
> applicability to system administration  tasks.

It will do both.  Start with the tutorial for general python info, and
something below should get your interest on the numerical analysis.

http://www.pfdubois.com/numpy/
http://scipy.org/
http://matpy.sourceforge.net/

--

Emile van Sebille
emile at fenx.com

---------



From trentm at ActiveState.com  Tue Jun 11 12:08:23 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Tue, 11 Jun 2002 09:08:23 -0700
Subject: Cross platform spawning
In-Reply-To: <659f82ff.0206110754.775490d9@posting.google.com>; from sketerpot@chase3000.com on Tue, Jun 11, 2002 at 08:54:55AM -0700
References: <659f82ff.0206110754.775490d9@posting.google.com>
Message-ID: <20020611090823.D3208@ActiveState.com>

[Peter Scott wrote]
> I am trying to make a script spawn a seperate process, like you would
> do like this on unix:
> 
> os.system('python myscript.py myargument &')
> 
> But the & at the end (to make a separate process) doesn't work on
> Windows. I want the script to run on both Linux (*BSD, whatever) and
> Windows. I didn't want to have to find out the path to the python
> interpreter, so os.spawnlp (with no blocking) sounded perfect until I
> saw that it wasn't available on Windows.
> 
> Does anyone out there know what I can do in this situation?

You could try to use os.spawnl(), which *is* availble on Windows. The
only pain is that you have to manually determine the full path to the
executable yourself via some which-like algorithm. I have attached a
script of mine that can do that on Windows and Linux (though I haven't
teseted it a *whole* lot.

Cheers,
Trent


-- 
Trent Mick
TrentM at ActiveState.com
-------------- next part --------------
#!/usr/bin/env python

"""
    Show full path of commands.

    Usage:
        which [<options>...] [<command-name>...]

    Options:
        -h, --help      Print this help and exit.
        -a, --all       Print *all* matching paths.
        -v, --verbose   Print out near misses on stderr.
        --version       Print the version info and exit.

    Show the full path to the program that would be run for each given
    command name, if any. Which, like GNU's which, returns the number of
    failed arguments, or -1 when no <command-name> was given.
"""
#REQUIREMENTS:
#   - Python >= 2.2 (because of the use of generators)
#
#TODO:
#   - improve the test suite
#   * launcher app? make the launcher app generic, it should be able to
#     use argv[0] to determine what to launch, then it can be built and
#     just renamed.
#   * explain what *near* misses are, perhaps that should be the default
#     and -q can be used to suppress them. I think so, yes.

from __future__ import generators
import os
import sys
import getopt
import stat


#---- exceptions

class WhichError(Exception):
    pass


#---- global data

_version_ = (0, 1, 2)
gVerbose = 0


#---- internal support stuff

def _getRegisteredExecutable(exeName):
    """Windows allow application paths to be registered in the registry."""
    if sys.platform.startswith('win'):
        if os.path.splitext(exeName)[1].lower() != '.exe':
            exeName += '.exe'
        import _winreg
        try:
            return _winreg.QueryValue(_winreg.HKEY_LOCAL_MACHINE,
                "SOFTWARE\\Microsoft\\Windows\\CurrentVersion\\App Paths\\"\
                + exeName)
        except _winreg.error:
            pass
    return None

def _samefile(fname1, fname2):
    if sys.platform.startswith('win'):
        return ( os.path.normpath(os.path.normcase(fname1)) ==\
            os.path.normpath(os.path.normcase(fname2)) )
    else:
        return os.path.samefile(fname1, fname2)

def _cull(potential, matches):
    """Cull inappropriate matches. Possible reasons:
        - a duplicate of a previous match
        - not a disk file
        - not executable (non-Windows)
    If 'potential' is approved it is returned and added to 'matches'.
    Otherwise, None is returned.
    """
    global gVerbose
    for match in matches:  # don't yield duplicates
        if _samefile(potential, match):
            if gVerbose:
                sys.stderr.write("duplicate: %s\n" % potential)
            return None
    else:
        if not stat.S_ISREG(os.stat(potential).st_mode):
            if gVerbose:
                sys.stderr.write("not a regular file: %s\n" % potential)
        elif not os.access(potential, os.X_OK):
            if gVerbose:
                sys.stderr.write("no executable access: %s\n" % potential)
        else:
            matches.append(potential)
            return potential

        
#---- public stuff

def which(command):
    """Generate a list of found apps for the given 'command' on the path."""
    matches = []
    path = os.environ.get("PATH", "").split(os.pathsep)
    if sys.platform.startswith("win"):
        path.insert(0, os.curdir)  # implied by Windows shell
    # Windows has the concept of a list of extensions (PATHEXT env var).
    exts = []
    if sys.platform.startswith("win"):
        exts = os.environ.get("PATHEXT", "").split(os.pathsep)
        # If '.exe' is not in exts then obviously this is Win9x and or a
        # bogus PATHEXT, then use a reasonable default.
        for ext in exts:
            if ext.lower() == ".exe":
                break
        else:
            exts = ['.COM', '.EXE', '.BAT']
    # File name cannot have path separators because PATH lookup does not
    # work that way.
    if os.sep in command or os.altsep and os.altsep in command:
        pass
    else:
        for dirName in path:
            # On windows the dirName *could* be quoted, drop the quotes
            if sys.platform.startswith("win") and len(dirName) >= 2\
               and dirName[0] == '"' and dirName[-1] == '"':
                dirName = dirName[1:-1]
            for ext in ['']+exts:
                absName = os.path.abspath(
                    os.path.normpath(os.path.join(dirName, command+ext)))
                if os.path.isfile(absName):
                    match = _cull(absName, matches)
                    if match:
                        yield match
        absName = _getRegisteredExecutable(command)
        if absName is not None:
            match = _cull(absName, matches)
            if match:
                yield match


#---- mainline

def main(argv):
    global gVerbose, _version_
    all = 0
    try:
        optlist, args = getopt.getopt(argv[1:], 'hav',
                                      ['help', 'all', 'version', 'verbose'])
    except getopt.GetoptError, msg:
        print "which: error: %s. Your invocation was: %s\n" % (msg, argv)
        print __doc__
        return 1
    for opt, optarg in optlist:
        if opt in ('-h', '--help'):
            print __doc__
            return 0
        elif opt == '--version':
            print "Which %s" % '.'.join([str(i) for i in _version_])
            return 0
        elif opt in ('-a', '--all'):
            all = 1
        elif opt in ('-v', '--verbose'):
            gVerbose = 1

    if len(args) == 0:
        return -1

    failures = 0
    for arg in args:
        #print "debug: search for %r" % arg
        npaths = 0
        for path in which(arg):
            print path
            npaths += 1
            if not all:
                break
        if not npaths:
            failures += 1
    return failures


if __name__ == "__main__":
    sys.exit( main(sys.argv) )



From eric.brunel at pragmadev.com  Fri Jun 28 09:46:40 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Fri, 28 Jun 2002 13:46:40 +0000
Subject: print() with unicode strings
References: <3D1AFC5C.2040708@thomas-guettler.de>
Message-ID: <afhhvd$n31$1@wanadoo.fr>

Thomas Guettler wrote:

> Hi!
> 
>  >>>print '?' #(German Umlaut)
> 
> works
> 
> but
> 
>  >>>print u'?'
> 
> does not.
> 
> Is there a reason why print does not accept unicode?
> Why not default to a latin-1 encoding?
> 
> Version:
> 
> Python 2.2 (#1, Dec 31 2001, 15:21:18)
> [GCC 2.95.3-5 (cygwin special)] on cygwin
> 
> thomas

To make latin-1 (a.k.a. iso8859-1) or any other your default 
encoding, create a file named sitecustomize.py somewhere in your PYTHONPATH 
and write in it:

import sys
sys.setdefaultencoding('iso8859-1')

NB: it won't work outside the sitecustomize module (cf. 
http://www.python.org/doc/current/lib/module-sys.html#l2h-264 )

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From kragen at pobox.com  Sun Jun  2 22:33:20 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 02 Jun 2002 22:33:20 -0400
Subject: My message text won't fit in a line
References: <b714b8de.0206021408.5398d0ef@posting.google.com>
Message-ID: <83vg91rszz.fsf@panacea.canonical.org>

akineko at pacbell.net (Aki Niimura) writes:
> Is there any better idea to deal with this problem.
> (Such as how to access a child widget of the Tk dialog) 

winfo_children() --- dunno if that will solve your problem though.



From i.linkweiler at gmx.de  Tue Jun 11 05:00:40 2002
From: i.linkweiler at gmx.de (Ingo Linkweiler)
Date: Tue, 11 Jun 2002 11:00:40 +0200
Subject: Poll: Readable languages
Message-ID: <3D05BC38.2020400@gmx.de>

Poll: How readable is the code of this languages?
1.)
-----
ABC:
ADA:
APL:
BETA:
C++:
COBOL:
DELPHI:
EIFFEL:
FORTH:
FORTRAN:
JAVA:
LISP:
MODULA:
PASCAL:
PERL:
PROLOG:
PYTHON:
SML:
SCHEME:
SMALLTALK:
TCL:
VISUAL BASIC:
other  (  *name of language * ):
-----
2.)
My most used language is:
-----

Please vote....
1 = Easy to read, like pseudocode :-)
2 = Good
3 = Average
4 = difficult
5 = nearly impossible, when not knowing the language :-(
blank = I do not know this language or have no opinion.

Please only vote, if you know at lease 2 languages.
Please reply to this newsgroup or mailto: i.linkweiler at gmx.net

This poll is done for my diploma paper at university of dortmund, germany.
I will not use your e-mail for any purpose.

Thanks for your help.
Ingo Linkweiler



From mcfletch at rogers.com  Tue Jun  4 13:44:51 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 04 Jun 2002 13:44:51 -0400
Subject: int, float and string...
References: <adhut1$10rtkl$1@ID-146704.news.dfncis.de> <mailman.1023206585.1785.python-list@python.org> <adirg9$11hv44$1@ID-146704.news.dfncis.de>
Message-ID: <3CFCFC93.7070803@rogers.com>

See object.h in the Python source code.  Basically, Python objects 
(PyObjects) are a C structure with book-keeping information (what is 
your class/type, and your reference count) + the actual data of the 
object.    Depending on the object type, that can be anywhere from 
(minimum) 8 bytes to 8+x for every object.

For ints, it's 8+4 (see intobject.h in Python source code), so 12 bytes, 
only 4 of which are data.

For doubles, it's 8+8, so 16 bytes, only a 33% increase in size over ints.

However, each object you're storing/accessing is also being held as a 
reference in another list/object (another 4 bytes for a pointer), the 
total becomes 16 and 20 bytes respectively, only a 25% increase between 
them, with 75% and 60% of the memory being Python object overhead.

BTW, I'm not a C person, so I may be mis-calculating some of those 
values, basic principle still applies: don't worry about differentiating 
between int and float based on storage concerns save when using huge 
numbers of them (preferably in Numeric arrays).

BTW2, storing two references to the _same_ int or float object only adds 
4 bytes to your memory overhead (i.e. references are cheap), so if you 
had a huge set of identical values, using references to the same object 
(not just objects with the same value) might save you significant memory 
(this is done automatically for ints from -1 to 100 in Python 2.2 if I'm 
not mistaken).  Still, this is probably premature optimisation unless 
you actually find yourself running out of memory.

HTH,
Mike

Shagshag13 wrote:
> "Mike C. Fletcher" <mcfletch at rogers.com> a ?crit dans le message de news: mailman.1023206585.1785.python-list at python.org...
...
>>machines if I recall correctly), but the overhead of Python objects is
>>huge compared to the size of the actual data, so don't worry about it
>>unless you're using Numeric Python and slogging around huge arrays of
>>numbers (where worrying about it might actually save you time).
> 
> 
> Thanks again for your replies !
> 
> Where could i go to see and understand what you call "overhead of
> Python objects" ?
> 
> thanks,
> 
> s13.
> 
> 
> 


-- 
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From amuys at shortech.com.au  Mon Jun 24 03:54:34 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 24 Jun 2002 00:54:34 -0700
Subject: Socket Disconnection
References: <3d1304d0$0$224$4d4ebb8e@news.nl.uu.net> <3d16817e$1@news.nz.asiaonline.net>
Message-ID: <7934d084.0206232354.52c2163d@posting.google.com>

"Colin Brown" <cbrown at metservice.com> wrote in message news:<3d16817e$1 at news.nz.asiaonline.net>...
> Hi Guyon
> 
> I assume here that you are talking TCP/IP.
> 
> If you are waiting on a receive [ data = connection.recv() ] then you will
> have an empty data string returned when the client disconnects.
> 
> If you are sending you will get an error telling you the link was shut down.
> 

Also note that the only reliable way to detect a client disconnect is
the latter.  If you want to demonstrate the failure of the recv()
approach, just yank out the ethernet cable on the client machine, and
notice that although the client is now disconnected, the server has no
idea.

Might I suggest taking a look at TCP/IP Illustrated Vol 1, by
W.Richard Stevens.  Definately a book to read if you are doing any
TCP/IP network programming.

Andrae Muys


From peter at engcorp.com  Thu Jun 20 00:09:46 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 20 Jun 2002 00:09:46 -0400
Subject: Variable Interpolation - status of PEP 215
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com>
Message-ID: <3D11558A.14451441@engcorp.com>

Ganesan R wrote:
> 
> I am a relative newbie to Python. I have a reasonably good perl background
> (don't worry I am not about to start a language flame war :-), and one of
> the things I really miss is variable interpolation in strings. I searched
> the net and found PEP 215 (http://www.python.org/peps/pep-0215.html). This
> seems to be exactly what I need. The proposal seems to be quite old, so I'd
> like to know if it's still in consideration or has it been dropped
> completely?

Evading the question: are you aware of the ability of the Python 
formatting operator to retrieve items from a dictionary rather 
than just by order in a sequence?

>>> foo = "trick"
>>> bar = 95.5
>>> print "This %(foo)s is loved by %(bar)s%% of programmers." % locals()
This trick is loved by 95.5% of programmers.

You can use any dictionary, but locals() or globals() can be handy
in many cases.

A search of the archives for the list/newsgroup should also turn up
a handful of implementations that pretty much do what the Perl 
version does with a dozen lines of Python code as setup.

-Peter


From siegfried.gonzi at kfunigraz.ac.at  Sun Jun 23 05:34:53 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Sun, 23 Jun 2002 11:34:53 +0200
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at> <uhb5ii5b16dle8@corp.supernews.com>
Message-ID: <3D15963D.5010702@kfunigraz.ac.at>

RPM1 wrote:
> "Siegfried Gonzi" <siegfried.gonzi at kfunigraz.ac.at> wrote
> 
>>I am not any longer willing to try the calculation on my Linux box and
>>Python 2.2 (20 Linux power off  shut downs are enough in a day; should I
>>destroy my laptop?).
>>
> 
> 
> Just curious, is your laptop a Hewlitt Packard?

No, it is a Gericom one. Only SuSE 8.0 can be installed on the Gericom 
Overdose without any problems. It has got a SiS 630 chipset and this 
causes some trouble with the X driver under SuSE 7.3,7.2,... (or one can 
compile the kernel if he wants; I find it easier to use SuSE 8.0). Btw: 
if someone has got a Gericom Overdose and want that he can use his 
Winmodem under Linux he should get the driver from the SMARTLINK guys. 
The Gericom Winmodem has got no brand name, but some nice guy assumed it 
is a Winmodem from Smartlink. Write to the SMARTLINK developers and they 
will send you a driver which works (at least for me).


S. Gonzi



From pinard at iro.umontreal.ca  Tue Jun 18 21:55:40 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 18 Jun 2002 21:55:40 -0400
Subject: procmail replacement in Python
In-Reply-To: <20020618235901.GA2197@lilith.my-fqdn.de>
References: <20020618235901.GA2197@lilith.my-fqdn.de>
Message-ID: <oqit4grq03.fsf@carouge.sram.qc.ca>

[Gerhard H?ring]

> I've recently looked into the procmail sources and run away screaming.

ROTFL!  If I remember well, the original author knew he had a peculiar
writing style, and was writing in some prominent file (either an FAQ or
the README, or something like) that his style was not to be discussed! :-)
I wondered if the new maintainer revised the sources or not, I did not check.

> I'd very much prefer a Python replacement to such badly written C code.

Is it bad style?  I never went as far as asserting that it was bad style,
because tastes may differ.  Yet in that case, I knew that tastes may differ
very widely. :-).

Surely, `procmail' successfully addressed difficult problems, and did it
rather solidly, with almost no bugs.  So, despite the writing style is rather
alien to mine, it surely had mysterious virtues.  I could not condemn it!

> Does anything exist that I can start from?

This has been discussed on 2000-02-07 on this list.  In particular:

Donn Cave <donn at u.washington.edu> says:

   Anyway, it might be interesting in this context to look
   at Sift-Mail, a Tcl application from Lawrence Lundblade:
   http://www.island-resort.com/sm.htm and the Sieve language proposed as
   an Internet standard for mail filtering, http://www.cyrusoft.com/sieve/

Bernhard Reiter <breiter at usf.Uni-Osnabrueck.DE> says:

   Just for completion, there is another newcomer in the mailfilter
   competion.  Still a greenhorn, but maildrop. Of course, who ever could
   get a C++ save. :) http://www.flounder.net/~mrsam/maildrop/README.html

Skip Montanaro <skip at mojam.com> says:

   Worth noting that the procmail tips page at
   http://www.procmail.org/jari/pm-tips.html mentions Lua as a possible
   replacement for the procmail language (see the end of question 3.5).

> I've even contemplated not inventing yet another filter language, but
> just using plain Python functions.

`procmail' syntax is a bit opaque, but it seems to be very efficient, and
there is a great deal of experience which has accumulated over this syntax.
It might be difficult to raise a challenging competitor.  I guess that
one should overview all tools of this kind before tackling such a project.
There is also a lot of experience and solved difficulties in such tools,
portable locking and solid error recovery not being the least.

> If I do this, I'd like to replace fetchmail with getmail as well.

What is `getmail'?

> - writing it as a getmail plug-in
> - writing it as a standalone program to deliver to mailboxes a la
>   procmail
> - let getmail deliver to a temporary folder, then using my procmail
>   replacement to filter into the right folders, delete spam, etc.

My experience with `procmail' as a SPAM filter told me that it may get
frustratingly slow, when you use hundreds of filtering rules.  A few years
ago, I really pondered many alternative approaches, this would be a long
story.  In these days, for SPAM and virus filtering, I gave in and wrote
a Python tool which is good for me, but which would not be comprehensive
enough for publication.  However, there are also many SPAM filtering systems
around which have been published.  You might want to give a comprehensive
look into all these things, before adding yet another one.  Maybe! :-)
-------------- next part --------------

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard

From Grant_member at newsguy.com  Tue Jun  4 12:14:02 2002
From: Grant_member at newsguy.com (Grant Griffin)
Date: 4 Jun 2002 09:14:02 -0700
Subject: getting windows processes list
References: <GCEDKONBLEFPPADDJCOEMEGHDGAA.whisper@oz.net> <mailman.1023145566.5739.python-list@python.org>
Message-ID: <adip0a0rv2@drn.newsguy.com>

In article <mailman.1023145566.5739.python-list at python.org>, "Mike says...
>
>If I'm not mistaken, win32pdh should be able to get the information 
>using a process counter.  WSH might also be able to get at it.
>
>HTH,
>Mike
>
>PS: Grant, your email address bounces.

Thanks, Mike (and all!)--win32pdh led me to the answer.

My purpose was to determine when a NetMeeting session had ended.  The problem
with NetMeeting is that program control returns after just a few seconds even if
NetMeeting is still running.  So, recognizing that Microsoft typically provides
more than one way to skin its cats, I decided to use NetMeeting's process name
to determine if it was running.

Based on Mike's tip, I hunted through win32pdh, which led me to win32pdhutil.  I
adapted the routine below from win32pdhutil's "GetPerformanceAttributes" to get
a list of the names of all processes; if "conf" (the secret code for
"NetMeeting") appears in the list, NetMeeting is presumed to still be running.

god-bless-microsoft-for-making-everything-so-complicated-<wink>-ly y'rs,

=g2
-----------
def GetProcessNames(object="Process", counter="ID Process",
                    format=win32pdh.PDH_FMT_LONG, machine=None,
                    bRefresh=1):
    if bRefresh: # PDH docs say this is how you do a refresh.
        win32pdh.EnumObjects(None, machine, 0, 1)
    items, names = win32pdh.EnumObjectItems(None, None, object, -1)
    return names

_________________________________________________________________________

Grant R. Griffin                                           g2 at dspguru.com
Publisher of dspGuru                               http://www.dspguru.com
Iowegian International Corporation                http://www.iowegian.com



From delaray at hotmail.com  Thu Jun 13 19:00:02 2002
From: delaray at hotmail.com (Raymond de Lacaze)
Date: Thu, 13 Jun 2002 23:00:02 GMT
Subject: International Lisp Conference 2002
Message-ID: <aeb70q$eoi$1@news.franz.com>

=================================================

INTERNATIONAL LISP CONFERENCE 2002 (ILC 2002)


SAN FRANCISCO, CALIFORNIA


Monday, October 28th - Thursday, October 31st


(Conference Announcement)


=================================================

Registration deadline:  September 31st, 2002


Paper Submission deadline: August 15th, 2002


http://www.international-lisp-conference.org/


=================================================


The Conference
--------------


At the start of the 21st century we see the challenges ahead:
Applications will be more dynamic than ever and will be part of the
fibers of the Web, software and hardware robots will be our
assistants, and biology will be source of both computational problems
and programming metaphors. Applications will be customized to
individuals who happen upon them, our needs and desires will be
served by artificial beings, and our software may more accurately
resemble a termite nest than Rube Goldberg inventions.

Will our mainstream concepts, approaches, and languages hold up?
Where might we turn for progress, inspiration, and technology? Was it
right in front of us all along? The International Lisp Conference
addresses these questions and more.

This four-day conference will bring together a mesh of brilliant minds
from  a diverse international community that involves or borders the
Lisp family of languages. The conference will be uniquely multi-themed in
nature, and each day will paint a different theme. The themes  are  Lisp
and Functional Languages, Web Applications, Robotics and BioInformatics.
We have already gathered an impressive list of distinguished
speakers from each of these communities that promise to deliver a
captivating a stimulating conference experience.



Papers
------

Authors desiring to submit a full paper should declare their intention by
submitting a title and abstract of their paper by July 15st, 2002. The
actual
paper submission dealine will be August 15th 2002. The Program Committee
will be annouced shortly. Papers should be submitted in electronic form as a
postscript or pdf file. Submitted papers must be written in English and not
exceed 15 pages. Accepted  papers will be presented at the conference and
published in the conference proceedings.

For additional information, please visit the conference website:

http://www.international-lisp-conference.org


 =================================================








From chris.gonnerman at newcenturycomputers.net  Wed Jun 19 08:38:29 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Wed, 19 Jun 2002 07:38:29 -0500
Subject: newbie: help please!
References: <aepltq$gl3$1@knossos.btinternet.com>
Message-ID: <002b01c2178e$3e6693e0$0101010a@local>

----- Original Message -----
From: "Steve" <steve.bond at btinternet.com>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Wednesday, June 19, 2002 5:20 AM
Subject: newbie: help please!


> hello,
>
> i have python 2.2.1 installed on my mandrake 8.1 system, which seems to
> work fine.
>
> The problem is im trying to install Tk exstensions, i have downloaded both
> tcl8.3.4 and tk8.3.4 files and installed using the ./configure, make and
> then make install commands. I then when into the setup file in the
> Python-2.2.1/Modules directory where i understand i have to enable the
> tkinter extension module, by uncommenting the approritae lines.
>
> Lastly i run 'make' at the top of my Python-2.2.1 directory but i get this
> error:
>
> gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
> -DHAVE_CONFIG_H   -DWITH_APPINIT -I/usr/local/include -I/usr/X11R6/bin -c
> ./Modules/_tkinter.c -o Modules/_tkinter.o
> In file included from ./Modules/_tkinter.c:49:
> /usr/include/tk.h:83:29: X11/Xlib.h: No such file or directory
>
> i have the Xlib.h in the Python-2.2.1/Include directory
>
> the tkinter part of my setup file looks like this and everything was
> installed in the default places

The standard Python 2.2 configuration system should automatically add tk
support if all components are there.  Your error is because you don't have
the X11 development RPM(s) installed.  Get out your Mandrake install set
and add any RPM's which are specifically for X11 development (I run SuSE
so I don't know what RPM's you need for Mandrake).

The Xlib.h in the Python-2.2.1/Include directory isn't the right one.  You
need the X11 version.

>
> # *** Always uncomment this (leave the leading underscore in!):
>  _tkinter _tkinter.c tkappinit.c -DWITH_APPINIT \
> # *** Uncomment and edit to reflect where your Tcl/Tk libraries are:
>         -L/usr/local/lib \
> # *** Uncomment and edit to reflect where your Tcl/Tk headers are:
>         -I/usr/local/include \
> # *** Uncomment and edit to reflect where your X11 header files are:
>         -I/usr/X11R6/bin \

X11 header files are conventionally in /usr/X11R6/include or in
/usr/include/X11, and on many distro's both exist, one being a link to the
other.  /usr/X11R6/bin is almost certainly NOT the right location.

> # *** Or uncomment this for Solaris:
> #       -I/usr/openwin/include \
> # *** Uncomment and edit for Tix extension only:
> #       -DWITH_TIX -ltix8.1.8.2 \
> # *** Uncomment and edit for BLT extension only:
> #       -DWITH_BLT -I/usr/local/blt/blt8.0-unoff/include -lBLT8.0 \
> # *** Uncomment and edit for PIL (TkImaging) extension only:
> #     (See http://www.pythonware.com/products/pil/ for more info)
> #       -DWITH_PIL -I../Extensions/Imaging/libImaging  tkImaging.c \
> # *** Uncomment and edit for TOGL extension only:
> #       -DWITH_TOGL togl.c \
> # *** Uncomment and edit to reflect your Tcl/Tk versions:
>         -ltk8.3 -ltcl8.3 \
> # *** Uncomment and edit to reflect where your X11 libraries are:
>         -L/usr/X11R6/lib \
> # *** Or uncomment this for Solaris:
> #       -L/usr/openwin/lib \
> # *** Uncomment these for TOGL extension only:
> #       -lGL -lGLU -lXext -lXmu \
> # *** Uncomment for AIX:
> #       -lld \
> # *** Always uncomment this; X11 libraries to link with:
>         -lX11
>
> Any help would be much appreciated!

Install the missing RPM's and correct your setup file as noted, and
all should build fine.

> Thanks

You're welcome.  Hope this helps.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net





From ark at research.att.com  Thu Jun  6 13:51:53 2002
From: ark at research.att.com (Andrew Koenig)
Date: Thu, 6 Jun 2002 17:51:53 GMT
Subject: for , ordering
References: <mailman.1023377649.29116.python-list@python.org> <yu99u1og1g9b.fsf@europa.research.att.com>
Message-ID: <yu99ptz41eiu.fsf@europa.research.att.com>

gabor> for obj1 in objects1:
gabor> objects2.append(obj2(obj1.x))


gabor> 1. isn't there a more elegant way to do this?

ark> Isn't this the same as saying

ark>         objects2 = objects1[:]

ark> ?  Am I missing something obvious?

The answer is yes:  I *am* missing something obvious:  I didn't
notice the obj2(obj1.x).  But why not

             objects2 = map(lambda x: obj2(obj1.x), objects1)

?
-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From logiplexsoftware at earthlink.net  Sun Jun 16 18:29:12 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: 16 Jun 2002 15:29:12 -0700
Subject: Visual Python
In-Reply-To: <337619fa.0206161337.546a958f@posting.google.com>
References: <337619fa.0206161337.546a958f@posting.google.com>
Message-ID: <1024266558.12682.0.camel@localhost.localdomain>

On Sun, 2002-06-16 at 14:38, Raphael Ribeiro wrote:
> When i go to tools/debug and the ms-dos prompt start appears ,  an
> error message appears , anyone knows why?

An error occurred.


;)








From BPettersen at NAREX.com  Tue Jun 11 19:16:53 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 11 Jun 2002 17:16:53 -0600
Subject: Can Distutils include .pyd files directly?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019221519C@admin56.narex.com>

I've built my extension modules using the MS Devstudio IDE and now I
would like to use Distutils to create a Windows installer for me that
puts both the debug and release versions of the extension in the
python22\DLLs directory. Is this possible? If not, is there a way to
have Distutils build my extension in both debug and release mode and
install both?

-- bjorn




From jeremy at jdyallop.freeserve.co.uk  Mon Jun 17 06:16:48 2002
From: jeremy at jdyallop.freeserve.co.uk (Jeremy Yallop)
Date: 17 Jun 2002 10:16:48 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com> <gYgP8.42356$GY.12361719@twister.nyroc.rr.com>
Message-ID: <aekcug$7js8n$1@ID-114079.news.dfncis.de>

* George Hester
| Well yours I haven't the slightest idea what you are getting at.

  http://www.python.org/doc/current/tut/tut.html


From johnroth at ameritech.net  Thu Jun 20 07:11:55 2002
From: johnroth at ameritech.net (John Roth)
Date: Thu, 20 Jun 2002 07:11:55 -0400
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com> <lkznxqjqx2.fsf@pc150.maths.bris.ac.uk> <3D11A21D.9010809@mxm.dk>
Message-ID: <uh3e1usajle42c@news.supernews.com>

"Max M" <maxm at mxm.dk> wrote in message news:3D11A21D.9010809 at mxm.dk...
> Michael Hudson wrote:
>  > "John Roth" <johnroth at ameritech.net> writes:
>  >
>  >
>  >>There's no equivalent of a goto command in Python, for good reason,
>  >>which I won't repeat here. Prof. Djikstra did it very well in 1974,
>  >>in his letter "Goto Considered Harmful."
>  >
>  >
>  > There are also technical difficulties implementing it (the
>  > blockstack), or I'd have probably done goto functionality in
>  > bytecodehacks :)
>
>
> What do you mean???
>
>
> Goto's are simple ;-)
>
>
> def goto(lable):
>      lable()
>
>
> def start():
>
>      print 'starting'
>      goto(lable2)
>
> def lable1():
>
>      print 'at lable 1'
>      goto(stop)
>
> def lable2():
>
>      print 'at lable 2'
>      goto(lable1)
>
> def stop():
>      print 'stopping'
>
> goto(start)
>
>  >>> starting
>  >>> at lable 2
>  >>> at lable 1
>  >>> stopping

Those aren't goto's in the classical sense. For a real goto,
you need to think assembler: if it's a statement anywhere,
you can branch to it, and who cares about the execution
history! Having the compiler manage your flow control is
for sissys!

John Roth
>




From markus.vonehr at ipm.fhg.de  Mon Jun 17 07:48:31 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Mon, 17 Jun 2002 13:48:31 +0200
Subject: Tkinter and threads
Message-ID: <3D0DCC8F.9A6D3B81@ipm.fhg.de>

Hi,

I'd like to have a progress bar increasing while a calculation process
calculates some results.
Unfortunately there is no mainloop processing while the calculation is
running and so the progress bar begins after the calculation.
Another thread neither can access the Tkinter resources, how can
I manage this problem?

I appreciate any help,

Markus

-------------- next part --------------
A non-text attachment was scrubbed...
Name: markus.vonehr.vcf
Type: text/x-vcard
Size: 245 bytes
Desc: Visitenkarte f?r Markus von Ehr
URL: <http://mail.python.org/pipermail/python-list/attachments/20020617/7493be5b/attachment.vcf>

From not.this at seebelow.org  Thu Jun 13 09:27:27 2002
From: not.this at seebelow.org (Grant Griffin)
Date: Thu, 13 Jun 2002 08:27:27 -0500
Subject: why not "'in' in 'in'"?
References: <ae8dg70217g@drn.newsguy.com> <3D080482.7A7FD514@alcyone.com> <3D082A62.CF800C75@seebelow.org> <Xns922C61724D96Dduncanrcpcouk@127.0.0.1>
Message-ID: <3D089DBF.349B7A1@seebelow.org>

Duncan Booth wrote:
> 
> Grant Griffin <not.this at seebelow.org> wrote in
> news:3D082A62.CF800C75 at seebelow.org:
> 
> >> Grant Griffin wrote:
> >>  This peculiar lack of generality seems a bit (dare I say
> >>  it?) "non-Pythonic".
> <lots of snips and then>
> > But anyway, let's "just say no" to the tuple case rather than confuse
> > ourselves; let's just make the string version of "in" a special case.
> 
> I see a slight lack of consistency in your approach here. First you
> complain that the way 'in' works lacks generality, then when everyone
> points out that to make it work the way you would expect would make it
> break on lists or tuples you suggest special casing.

(You forgot "QED" <wink>.)

> 
> The Pythonic way is to maintain generality, not to introduce special cases.

See, that's where we bump into "practicality beats purity".

note-that-every-one-of-tim's-principles-is-subjective-ly y'rs,

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation            http://www.iowegian.com


From maxm at mxm.dk  Thu Jun 20 15:37:11 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 20 Jun 2002 21:37:11 +0200
Subject: application development - separating content from function
References: <yroQ8.260949$i6.17986369@amsnews02.chello.com>
Message-ID: <3D122EE7.5000501@mxm.dk>

Henry Baumgartl wrote:

> The more I have tried to solve these requirements in an elegant and
> efficient manner, the further I seem to stray from anything remotely
> resembling a viable solution. In other words, I can't see the forest for the
> trees anymore.


Do the simplest thing that can possible work for each part in your 
system. This is to get a system up and running as quickly as possible.

Take this system and refactor it until there is no redundancy in your 
code. But remember to allway do the simplest thing possible.

Add new convenience features on top of this. Then repeat all over.

I most often find myself in trouble when I try to design overly clever 
frameworks without understanding the area for which the software is 
developed.

The problem with this approach though is that it can be difficult to 
take the high level view needed to modularize your code.

In time it will probably be a good idea to read a few books about 
patterns. It can be a pretty advanced topic that can be hard to fit to 
the problem currently at hand. So it's better to have developed a fair 
amount of code before reading up on patterns.


regards Max M



From sholden at holdenweb.com  Sat Jun  8 16:18:11 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sat, 8 Jun 2002 16:18:11 -0400
Subject: Efficient python programming...
References: <mailman.1023452786.27923.python-list@python.org> <3D013FB1.3D7477EA@engcorp.com>
Message-ID: <xItM8.382470$%u2.167544@atlpnn01.usenetserver.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3D013FB1.3D7477EA at engcorp.com...
> Roman Suzi wrote:
> >
> > On Fri, 7 Jun 2002, Peter Hansen wrote:
> > > You forgot the even more important first thing for a beginner:
> > > get it correct!  The only good way to do this is to write unit
> > > tests that prove it.  If you don't have tests, you don't have
> > > working code (or how can you prove it?).
> >
> > To prove that code is correct is very difficult (practically
impossible).
> > Unittests are there to prove that you have errors, not the other way
> > around.
> >
> > Of course, they can increase your confidence in that optimized code is
> > equivalent to what you had before.
>
> I disagree, if the tests are written first (and run, proving that
> they will fail when the code is wrong or missing).
>
Clearly a test that succeeds when the code is demonstrably wrong is an
incorrect test. But a test that succeeds when the code *isn't* demonstrably
wrong doesn't guarantee the code is correct. Most tests aren't complete
enough to guarantee that the tested code runs to specification over all
possible inputs.

Please don't think I'm arguing that test-before-code is bad - I think it's
great, and very helpful for collaborators.

> The tests are executable requirement specifications.  Running
> them "proves" (yes, not 100%... nothing could) that the code
> meets the specs.  The specs could be wrong of course, but that
> doesn't mean the code is broken...
>

That's pretty much where I fall out with the formal methods crowd (who
invaded the academic world in the 1980's, putting paid to some of *my*
fun...). They don't seem to realise that for practical development projects
all they really do is replace bugs in the implementation with bugs in the
specification, since the specifications become mathematical notations
incomprehensible to the average user. Ultimately it's the users'
requirements that must dictate the specification.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From vjovanov at stevens-tech.edu  Thu Jun  6 10:34:09 2002
From: vjovanov at stevens-tech.edu (Vojin Jovanovic)
Date: Thu, 6 Jun 2002 10:34:09 -0400
Subject: self
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu> <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com> <mailman.1023294728.2322.python-list@python.org> <sWsL8.82$v_7.4260962@newnews.cc.stevens-tech.edu> <mailman.1023323409.9676.python-list@python.org>
Message-ID: <jpKL8.83$HV1.4324388@newnews.cc.stevens-tech.edu>

I don't think you tested your code, because I can't get the output you
specified.  In fact it seems that you broke your own code.

Vin




From aahz at pythoncraft.com  Thu Jun 20 19:24:53 2002
From: aahz at pythoncraft.com (Aahz)
Date: 20 Jun 2002 19:24:53 -0400
Subject: do...while
References: <slrnah40l1.12v.msoulier@pmerd071.ca.nortel.com>
Message-ID: <aeto85$pjf$1@panix1.panix.com>

In article <slrnah40l1.12v.msoulier at pmerd071.ca.nortel.com>,
Michael P. Soulier <msoulier at nortelnetworks.com_.nospam> wrote:
>
>    line = filehandle.readline()
>    while len(line) > 5:
>        line = filehandle.readline()

Use Python 2.2:

    for line in filehandle:
        if len(line) <= 5:
            break
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From martin at v.loewis.de  Sun Jun  9 04:09:26 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 09 Jun 2002 10:09:26 +0200
Subject: libpython as a shared library?
References: <20020604143402.GB22993@hal.mediasupervision.de> <Pine.OS2.4.32.0206051931120.1678-100000@tenring.andymac.org> <mailman.1023348398.27744.python-list@python.org> <j4u1og6bvt.fsf@informatik.hu-berlin.de> <slrnag5k5g.17b.quinn@lira.ugcs.caltech.edu>
Message-ID: <m3n0u4rhzd.fsf@mira.informatik.hu-berlin.de>

quinn at lira.ugcs.caltech.edu (Quinn Dunkan) writes:

> I hope 2.3 doesn't dynamicly link libpython by default.  

Certainly not. I have the same reservations towards shared libraries
as you, so it was an absolute prerequisite that this is a configure
option, and that it is off by default.

Regards,
Martin


From rs at onsitetech.com  Mon Jun 17 13:51:56 2002
From: rs at onsitetech.com (Robb Shecter)
Date: 17 Jun 2002 17:51:56 GMT
Subject: Python interface to ImageMagick
References: <aec8qd$19p$05$1@news.t-online.com>
Message-ID: <3D0E2159.50602@onsitetech.com>

Achim Domma wrote:
> Hi,
> 
> I just uploaded a first alpha version of my python wrapper for ImageMagick...

This sounds really interesting, but could you give a link to some other 
resource besides a .exe file?  A simple readme file, or docs or 
something?  (I looked around the procoders.net website but didn't see 
anything.)

(I think that simply putting a .exe file out there requires a lot of 
faith on the part of people to download it and execute it.)

Robb





From paul at boddie.net  Wed Jun  5 13:08:47 2002
From: paul at boddie.net (paul at boddie.net)
Date: 5 Jun 2002 17:08:47 -0000
Subject: [ANN] pyPgSQL 2.1 released
Message-ID: <20020605170847.2631.qmail@www4.nameplanet.com>

On Wed, 5 Jun 2002 18:40:11 +0200 Gerhard H?ring <gerhard.haering at gmx.de> wrote:
>CC-ing to pypgsql-users list, maybe other users have comments on this issue,
>too.

[Arrays vs. 'IN']

>QUERY: DECLARE PgSQL_08120A6C CURSOR FOR select * from cages where name
>       in '{"mammals","apes"}'
>
>... which obviously cannot work. The reason is that if pyPgSQL encounters a
>tuple or a list in the statement parameters, it will quote them suitable for
>use as an ARRAY. Unfortunately, the ARRAY syntax is different from what you
>would need for the IN operator in SQL.

Right, and there's no way of knowing what the database engine expects for that 
parameter until it has parsed the statement. I suppose there isn't a way of 
getting type information from PostgreSQL about parameters - indeed, does 
PostgreSQL actually support parameters in the way that other database systems 
do? I got the impression that PostgreSQL database modules have to bundle the 
parameter values into the query as shown above, rather than passing those 
values separately.

>I'm not sure if something can be done about this problem. It seems to me that
>it's an either/or with ARRAY vs. your desired conversion for the IN operator.

That's an unfortunate limitation, but thankfully I didn't need to use 'IN' 
after all. :-)

>I've fiddled with adding Unicode support for some time now, but I didn't get
>enough testers yet to make me comfortable enough with it for adding it to a
>pyPgSQL release. On the Sourceforge patches section, I've uploaded a patch for
>full Unicode support, including getting Unicode results back from the database,
>insert Unicode strings, and automatic encoding conversions. I'll try to update
>this patch today to work with the latest pyPgSQL 2.1 release and drop you a
>note when it's ready.

That would be great. Where XML data and PostgreSQL come together, it quickly 
becomes a requirement to handle the data gracefully or transparently, although 
I suspect that many database systems still struggle with Unicode.

Thanks for the explanation and the great work!

Paul




From stephen at theboulets.net  Wed Jun  5 12:59:52 2002
From: stephen at theboulets.net (Stephen Boulet)
Date: Wed, 05 Jun 2002 11:59:52 -0500
Subject: os.popen and place in file
References: <3CFE2762.519E4344@motorola.com> <kud6v5ogny.fsf@lasipalatsi.fi>
Message-ID: <ufsgqu1atq9313@corp.supernews.com>

Thanks to all for the responses. Here's what I was able to do, using the 
StringIO module/class:

>>> import os
>>> from StringIO import *
>>> f = os.popen('ls /usr/local/bin')
>>> f = f.read()
>>> f=StringIO(f)
>>> for i in range(5):
...     print f.readline()
...
acroread

audacity

cabextract

demo

dos2unix

>>> f.seek(0)
>>> for i in range(5):
...     print f.readline()
...
acroread

audacity

cabextract

demo

dos2unix


-- Stephen

Erno Kuusela wrote:

> In article <3CFE2762.519E4344 at motorola.com>, Stephen Boulet
> <stephen.boulet at motorola.com> writes:
> 
> | I'm running windows NT.
> | The command:
> |   f = os.popen('dir')
> | returns a file object 'f'. I can't however use the seek command on it
> | [...] Anyone know why?
> 
> it is because the text output by the "dir" program is relayed directly
> to your program. the operating system does not store it away anywhere
> in case you want to revisit parts of it later. sockets and many
> unix device files work the same way. otherwise there would
> be problems when transferring large amounts of data through
> pipes.
> 
> if you want to access the data like a regular file, you could redirect
> the output from the command to a disk file and open that. or, read
> all of the data from the pipe and put it in a StringIO object.
> 
>   -- erno



From theller at python.net  Mon Jun  3 04:01:42 2002
From: theller at python.net (Thomas Heller)
Date: Mon, 3 Jun 2002 10:01:42 +0200
Subject: Where can I find .join() in the docs
References: <mailman.1022786331.17174.python-list@python.org> <wwyJ8.72901$cQ3.2613@sccrnsc01>
Message-ID: <adf7rt$10uj6e$1@ID-59885.news.dfncis.de>

> But indexing isn't the full answer.  Ask a newbie to wade through
>
> http://web.pydoc.org/2.2/search.cgi?key=join
>
But then there's the Python Manual Index search:
http://starship.python.net/crew/theller/pyhelp.cgi?keyword=join

Thomas




From gumuz at looze.net  Fri Jun 28 05:37:27 2002
From: gumuz at looze.net (Guyon Morée)
Date: Fri, 28 Jun 2002 11:37:27 +0200
Subject: self destruction :)
References: <f4KS8.1326$qq1.14832@castor.casema.net> <7x6604ctb1.fsf@ruckus.brouhaha.com>
Message-ID: <3d1c2d6a$0$231$4d4ebb8e@news.nl.uu.net>

>   x = [a, b, c]
>
> and you want to remove b from the list?
>
> Just say "del x[1]".


yep this is actually what i mean, but the problem is is that this action
should come from the object within the list. this is object 'knows' when
it's done, the list doesn't

see my problem?

ps. thanx everyone who responded so quickly




From jdhunter at nitace.bsd.uchicago.edu  Sun Jun 30 10:18:56 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Sun, 30 Jun 2002 09:18:56 -0500
Subject: can't import generators
References: <mailman.1025400278.15389.python-list@python.org>
Message-ID: <m2wusgvohb.fsf@mother.paradise.lost>

>>>>> "Tim" == Tim Peters <tim.one at comcast.net> writes:

    Tim> Start Python with the -v switch to see how imports are
    Tim> getting resolved.  __future__.py is an actual module that
    Tim> actually gets imported.  If you're picking up a wrong version
    Tim> of the Python libraries, that would explain it (maybe due to
    Tim> a bad PYTHONPATH setting, or some local module named
    Tim> __future__.{py,pyc,pyo}).

Bingo -- thanks.  Old musty, crusty PYTHONPATH setting was the
culprit.

JDH


From edream at tds.net  Wed Jun  5 13:26:55 2002
From: edream at tds.net (Edward K. Ream)
Date: Wed, 05 Jun 2002 17:26:55 GMT
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet> <3CFDF148.AA2035FF@info.unicaen.fr>
Message-ID: <3CFE49CA.46A5FEE1@tds.net>

Jerzy Karczmarczuk wrote:

> > My question is this; I realize that Python compiles to bytecode automatically
> > as part of the interpreters tokenization run, but is there a way to permanently
> > compile python code outputting to an executable binary?
> 
> Once upon a time Guido Van Rossum thought "this seems very difficult,
> perhaps impossible".
> 
> He changed his mind a little bit now...
> 
> See:
> 
> http://www.onlamp.com/pub/a/python/2002/06/04/guido.html
> 
> and the references cited by The Master:
> 
> http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/
> http://homepages.ulb.ac.be/~arigo/psyco/

Thanks for these links.

Earlier this year I looked into this question in some depth and
concluded:

a) compiling Python to, say, C or C++ is in fact doable much of the
time,
b) the hard issues are more related to library compatibility (what is a
compiled object?)
c) such a project probably would be reasonable only if part of
"official" Python.

As far as a) goes, it should be possible to "propagate" type information
from constants and arguments and return values from known library
routines, so that in fact a lot more type information may be present
than one might think.  I saw this first this way: there is an
isomorphism between the C++ and Python versions of Leo, so it ought to
be possible to back translate from leo.py to leo.cpp.  The more I looked
at this, the more plausible it became.  However, because of b), it is
not clear what kind of performance improvements can be expected.  Point
b) also implies a close connection between compiler, interpreter and
libraries, which is why I say c).

I offer the following existence proof for the possibility of compiling
Python.  One could simply "unwrap" the Python interpreter.  That is, we
could replace byte codes with the equivalent portions of the
interpreter's main loop.  Obviously, nothing practical is gained by
this.  However, it does provide a mental picture of what is going on,
and it hints at what might need to be optimized in order to make a
compiler practical.

I found this preliminary design work to be great fun, and I am excited
that Guido has shown some interest in this area.  I wonder whether there
would be any interest in a sig related to this subject?  I have lots of
experience and ideas on this subject and I would love to toss ideas
around.

BTW, a compiled version of a Python program would be valuable
commercially, as it would allow companies to protect their source code,
as well as (possibly) improve performance.

If anyone knows of any reasons why a Python to C++ compiler clearly can
not be made to work, I would be very interested in hearing them.

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


From ark at research.att.com  Thu Jun 13 11:57:13 2002
From: ark at research.att.com (Andrew Koenig)
Date: Thu, 13 Jun 2002 15:57:13 GMT
Subject: modifying string objects
References: <MPG.1772db35d89f7b9c989685@news.rz.uni-karlsruhe.de>
Message-ID: <yu99ofefgoiu.fsf@europa.research.att.com>

Uwe> strings are immutable in python. so what do i do if i need to pass a 
Uwe> function a string as parameter and need it to be modified within the 
Uwe> function so that changes are reflected outside of the function body?

Uwe> f.e. a class constructor must return None

Uwe> class foobar:
Uwe>   def __init__(text):
Uwe>     text = text[-1]			#this is what i actually want

Instead of a string, use a one-element list:

       def __init__(text):
         text[0] = text[0][-1]

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From rajarshi at presidency.com  Fri Jun 28 15:06:33 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Fri, 28 Jun 2002 15:06:33 -0400
Subject: joining 2 lists together - an elegant way?
References: <pan.2002.06.28.15.03.29.911402.1787@presidency.com>
Message-ID: <pan.2002.06.28.15.06.32.33132.1787@presidency.com>

On Fri, 28 Jun 2002 15:03:29 -0400, Rajarshi Guha wrote:

> Hi,
>   is there a more elegant way to join to lists together?
> 
> Right now I do:
> 
> newlist = []
> for i in l1:
>   newlist.append(i)
> for i in l2:
>   newlist.append(i)
> 
> It does the job - but a one liner (if possible) would be nice :) TIA,

Sorry :( Found it!


From irmen at NOSPAMREMOVETHISxs4all.nl  Mon Jun 10 14:16:13 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Mon, 10 Jun 2002 20:16:13 +0200
Subject: M2Crypto: select() behaves weird on SSL sockets
Message-ID: <3D04ECED.2030202@NOSPAMREMOVETHISxs4all.nl>

Hi
I'm using m2crypto 0.07 snapshot 3 with Python 2.2.1
and I'm experiencing weird behavior when using select().

Basically, it seems that a SSL socket is no longer in ready
state when entering select() again, after reading *a part*
of the available data on the socket.

Using plain sockets, the select() will keep reporting that
the socket is ready for reading, until all data has been read.
With SSL sockets, the second select() blocks eternally...

Any clue? I really need the regular select() behavior,
also on SSL sockets...


Thanks,
Irmen de Jong

(test code available on request.)



From phlip_cpp at yahoo.com  Wed Jun 19 10:13:00 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 19 Jun 2002 14:13:00 GMT
Subject: imp.loadmodule creates compiled file
References: <zv%P8.332187$o66.856939@news-server.bigpond.net.au>
Message-ID: <aeq3hc$6ii@dispatch.concentric.net>

Neil Hodgson wrote:

>    I've been using imp.loadmodule with Python 2.2 (on both Windows and
> Linux) to load changed versions of source code into a long running process
> where they will run alongside objects already created from earlier
> versions of the source code. An unexpected result of the call is to create
> a compiled module on disk with a name based on the filename argument with
> "c" appended.
> 
>>>> f = file("x.py")
>>>> import imp
>>>> imp.load_module("y", f, "z", (".py", "r", 1))
> <module 'y' from 'z'>
> 
>    At this point, a "zc" file is created. I would prefer no compiled file
>    be
> created. The documentation indicates that file and filename arguments can
> be None and '' when the module is not being loaded from a file but this
> makes litle sense as there is no other argument to provide the module
> contents from other than a file. Trying '' as the filename argument leads
> to a compiled file called "c". The compiled files can be unlinked but it
> would be better to have a solution that avoided writing to disk at all as
> there may be problems such as name clashes with other running instances or
> with being in a protected directory.
> 
>    Any other techniques for creating mdoules from source code?

PyUnitTestBrowser uses find_module then load_module; it appears to only 
create pyc files.

    folder, mod = path.split(filePath)    
    mod = mod[:-3]   
    got = imp.find_module(mod, [folder])
    module = apply(imp.load_module, [mod] + list(got))

http://c2.com/cgi/wiki?PyUnitTestBrowser

-- 
  Phlip
          http://www.greencheese.org/ParodyMode
  --  Because I'm the sysadmin. That's why.  --



From phd at phd.pp.ru  Wed Jun 19 08:58:03 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 19 Jun 2002 16:58:03 +0400
Subject: mimedecode.py version 1.1.3
Message-ID: <20020619165803.X4127@phd.pp.ru>

Hello!

                                mimedecode.py

WHAT IS IT

   Mail users, especially in non-English countries, often find that mail
messages arrived in different formats, with different content types, in
different encodings and charsets. Usually this is very good because it allows
us to use apropriate formats/encodings/whatever. Sometimes, though, some
unification is desireable. For example, one may want to put mail messages into
an archive, make HTML indicies, run search indexer, etc. In such situations
converting messages to text in one character set and skipping some binary
atachmetnts will be much desireable.

   Here is the solution - mimedecode.py.

   This is a program to decode MIME messages. The program expects one input
file (either on command line or on stdin) which treated as an RFC822 mesage,
and decoded to stdout. If the file is not an RFC822 message the file just
piped to stdout one-to-one. If it is a simple RFC822 message it is just
decoded as one part. If it is a MIME message with multiple parts
("attachments") all parts decoded. Decoding can be controlled by command-line
options.


WHAT'S NEW in version 1.1.3
   Fixed a minor bug. The program should never raise UnicodeError anymore.


WHERE TO GET
   Master site: http://phd.pp.ru/Software/Python/#mimedecode

   Faster mirrors: http://phd.by.ru/Software/Python/#mimedecode
      http://phd2.chat.ru/Software/Python/#mimedecode

   Requires: Python 2.0+ (actually tested with 2.1 and 2.2),
      configured mailcap database.

   Documentation (also included in the package):
      http://phd.pp.ru/Software/Python/mimedecode.txt
      http://phd.by.ru/Software/Python/mimedecode.txt
      http://phd2.chat.ru/Software/Python/mimedecode.txt


AUTHOR
   Oleg Broytmann <phd at phd.pp.ru>

COPYRIGHT
   Copyright (C) 2001-2002 PhiloSoft Design

LICENSE
   GPL


NAME
   mimedecode.py - decode MIME message


SYNOPSIS
   mimedecode.py [-h|--help] [-V|--version] [-cCDP] [-f charset] [-d header] [-p header:param] [-beit mask] [filename]


DESCRIPTION
   First, Subject and Content-Disposition headers are examined. If any of
those exists, they decoded according to RFC2047. Content-Disposition header
is not decoded - only its "filename" parameter. Encoding header's
parameters is in violation of the RFC, but widely deployed anyway,
especially in the M$ Ophice GUI (often referred as "Windoze") world, where
programmers are usually ignorant lamers who never even heard about RFCs.
Correct parameter encoding specified by RFC2231. This program decodes
RFC2231-encoded parameters.

   Then the body of the message (or current part) decoded. Decoding starts
with looking at header Content-Transfer-Encoding. If the header specifies
non-8bit encoding (usually base64 or quoted-printable), the body converted
to 8bit. Then, if its content type is multipart (multipart/related or
multipart/mixed, e.g) every part recursively decoded. If it is not
multipart, mailcap database is consulted to find a way to convert the body
to plain text. (I have no idea how mailcap could be configured on said M$
Ophice GUI, please don't ask me; real OS users can consult my example at
http://phd.pp.ru/Software/dotfiles/mailcap.html). The decoding process uses
first copiousoutput filter it can find. If there is no filter the body just
passed as is.
   Then Content-Type header consulted for charset. If it is not equal to
current default charset the body text recoded using Unicode codecs. Finally
message headers and body flushed to stdout.


OPTIONS
   -h
   --help
      Print brief usage help and exit.

   -V
   --version
      Print version and exit.

   -c
      Recode different character sets in message body to current default
      charset; this is the default.

   -C
      Do not recode character sets in message body.

   -f charset
      Force this charset to be the current default charset instead of
      sys.getdefaultencoding().

   -d header
      Add the header to a list of headers to decode; initially the list
      contains headers "From" and "Subject".

   -D
      Clear the list of headers to decode (make it empty).

   -p header:param
      Add the (header, param) pair to a list of headers' parameters to
      decode; initially the list contains header "Content-Disposition",
      parameter "filename".

   -P
      Clear the list of headers' parameters to decode (make it empty).

   -b mask
      Append mask to the list of binary content types; if the message to
      decode has a part of this type the program will pass the part as is,
      without any additional processing.

   -e mask
      Append mask to the list of error content types; if the message to
      decode has a part of this type the program will raise ValueError.

   -i mask
      Append mask to the list of content types to ignore; if the message to
      decode has a part of this type the program will not pass it, instead
      a line \nMessage body of type `%s' skipped.\n" will be issued.

   -t mask
      Append mask to the list of content types to convert to text; if the
      message to decode has a part of this type the program will consult
      mailcap database, find first copiousoutput filter and convert the
      part.

   The last 4 options (-beit) require more explanation. They allow a user
to control body decoding with great flexibility. Think about said mail
archive; for example, its maintainer wants to put there only texts, convert
Postscript/PDF to text, pass HTML and images as is, and ignore everything
else. Easy:

   mimedecode.py -t application/postscript -t application/pdf \
       -b text/html -b 'image/*' -i '*/*'

   When the program decodes a message (or its part), it consults
Content-Type header. The content type is searched in all 4 lists, in order
"text-binary-ignore-error". If found, appropriate action performed. If not
found, the program search the same lists for "type/*" mask (the type of
"text/html" is just "text"). If found, appropriate action performed. If not
found, the program search the same lists for "*/*" mask. If found,
appropriate action performed. If not found, the program uses default
action, which is to decode everything to text (if mailcap specifies
filters).
   Initially all 4 lists are empty, so without any additional parameters
the program always uses the default decoding.


ENVIRONMENT
   LANG
   LC_ALL
   LC_CTYPE
      Define current locale settings. Used to determine current default
      charset (if your Python is properly installed and configured).


BUGS
   The program may produce incorrect MIME message. The purpose of the
program is to decode whatever is possible to decode, not to produce
absolutely correct MIME output. The incorrect parts are obvious - decoded
Subject headers and filenames. Other than that output is correct MIME
message (tested with mutt mail reader).
   The program does not try to guess whether the headers are correct. For
example, if a message header states that charset is iso8859-5, but the
body is actually in koi8-r - the program will recode the message to wrong
charset.


NO WARRANTIES
       This  program  is  distributed in the hope that it will be
       useful, but WITHOUT ANY WARRANTY; without even the implied
       warranty  of  MERCHANTABILITY  or FITNESS FOR A PARTICULAR
       PURPOSE.  See the GNU  General  Public  License  for  more
       details.


Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From claird at starbase.neosoft.com  Fri Jun  7 15:07:00 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 7 Jun 2002 14:07:00 -0500
Subject: Comparison: HTTP with Java, Perl or Python
References: <3D009911.9000904@gmx.de>
Message-ID: <48E4942E9705DCC0.9D29F3BA7006C138.13F10388C543A6D3@lp.airnews.net>

In article <3D009911.9000904 at gmx.de>,
Ingo Linkweiler  <i.linkweiler at gmx.de> wrote:
>Hello,
>
>for a comparison of languages I need a complete "translation" of this short skript into Perl, Tcl and Java:
>
> myurl = "http://www.uni-dortmund.de"
> import urllib
> file  = urllib.urlopen(myurl)
> text  = file.read()
> file.close()
> print text
>
>Can you send me some solutions?
>The skripts must be able to run, please do not post "partial" solutions.
			.
			.
			.
  set myurl http://www.uni-dortmund.de
  package require http
  set token [::http::geturl $myurl]
  set text [::http::data $token]
  ::http::cleanup $token
  puts $text
Do you already have Perl and Java covered?
-- 

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


From kragen at pobox.com  Sun Jun  2 23:41:04 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 02 Jun 2002 23:41:04 -0400
Subject: Extracting data from HTML
References: <82096df4.0205311152.5891a17b@posting.google.com> <1022883953.6150.5802.camel@localhost> <mailman.1022908262.30070.python-list@python.org>
Message-ID: <837klhrpv3.fsf@panacea.canonical.org>

Geoff Gerrietts <geoff at gerrietts.net> writes:
> Both techniques are worth knowing -- but better than either would be
> finding a way to get the information you're after via XML-RPC or some
> other protocol that's designed to carry data rather than rendering
> instructions.

You seem to imply that XML-RPC is better suited to carrying data
rather than rendering instructions than HTTP is.  I disagree with this
implication, and I adduce the following evidence:
- the thousands of RSS feeds (see www.syndic8.com) using HTTP
- people downloading Python via HTTP
- the fact that XML-RPC runs over HTTP

XML-RPC is better suited to expressing RPC than HTTP is, but "getting
some data" is probably better done over HTTP GET, where you can take
advantage of things like caching and URL linkability.

There's a *reason* we added MIME headers in HTTP 1.0 about ten years
ago, boy.



From dbasch at yahoo.com  Sat Jun 29 01:07:32 2002
From: dbasch at yahoo.com (Derek Basch)
Date: Fri, 28 Jun 2002 22:07:32 -0700 (PDT)
Subject: Win32all serial port event character problem.
Message-ID: <20020629050732.98977.qmail@web20809.mail.yahoo.com>

Hi,

I use the following code to read data from the serial
port when a certain event character ('\n') is
received:
http://www.geocities.com/dbasch/temp.txt

It seems to work fine except for that when I send the
following data via the serial port:
http://www.geocities.com/dbasch/testdata.txt

the output is:

Comport Thread:test 1
Comport Thread:
test 2
te
Comport Thread:st 3
tes
Comport Thread:t 4
test 5

It seems to me that in the time between the event
character being received and the ReadFile operation
occurring additional characters are being placed into
the windows OS serial buffer.
Has anyone ever dealt with this? Is using the event
character not a reliable method in windows?
I know the event character is being triggered because
I can change the event character to 'g' and nothing
happens unless a 'g' is in the received data.
I have obviously left out some of the code in the
example but I hope everyone will understand it ok.
This is driving me batty so any help is greatly
appreciated.
Thanks,
Derek Basch
1 very unskilled win32 programmer.


__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com




From peter at engcorp.com  Mon Jun 10 09:27:20 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 10 Jun 2002 09:27:20 -0400
Subject: Newbie question communicating between 2 progs
References: <3D046BF0.DC36542C@raplounge.de>
Message-ID: <3D04A938.627EE705@engcorp.com>

Ben Thursfield wrote:
> 
> I am running two python scripts in 2 different frames simultanously, and
> one of the programm needs to access a list defined in the other
> programm. How do i manage to access that list?

The question is ambiguous.  When you say "frames", the only thing
that comes to mind is that you are running Python on client-side
Internet Explorer under Windows, as scripts in two different parts
of a window (the frames) simultaneously.  I think you must have
meant something different however.

Could you please describe more about your environment?  The
term "frame" can mean a number of things.

-Peter


From fperez528 at yahoo.com  Mon Jun 17 18:31:48 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 17 Jun 2002 16:31:48 -0600
Subject: ImageMagick for Python incl. Source
References: <aelmhc$35m$03$1@news.t-online.com>
Message-ID: <aelo0t$sog$1@peabody.colorado.edu>

Achim Domma wrote:

> Hi,
> 
> I have just updated http://www.procoders.net/imagemagick with some
> documentation and small examples. There is also source code to compile
> ImageMagick for Python on non-windows plattforms, but this might be a bit
> tricky. Feel free to ask, if you have problems.

Sorry to display my ignorance here, but what on earth is a "Jamfile"? You 
should either specify that to build this on other platforms you need 'bjam' 
(non-standard in the Unix world) giving a link to it, or (more likely to be 
useful) provide a normal Makefile for the build. I tried google(bjam) and got 
all sorts of irrelevant links. Under Linux, 99% of the people will try to 
build this using gcc/make, so a simple makefile will probably do.

An unrelated question: do you know what advantages ImageMagick has over PIL? I 
ask out of honest curiosity. I've long been an ImageMagick fan, but since 
Python has sort of 'builtin' imaging support via PIL, I wonder what the 
differences are. Anyway good luck with this, I love ImageMagick so I hope 
your project can be built under Unix soon.

cheers,

f


From skip at pobox.com  Tue Jun 11 12:36:11 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 11 Jun 2002 11:36:11 -0500
Subject: getopt
In-Reply-To: <3D061A9F.7CDC7132@3captus.com>
References: <3D061A9F.7CDC7132@3captus.com>
Message-ID: <15622.9979.675543.351236@12-248-41-177.client.attbi.com>

    Bernard> I was expecting an exception from the following getopt.getopt()
    Bernard> called, but didn't.  Is it normal?  

Yes.  There's nothing invalid about "-d" as an argument to the "-c" flag, so
getopt gobbles it up.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html





From dw133 at cs.york.ac.uk.go.away  Fri Jun  7 10:35:21 2002
From: dw133 at cs.york.ac.uk.go.away (Darren Winsper)
Date: Fri, 07 Jun 2002 15:35:21 +0100
Subject: Detecting OS and Window Manager
References: <mailman.1023304388.9659.python-list@python.org> <ufvsco8mlhf213@corp.supernews.com> <3d00bfb5$1_2@hpb10302.boi.hp.com>
Message-ID: <ug1h5q29aulga9@corp.supernews.com>

In <3d00bfb5$1_2 at hpb10302.boi.hp.com>, Daniel Fackrell wrote:

> Correct me if I'm wrong, but isn't this exactly the pythonic way of handling
> dynamic loading of modules?
> 
> try:
>     import kdecore
> except ImportError:
>     # do something else

OK, fair enough.  I did only read and write my first line of Python a
few days ago :)

> Of course, this will probably not do what you want in this case, as kdecore
> will probably load if it is installed, without regard to whether the current
> user is running KDE.  I don't have much experience with this particular
> topic yet.

Well, I could attempt to see if something like "kdesktop" is running,
since if it is, the chances of running KDE are very high.  It's kludgy,
but I can't really think of anything else, since KDEINIT would be running
if a single kde application is running IIRC.

-- 
ICQ: 8899775 Jabber ID: darren at rooter.dyndns.org
Stop the EUCD before it's too late:
http://uk.eurorights.org/issues/eucd/
http://www.tomchance.uklinux.net/articles/darkages.shtm



From Lutz.Schroeer at kybernetik-manufaktur.de  Sat Jun  8 08:39:32 2002
From: Lutz.Schroeer at kybernetik-manufaktur.de (Lutz Schroeer)
Date: 8 Jun 2002 12:39:32 GMT
Subject: pygoogle search restricted by date
References: <adfmsk$205i$1@ariadne.rz.tu-clausthal.de> <Xns92257522D50DCduncanrcpcouk@127.0.0.1>
Message-ID: <Xns922795C82CE35Latzikatz@139.174.2.56>

Duncan Booth <duncan at NOSPAMrcp.co.uk> wrote in
news:Xns92257522D50DCduncanrcpcouk at 127.0.0.1: 

>> restrict_query = 'daterange:%d-%d' % (yesterday, today)
>> 
>> data = google.doGoogleSearch('python', restrict = restrict_query)
> 
> restrict is for searching by country or topic. I think you just want
> to put the daterange in as part of the main query string:
> 
>> data = google.doGoogleSearch('python '+restrict_query)

Thanks, that did the thing. 


From gh_pythonlist at gmx.de  Wed Jun  5 05:43:11 2002
From: gh_pythonlist at gmx.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Wed, 5 Jun 2002 11:43:11 +0200
Subject: [ANN] pyPgSQL 2.1 released
Message-ID: <20020605094311.GB31203@gargamel.hqd-internal>

Announce: pyPgSQL - Version 2.1 is released.
===========================================================================

pyPgSQL v2.1 has been released. It is a bug fix release to version 2.0, but
also includes some enhancements.

It is available at http://pypgsql.sourceforge.net.

pyPgSQL is a package of two (2) modules that provide a Python DB-API 2.0
compliant interface to PostgreSQL databases. The first module, libpq,
exports the PostgreSQL C API to Python. This module is written in C and
can be compiled into Python or can be dynamically loaded on demand. The
second module, PgSQL, provides the DB-API 2.0 compliant interface and
support for various PostgreSQL data types, such as INT8, NUMERIC, MONEY,
BOOL, ARRAYS, etc. This module is written in Python and works with
PostgreSQL 7.0 or later and Python 2.0 or later.

Note: It is highly recommended that you use PostgreSQL 7.1 or later and
Python 2.1 or later.

PostgreSQL is a sophisticated Object-Relational DBMS, supporting almost all
SQL constructs, including sub-selects, transactions, and user-defined types
and functions. It is the most advanced open-source database available any-
where More information about PostgreSQL can be found at the PostgreSQL home
page at http://www.postgresql.org.

Python is an interpreted, interactive, object-oriented programming lang-
uage. It combines remarkable power with very clear syntax. It has mod-
ules, classes, exceptions, very high level dynamic data types, and dynamic
typing. There are interfaces to many system calls and libraries, as well
as to various windowing systems (X11, Motif, Tk, Mac, MFC). New builtin
modules are easily written in C or C++. Python is also usable as an exten-
sion language for applications that need a programmable interface. Python
is copyrighted but freely usable and distributable, even for commercial
use. More information about Python can be found on the Python home page at
http://www.python.org.

---------------------------------------------------------------------------
ChangeLog:
===========================================================================

Changes since pyPgSQL Version 2.0
=================================

pyPgSQL 2.1 is now compatible and tested to work with PostgreSQL 7.2.x.

Changes to README
-----------------
* Added documentation for the new TransactionLevel attribute of
the Connection object.

Changes to PgSQL.py
-------------------
* Added code to implement support for setting PostgreSQL transaction
levels. [Feature Request #481727].

* Got rid of redundant building and storing of the
mapping of column names to column positions in the PgResultSet class. Now,
rows of the same query are instances of a dynamically created class, which
has this mapping as a class attribute instead of an attribute of the
instance. This saves a lot of space, and also slightly increases
performance of cursor fetches.
	
* Fixed the array parsing so it also works with PostgreSQL versions 7.2.x.
The PostgreSQL developers changed the quoting in the string representation
of arrays in 7.2 beta cycle: strings are now only quoted when otherwise
the array representation would be ambiguous. The new parseArray() method
should handle the old and new way of quoting in arrays. [Bug #539769].
Also added a new testcase for the ARRAY type.

* Improved the array parsing, so that it now passes all
the new mean testcases. Added support for parsing multidimensional arrays.
Eventually, the array parsing code should go as a support function into
libpq.

* Replaced all typechecks with "is" operators instead
of equals. Mark McEahern had a problem with using pyPgSQL in combination
with a FixedPoint class where the reason was that the FixedPoint class was
not comarable to None. The consensus on python-list was that None and all
types are singletons, so they should be checked using "is", which is also
faster, because it only checks for object identity.

* Fixed a couple of problems found by Steven D. Arnold:
1. A undefined variable was used in a few places where notices were popped
   from the notice list.
2. Comparisons between 2 PgNumerics gave the wrong result.

* Fixed problem that occurs when the sum() aggregate returns a
NULL. [Bug #505162].

Changes to pgversion.c
----------------------
* Allow for development and beta versions of PostgreSQL

Changes to libpqmodule.c
------------------------
* Removed special escaping of control characters in arrays.

* Added support for two missing OID types: aclitem and macaddr. 

* Applied patch by Chris Bainbridge [Patch #505941] "fix for null bytes in
bytea".

Changes to pgresult.c
---------------------
* Change the point at which an OID is tested to see if it is a
Large Object from 1700 to 16383.




From kragen at pobox.com  Mon Jun 10 17:43:30 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 17:43:30 -0400
Subject: Embedding and threads ?
References: <mailman.1022667286.17372.python-list@python.org> <j44rgrnkfm.fsf@informatik.hu-berlin.de> <mailman.1022740196.32586.python-list@python.org> <j4d6vdwvyk.fsf@informatik.hu-berlin.de> <1022771667.689.136.camel@netgroup> <j4znyhzguh.fsf@informatik.hu-berlin.de> <1022830475.893.17.camel@netgroup> <j4znygd7iv.fsf@informatik.hu-berlin.de> <1022839266.823.66.camel@netgroup> <j4d6vac9b7.fsf@informatik.hu-berlin.de> <mailman.1023091388.3109.python-list@python.org> <j4wutg8shk.fsf@informatik.hu-berlin.de> <mailman.1023108789.25792.python-list@python.org>
Message-ID: <83ofeiq071.fsf@panacea.canonical.org>

Bo Lorentsen <bl at netgroup.dk> writes:
> Ok, I think my problem is that I try to model my code around a thread
> model (to keep things clear), and I got confussed about the way Python
> did things, as I somehow assumed it had a more strit forward thread
> model (ala : 1 thread =3D 1 interpreter state).
> 
> I will use Python for this project as long as what I do only need a
> single thread model, and when I need more than that I'll se what to do
> about it :-)

It sounds like you want processes, not threads.  Assuming you're not
on Microsoft Windows, os.fork() will give you one Python interpreter
state per process.  Different processes have to make special
arrangements to communicate, though, through explicitly shared memory
or explicit channels.



From jb at cascade-sys.com  Sun Jun 30 03:11:14 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Sun, 30 Jun 2002 00:11:14 -0700
Subject: Python needs better error reporting
References: <LNBBLJKPBEHFEDALKOLCEEOKAAAB.tim.one@comcast.net>
Message-ID: <3D1EAF12.84CE5914@cascade-sys.com>

I rather think Dave has a good point.

A better parser could give a more informative diagnostic in this case and I
suspect others.

Tim Peters wrote:

> Well, "xor" isn't legit there, but other tokens that are OK after
>
>     if s == ' '
>
> include
>
>    [a ong list]

The question is NOT be what all may follow the string literal.  Rather it should
be what tokens are legal between the string literal and end of line?  Very
different questions.

If I'm not mistaken, only ":" or "\" are legal in that narrow context.  While
additional tokens may be legal in terms of following the string literal, they
still result in an error, so there's no point in listing them.  Line
continuation is a special case all its own.  Either it's illegal for some
reason, in which that situation constitutes a diagnostic all its own.  Otherwise
it can be completely ignored, as any errors/diagnostics would involve the next
following token, not the line continuation itself.

So it would be perfectly reasonable for the parser to say "semicolon missing" or
"semicolon expected" in this case.

Whether or not it's worth the effort to implement is, as always, debatable and
furthermore, I suppose, moot unless somebody is offering to do the work.

Regards

--jb

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From maxm at mxm.dk  Fri Jun 14 04:40:11 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 14 Jun 2002 10:40:11 +0200
Subject: Date formats
References: <3D09A1B1.6040705@mxm.dk> <Xns922D5E1FF60FEduncanrcpcouk@127.0.0.1>
Message-ID: <3D09ABEB.4000209@mxm.dk>

Duncan Booth wrote:

> The documentation says that on some platforms an optional field width and 
> precision specification can follow the initial '%'. This is obviously 
> highly non-portable, but might provide a suitable solution if you are not 
> desperate for portability. For example, if you are running under Windows 
> using "%#d-%#M-%Y" as the format string gives the format you require.


Oh ... I cannot seem to find it in the docs. The closest I come is in:

2.2.6.2 String Formatting Operations

where it says that:

"# The value conversion will use the ``alternate form'' (where defined 
below)."

So I guess that single digit is the "alternate form" here.

Well anyhoo it solves my problem as it' will only run on a windows 
server, and even if it's ever moved to another platform my code has no 
trouble reding the two digit form, so I guess I will settle for that.

thanks Max M



From huaiyu at gauss.almadan.ibm.com  Thu Jun 27 15:27:44 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Thu, 27 Jun 2002 19:27:44 +0000 (UTC)
Subject: Reminder of a python Project.
References: <2218.941T781T12224051threeseas@earthlink.net> <afc2mn$2kn$1@verence.demon.co.uk> <2948.942T2959T5796077threeseas@earthlink.net>
Message-ID: <slrnahmppf.9av.huaiyu@gauss.almadan.ibm.com>

Timothy Rue <threeseas at earthlink.net> wrote:
>Anyway, the flamers didn't have to respond at all. It was a simple post
>and the reference to research indicating that GPL projects have a low
>count of developers, with what I believe is 1 as a mean, is as well what
>research found (sorry no link - search for it yourself - /. had it linked)
>
>It was an informational post, that all.
>
>Amazing who all something like that get a flaming response from. Good
>thing the world is not made up of nothing but, for we wouldn't have any
>research then.

It looks like nobody else is talking to you in plain English this time.
I'll make one last effort in the hope of making it clear to you why this is
so.  There's no point in replying if you do not agree.  Indeed, that would
be an indication that I failed.

First off, there is no point in explaining the cost of software bugs.
Everybody knows it, including those who said, tongue in cheeck, that all
such problems are solved.  It's the same if someone starts to say that road
accident is a big problem and we should reduce it.  (No, please don't start
to analyse how inadequate this analogy is; that would miss the point
entirely.  It is only intended to convey the kind of impression people have
about such claims.  Either you get the idea, or you don't.)

Second, many poeple are convinced that a better theory/framework/paradigm is
needed to go to the next stage. (We are living in an exciting time, don't
we?)  However, no one here is convinced in the slightest bit that you've got
_it_.  It is like claiming, in the 1940s, that machines should be capable of
controlling themselves, and that all this is reduced to the basic principles
of "action and reaction".  That won't cut it.  On the other hand, someone
who actually developed the theory of transfer functions, frequency analysis
and various associated techniques, and identified a specific class of
machines and a specific class of goals for which it is useful, would have
been ranked among the pioneers of control theory.  Again, this is just an
analogy to convey the idea.  Either you get the idea, or you don't.  

Thirdly, if you really want to convince people of something, it is to your
advantage to focus on the positive things instead of the negative.
Retorting people who discounts your "discovery" just makes more people
annoyed at you.  Touting the importance of research that claims the
importance of some problem does not in any way advance the importance of
your approach to solving that said problem.  Instead of fighting people in
this group (ok, this time they started to fight you first, but that's beside
the point), it is more to your advantage to do some solid research and come
up with at least one real solution to one real prolem (in the science sense,
not in the marketing sense).

Finally, get really worried if slashdot becomes your primary feed of
"research".

why-am-i-wasting-my-time-ly y'rs

Huaiyu



From Alexandre.Fayolle at logilab.fr  Fri Jun  7 12:50:52 2002
From: Alexandre.Fayolle at logilab.fr (Alexandre)
Date: Fri, 7 Jun 2002 18:50:52 +0200
Subject: [ANN] constraint-0.2
Message-ID: <20020607165052.GG7579@orion.logilab.fr>

Logilab has released constraint-0.2.

The constraint package is an extensible Constraint Satisfaction Problem
(CSP) solver using constraint propagation algorithms written in 100%
pure Python (tested with Python 2.1)

You can download it from ftp://ftp.logilab.org/pub/constraint/, read a
bit more about it on
http://www.logilab.org/python-logic/constraint.html,
and discuss it on the logic-sig mailing list (see
http://lists.logilab.org/mailman/listinfo/python-logic to subscribe)

The package is still young, but we find it useful enough right now, so I
thought I might as well publish it.

What can it do?
 * solve some finite-domain CSP
 * handle n-ary constraints

How can you extend it:
 * by implementing new Domain classes
 * by implementing new constraint classes
 * by implementing new Distributor classes

How well does it perform?
 * 8-queens problems solved in 15 seconds on my 1GHz PC
 * SEND+MORE=MONEY problem solved in 8 seconds on the same machine

What has changed?
 * a BasicConstraint class was added (these constraint affect only a single
 variable, and can be entailed in one pass), with several example
 implementations
 * The Repository class was refactored, and a Solver class was added as
 a result, which gives a better separation of the solving logic and
 storage logic

If you're interested by constraint propagation programming in python,
please join the logic-sig mailing list.
       
-- 
Alexandre Fayolle
-- 
LOGILAB, Paris (France).
http://www.logilab.com   http://www.logilab.fr  http://www.logilab.org
Narval, the first software agent available as free software (GPL).





From candiazoo at attbi.com  Sat Jun 29 10:44:25 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Sat, 29 Jun 2002 14:44:25 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com> <mailman.1025336262.12245.python-list@python.org>
Message-ID: <3d1dc547.468619140@netnews.attbi.com>

Basically, Erv, what I did was to try optimize any code referenced in the major
loop of the function that was taking so long.  I ended up getting that one piece
down from 8 hours to 5 hours of processing time.  Still forever, but not any
longer than a VB app that I am replacing took to run...  in fact it is a bit
quicker.  Total run-time of the program is now 7 hours 30 minutes down from much
longer (the original VB program ran for 24 hours, but it also generated a
fulcrum database along with the text files and also pulled back a few hundred
thousand more records for another product we market).  Part of that is replacing
a MS Access database with mySQL.  Part of that is my aforementined
optimizations...

1) "import sys" became "import exit from sys"
2) "import os" became "import makedires from os"
3) "import time" bacame "import asctime from time" (this one is used frequently)
4) I localized constants I was using in boolean tests (there were declared at
the beginning of the module, I moved them to the function call just before the
loop).
5) I localized self.__source_id...  source_id = self.__source_id before using it
in every loop iteration.
6) I localized self.fix_date_range()... fix_date = self.fix_date_range before
using it in every loop iteration/
7) I changed my method that built the output string by changing...

return self.val_a + self.value_b + etc..

...to...

return "%-10.10s%-200.200s (etc)" % (self.val_a, self.val_b, etc.)

...and those were basically the changes I made!  They made a huge difference!

Mike J.



From fredrik at pythonware.com  Tue Jun 11 17:01:40 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 11 Jun 2002 21:01:40 GMT
Subject: Making an About Box...
References: <Xns9229BE9B416A2endymion84charternet@216.168.3.40>
Message-ID: <UytN8.42842$n4.9774115@newsc.telia.net>

"TuxedoKamen" wrote:

> Hello.  I'd like to make an About Box using Tkinter.  I'm using a grid
> inside a Toplevel widget, but I don't really like this solution.  I'd like
> a window that did not have the minimize and maximize buttons, like the
> tkMessageBox derived boxes showinfo, showwarning, etc.

create a toplevel widget, and use the "transient" method
to associate it with a parent window.  for more info, see:

    http://www.pythonware.com/library/tkinter/introduction/dialog-windows.htm

</F>




From n.larosa at araknos.it  Tue Jun 11 09:46:25 2002
From: n.larosa at araknos.it (Nicola Larosa)
Date: Tue, 11 Jun 2002 15:46:25 +0200
Subject: [OT] Antispamming the email address
References: <Xns9229E662C1E0AAusYourStandingInIt@130.133.1.4> <rkcbgu8j0pi93be4m2oej7q18gpigc36sg@4ax.com>
Message-ID: <3D05FF31.4060701@araknos.it>

 > Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')

That's nifty. :^)

You could encode it in rot13 format instead, so it'll keep the
shape of an email address and *will* be harvested, while still being
invalid. Let's throw more wrenches in the wheels of spammers!

Like this:

 > Real email address: 'gmbg at fvy-grp.te'.decode('rot13')


-- 
"Too much cleverness in the parser can turn against you."
     Guido Van Rossum

Nicola Larosa - nico at tekNico.net






From nicolas.torzec at rd.francetelecom.com  Mon Jun  3 04:03:45 2002
From: nicolas.torzec at rd.francetelecom.com (Nicolas Torzec)
Date: Mon, 3 Jun 2002 10:03:45 +0200
Subject: How to suspend execution for at least x seconds in python ? (ie does "sleep" exist in Python ?)
Message-ID: <adf7t3$ki71@news.rd.francetelecom.fr>

Dear all,
I was wondering if an utility like "sleep x" exist in python ?
Under Unix-like operating systems, the sleep utility suspends execution for at least x seconds under...


Thank you in advance,
Nicolas Torzec. 




From insaney at ufl.edu  Mon Jun  3 13:03:57 2002
From: insaney at ufl.edu (Daniel)
Date: 3 Jun 2002 10:03:57 -0700
Subject: UnboundLocalError: local variable part 2
Message-ID: <40a8aa92.0206030903.466fc3ff@posting.google.com>

That last example was a bit too contrived... instead of blah='abc',
blah is actually a list and i'm appending and popping elements so
func1 actually truly does affect the global blah.  So a better example
of the logic is:

def func1():
   blah.pop()

def func2():
   print blah
   blah.pop()

def main():
   func1()
   try:
      func2()
   except:
      pass
   print blah

blah=[1,2,3]
main()

except instead of:
[1, 2]
[1]

it's getting:
[1, 2]
(trap the UnboundLocalError)
[1, 2]
 
I've looked through and cannot find anything that vi might have put in
that is throwing the python interpreter off.

Any thoughts on what might cause this behavior?
Thanks,
Daniel


From ken at hotmail.com  Mon Jun  3 11:21:45 2002
From: ken at hotmail.com (Ken)
Date: Tue, 4 Jun 2002 01:21:45 +1000
Subject: Automated Email
Message-ID: <adg1do$10nmr5$1@ID-49758.news.dfncis.de>

How to do automated emailing when results being processed?

Thanks




From peter at engcorp.com  Tue Jun 25 22:51:42 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 25 Jun 2002 22:51:42 -0400
Subject: Switch from 2.1 to 2.2
References: <f29c2586.0206250600.48e0a8e3@posting.google.com> <3D188302.1030109@thomas-guettler.de> <zv0S8.3036$%N4.731526@news20.bellglobal.com>
Message-ID: <3D192C3E.6926BE61@engcorp.com>

Mathieu Belanger wrote:
> 
> > Plese send the stracktrace, too
> >
> 
> Yeah, I know it will be easier, but I can't have a complete one for any
> reason !?!
> 
[...]
>   File "D:\Python22\Lib\site-packages\_xmlplus\dom\ext\Printer.py", line
> 381, in step
>     self.visitor.visit(self.start_node)
>   File "D:\Python22\Lib\site-packa

Do you really mean that your stack trace, as print on the stdout of
your machine, is truncated like that, in the middle of the word
"site-packages"?  That's a pretty serious problem, I think.  Or was
it just a cut-and-paste error?

Without the final line of that stacktrace it might be hard to see
what the problem really is...

-Peter


From mike at radiomag.com  Tue Jun 18 02:51:12 2002
From: mike at radiomag.com (Michael Steuer)
Date: Tue, 18 Jun 2002 16:51:12 +1000
Subject: Listing all available (not just loaded) modules
References: <aem9om$888fb$1@ID-140151.news.dfncis.de> <3D0EB342.43DB0FE3@engcorp.com>
Message-ID: <aeml81$862pn$1@ID-140151.news.dfncis.de>

Thanks for that link. The "sakecharmer.cgi" at that page was of great help.:)

Cheers,

Mik.e




From duncan at NOSPAMrcp.co.uk  Fri Jun 14 04:34:21 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Fri, 14 Jun 2002 08:34:21 +0000 (UTC)
Subject: Python 2.2.1
References: <II7O8.622$e27.275217555@twister1.starband.net>
Message-ID: <Xns922D5FE7B72D8duncanrcpcouk@127.0.0.1>

"Dave" <talbotdr at starband.net> wrote in
news:II7O8.622$e27.275217555 at twister1.starband.net: 

> Hi, I am 12 years old and would like to begin learning Python over the
> summer. I have Windows 98 SE. I have python 2.2.1 and all of its
> utilities on my computer. I am taking a tutorial from 
> http://softwaredev.earthweb.com/sdopen/article/0,,12382_626311,00.html

This isn't related to your question (there seem to be plenty of answers to 
that already), but if you are teaching yourself Python you might also want 
to look at http://www.livewires.org.uk/python/

They have a Python course using a set of worksheets which start off simple 
but include writing Asteroids and Pacman games.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From shalehperry at attbi.com  Fri Jun  7 22:41:31 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 07 Jun 2002 19:41:31 -0700 (PDT)
Subject: Efficient python programming...
In-Reply-To: <3D015771.4BDCB22F@engcorp.com>
Message-ID: <XFMail.20020607194131.shalehperry@attbi.com>

> 
> I'll step back from any claim that unit tests really prove anything,
> but in a very Pythonic and pragmatic fashion they are Good Enough to 
> make them a /necessity/ for anyone trying to deliver reliable software.
> 

my current project is a window manager for X Windows.  It is all X lib calls
and UI events.  I can maybe test 5% of it.  Seems most of the programs I have
worked on turn out that way.  I write as many tests as I can but usually it
just doesn't work out for me.

But you are right, test as early and as often as possible and the code can't be
all bad (-:




From missive at frontiernet.net  Fri Jun 21 18:36:49 2002
From: missive at frontiernet.net (Lee Harr)
Date: Fri, 21 Jun 2002 22:36:49 -0000
Subject: Generating SWFs using Ming via CGI
References: <2wvQ8.19$eH2.59629@ruti.visi.com> <r35uea.rc.ln@grey.aerojockey.localdomain>
Message-ID: <uh7ak1qa08psa7@corp.supernews.com>

>> Content-type: applicaton/x-shockwave.flash
>> 
> First, check your spelling.  The MIME type of SWF files is
> application/x-shockwave-flash (no period in it).
> 

right, and application, not applicaton



From mf at mrinfo.de  Fri Jun 28 11:26:01 2002
From: mf at mrinfo.de (Marko Faldix)
Date: Fri, 28 Jun 2002 17:26:01 +0200
Subject: xslt on windows
Message-ID: <afhv32$emfjh$1@ID-108329.news.dfncis.de>

Hello,

I tried installing 4suite's xslt on Win 2000, but I really don't know how to
deal with pathes.

I used

  Python-2.2.1.exe
  win32all-146.exe
  4Suite-0.12.0a2.win32-py2.2.exe

All scripts, wether if command-line or through web server, doesn't know
module xml.xslt.

Unfortunately, you can't import a module   *.site-packages.*, where * is
anything, because "-" is forbidden as part of module names.

Also, you can't copy directories from site-packages to c:\python22, they
don't work properly then.

Is there anybody using xslt with python? Maybe better with linux in place of
windows?

Best wishes,
--
Marko Faldix
M+R Infosysteme
Hubert-Wienen-Str. 24     52070 Aachen
Tel.: 0241-93878-16     Fax.:0241-875095
E-Mail: marko.faldix at mplusr.de




From walter at livinglogic.de  Sat Jun 29 07:50:19 2002
From: walter at livinglogic.de (=?windows-1252?Q?Walter_D=F6rwald?=)
Date: Sat, 29 Jun 2002 13:50:19 +0200
Subject: Python needs better error reporting
References: <LNBBLJKPBEHFEDALKOLCEEOKAAAB.tim.one@comcast.net>
Message-ID: <3D1D9EFB.6050607@livinglogic.de>

Tim Peters wrote:

> [David LeBlanc]
> 
>>Cute anthropomorphism is lots of fun, but doesn't _quite_ obscure the fact
>>that computers don't really "tell" anybody anything, programmers do.
> 
> 
> When we ship Guido with every compiler, the world will be a better place for
> sure.

This is easy: Use the _G_uido _C_loning _C_ommand! ;)

Bye,
    Walter D?rwald






From gerhard.haering at gmx.de  Thu Jun 27 13:29:52 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 27 Jun 2002 17:29:52 GMT
Subject: crypt for windows?
References: <mailman.1025187895.14828.python-list@python.org>
Message-ID: <slrnahmisf.3eo.gerhard.haering@lilith.my-fqdn.de>

Klaus Reinhardt wrote in comp.lang.python:
> Hi
> Is there a crypt for windows-python, so
> I can manage some password-files?

Probably not. I believe Cygwin's Python has this module. Is that an
option for you?

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From phd at phd.pp.ru  Tue Jun 11 15:44:24 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 11 Jun 2002 23:44:24 +0400
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Jun 10)
In-Reply-To: <ae5cce$hpd$1@panix1.panix.com>; from aahz@pythoncraft.com on Tue, Jun 11, 2002 at 01:35:10PM -0400
References: <EE3CDB256E0A231C.577F9FCCC77B96EC.B451C214671092A5@lp.airnews.net> <ae5cce$hpd$1@panix1.panix.com>
Message-ID: <20020611234424.C23356@phd.pp.ru>

On Tue, Jun 11, 2002 at 01:35:10PM -0400, Aahz wrote:
> >    Oleg BroytMann posted glue that automatically moves vim users to
> 
> Broytman

   It is arguable :) This is transliteration from Russian, so there are too
many way to do it. When my sister moved to Israel, they wrote in her
passport "Brojtman"!
   When I was considering my way of the transliteration I decided to use
German-like "mann" instead of English-like "man"; after all these are
German-like surnames. And in programs I am almost always do
DoubleCapitalization and write "BroytMann"! :)

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From phr-n2002b at NOSPAMnightsong.com  Tue Jun 18 22:13:02 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 18 Jun 2002 19:13:02 -0700
Subject: tomorrow in yyyymmdd format
References: <qpbvguscmblone2r92o8sfc7evo5709mq8@4ax.com>
Message-ID: <7xznxsf235.fsf@ruckus.brouhaha.com>

Giulio Cespuglio <have.a.look at the-body.com> writes:

> We all know that if I want to get the present GMT date in yyyymmdd
> format I should say
> 
> 	print time.strftime("%Y%m%d", time.gmtime() )

print time.strftime("%Y%m%d", time.gmtime(time.time() + 24*60*60))


From twisted at itamarst.org  Sun Jun  2 03:06:18 2002
From: twisted at itamarst.org (Itamar Shtull-Trauring)
Date: Sun, 02 Jun 2002 09:06:18 +0200
Subject: ANN: Twisted 0.18.0
Message-ID: <3CF9C3EA.3050603@itamarst.org>

Twisted is an event-based framework for internet applications, written
in Python and licensed under the LGPL. For more information, visit:

      http://www.twistedmatrix.com

or join the list at:

      http://twistedmatrix.com/cgi-bin/mailman/listinfo/twisted-python


What's New since 0.17.4
=======================

0.18.0 is a major release, cleaning up many of the main APIs in 
twisted.internet and moving us much closer to a 0.99.x series leading up 
to Twisted 1.0.

- The networking core of Twisted - twisted.internet - was totally 
refactored, cleaned up, documented and improved, using explicit 
interfaces. The API is now much improved, and mostly stable. The main 
change was to create an explicit reactor object, using which you can 
schedule events, connect or listen on TCP/UDP/SSL ports, run methods in 
a thread pool, and so on. The tools that come with Twisted have been 
extended to allow choosing the reactor at runtime. Currently supported 
are select(), poll(), GTK+, Qt, Win32 event API, and Java thread based 
reactors, with wxPython and Tkinter support integrating with whichever 
reactor is chosen.

- Perspective Broker, Twisted's remote object protocol, now supports 
returning the equivalent of exceptions for failed remote method calls. 
Additionally, resource limiting was added to the implementation.

- Applications can now be persisted to XML, using marmalade, a XML 
pickle equivalent. This allows you to edit low level application 
settings using a text editor.

- Added a sample IRC bot - it listens on a channel and logs it to a file.

- Improvements to twisted.mail - the SMTP server now has basic support 
for bouncing.

- Improvements to t-im, Twisted's instant messaging client.

- LDAP was moved out into a separate package:
   http://twistedmatrix.com/users/tv/ldaptor/

- Many bug fixes, feature enhancements and additional documentatiom.


What is Twisted?
================

Twisted is an event-based framework for internet applications.  It
includes a web server, a telnet server, a chat server, a news server, a
generic client and server for remote object access, and APIs for
creating new protocols and services. Twisted supports integration of the
Tk, GTK+, Qt or wxPython event loop with its main event loop. The Win32
event loop is also supported, as is basic support for running servers on
top of Jython.

Twisted currently supports the following protocols, all implemented in
pure python, most of them as both servers and clients:

       - NNTP
       - FTP
       - HTTP
       - SOCKSv4
       - SMTP
       - IRC
       - telnet
       - POP3
       - AOL's instant messaging TOC
       - OSCAR, used by AOL-IM as well as ICQ (client only)
       - DNS
       - finger
       - Echo, discard, chargen and friends
       - Twisted Perspective Broker

For more information, visit http://www.twistedmatrix.com, or join the
list at http://twistedmatrix.com/cgi-bin/mailman/listinfo/twisted-python






From martin at v.loewis.de  Wed Jun  5 16:44:39 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 05 Jun 2002 22:44:39 +0200
Subject: Urllib thread safe?
In-Reply-To: <20020605231131.A30904@phd.pp.ru>
References: <mailman.1023280808.19305.python-list@python.org>
	<m3it4xobkb.fsf@mira.informatik.hu-berlin.de>
	<20020605231131.A30904@phd.pp.ru>
Message-ID: <m33cw1o3pk.fsf@mira.informatik.hu-berlin.de>

Oleg Broytmann <phd at phd.pp.ru> writes:

> On Wed, Jun 05, 2002 at 07:55:00PM +0200, Martin v. Loewis wrote:
> > Yes, that is thread-safe. urllib does not maintain any per-url state,
> 
>    Hah? It maintains global _urlopener...

Sure, but the OP said he uses

urllib.URLopener().open(...)

which does not make use of _urlopener. However, I believe that even
the global _urlopener is thread-safe, since it does not maintain any
per-url state. About the only thing that changes in this object is
__tempfiles, and that is also thread-safe (since .append is atomic).

Regards,
Martin




From joonas.paalasmaa at iki.fi  Thu Jun 13 06:42:11 2002
From: joonas.paalasmaa at iki.fi (Joonas Paalasmaa)
Date: Thu, 13 Jun 2002 10:42:11 GMT
Subject: How to check version of Python under CGI
References: <rbmggus7ln1uiutktf5libf4bbngu29ji7@4ax.com> <873cvrin1q.fsf@tux.ntw23.fr> <cgnggu40ethjhpqcmltnfb06d5qvuf078k@4ax.com>
Message-ID: <3D0877FF.3090401@iki.fi>

Johann wrote:
> Thanx. Do you know how to check what modules are installed? E.g. I
> would like to check if PYANA or other xml modules are installed.

try:
     import somemodule
     print "somemodule installed"
except ImportError:
     print "somemodule not installed"



From jwbaxter at spamcop.net  Thu Jun 20 00:30:03 2002
From: jwbaxter at spamcop.net (John Baxter)
Date: Wed, 19 Jun 2002 21:30:03 -0700
Subject: Unicode hexadecimal characters
References: <3eeda89d.0206191615.71b0629d@posting.google.com> <3D112C60.9C3D2F9F@engcorp.com> <njaQ8.101599$6m5.87417@rwcrnsc51.ops.asp.att.net> <3D1137E2.7CB8F814@engcorp.com> <jwbaxter-CB588C.20463919062002@corp.supernews.com>
Message-ID: <jwbaxter-01A379.21300319062002@corp.supernews.com>

In article <jwbaxter-CB588C.20463919062002 at corp.supernews.com>,
 John Baxter <jwbaxter at spamcop.net> wrote:

> No representation I've seen is less pretty than NCR's for the C-315 (ca 
> 1963).  The 6 digits for the hex values were simply the six characters 
> which happened to fall after the digits in the four-byte world.  I think 
> 10 was @...12 was space, and others were , . and two more.
> 
> Memory dumps in which spaces are either significant data or whitespace 
> depending on column "can be a nuisance."  ;-)

Some idiot name Baxter wrote the above.  Of course, he meant the 
"four-bit world".  The C-315 didn't have bytes, it had "slabs" of 12 
bits containing two characters or three digits depending on what one 
meant.

But all of you figured that out for yourselves and laughed at Baxter, 
anyhow.

Sorry about that.  --John

Short for syllable because someone felt that 12 bits was too short for a 
"word."


From logiplexsoftware at earthlink.net  Tue Jun 11 14:11:01 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 11 Jun 2002 11:11:01 -0700
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
In-Reply-To: <15622.26.84312.207270@12-248-41-177.client.attbi.com>
References: <Xns9229EE722467Fgustafl@195.100.94.182>
	<mailman.1023746539.13373.python-list@python.org>
	<Xns922AD3C74280gustafl@195.100.94.182>
	<15622.26.84312.207270@12-248-41-177.client.attbi.com>
Message-ID: <20020611111101.420d7b9a.logiplexsoftware@earthlink.net>

On Tue, 11 Jun 2002 08:50:18 -0500
Skip Montanaro wrote:

> You could also fold the int case into the tuple case (making it a bit more
> uniform), but at a slight cost in performance if you have lots of ints.  In
> this case, all the try/except statement is responsible for is setting s and
> e:
> 
>     def isin(n, ranges):
>         for r in ranges:
>             try: 
>                 s, e = r
>             except TypeError:
>                 s, e = n, n
>             if s <= n <= e:
>                 return 1
>         return 0

I do think this is a bit clearer, however I know you meant to write

except TypeError:
    s, e = r, r



-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From shagshag13 at yahoo.fr  Tue Jun  4 06:23:10 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Tue, 4 Jun 2002 12:23:10 +0200
Subject: Looking for a list subclassing example...
Message-ID: <adi4dv$118tko$1@ID-146704.news.dfncis.de>

Hello,

I'm looking for a list subclassing example, and i can't find it...

Do you have one to show ?

Thanks,

s13.





From mgerrans at mindspring.com  Sat Jun 29 01:15:46 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Fri, 28 Jun 2002 22:15:46 -0700
Subject: I'm an idiot
References: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16> <Xns923BD424CF53Cdavidgriswold1netsca@204.127.68.17> <afjfdg$mh2$1@slb7.atl.mindspring.net>
Message-ID: <afjfti$e27$1@slb0.atl.mindspring.net>

Oops, silly me, it was posted, but I think there was a typo in it:

> for line in open('c:/temp/temp.txt'):

is missing the ".readlines()" part.




From rivers at cars.uchicago.edu  Sun Jun 23 20:35:33 2002
From: rivers at cars.uchicago.edu (Mark Rivers)
Date: Sun, 23 Jun 2002 19:35:33 -0500
Subject: Problems with Pmw
Message-ID: <QUtR8.20$U4.2298@news.uchicago.edu>

Folks,

In a previous post I mistakenly attributed one problem to Pmw.Blt, while
subsequent testing has shown that the problem has nothing to do with Blt, it
is a problem (?) with Pmw.

The problem is illustrated by the following short script:
########################################
from Tkinter import *
import Pmw
t1 = Tk()
t2 = Tk()
l = Label(t2, text='A label'); l.pack()
b = Button(t2, text='Press me'); b.pack()
e = Pmw.EntryField(t2, labelpos=W, label_text='Enter something'); e.pack()
########################################

Executing this script as shown results in the first Tk() window (t1) being
created, and the second Tk() window (t2) and the Label and Button widgets
being created, and then the following error when creating the
Pmw.EntryField:

Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "testBlt2.py", line 7, in ?
    e = Pmw.EntryField(t2, labelpos=W, label_text='Enter something');
e.pack()
  File "C:\PROGRA~1\Python21\Pmw\Pmw_0_8_5\lib\PmwEntryField.py", line 72,
in __
init__
    sequences = root.bind_class(tag)
  File "C:\PROGRA~1\Python21\lib\lib-tk\Tkinter.py", line 923, in bind_class
    return self._bind(('bind', className), sequence, func, add, 0)
  File "C:\PROGRA~1\Python21\lib\lib-tk\Tkinter.py", line 858, in _bind
    return self.tk.splitlist(self.tk.call(what))
TclError: bad window path name ".15422476.15422044"

If the line "t1 = Tk()" is commented out then it works fine.

This happens both on Win32 and Linux with Python 2.1, and Pmw 0.8.5.

I think it should be legal to create multiple Tk() windows in an
application, so this seems like a serious bug?

Thanks,

              Mark Rivers
              Dept. of Geophysical Sciences, University of Chicago
              GSECARS, Advanced Photon Source, Argonne National Laboratory





From jj at void.si  Thu Jun  6 18:25:42 2002
From: jj at void.si (jj)
Date: Fri, 7 Jun 2002 00:25:42 +0200
Subject: win32com map attributes to methods
Message-ID: <3cffe1db@news.perftech.si>

Hi, I am trying to map attributes in COM object to methods, so when
attribute (for example: Expires) is set a "set_Expires" method will be
called, and when the attribute is read a method prefixed with "get_"
(get_Expires) method should be called.

I think I  am on a good path, because my example is working 50%.
get_Expires is called when reading attr., but set_Expires is never called.
their implementation is so similar and simple, that there must be a simple
solution.

Thanks for your help


Janez



import sys, string
import pythoncom
import win32com.server.util
import win32com.client
from win32com.server.policy import DesignatedWrapPolicy

class PropWrapPolicy(DesignatedWrapPolicy):
    'policy that will handle properties correctly'
    def __init__(self, object):
        DesignatedWrapPolicy.__init__(self, object)
    def _wrap_(self, ob):
        DesignatedWrapPolicy._wrap_(self, ob)

        for name in ob._public_attrs_:
            lname = string.lower(name)
            id = self._name_to_dispid_[lname]
            if hasattr(ob, 'set_' + name):
                print 'found set_'+ name, id
            self._dispid_to_put_[id] = 'set_' + name

            if hasattr(ob, 'get_' + name):
                print 'found get_' + name, id
            self._dispid_to_get_[id] = 'get_' + name


class ResponseIStream:
    '''Istream for response'''
    _public_methods_ = ['Write', 'Flush']
    _public_attrs_= ['ContentType', 'Expires']

    _com_interfaces_ = [pythoncom.IID_IStream, pythoncom.IID_IDispatch]

    def __init__(self):
        pass
    def set_Expires(self, value):
        print 'set_expires', value
    def get_Expires(self):
        print 'get_expires'
        return 10
    def write(self, x):
        print 'write:', x
    def Write(self, s):
        self.write(s)
        return len(s)
    def Flush(self):
        print 'Iresponse.Flush'


if __name__=="__main__":

    o = win32com.server.util.wrap(ResponseIStream(),
pythoncom.IID_IDispatch, usePolicy = PropWrapPolicy)
    o = win32com.client.Dispatch(o)
    o.Flush()
    o.Write('test')
    print 'set'
   # !!!!! set_Expires should be called !!!!
    o.Expires = -1
    print 'get'
    print o.Expires

    xml = win32com.client.Dispatch("Msxml2.DOMDocument")
    xml.loadXML('<x>streaming</x>')
    xml.save(o)




From op73418 at mail.telepac.pt  Tue Jun 18 15:22:50 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Tue, 18 Jun 2002 20:22:50 +0100
Subject: what *is* a class?
References: <B43D149A9AB2D411971300B0D03D7E8BF0A396@natasha.auslabs.avaya.com> <mailman.1024398742.13711.python-list@python.org> <jp7uguoqjs7t23itb7vihai30cnv3qr7sm@4ax.com> <mailman.1024406543.4333.python-list@python.org>
Message-ID: <s71vgu8cfe4gn41m38gl6q75mo42i64c27@4ax.com>

On Tue, 18 Jun 2002 15:21:38 +0200, holger krekel
<pyth at devel.trillke.net> wrote:

>Gon?alo Rodrigues wrote:
>> Here goes my personal version of this module. It is largely based off
>> Sami Hangaslammi recipe at Active State plus some additions/differences
>> of mine.
>> ...
>
>thanks. that's a good starting point to implement an iterable module. 
>I guess we all agree that iterator versions of the builtin counterparts
>should at least accept the same arguments. 
>
>    holger
>

You are absolutely right. Here goes an improved version, where besides
some errors corrected, the iterator versions now behave as the builtin
ones.

All the best,
Gon?alo Rodrigues

"""
Iterator utilities module.

Acknowledgements: Largely inspired by Iterator utilities by Sami
Hangalasmmi.
"""

#Import generators.
from __future__ import generators


__version__ = 1.1
__author__ = "G. Rodrigues"


def iterzip(*iterators):
    """Iterator version of builtin zip."""
    #Initialize variables.
    iterators = [iter(iterator) for iterator in iterators]
    while 1:
        yield tuple([i.next() for i in iterators])


def iterzipper(*iterators):
    """Second version of the builtin zip, exiting only when all the
iterators are exausted."""
    #Initialize variables.
    iterators = [iter(iterator) for iterator in iterators]
    ret = []

    while 1:
        #Emulate map behaviour.
        none_added = 0
        ret = []

        for iterator in iterators:
            try:
                ret.append(iterator.next())
            except StopIteration:
                ret.append(None)
                none_added += 1

        #Return value or exit.
        if none_added == len(iterators):
            raise StopIteration
        else:
            yield tuple(ret)


def iterfilter(func, iterator):
    """Iterator version of builtin filter."""
    #Initialize variables.
    if func is none:
        func = lambda x: x

    for elem in iterator:
        if func(elem):
            yield elem


def iterwhile(func, iterator):
    """Iterate while func(elem) is true."""
    for elem in iterator:
        if func(elem):
            yield elem
        else:
            raise StopIteration


def itermap(func, *iterators):
    """Iterator version of builtin map."""
    #Initialize variables.
    if func is None:
        func = lambda *x: x

    for elems in iterzipper(*iterators):
        yield func(*elems)


def iterconcat(*iterators):
    """Concatenates all the iterators passed as arguments."""
    for iterator in iterators:
        for elem in iterator:
            yield elem


def iterapply(iterator, *args):
    """Applies an iterator of callables iteractively to args."""
    #Initialize variables.
    arg = args
    for func in iterator:
        arg = func(*arg)
        yield arg


def iterreduce(func, iterator, default = None):
    """Lazy iterator version of reduce builtin."""
    #Initialize variables.
    iterator = iter(iterator)

    #Get first value.
    try:
        prev = iterator.next()
    except StopIteration:
        if default is not None:
            yield default
        raise StopIteration

    #Get second value.
    try:
        next = iterator.next()
    except StopIteration:
        yield prev
        raise StopIteration

    prev = func(prev, next)
    yield prev

    #Continue looping.
    for next in iterator:
        prev = func(prev, next)
        yield prev


def iterslice(iterator, slice):
    """Lazy slice of an iterator."""
    #Initialize variables.
    iterator = iter(iterator)

    if isinstance(slice,int):
        stop = slice
    else:
        stop = slice.stop
        start = slice.start or 0
        step = slice.step or 1
    
    #Skip until start if start different from 0.
    if start:
        for i in range(start):
            iterator.next()
    else:
        i = 0

    while i < stop:
        yield iterator.next()
        i += 1
        #Skip step values.
        for i in range(i + 1, i + step):
            iterator.next()
        

#More specialized iterator utilities.
def itergroup(iterator, n):
    """Iterates through iterator, yielding a list of n iterations at a
time."""
    #Initialize variables.
    ret = []

    #Read n elems at a time.
    for elem in iterator:
        ret.append(elem)
        if len(ret) == n:
            yield ret
            ret = []

    #Yield what is left.
    if not ret:
        yield ret


def itersum(iterator):
    """Returns the sum of all the elements in the iterator, None if the
iterator has no elements."""
    #Initialize variables.
    iterator = iter(iterator)
    try:
        ret = iterator.next()
    except StopIteration:
        return None
    
    for elem in iterator:
        ret += elem

    #Return result.
    return ret





From jonathan.gilligan at vanderbilt.edu  Tue Jun 25 13:01:21 2002
From: jonathan.gilligan at vanderbilt.edu (Jonathan Gilligan)
Date: Tue, 25 Jun 2002 12:01:21 -0500
Subject: Active Python, ASP, and security
Message-ID: <afa8ac$sqi$1@news.vanderbilt.edu>

I have a problem with ActivePython scripting on ASP pages. If I set IIS to
use the low-security (in-process) settining, the demo files (e.g.,
Python21\win32comext\axscript\demos\client\asp\tut1.asp) work fine. If I set
IIS to use Medium (Pooled) security, I get a 500 error message, and the
following Event in the event viewer:

   The COM Application '{3D14228D-FBE1-11d0-995D-00C04FD919C1}' at
   '/LM/W3SVC/1/Root/TestPyAsp' failed to activate out of process.
   For additional information specific to this message please visit
   the Microsoft Online Support site located at:
   http://www.microsoft.com/contentredirect.asp.

I am running Windows 2000 Pro, SP2 with IIS 5.0, and ActiveState Python
2.1.2 build 2.1.212 (PythonWin 2.1.1 (#20, Jul 26 2001, 11:38:51)).

Can anyone help me understand how to fix the problem with OOP activation of
the Python Scripting engine?




From owen at nojunk.invalid  Sun Jun  2 00:57:26 2002
From: owen at nojunk.invalid (Russell E. Owen)
Date: Sat, 01 Jun 2002 21:57:26 -0700
Subject: Help for Tkinter and python's programming
References: <mailman.1022985123.22656.python-list@python.org>
Message-ID: <adc8kj$1ovk$1@nntp6.u.washington.edu>

In article <mailman.1022985123.22656.python-list at python.org>,
 aurel-le-gnou <aurelienviolet at noos.fr> wrote:

>...
>button_go.bind("<Button-1>", make_call(frame_left, frame_right,
>list_one, list_two, button_number, button_search, button_language,
>button_rank, button_engine))  
># the arguments of make_call() are visible and defined in the main when
>I do this "bind".

This is a common problem with callbacks in Python. When you supply 
make_call(frame_left...) as a callback argument, you are actually 
executing the function at that very moment and passing the result of its 
execution to button_go.bind. In other words, you aren't passing a 
function, you are passing a result of calling a function. Thus nothing 
at all happens later when you press the button.

You must pass in the name of a function WITHOUT the parenthesis, e.g. 
make_call, not make_call(frame_left...). But there is a hitch. 
button_go.bind requires a callback function that takes exactly one 
argument: an event. Yet you want to pass in a lot of extra information. 
What you want is some way of saying "here is a function with some stuff 
I know now and some other stuff I am going to specify later".

The thing I usually do in this situation is to create a callable object. 
This is handy because an object can hold state, i.e. the "stuff I know 
now", yet can also have a __call__ method which allows you to call the 
object just like a function.

This comes up often enough that I have a class GenericCallback which 
makes it easy (see code at the end, with thanks to Eric Brunel). Here is 
an example, a simplified version of your code:

Suppose I want to call a function from a button press event, and that 
function has two pieces of info I know in advance: frame_left and 
frame_right. A callback function bound to a button press event must take 
exactly one argument: the event (which in your case you don't care 
about, but you still have to accept it!).

First define a function that takes all arguments, starting with the ones 
you know in advance (frame_left and frame_right) and ending with the 
ones that will be sent to the callback (evt):

def myFunc(frame_left, frame_right, evt):
    ...your code here...

Now create a GenericCallback object that can be called by bind. It must 
take just one argument (evt), so you specify all the other arguments to 
GenericCallback. This code is complete:

myCallback = GenericCallback(myFunc, frame_left, frame_right)
button_go.bind("<Button-1>", myCallback)

That's all there is to it! I've appended my GenericCallback.py. I save 
it as a file and import it when I need it, and I use it a lot. There are 
other approaches. Perhaps if I understood Python's scoping rules better 
I could use non-local variables in the function, but I find the rules 
non-obvious (whether the variable is local or global depends critically 
on how it first appears in the function--on the right or left of an 
assignment) and prefer to make things more explicit. Some folks use 
lambdas, but I personally find them ugly.

-- Russell

Here is GenericCallback.py:

"""From Scott David Daniels 
<http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52549>.
My changes include:
- Used a name suggested by Eric Brunel
- Created a simpler and faster class that does not handle initial 
keyword arguments
- Created a factory function to return the fastest callable object.
"""
def GenericCallback(callback, *firstArgs, **firstKWArgs):
    """Returns a callable object that when called with (*args, **kwArgs),
    calls a callback function with arguments: (*(firstArgs + args), 
**allKWArgs),
    where allKWArgs = firstKWArgs updated with kwArgs
    
    Note that if a given keyword argument is specified both at 
instantiation
    (in firstKWArgs) and when the object is called (in kwArgs),
    the value in the call (kwArgs) is used. No warning is issued.
    """
    if firstKWArgs:
        return GC(callback, *firstArgs, **firstKWArgs)
    else:
        return GCNoKWArgs(callback, *firstArgs)


class GC:
    """A generic callback class."""
    def __init__(self, callback, *firstArgs, **firstKWArgs):
        self.__callback = callback
        self.__firstArgs = firstArgs
        self.__firstKWArgs = firstKWArgs

    def __call__(self, *lastArgs, **kwArgs):
        if kwArgs:
            netKWArgs = self.__firstKWArgs.copy()
            netKWArgs.update(self.__kwArgs)
        else:
            netKWArgs = self.__firstKWArgs 
        return self.__callback (*(self.__firstArgs + lastArgs), 
**netKWArgs)


class GCNoKWArgs:
    """A generic callback class optimized for the case
    of no stored keyword arguments. This has the potential
    to significantly speed up execution of the callback.
    """
    def __init__(self, callback, *firstArgs):
        self.__callback = callback
        self.__firstArgs = firstArgs

    def __call__(self, *args, **kwArgs):
        return self.__callback (*(self.__firstArgs + args), **kwArgs)


From denpeterson at yahoo.com  Wed Jun 19 20:48:47 2002
From: denpeterson at yahoo.com (Dennis Peterson)
Date: Wed, 19 Jun 2002 20:48:47 -0400
Subject: My Python/MySQL code is too slow
References: <aer2s4$mp7$1@news7.svr.pol.co.uk>
Message-ID: <aer8qm024re@enews2.newsguy.com>

I'm new to python and mysql specifically, but...your join's going to be slow
if you don't have an index on var1,var2,var3 in both tables. But one of the
tables is a temporary, no indexes, so if it's big, you have a problem.

I don't know if you can apply indexes to a temporary table in mysql - if
not, you could use a regular table, just build it like your temp, add an
index, run your join, that'll probably be faster. If this would be a
concurrency problem, add a userid to the key for your "temporary."

Also, I don't know for sure that mysql optimizes this way, but in some
databases at least it's best if you put your most selective column first in
the index. If there are 5 distinct var1 values, and 1000 distinct var3, make
your index var3,var2,var1. Probably you want it in the same order, both
tables.

There might be a better way, though. Same index hints apply, but if dbB has
a unique index on var1,var2,var3, you could just do this:

SELECT dbB.var1, dbB.var2, dbB.var3 FROM dbB inner join dbA
on dbB.var1=dbA.var1 AND dbB.var2=dbA.var2 AND dbB.var3=dbA.var3
group by dbB.var1,dbB.var2,dbB.var3 having count(*)=2

Hope this helps,
Dennis

"Duncan Smith" <buzzard at urubu.freeserve.co.uk> wrote in message
news:aer2s4$mp7$1 at news7.svr.pol.co.uk...
> I have already posted this problem on mailing.database.mysql as I see it
as
> a problem with my lack of knowledge of SQL.  But no responses so far, and
> maybe Python could do more of the work.  I repost the original question
> below (SQL easier to follow), and the Python code below that.  Any help
> appreciated.  TIA.
>
> ---------------------------------------------------------------------
> I have two tables (say dbA and dbB) and need to find the number of rows in
> dbB which have exactly 2 matching rows in dbA (on the selected columns).
> I'm using Python/MySQL, but the relevant SQL is below.  This is slow, and
> I'm sure it's possible without creating the temporary table and without
the
> join.  But everything I try fails (because of my limited knowledge of
SQL).
> Can anyone show me an efficient way of getting the answer?  All I need is
> the number of rows in dbB which have exactly 2 matching rows in dbA.
Thanks
> in advance.
>
> CREATE TEMPORARY TABLE tmp SELECT var1, var2, var3 FROM dbA GROUP BY var1,
> var2, var3 HAVING COUNT(*)=2;
>
> SELECT dbB.var1, dbB.var2, dbB.var3, tmp.var1, tmp.var2, tmp.var3 FROM
dbB,
> tmp WHERE dbB.var1=tmp.var1 AND dbB.var2=tmp.var2 AND dbB.var3=tmp.var3;
>
> Cheers.
>
> Duncan Smith
>
> ------------------------------------------------------------------
> import sys
> import string
> import MySQLdb
>
> def MyFunc(dbname, tblname, pert_table, variables=()):
>
>     """'dbname' is a MySQL database containing tables
>     'tblname' and 'pert_table'.  'variables' is a
>     sequence listing the relevant columns"""
>
>     variables = list(variables)
>
>     try:
>         conn = MySQLdb.connect()
>         curs = conn.cursor()
>         curs.execute('USE %s' % (dbname,))
>         if not variables:
>             curs.execute('DESCRIBE %s' % (tblname,))
>             variables = [x[0] for x in list(curs.fetchall())]
>         vars = string.join(variables, ', ')
>
>         #produce a temporary table containing all records
>         #of tblname which occur in pairs and find P
>
>         curs.execute("""CREATE TEMPORARY TABLE tmp
>                     SELECT %s FROM %s GROUP BY %s
>                     HAVING COUNT(*)=2""" % (vars, tblname, vars))
>
>         def f(s): return string.join([s + var for var in variables], ', ')
>
>         P = curs.execute("""SELECT %s, %s FROM %s, tmp WHERE %s""" %
> (f(pert_table+'.'), f('tmp.'), pert_table, string.join([pert_table + '.' +
> variables[i] + '=' 'tmp.' + variables[i] for i in range(len(variables))],
'
> AND ')))
>
>         curs.execute('DROP TABLE tmp')
>
>         curs.close()
>         conn.close()
>
>         return P/2
>
>     except MySQLdb.Error, e:
>         print "Error %d: %s" % (e.args[0], e.args[1])
>         sys.exit(1)
>
>




From gnufnork at fuckspam.hetdigitalegat.nl  Mon Jun 10 22:56:51 2002
From: gnufnork at fuckspam.hetdigitalegat.nl (Wladimir)
Date: Tue, 11 Jun 2002 04:56:51 +0200
Subject: urllib question
References: <644f6688.0206101835.5dc793b1@posting.google.com>
Message-ID: <pan.2002.06.11.02.56.51.203503.25803@fuckspam.hetdigitalegat.nl>

On Tue, 11 Jun 2002 04:35:07 +0200, rdack wrote:

> this works but i don't understand it. it retrieves data from my router.
>   params = urllib.urlencode({'RC': '@D', 'ACCT' : "root", 'PSWD' :
> "71:29:26", 'URL': 'admin' })
>   ipurl = "http://" + iphost + "/cgi-bin/logi"
<   urlfp = urllib.urlopen(ipurl, params)
> 
> i was wondering exactly what those params mean. i have some other code
> that wants a user name and password and i don't know what to give it. is
> the username = 'root'? and what is the password? a string with 3 two
> digit numbers separated by colons?

So it passes four parameters to the script 

[name]		[value]
RC		@D
ACCT		root
PSWD		71:29:26
URL		admin

What those parameters exactly mean is not a python issue but propetiary to
the router in question. (see router docs for that)

> what is the 'URL' param mean?
> somewhere to find documentation?
 
http://www.python.org/doc/current/lib/module-urllib.html would be a good
place to start :D

Wladimir


From dom at edgereport.put_a_c_o_m_here  Fri Jun  7 16:39:59 2002
From: dom at edgereport.put_a_c_o_m_here (Domenic R. Merenda)
Date: Fri, 07 Jun 2002 20:39:59 GMT
Subject: Medium to Large Scale Python Deployments
References: <mailman.1023465334.24936.python-list@python.org>
Message-ID: <zS8M8.21881$EM.2548365331@newssvr10.news.prodigy.com>

"Simon Brunning" <SBrunning at trisystems.co.uk> wrote in message
news:mailman.1023465334.24936.python-list at python.org...
> Wacky? Hmm, well, it's pretty different to anything else that I've come
> across. Coming from an OS/400 background, Windows and Unix look like
> slightly different flavours of the same thing to me. (What do you mean a
> program is a file? In OS/400, a file is a file and a program is a program,
> dammit, not the same thing at all.)

   Yes, it is different, to say the least.  I enjoy the mixed environment,
though.  When I get sick of staring at a dumb terminal, I can hop over to a
PC and make a connection, turning my PC into a revolutionary, new ... dumb
terminal. :-)

> RPG II? Now that *is* wacky! No structured op-codes at all - you had to do
> everything with IFs and GOTOs. Yuck. I hope that anything new is done in
> RPGIII at least! (We use RPGIV exclusively for new stuff - the free
> formatting rocks. Well, compared to fixed formatting, that is - I still
hate
> semi-colons and braces everywhere.)

   Ok, you want a laugh?  The reason for the RPG II is twofold (no pun).
First, when the system was first begun, RPG II was a newfangled
thingamabobber, and the programmers really wanted to use it.  We have
hand-writted documentation like you wouldn't believe.  Second, we have a
support contract for RPG II which runs out in July of 2004, so management is
convinced it's the safest way for us to go, in case it somehow ... breaks?


--
Domenic R. Merenda
Editor, The Edge Report
http://www.edgereport.com




From boud at valdyas.org  Mon Jun 17 09:13:39 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Mon, 17 Jun 2002 15:13:39 +0200
Subject: pychecker + XEmacs
Message-ID: <3d0de483$0$3881$e4fe514c@dreader4.news.xs4all.nl>

Has anyone succeeded in combining pychecker and XEmacs (or emacs)? I'm 
checking a large project I began in 1999, and which is full of the most 
atrocious errors, and I'd like to click on the line that mention the error 
and be transported there, just like when you run a python program from 
XEmacs and get a stacktrace.

-- 
Boudewijn Rempt | http://www.valdyas.org


From martin at v.loewis.de  Sat Jun 29 15:00:24 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 29 Jun 2002 21:00:24 +0200
Subject: incorrect upper()/lower() of UTF-8
References: <afiecp$oss$1@web1.cup.hp.com> <3phT8.45577$n4.10823047@newsc.telia.net>
Message-ID: <m3ofdt9agn.fsf@mira.informatik.hu-berlin.de>

"Fredrik Lundh" <fredrik at pythonware.com> writes:

> It's not supposed to: the 8-bit string type can hold either
> text strings or binary buffers.  The encode method takes
> a text string and returns a binary buffer.
> 
> For reliable processing of Unicode data, use Unicode strings.

It might be worth noting that the .upper and .lower methods of the
Unicode string are not locale-aware. Instead, they work on all
characters independent of the locale (*). 

Whether this is a good or bad thing probably depends on the
application.

Regards,
Martin

(*) Actually, if Python uses wchar_t for the Unicode type, it also
uses the C library for .upper/.lower; I would consider that a bug.


From chris at blaue-elise.net  Sun Jun 23 07:13:52 2002
From: chris at blaue-elise.net (Christian Guenther)
Date: 23 Jun 2002 13:13:52 +0200
Subject: bloody freshman question regarding strings
In-Reply-To: <1024821351.3424.12.camel@dorian.blaue-elise.net>
References: <1024821351.3424.12.camel@dorian.blaue-elise.net>
Message-ID: <1024830832.3424.21.camel@dorian.blaue-elise.net>

Wow, that was quick. Many thanks for the responses. I suppose 
I have to learn a lot about pythons sysntax and way it handles 
objects and stuff.

I always had bit of a problem understanding OO-concepts.

	chris

Am Son, 2002-06-23 um 10.35 schrieb Christian Guenther:
> Hi folks,
> 
> I'm absolutely new to python - also I'm not a beginner in programming.
> I bought myself a book called "Einf?hrung in python" - should be
> "introduction to python" in the original - by Markus Lutz and David
> Asher published by O'Reilly. This book is good, nevertheless I'm a bit
> in a hurry because I have to write a script that is supposed to populate
> an LDAP-directory for user-authentication. 
> By the way: Is there a module for python -> ldap? 
> 
> My biggest question at the moment is why can't I do something like this:
> 
> def mailmsg(severity message opts target recipient):
> 	# first create the message body
> 	msgbody="computer: " + COMPUTER + "\n
> 	         system:   " + SYSTEMTYPE + "\n
> 		 cpu:      " + CPUTYPE + "\n
> 		 kernel:   " + KERNELVER + "\n
> 		 date:     " + DATE + "\n
> 		 \n
> 		 severity: " + severity + "\n
> 		 message:  " + message + "\n\n"
> 
> 
> I want to generate messages - to concole and to mail-recipients - that
> show the status of user-processing. This message should include all the
> information you see above - they are stored in global vars - a severity
> and of course a message itself. Let alone, it just doesn't work and I
> know I do this in a nodo way, I just don't have alot of time left ;-)
> 
> Could someone point me to the right direction????
> 
> 	chris
> 
> 
> -- 
>  __    __  __     __  __ ___    ___                      
> |  |  |  ||  \   |  ||  |\  \  /  /              chris Guenther
> |  |  |  ||   \  |  ||  | \  \/  /               chris at blaue-elise.net
> |  |  |  ||  . \ |  ||  |  >    <                Wuppertal / Germany
> |  |__|  ||  |\ \|  ||  | /  /\  \  
>  \______/ |__| \____||__|/__/  \__\ 
> 
> UNIX _IS_ user friendly, it's just selective about who its friends are
>   
> ----------------------------------------------------------------------
>   UNIX was not designed to stop you from doing stupid things, 
>   because that would also stop you from doing clever things.
>                                                          ...Doug Gwyn
> ----------------------------------------------------------------------
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 
-- 
 __    __  __     __  __ ___    ___                      
|  |  |  ||  \   |  ||  |\  \  /  /              chris Guenther
|  |  |  ||   \  |  ||  | \  \/  /               chris at blaue-elise.net
|  |  |  ||  . \ |  ||  |  >    <                Wuppertal / Germany
|  |__|  ||  |\ \|  ||  | /  /\  \  
 \______/ |__| \____||__|/__/  \__\ 

UNIX _IS_ user friendly, it's just selective about who its friends are
  
----------------------------------------------------------------------
  UNIX was not designed to stop you from doing stupid things, 
  because that would also stop you from doing clever things.
                                                         ...Doug Gwyn
----------------------------------------------------------------------





From gorny0 at zonnet.nl  Sun Jun 16 05:56:12 2002
From: gorny0 at zonnet.nl (Gorny)
Date: Sun, 16 Jun 2002 11:56:12 +0200
Subject: Args not set in extension class
References: <aefcav$8ef$1@reader05.wxs.nl> <aefs7n$4cc$1@reader09.wxs.nl> <83sn3otlkf.fsf@panacea.canonical.org>
Message-ID: <aehngd$22l$1@reader06.wxs.nl>

"Kragen Sitaker" <kragen at pobox.com> schreef in bericht
news:83sn3otlkf.fsf at panacea.canonical.org...
> "Gorny" <gorny0 at zonnet.nl> writes:
> > I've already dealt with the problem... Just create a function like:
> >
> > static PyObject *
> > wrap_pkt_object(PyObject *self, PyObject *args)
> > {
> >         PyObject * return_object;
> >         return_object = new_pkt_object(args);
> >         return return_object;
> > }
> >
> > But it would be very nice if someone can explain to me, why this
> > is happening and it isn't possible to *directly* call new_pkt_object
even
> > if you prototype the latter function with two arguments like
> > wrap_pkt_object!
>
> If you prototype new_pkt_object with two args, make sure you're using
> the second one.

That doesn't seem to work as well for some vague reason... It complains
it's not a valid tuple

Gorny

--
"Spread the sources"
http://gorny.cjb.net




From matthew.russell at securetrading.com  Mon Jun 10 09:26:36 2002
From: matthew.russell at securetrading.com (Matt Russell)
Date: 10 Jun 2002 06:26:36 -0700
Subject: Compiling python2.2.1 on redHat 6.1 - doesn't work
References: <bb06817.0205210533.d8d5ffb@posting.google.com> <23891c90.0205220343.4aa581c8@posting.google.com> <bb06817.0205230212.7b6a388c@posting.google.com> <23891c90.0205230743.7ea59255@posting.google.com> <23891c90.0206030901.26402dd0@posting.google.com>
Message-ID: <bb06817.0206100526.55f36028@posting.google.com>

Thanks Paul, 
am afraid being a "c" newbie, I have no idea on how to apply the
patch, would be great if you could provide an example.

thanks for all your help.

Matt
paul at boddie.net (Paul Boddie) wrote in message news:<23891c90.0206030901.26402dd0 at posting.google.com>...
> paul at boddie.net (Paul Boddie) wrote in message news:<23891c90.0205230743.7ea59255 at posting.google.com>...
> >
> > matthew.russell at securetrading.com (Matt Russell) wrote in message news:<bb06817.0205230212.7b6a388c at posting.google.com>...
> > >
> > > has anyone else had this problem? I would be very suprised if no one
> > > else out there has not tried to compile Python2.2.1 on RedHat 6.1!!
> > 
> > I'll try and have a look in the next few days, at least.
> 
> OK. My suggestion can be found at the following location:
> 
> http://sourceforge.net/tracker/?group_id=5470&atid=305470&func=detail&aid=563954
> 
> That's patch #563954 in the Python project at SourceForge
> (http://sf.net/projects/python).
> 
> Paul


From jadestar at idiom.com  Tue Jun 18 04:02:46 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 18 Jun 2002 08:02:46 GMT
Subject: Is there a Text to HTML conversion module?
References: <mailman.1023132485.29235.python-list@python.org> <adgn01$d18$1@slb5.atl.mindspring.net> <3CFBDF95.92099C9E@alcyone.com>
Message-ID: <aempf6$1nq2$1@news.idiom.com>

Erik Max Francis <max at alcyone.com> wrote:
> Andrew Dalke wrote:

>> <wink/>

>>>(sorry... couldn't resist)
>> Neither could I.  :)
> Ah, but that's XML, not HTML.  :-)

 Of course it might be some future extension to XHTML 



From karczma at info.unicaen.fr  Fri Jun 21 05:15:22 2002
From: karczma at info.unicaen.fr (Jerzy Karczmarczuk)
Date: Fri, 21 Jun 2002 11:15:22 +0200
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com> <lkznxqjqx2.fsf@pc150.maths.bris.ac.uk> <3D11A21D.9010809@mxm.dk> <uh3e1usajle42c@news.supernews.com> <3D11DA9B.163A7BD0@info.unicaen.fr> <uh446elev7af68@news.supernews.com>
Message-ID: <3D12EEAA.9C3A3D29@info.unicaen.fr>

John Roth comments my observation concerning the relation between gotos
and continuations:

> >    ... in order to speak about genuine gotos *you don't need to think
> >    assembler*.
> >    What you need is to understand correctly the notion of continuation
> >    (something quite standard - again - in the Functional Programming
> >    world). This is a high-level model of "flat" branching, but may be
> >    used to more exquisite constructions, such as backtracking.
> 
> That's still not a goto in the classical sense. In the classical sense,
> a goto allows you to go to any statement. The point you're missing is
> that the goto statement is a denial of structure - since it can go
> anywhere, you can't make any assumptions about the program
> structure. The piece of code you're looking at can be mysteriously
> activated from somewhere else at any time. 

I am not sure I am "missing a point". Perhaps I have a different vision
of the computational process, hm? What is the "classical sense" for one
may be just an orthodox, restricted view for somebody from another galaxy.
I tend to apply often a Functionalist, sometimes too formal view on the
semantics of computations, you don't need to accept it, but I assure you
that for many people gotos *ARE* continuations. Continuation in a broad
sense is just the *specification* of what your <<future>> is.

Gotos and continuations may be as structured or as unstructured as the
programmers wishes, and the language permits. If all labels are static (no
jump address computed out of scratch), this is not true that a piece of code
can be "mysteriously activated" from "somewhere else", at "any" time. If,
in a disciplined and austere Scheme you permit yourself to plug into an 
*expression* (not even a "statement") the call-with-current-continuation
device, you may produce a bedlam which nobody understands, including the
author. This doesn't prove anything, but it suggests that your "classical
gotos" from the operational point of view are not restricted to classical
imperative code sequencing.

(Oh, I adored that wonderful piece of programming madness, an "article"
where somebody proposed a new super-Fortran where GO TO was replaced by
CAME FROM statement, and the language had an inspiring non-deterministic
conditional:   If <<condition>> THEN WHAT?.
Any references, anybody?)

> Continuations, on the other hand, have been discussed, and may
> very well wind up in Python some day. See the discussion around
> a variant called 'stackless.'

Yeah, I know it. I follow the Mission Impossible Software Team (Tismer
et al.) for some time. I am interested in scientific computing, and I 
believe that generators, coroutines etc. are essential for the simu-
lation. So, the idea to use continuations was for me always very
appealing (even if sometimes appalling...). Still, the notion of a
full continuation, as implicit in the letter of GvR
http://www.stackless.com/continuations.guido.van.rossum.html,
the machinery which makes a snapshot of the actual state, and which
is implemented in Scheme through call/cc, is not necessarily the
very essence of the concept itself, which is simpler. My turn to send
you to some references. I suggest the Andrew Appel's book 'Compiling
with Continuations'.


Thank you for this interesting exchange.

Jerzy Karczmarczuk
Caen, France


From candiazoo at attbi.com  Fri Jun 21 21:43:43 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Sat, 22 Jun 2002 01:43:43 GMT
Subject: Help with MySQLdb insert with data extracted from odbc <-- Oracle
Message-ID: <3d13d5fb.651829078@netnews.attbi.com>

Can someone tell me what this means?  I mean, I guess I *know* what it means,
but why?

The insert statement into a mysql database is...

'insert into works values(?, ?, ?, ?, ? ,?)'

The invocation of the dynamic sql is...

    status = cur_mysql.execute(sql, 
        (wrk_inst, ZERO, trs_inst, aas_inst, title, NADA))

(ZERO = 0L and NADA = 'N/A')

I do a print of the wrk_inst, trs_inst, aas_inst (and zero and nada) followed by
a print of title before the error...

80956 177802 0 0 N/A
WANTED ... A BROTHER

...here is the error...

Exception in Tkinter callback
Traceback (most recent call last):
  File "C:\Python22\lib\lib-tk\Tkinter.py", line 1292, in __call__
    return apply(self.func, args)
  File "C:\Documents and Settings\mjessop\My Documents\Python Development\Catalo
g\catalog_gui.py", line 82, in engage
    catalog.extract_works()
  File "C:\Documents and Settings\mjessop\My Documents\Python Development\Catalo
g\catalog_model.py", line 249, in extract_works
    (wrk_inst, ZERO, trs_inst, aas_inst, title, NADA))
  File "C:\Python22\Lib\site-packages\MySQLdb\cursors.py", line 70, in execute
    raise ProgrammingError, m.args[0]
ProgrammingError: not all arguments converted

======
here is a description of the database table...

mysql> describe works
    -> ;
+----------+--------------+------+-----+---------+-------+
| Field    | Type         | Null | Key | Default | Extra |
+----------+--------------+------+-----+---------+-------+
| wrk_inst | int(11)      |      | MUL | 0       |       |
| sinker   | int(11)      |      | MUL | 0       |       |
| trs_inst | int(11)      | YES  |     | 0       |       |
| aas_inst | int(11)      | YES  |     | 0       |       |
| title    | varchar(255) | YES  |     | NULL    |       |
| pub_name | varchar(80)  | YES  |     | NULL    |       |
+----------+--------------+------+-----+---------+-------+
6 rows in set (0.03 sec)

Any thoughts?

Thank you, in advance...

+++-------------+++--------------+++
+ Michael S. Jessop                +
| Senior Web Developer             |
+ Copyright Clearance Center, Inc. +
+++-------------+++--------------+++



From johnroth at ameritech.net  Thu Jun 20 09:39:29 2002
From: johnroth at ameritech.net (John Roth)
Date: Thu, 20 Jun 2002 09:39:29 -0400
Subject: CR+LF problem
References: <3d11bc33.1129927479@news.mad.ttd.net>
Message-ID: <uh3mmkc7cugr82@news.supernews.com>

"Manuel Ruiz" <mruiz at safa.es> wrote in message
news:3d11bc33.1129927479 at news.mad.ttd.net...
>
> Hi,
>
> ------------------------
> .
> char * txt = "print 123\r\n";
> result = PyRun_Simple( txt, Py_file_input, PyDict_New(), NULL );
> .
> -----------------------
>
> this code fails to exec txt, I got a SyntaxError exception,
> however I if save to file the statement "print 123\r\n" and try to run
> it ussing python interprete (version 2.2.1) It run fine.
>
> Can anyone tell me anything about this rare behaviour
>
> Regards
> Manuel Ruiz

 \r is not a valid character for Python input.

I suspect you're running on a Windows system. When
you saved this to file, the C library would have converted
the \r\n sequence to \n on the readback.

To insert a return, just use \n, regardless of the
system. Python (and C) takes care of the rest, unless
you really need the \r because you're using a network
rather than the file.

Are you sure you didn't mean "print '123'" ?

John Roth




From Adndrew at novaresponse.com.au  Thu Jun  6 02:47:12 2002
From: Adndrew at novaresponse.com.au (Adndrew at novaresponse.com.au)
Date: Thursday, 06 Jun 2002 00:47:12 -0600
Subject: Urgent News Flash
Message-ID: <06060200.4712@novaresponse.com.au>

URGENT NEWSFLASH

The government has installed BLACK BOXES into ISPs to watch your
Internet surfing.

When you are targeted, the first you will know is the KNOCK AT YOUR
DOOR and the JAIL SENTENCE.

If you think you have cleaned your computer - you are WRONG.

They can recover from your PC every single picture or item you have
ever watched on the Internet.

You need protection. You need an Evidence Eliminator. 

Click Here.

http://www.geocities.com/testee234/




From Chris.Barker at noaa.gov  Wed Jun 19 18:15:36 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Wed, 19 Jun 2002 15:15:36 -0700
Subject: How to Store Binary data in a *.py module
References: <3D10E0E9.906DF517@noaa.gov> <3d10e411$0$31829$afc38c87@news.optusnet.com.au>
Message-ID: <3D110285.50F18058@noaa.gov>

Steven Adams wrote:
> I've stored GIFs using base64 encoding via the base64 module and it seemed
> to work fine.

Thanks, that's just what I was looking for.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From pyth at devel.trillke.net  Tue Jun 18 06:35:07 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 18 Jun 2002 12:35:07 +0200
Subject: python version?
In-Reply-To: <yXBP8.44417$GY.13865423@twister.nyroc.rr.com>; from hesterloli@hotmail.com on Tue, Jun 18, 2002 at 08:12:14AM +0000
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <yXBP8.44417$GY.13865423@twister.nyroc.rr.com>
Message-ID: <20020618123507.Y15079@prim.han.de>

George Hester wrote:
> Listen you aren't my daddy.  If you do not want to help constructively then
> please get off my boat.

that's very harsh. Peter was helping very constructively. He basically said
don't drive if you don't know what brakes, tires and a wheel are for.
And he showed you how to get into the car.

*Everybody* of us has to go through learning sessions, getting
accustomed to new stuff and finally (hopefully) understanding it. 

again-and-again-ly y'rs holger


> --
> George Hester
> _________________________________
> "Peter Hansen" <peter at engcorp.com> wrote in message
> news:3D0E8AC8.6C5C96EE at engcorp.com...
> > George Hester wrote:
> > >
> > > "Gerhard H?ring" <gerhard at bigfoot.de> wrote:
> > > > You did not import the sys module.
> >
> > > Probably not.  Where do I import that from and how do it?  I am trying to
> put a
> > > script in ASP using Python as the script language to give me the version of
> > > Python that I am running when I call the ASP.  The more the script works out
> of
> > > the box the better I'll be able to catch on to this language.  Thanks.
> >
> > It sounds to me like you should probably crawl before you walk.
> > I suggest downloading the tutorial and installing Python on
> > your own machine and using it interactively from the prompt
> > before you advance to something like ASP scripting.  Your
> > responses clearly show you aren't ready to write code in an
> > environment which is doubtless going to mask from you many
> > of the errors and messages which you will encounter as you learn.
> >
> > In the above case you would be able to experiment and learn about
> > these things:
> >
> > C:\>python
> > Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
> > >>> sys.version
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> > NameError: name 'sys' is not defined
> >
> > Here's where you would pause to learn a little more about Python....
> >
> > >>> import sys
> > >>> sys.version
> > '2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)]'
> >
> > -Peter
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From amckay at merlintechnologies.com  Tue Jun 11 19:27:09 2002
From: amckay at merlintechnologies.com (Andy McKay)
Date: Tue, 11 Jun 2002 16:27:09 -0700
Subject: html parser etc help
In-Reply-To: <7fe97cc4.0206111445.43b196de@posting.google.com>
References: <7fe97cc4.0206111445.43b196de@posting.google.com>
Message-ID: <20020611232709.GA15300@mckay.merlintechnologies.com>

On 11/06/02 15:45 -0700, Xah Lee wrote:
> i'm learning python in hoping to do switch web programing from perl.
> can anyone show me code snipets for the following problem to get
> me started?
> 
> * snipet to traverse a directory and list files by typer or suffix.

have a look at glob or listdir

import glob
glob.glob('*.py')

> * sample snipet that parse a given html file and print out all <a href> links.

a snippet that will read <link> tags
http://www.agmweb.ca/Scripts/rssAutoDiscovery

> * a sample snipet that fetch and print out the content of a given url.

as above

> --
> # fetch a given webpage
> use LWP::Simple;
> $content = get('http://xahlee.org/');
> print $content;

hey its even shorter in python :)

from urllib import urlopen
print urlopen('http://xahlee.org').read()

-- 
  Andy McKay
  Merlin Technologies




From candiazoo at attbi.com  Thu Jun 27 14:59:44 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Thu, 27 Jun 2002 18:59:44 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com>
Message-ID: <3d1b6089.311756812@netnews.attbi.com>

Is there a way for me to derive more data from Profile?  Can I place some sort
of stops or counters between sections of code?

Mike


From sholden at holdenweb.com  Fri Jun 28 09:18:31 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 28 Jun 2002 09:18:31 -0400
Subject: Interactive session via tcp
References: <20020627120602.3659.37119.Mailman@mail.python.org> <mailman.1025184175.27162.python-list@python.org>
Message-ID: <nlZS8.106926$mF.68005@atlpnn01.usenetserver.com>

"Andreas Ames" <andreas.ames at Tenovis.com> wrote in message
news:mailman.1025184175.27162.python-list at python.org...
> Hi,
>
> > Sorry, if this if wrong, I read your post quickly.
> > Doesn't "telnet jabber_port" do what you want?
>
> Nope, it doesn't.  My main problem is the python (interpreter) side.
> And while it's possible to jabber via telnet it's not exactly what I
> call fun (I've tried it).
>

Take a look at Sam Rushing's Medusa code. He implements a server-control
connection that allows you to telnet in to the server environment and
basically exposes the whole thing via Python. To the extent that if your
server crashes you can replace the crashed instances while still retaining
the old server for debugging.

    www.nightmare.com/medusa/

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From emile at fenx.com  Thu Jun  6 16:22:37 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 06 Jun 2002 20:22:37 GMT
Subject: numeric limits?
References: <adofuv$6v$1@bob.news.rcn.net>
Message-ID: <gwPL8.124971$cQ3.3841@sccrnsc01>

"David Abrahams" <david.abrahams at rcn.com> wrote in message
news:adofuv$6v$1 at bob.news.rcn.net...
> Is there some straightforward way to find out the range of a Python
Int (in
> code)?
> The best I could do was to probe it using tricks like:
>
> x = 1
> while (x << 1) > 0:
>     x <<= 1
>

Do you mean sys.maxint ?


--

Emile van Sebille
emile at fenx.com

---------



From achim.domma at syynx.de  Fri Jun 14 04:26:15 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Fri, 14 Jun 2002 10:26:15 +0200
Subject: Python interface to ImageMagick
Message-ID: <aec8qd$19p$05$1@news.t-online.com>

Hi,

I just uploaded a first alpha version of my python wrapper for ImageMagick.
Currently there is only a windows version, which is tested on Win2K with
ActiveState Python 2.2.1. Other version, more documentation and examples
will follow hopefully soon. Every feedback or bugreport is wellcome, because
I don't have different testing environments.

If you want to try it, you can download it here:
http://www.procoders.net/imagemagick/ImageMagick-0.1a.win32.exe

An working ImageMagick (www.imagemagick.org) Installation is required. The
wrapper is linked with ImageMagick 5.4.5-1.

greetings
Achim




From aurelienviolet at noos.fr  Sat Jun  1 22:51:11 2002
From: aurelienviolet at noos.fr (aurel-le-gnou)
Date: Sun, 02 Jun 2002 04:51:11 +0200
Subject: HTML links in a GUI tkinter
Message-ID: <3CF9881F.FC8D4116@noos.fr>

Hi,


If you can also help me to realize this, it would be really fine:


How can I put in a frame or in any widget (?) an html link so that the
browser opens at the url indicated when the user clicks ?
Just simply as if it were an HTML link in an HTML document but in a
tkinter GUI ?

What are the packetages to import ?
In what king of widget can I put such a link ?
What kind of lines of code should i put in this widget the link to be
active ?


Thanks a lot for your help


Aur?lien Violet
aurelienviolet at noos.fr




From ngiven at lanl.gov  Thu Jun 20 11:16:02 2002
From: ngiven at lanl.gov (Nathan Given)
Date: Thu, 20 Jun 2002 09:16:02 -0600
Subject: logical statements (in Python and Numeric)
Message-ID: <3D11F1B2.21587F31@lanl.gov>

I am graphing piecewise functions, and I have a method of getting the
proper y values in a vector that works in MATLAB... I was trying to get
it to work in Python, but it wasn't...

it has to do with a using an and statement...


here is my code...

x = arrayrange(-4.,2.+.001,.001, Float)
y = zeros(len(x),Float)

#------ Create Piecewise Function ------#
y = y + (x < -2)*(((x + 2.)**2) + 4)
y = y + (x >= -2. and x < 0.) * (x + 6.)
y = y + (x >=0) * (x ** 2. + 6.)

Now, the problem is with the second piece (where x is BETWEEN -2 and
0)...the y statement should only affect x values less then -2, the
second y statement should only affect x values between -2 and 0, but it
doesn't, it affects EVERYTHING less then zero and greater than negative
2.

How do I make this statement work properly?

thanks
--
Nathan

PS.  Here is the matlab code...

x = -4:.001:2;
y = zeros(1,length(x));

y = y + (x < -2) .* ((x+2) .^ 2 + 4);
y = y + (x >=-2 & x < 0) .* (x + 6);
y = y + (x >= 0) .* (x .^ 2 + 6);


From tim.one at comcast.net  Fri Jun 21 19:42:54 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 21 Jun 2002 19:42:54 -0400
Subject: RE strings (was: Variable Interpolation - status of PEP 215)
In-Reply-To: <29e28c51.0206210944.400eaedc@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEDPPPAA.tim.one@comcast.net>

[Cimarron Taylor]
> ...
>    import re
>    emp_re = re.compile(r'^EMP:([^,]*),([^,]*),([^,]*),([^,]*)$')
>
>    def parseline2(line):
>       m = emp_re.match(line)
>       if m:
>           ...
> ...
>
>    import re
>    def parseline3(line):
>       m = re.match(r'^EMP:([^,]*),([^,]*),([^,]*),([^,]*)$', line)
>       if m:
>           ...
>
> Unless I'm mistaken about how the re module works, parseline3 will
> recompile the regex each time it is called.

It does not.  re maintains internal caches of compiled regexp objects, and
has done so in all versions of Python to date.

There's a still a good reason to compile your regexps at module level,
though:  then you've got lots of room to use re.VERBOSE mode and add useful
comments to these dreadful beasts <0.7 wink>.





From whisper at oz.net  Tue Jun 11 12:51:23 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 11 Jun 2002 09:51:23 -0700
Subject: Cross platform spawning
In-Reply-To: <659f82ff.0206110754.775490d9@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEIEDIDIAA.whisper@oz.net>

On windows use os.system("start python myscript.py myargument") to get a
seperate process going. This is analgeous to the trailing & on Linux.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Peter Scott
> Sent: Tuesday, June 11, 2002 8:55
> To: python-list at python.org
> Subject: Cross platform spawning
>
>
> I am trying to make a script spawn a seperate process, like you would
> do like this on unix:
>
> os.system('python myscript.py myargument &')
>
> But the & at the end (to make a separate process) doesn't work on
> Windows. I want the script to run on both Linux (*BSD, whatever) and
> Windows. I didn't want to have to find out the path to the python
> interpreter, so os.spawnlp (with no blocking) sounded perfect until I
> saw that it wasn't available on Windows.
>
> Does anyone out there know what I can do in this situation?
> --
> http://mail.python.org/mailman/listinfo/python-list





From rdacker at pacbell.net  Sat Jun 22 12:30:36 2002
From: rdacker at pacbell.net (rdack)
Date: 22 Jun 2002 09:30:36 -0700
Subject: urllib POST question
References: <644f6688.0206200930.48110837@posting.google.com> <slrnah4dbq.r77.insyte-clp@petra.bitstream.net> <644f6688.0206210900.68dd930c@posting.google.com> <3D13D62B.D7AE6554@engcorp.com>
Message-ID: <644f6688.0206220830.46583e8f@posting.google.com>

Peter Hansen <peter at engcorp.com> wrote in message news:<3D13D62B.D7AE6554 at engcorp.com>...
> rdack wrote:
> > 
> > Ben Beuchler <insyte-clp at emt-p.org> wrote in message news:<slrnah4dbq.r77.insyte-clp at petra.bitstream.net>...
> > > In article <644f6688.0206200930.48110837 at posting.google.com>, rdack wrote:
> > > > i am doing a POST to a server using urllib and it works.
> > > > is there some way i can see exactly what is being posted?
> > > > a way to get the post object out of urllib - the exact lines it is sending?
> Howzabout this?
> 
> >>> import httplib
> >>> httplib.HTTPConnection.debuglevel = 1
> >>> import urllib
> >>> page = urllib.urlopen('http://web.engcorp.com:8080/?test=1').read()
> connect: (web.engcorp.com, 8080)
> send: 'GET /?test=1 HTTP/1.0\r\n'
> send: 'Host: web.engcorp.com:8080\r\n'
> send: 'User-agent: Python-urllib/1.15\r\n'
> send: '\r\n'
> reply: 'HTTP/1.0 200 OK\r\n'
> header: Server: Zope/(Zope 2.5.1 (binary release, python 2.1, linux2-x86), pytho
> n 2.1.3, linux2) ZServer/1.1b1
> header: Date: Sat, 22 Jun 2002 01:25:46 GMT
> header: Connection: close
> header: Content-Type: text/html
> header: Etag:
> header: Content-Length: 174
> >>> len(page)
> 174
> 
> -python-rocks-ly yr's,
>  Peter

perfect! thanks. although i was able to install ethereal and see the
packet-it was the 100 megabyte solution.
this was the zero byte solution.
of course, now i have the roll royce of packet sniffers.
anyway - you are right - python-rocks - when you know which button to
push. :-).


From ianb at colorstudy.com  Sat Jun  1 17:25:13 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 01 Jun 2002 16:25:13 -0500
Subject: How to open a HTML file when the python cgi program is
	executing?
In-Reply-To: <adajh1.13g.1@kserver.org>
References: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de>
	<ad95l1.i0.1@kserver.org> <ad9vsm$vhjnu$1@ID-49758.news.dfncis.de>
	<9m6ada.31b.ln@10.0.0.1> <adaeh8$uum7t$1@ID-49758.news.dfncis.de>
	<ada3k7.lo.1@kserver.org> <83n0uevjvx.fsf@panacea.canonical.org> 
	<adajh1.13g.1@kserver.org>
Message-ID: <1022966713.10680.8618.camel@localhost>

On Sat, 2002-06-01 at 15:51, Sheila King wrote:
> On 01 Jun 2002 16:12:02 -0400, Kragen Sitaker <kragen at pobox.com> wrote in
> comp.lang.python in article <83n0uevjvx.fsf at panacea.canonical.org>:
> 
> > Sheila King <usenet at thinkspot.net> writes:
> > > print "Location: http://replacement-url\n"
> > > 
> > > You should have nothing before this, and there's no point in having
> > > anything after it, since the page is being redirected.
> > 
> > Some browsers (in particular old versions of Opera) have problems with
> > redirects every once in a while; that's why redirects traditionally
> > contain HTML, in case the browser is broken.
> 
> I posted a short, quick, quip of a reply to your message. Although I've
> cancelled it now, it will probably still show up on most news servers.
> 
> In any case, a Google Groups search does find support for your statements
> (which I'm sure is no surprise to you).
> 
> As an example:
> http://groups.google.com/groups?as_umsgid=DoBG3t.LMv%40world.std.com

Well, that's HTTP -- apache CGI has a few more details.  If you don't
include a body, Apache will create one for you (using it's standard
mechanisms and text).

*However* you should send a Status header as well, because if you don't
*and* you don't give a fully qualified URL (i.e., with http://), then
Apache will do a local redirect.  In the case of a local redirect, the
browser will not be aware of the redirect, but the output will be
replaced with the target location.  Relative links in the target HTML
might be broken because of this.

  Ian






From gnufnork at fuckspam.hetdigitalegat.nl  Tue Jun 11 11:39:21 2002
From: gnufnork at fuckspam.hetdigitalegat.nl (Wladimir)
Date: Tue, 11 Jun 2002 17:39:21 +0200
Subject: Hardware acceleration?
References: <ae51s6$1vs$1@mailgate2.lexis-nexis.com>
Message-ID: <pan.2002.06.11.15.38.16.296810.4300@fuckspam.hetdigitalegat.nl>

On Tue, 11 Jun 2002 16:34:36 +0200, rzed wrote:

> It makes me wonder whether any such thing has been tried for the Python
> VM. Have there been any hardware projects taking a look at revving up
> Python?

Great idea.
But I think the python (same for perl) bytecode format is too much in
state of flux (not to mention undocumented) to make a hardware
implementation really feasible.
On the other hand, I remember something about the ARM series having
relatively easily customizable opcodes, then it would be great. *although
modern processors already have lots op optimalisations for interpreted
code so I doubt it would serve much of a purpose, maybe for embedded
devices?*

Wladimir


From loredo at astro.cornell.edu  Thu Jun 20 17:13:25 2002
From: loredo at astro.cornell.edu (Tom Loredo)
Date: Thu, 20 Jun 2002 17:13:25 -0400
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <aer2fn$frc$1@peabody.colorado.edu> <3D122E79.A189C72E@astro.cornell.edu> <aetapt$kld$1@peabody.colorado.edu>
Message-ID: <3D124575.2A1393FF@astro.cornell.edu>

Fernando P?rez wrote:
> 
> Don't know if there's a misunderstanding here: 

Very likely!  (On my part!)

> cygwin is a unix-like
> environment under Windows. 

Okay, my concern was that it wouldn't work with or be able to produce
Windows binaries---like if I had to produce binary extension modules
to ship to others, or a McMillan-ized install that used PythonWin
and Tk or wxWindows.  Cygwin is so huge that I got the impression
it was a self-contained "subenvironment" so to speak, so I didn't
realize I could use it to produce stuff that non-cygwin Windows users
could use.

> Well, if you do have to live with windows, I'd recommend at least using XP.

Well, there is now a recommendation each for WinXP and Win2k in this
thread.  If anyone has tried both and can offer a comparison/recommendation,
I'm all ears.  If I'm going to upgrade, I'd rather do it sooner than later.

Thanks again,
Tom


From aahz at pythoncraft.com  Sat Jun  1 13:09:36 2002
From: aahz at pythoncraft.com (Aahz)
Date: 1 Jun 2002 13:09:36 -0400
Subject: Generating unique numbers?
References: <3CF7EBC8.4080802@verio.net> <3CF8FC1C.7060903@lindbergs.org>
Message-ID: <adav4g$o2b$1@panix1.panix.com>

In article <3CF8FC1C.7060903 at lindbergs.org>, VanL  <news at lindbergs.org> wrote:
>
>The reason why I am wanting keys that are more likely to be globally
>unique is because I will be building in hooks to make this outliner
>network transparent (using ZEO), making it possible for multiple
>outliners to be used together as a real-time collaborative tool.  I
>want to be able to be able to have people merge stuctures that people
>have made in the standalone instance into a shared instance without
>tromping on another nodes data.
>
>Based on what people have said, I was thinking of using some
>combination of DateTime.gmticks and the id of the obj, making it (I
>think) extremely unlikely that there would be a collision.

id() is very, very bad for any kind of persistent ID; I still think it's
at least very bad even if you add in any kind of time.

No, if you want to trade IDs across a network, you need GUIDs (DCE
UUIDs).  There's just no way around that.  It's not the most perfect
system in the world, but it's better than any other generally available
scheme.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"In the end, outside of spy agencies, people are far too trusting and
willing to help."  --Ira Winkler


From rhaper at houston.rr.com  Fri Jun 14 15:55:31 2002
From: rhaper at houston.rr.com (Rod Haper)
Date: Fri, 14 Jun 2002 19:55:31 GMT
Subject: Writing a simple linux daemon type thing...
References: <K9eO8.393$LN6.261@newsread1.prod.itd.earthlink.net> <rn3cea.rt4.ln@127.0.0.1> <yylO8.916$TP4.88596@newsread2.prod.itd.earthlink.net> <aecrfe$a4l$1@minotaurus.cip.informatik.uni-muenchen.de>
Message-ID: <3D0A4A07.8070109@houston.rr.com>

Johannes Stiehler wrote:
> The point is to completely unconnect the process from the console where it
> was started.


I use setsid to do this.  For example:


setsid myprog.py [args] &


-- 
Rod

+----------------------------------+
| Rod Haper, rhaper at houston.rr.com |
|     There is a better way ...    |
|   LAP => Linux + Ada95 + Python  |
+----------------------------------+



From cliechti at gmx.net  Wed Jun 12 19:27:59 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 13 Jun 2002 01:27:59 +0200
Subject: Compiling to .exe
References: <UwQN8.43539$zh2.13069449@twister.neo.rr.com>
Message-ID: <Xns922CF52079FAcliechtigmxnet@62.2.16.82>

"Dylan Appleman" <tlirobot at tlirobotics.cc> wrote in
news:UwQN8.43539$zh2.13069449 at twister.neo.rr.com: 

> Is it possible to compile Python to .exe?

not in the sense of 'compile' like known from C and other programming 
languages. 

>If so, what program(s) do I
> need? Windows XP, no other programming experience but Python.

but you might well get happy with py2exe or McMillans' Installer.
they both create an exe with an embedded python interpreter plus all the 
files that are needed in a directory, ready to copy to any machine with or 
without python.

http://www.mcmillan-inc.com/install1.html
http://starship.python.net/crew/theller/py2exe/

chris

-- 
Chris <cliechti at gmx.net>



From maxm at mxm.dk  Tue Jun 18 02:32:19 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 18 Jun 2002 08:32:19 +0200
Subject: Creating Dynamic Web Pages (ZOPE)
References: <46ca81a0.0206062016.789d3af6@posting.google.com> <J29M8.358696$%u2.156726@atlpnn01.usenetserver.com> <46ca81a0.0206141201.23db4a8c@posting.google.com> <kh3tgus5d6hlrt4qjsdg1odr6j9ft0nue2@4ax.com>
Message-ID: <3D0ED3F3.2020008@mxm.dk>

robin at execulink.com wrote:

> I would very much advise against DreamWeaver and similar tools that
> pretend designing web pages is a WYSIWYG activity, which it certainly
> is not.


I beg to differ Sire!

Dreamweaver is a very fast and convenient way to edit html. If you know 
what you do, your code is just as clean as handcrafted html. There is 
nothing in html that you cannot tweak in Dreamweaver.

regards Max M



From aurelienviolet at noos.fr  Thu Jun 20 23:52:04 2002
From: aurelienviolet at noos.fr (aurel-le-gnou)
Date: Fri, 21 Jun 2002 05:52:04 +0200
Subject: Tkinter & Python: how to get the return of a call associated to an event 
 on my gui
Message-ID: <3D12A2E4.6A566789@noos.fr>

Hi


I would like to thank all the people who have helped me these last days
on tkinter programming. (HTML links in a GUI tkinter/ Help for Tkinter
and python's programming)
I'm surprised for being helped with so kindless. 
Special thanks to Russell E. Owen whose response was really excellent
and who also learned me lots of things interessants.

I have another problem with my GUI :/


Here are some extract of my source code:



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

class Result:

    def __init__(self):
        # code d'erreur
        self.result = 0
        # nombre de pages non touv?es
        self.num_pages = 0
        # les pages non trouv?es
        self.pages_not_found = []
        # diverses exceptions
        self.exceptions = []
        # la liste des urls et leur rang associ?
        self.results = []


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

def GenericCallback(callback, *firstArgs, **firstKWArgs):

     if firstKWArgs:
         return GC(callback, *firstArgs, **firstKWArgs)
     else:
         return GCNoKWArgs(callback, *firstArgs)

class GCNoKWArgs:

     def __init__(self, callback, *firstArgs):
         self.__callback = callback
         self.__firstArgs = firstArgs
         
     def __call__(self, *args, **kwArgs):
         return self.__callback (*(self.__firstArgs + args), **kwArgs)

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

def make_call(frame_left, frame_right, button_number, button_search,
button_language, button_rank, button_engine, event):  

 ...

    result = run(button_language_content, button_number_content,
button_search_content, button_rank_content, button_engine_content) 
 ...

    return result

---------------------------------------------------------------------
IN MY MAIN:


myCallback1 = GenericCallback(make_call, window_left, window_right,
button_number, button_search, button_language, button_rank,
button_engine)

button_go.bind("<Button-1>", myCallback1)
button_search.bind("<Return>", myCallback1)


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


here is my question, if you could help me, it would be really fine:


"result" is an instance of the class "Result" and is returned by the
call of 'run' without any problem. It's also returned by the call of
"make_call".
But as you can see, "make_call" is binded with an event.
I would like to know how to get the object "result" returned by this
call in my main.
I need this object "result" in my main, but i don't success in getting
it after the event <button-1> on the button_go and the execution of
"make_call"


Thanks a lot for all this help


Aur?lien Violet
aurelienviolet at noos.fr




From whisper at oz.net  Wed Jun 19 19:43:24 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 19 Jun 2002 16:43:24 -0700
Subject: old fashioned COM
In-Reply-To: <uh20956iiu8q52@corp.supernews.com>
Message-ID: <GCEDKONBLEFPPADDJCOEKEEMDKAA.whisper@oz.net>

Windows serial extensions for Python:
http://py.vaults.ca/parnassus/apyllo.py/990075885.195097684

I think on Linux, it's built in (just access the device with open?).

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of mark
> Sent: Wednesday, June 19, 2002 15:20
> To: python-list at python.org
> Subject: old fashioned COM
>
>
> I am a newbie at this, and haven't done a lot of programming in some
> years.  The book I have on Python describes a lot about programming
> internet and email apps.  I need to just select a COM port, set
> parameters like baud rate, parity, etc. and do a file capture in dumb
> terminal mode.  This may seem to simple in todays world, but I'm trying
> to log some events from a piece of equipment that uses serial I/O.
>
> Help?
>
> Thanks
> Mark
> --
> http://mail.python.org/mailman/listinfo/python-list





From tdelaney at avaya.com  Tue Jun 18 20:10:08 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 19 Jun 2002 10:10:08 +1000
Subject: Iterator module (was: what *is* a class?)
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A39E@natasha.auslabs.avaya.com>

> From: holger krekel [mailto:pyth at devel.trillke.net]
> 
> Delaney, Timothy wrote:
> > > From: holger krekel [mailto:pyth at devel.trillke.net]
> >
> > iterable.sort(iterable [, func])
> > iterable.reverse(iterable)
> 
> that would be quite expensive for common cases (like lists).  
> reverse and 
> sort should also allow to operate on complete sequence types. 
> iterable.reverse could be cheap again then and iterable.sort would be
> as expensive as it is now (but not requiring any inplace-steps).

I disagree. The module I would propose is specifically for operating on
iterable objects (i.e. it is a general mechanism). Although there's nothing
stopping special-casing for common cases (such as tuples and lists) for
performance reasons if the semantics don't change.

It is also a *functional* module - nothing is modified in-place.

I'm also more inclined to call the module 'iterator' because ...

> > iterable.map(func, iterable, ...)
> > iterable.filter(func, iterable, ...)
> > iterable.reduce(func, iterable, ...)
> > iterable.zip(iterable, ...)
> > iterable.range([start,] stop[, step])
> 
> yes, indeed! i'd probably do 'from iterable import *' :-)

... these are *not* drop-in replacements for the similarly-named functions
in builtins. Specifically, every method returns an iterator (which excludes
reduce from the list).

    l = [ 3, 1, 2, ]
    s = iterator.sort(l)
    r = iterator.reverse(s)

    print l
    print s
    print r

What output do you expect?

> seriously, i'd include something like concat as an iterable, too: 
> 
>     for item in iterable.concat(iter1,iter2):
>         # items iterates with iter1, then iter2 
> 
> All the the iterable functions should accept sequence and 
> possibly mapping types (same semantics as for 'for'-loops now).

Nope - they would all accept iterable objects (where appropriate). Same
semantics as for 'for'-loops now (they call iter() on appropriate
parameters).

Things like extend() and concat() could be appropriate, but you could end up
calling:

    t = (1, 2, 3,)
    i = iterator.extend(t, 4)

which may or may not be considered a good thing.

> > There have been suggestions like this before, but I don't 
> recall one for
> > sort() and reverse(), and these two have fired me :)
> 
> I think an iterable module should be done. Although there would 
> inevitably be overlapping functionality with the current builtins.

Some overlapping, but the majority (all?) builtins return restartable
iterables (such as a list) rather than an iterator.

Tim Delaney




From kragen at pobox.com  Mon Jun 10 20:44:56 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 20:44:56 -0400
Subject: Iterators vs. Generators
References: <ae3a6f$9d9$1@panix1.panix.com>
Message-ID: <83elfemynr.fsf@panacea.canonical.org>

aahz at pythoncraft.com (Aahz) writes:
> So when would one actually write an iterator instead of a generator?
> I've been trying to think of an example and failing.

Dunno.  Here are some ideas:
- when it needs to run in Python 1.5.2 (really!  MetaPy.Iter lets you
  write iterators that work with 1.5.2 and work fast with 2.2)
- when it is written in C or OCaml instead of Python




From steve.bond at btinternet.com  Wed Jun 19 09:59:43 2002
From: steve.bond at btinternet.com (Steve)
Date: Wed, 19 Jun 2002 13:59:43 +0000 (UTC)
Subject: newbie: help please!
References: <aepltq$gl3$1@knossos.btinternet.com> <mailman.1024490396.5209.python-list@python.org>
Message-ID: <aeq2of$a1d$1@knossos.btinternet.com>

Cheers Chris it was exactly as you said i'm so happy now! i just installed 
the two RPM's X11R6-contrib & XFree86-devel and updated my setup file and 
it worked!, thanks again

Steve


From gcordova at hebmex.com  Thu Jun 27 18:29:16 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 27 Jun 2002 17:29:16 -0500
Subject: Be the first on your block to register your computer.
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233015DEB8C@mail.hebmex.com>

> 
> "David LeBlanc" <whisper at oz.net> wrote in message
> news:mailman.1025211444.21440.python-list at python.org...
> > Yes! exciting times are coming again as Intel, AMD and 
> > Microsoft (among quite a few others) are banding together
> > to once more put ID numbers (and now basic encryption)
> > onto next generation silicon.
> 
> Not that I _want_ to sound crabby, but: this has relevance to 
> Python _how_ exactly?
> 
> James
> 

Well, for one, Python is one of those evil Open Sores
software viruses, unstable, dangerous to the health of
your machine.

You can bet your snickers that Python will not be one of
those "M$ Approved" software packages that you may freely
install, once this comes through.

-gus




From gerhard at bigfoot.de  Tue Jun  4 12:27:53 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 4 Jun 2002 16:27:53 GMT
Subject: self
References: <BI5L8.78$Vv1.4155449@newnews.cc.stevens-tech.edu>
Message-ID: <slrnafpq5k.os.gerhard@developer2.opus-gmbh.net>

Vojin Jovanovic wrote:
> I have been working on a project trying to implement some advanced
> functionality.  Now, one thing that has been a constraint for me is the
> concept of self.  Why is it that within a class one has to use self. (or
> "any word". ) to refer to instance variables? Would it not be more logical
> to just use the name of the variable like it is for example in C++?
 
It's good practice in C++ and Java too, to always use this->myvar resp.
this.myvar.

I hate to have to look at code where the scope of the variable isn't obvious at
once. And I also hate conventions like m_mymembervar to distinguish members
from locals.

Gerhard


From derek at wedgetail.com  Fri Jun 21 08:13:55 2002
From: derek at wedgetail.com (Derek Thomson)
Date: Fri, 21 Jun 2002 12:13:55 GMT
Subject: I'd give up Perl tomorrow if only...
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au> <3D12F6FC.3000608@wedgetail.com> <pan.2002.06.21.10.09.01.691351.4527@NOSPAM.optushome.com.au>
Message-ID: <3D131881.3070200@wedgetail.com>

David Mitchell wrote:

> 
> Thanks Derek, I'll check these out over the weekend.  I knew of the
> existence of Fnorb, but not perlorb

Well, it's kind of an open secret. I'll announce it properly in the 
appropriate places when I'm happier with it!

> 
> I've thought several times "why don't I just start porting e.g. DBI and
> DBD::Oracle myself, and maybe someone else will do DBD::*", and we'd then
> be on our way.  Time, as always, is the big problem - I just don't have
> the time on my hands to take on such a big task myself, and commit to
> supporting and updating it afterwards, although I'd happily commit small
> chunks of time to such a project.

Isn't there a DBI-alike project for Python? Perhaps you could help with 
that?

> 
> I'm busy playing with Jython at the moment to see if I can deal with the
> problem from another angle.  Like Perl, Java's also got some nice
> "generic" solutions (e.g. JDBC) to common problems, and Jython looks like
> potentially a really nice fit for a bunch of people I know who are struggling
> with WebSphere at present.  WebSphere on sites that need to change rapidly is a
> particularly ugly mix...

I can't understand why anyone bothers with anything other than Zope ;)

Regards,
Derek.



From tismer at tismer.com  Sat Jun 22 18:17:38 2002
From: tismer at tismer.com (Christian Tismer)
Date: Sun, 23 Jun 2002 00:17:38 +0200
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14D3B4.7050709@kfunigraz.ac.at>
Message-ID: <3D14F782.9080306@tismer.com>

Siegfried Gonzi wrote:
> Bengt Richter wrote:
> 
>> Have you never chased a bug that only shows up in secondary effects?
>> In C code, the most common way for that to happen is storing something
>> through a bad pointer, but instead of getting a memory fault, it corrupts
>> some data structure or data value. The value could be used 20 minutes 
>> later
>> and cause GPF then, or e.g., it could be data that's part of a 
>> calculation,
>> but maybe it's only overwriting the least significant half of a floating
>> point number, so you get the effect systematic bias or random error 
>> injected
>> into your calculation, but nothing else. If you don't look for 
>> explanations
>> of last-decimal errors, this kind of thing can persist a long time. OTOH,
>> you could be corrupting numbers that are used in an iterative calculation
>> that mysteriously starts to take longer to converge, and does so in a 
>> systematically
>> increasing way, because the corrupting data varies systematically with 
>> time.
>> Or you could be corrupting data structures and sabotaging memory 
>> management
>> without causing immediate GPF, etc. etc.
> 
> 
> The explanation is good and could explain the bad behavior. But note: I 
> use the same C code on Windows and on Linux! On Windows I used SWIG and 
> Visual C++. On Linux I used gcc and SWIG.

So what?
You get similar behavior under Windows and Linux. The program
goes astray. Windows can kill the job. Linux can, too; you are
just not used to Linux.
How the program misbehaves is surely slightly different due to
different OSes and compilers.

But you tend to draw conclusions from nothing.
Using the same tool (Python) on different machines
with your programs gives you crashes. That makes you
damn the tool.

The first one a programmer should never trust is yourself.
If you count the errors and crashes of your life's
programming, you will have to conclude that most of the
time it is *you* who makes errors. This is fine, since
everybody has this problem.
But your program has only undergone your own testing, while the
tools you are using are well-approved hundreds of thousands
of times. They have grown the same way as your own stuff, but
they had orders of magnitude more chances to ripen.

Please keep that in mind, whether you leave Python and stick with
F90 or not. F90 has bugs, and Python has, but before searching
for them, it is much more efficient to look for your own.

> Theoretically I could do the following: trying to use the Fortran code 
> instead of the C code. The C code relies on a Fortran 77 code. I pinched 
> the C code from the netlib (or so). But I could try to use F2PY in 
> combination with the Fortran 77 code.
> 
> The F2PY would have the advantage that I am not forced to write 
> explicitely pointer-wrapper functions in Python in order to use the code 
> which comes from SWIG.

This might in fact be simpler.
What I'd like to suggest:
Go through you code and add comments. Think that you
have to explain your code to somebody who has no idea
about it at all. Then, give it realy to a colleague,
let him read the code and ask questions. Explain everything,
and defend every line of your code. Why is it so and not
different, what is the reason ebhind this, what happens if
that condition isn't met -- oops, I didn't check. ???

You will find the bugs, finally, probably also those, which
are nicely hiding in the original code for a long time.

> The above could also explain why under Windows the system sucks down, 
> but it is not destroyable!

Run it under restricted resources with Linux, and you will
be able to kill the job, too.
Or just stick with Windows. The problem is not the OS!

or-make-me-an-offer-to-solve-this-for-you - ly y'rs - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From gmcm at hypernet.com  Tue Jun  4 09:55:20 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 04 Jun 2002 13:55:20 GMT
Subject: ?Tcl's embedded virtual filesystem and Scripted documents
References: <3CFBF133.7DF60BA4@email.sps.mot.com>
Message-ID: <Xns9223651854560gmcmhypernetcom@199.171.54.214>

Norman Shelley wrote:

> Tcl now has an embedded virtual filesystem that looks very useful.
> http://mini.net/tcl/2138.html
> 
> Is there any motivation in the Python camp to have something similiar?

There seems to be some interest (it has come up before), but
this would be much harder to pull off in Python than it is 
in Tcl. There are many places in the C code where a c runtime 
FILE* is required.
 
> Also, a "scripted document" like methodology looks to be very useful
> also. http://www.equi4.com/jcw/scripdoc.html
> 
> Any work going on in this arena?

I put together a crude one a few years ago (using Metakit and
tools from Installer). I've never seen any interest in this in
the Python world. But Tcl doesn't have anything as slick as
py2exe / Installer, so to a certain extent, it may be that 
scripted docs fill that role.

-- Gordon
http://www.mcmillan-inc.com/


From robin at jessikat.fsnet.co.uk  Sat Jun 29 18:43:54 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 29 Jun 2002 23:43:54 +0100
Subject: yield equivalent in C/JavaScript?
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E201922151B8@admin56.narex.com>
References: <60FB8BB7F0EFC7409B75EEEC13E201922151B8@admin56.narex.com>
Message-ID: <y3QdaIAqgjH9EwBR@jessikat.fsnet.co.uk>

In message <60FB8BB7F0EFC7409B75EEEC13E201922151B8 at admin56.narex.com>,
Bjorn Pettersen <BPettersen at NAREX.com> writes
>> From: Robin Becker [mailto:robin at jessikat.fsnet.co.uk] 
>> 
>> Is there any way to do yielding easily in C? I have a bunch 
>> of programs which are normally connected together using unix 
>> pipes. I would like to merge them and the problem is then how 
>> to get them to communicate nicely using a single process.
>
>Using threads, you can simulate yield by having the yielding thread
>blocking until the thread calling it tells it to restart. This is rather
>obscure however, so I wouldn't suggest that you do it. If you are going
>to use threads, I would suggest using a Queue to communicate
>work/results (I can send you our C++ version of Python's Queue.Queue()
>implementation if that would help...)
>
>-- bjorn
thanks I know how to do this with threads, but I suspect I will end up
doing this in JavaScript and threading is a complete mystery there (if
it exists).
-- 
Robin Becker




From ChrisE at lantech.com  Fri Jun 28 09:28:57 2002
From: ChrisE at lantech.com (Christopher Encapera)
Date: Fri, 28 Jun 2002 09:28:57 -0400
Subject: [Application] Porn Toolkit/0.2.0
Message-ID: <BC74F8F7E87DD511990C000103E2FEFA30E4@XCHG>

Gezz, and I thought I wrote a script for everything...Maybe we could all
chip in & get this guy a mail order bride - or perhaps someone could
recommend a nice community organization in his area where he could meet
others with (hopefully) slightly different hobbies ;)



Christopher Encapera
Network Administrator
Lantech.Com




-----Original Message-----
From: Iggeres Bet [mailto:iggeres at yahoo.es]
Sent: Wednesday, June 26, 2002 11:18 PM
To: python-announce-list at python.org
Subject: [Application] Porn Toolkit/0.2.0


                           Porn Toolkit/0.2.0                           
                           ------------------                           

A script to download porn.

Porn Toolkit is a python script to download porn from Internet. It's
oriented to the lazy Internet navigator confused with million of banners
that points to nowhere.

Any idea to improve this software is welcomed.

       URL:  http://porntoolkit.sourceforge.net
  Download:  http://porntoolkit.sourceforge.net/porntoolkit-0.2.0.tar.gz

   License:  GPL

  Categories:  Net Applications

Iggeres Bet (iggeres at yahoo.es)
http://porntoolkit.sourceforge.net

--
<a href="http://porntoolkit.sourceforge.net">Porn Toolkit/0.2.0</a> -- A
script to download porn.



-- 
http://mail.python.org/mailman/listinfo/python-announce-list




From paul at boddie.net  Wed Jun  5 11:21:32 2002
From: paul at boddie.net (Paul Boddie)
Date: 5 Jun 2002 08:21:32 -0700
Subject: [ANN] pyPgSQL 2.1 released
References: <mailman.1023270308.20913.python-list@python.org>
Message-ID: <23891c90.0206050721.49b31f50@posting.google.com>

Gerhard =?unknown-8bit?Q?H=E4ring?= <gh_pythonlist at gmx.de> wrote in message news:<mailman.1023270308.20913.python-list at python.org>...
> 	
> * Fixed the array parsing so it also works with PostgreSQL versions 7.2.x.

I doubt that this is related, but does the 'IN' operator work with
statement parameters (bind variables) in pyPgSQL yet? I recently found
that doing something like this...

  cursor.execute("SELECT * FROM CAGES WHERE NAME IN %s",
    (("mammals", "apes"),))

...gives some kind of parsing error inside pyPgSQL. Just to verify
that I wasn't doing anything too strange, I tried the same thing with
psycopg and it worked as expected.

Another thing: pyPgSQL seems to be fine with Unicode query strings
(unlike psycopg, apparently), but it doesn't like Unicode parameter
values. Is there some explicit encoding step I should be performing?

Paul


From dwelch91 at home.nospam.com  Mon Jun 17 14:45:09 2002
From: dwelch91 at home.nospam.com (DJW)
Date: Mon, 17 Jun 2002 11:45:09 -0700
Subject: Long pathes on windows
References: <mailman.1024334302.2943.python-list@python.org>
Message-ID: <aele3j$tq6$1@news.vcd.hp.com>

Use:
short_path = win32api.GetShortPathName( long_path )

Example:

>>> import win32api
>>> s = win32api.GetShortPathName( "c:\\program files" )
>>> print s
c:\PROGRA~1
>>>


Don

"Roman Yakovenko" <romany at actimize.com> wrote in message
news:mailman.1024334302.2943.python-list at python.org...
Good evening. I have big problem Windows 2000 and small one.
The small one is long pathes on windwos. How can I convert long path with '
'
into short path?

Thanks for the help. Roman.





From tim.one at comcast.net  Fri Jun 21 23:53:32 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 21 Jun 2002 23:53:32 -0400
Subject: How to represent the infinite ?
In-Reply-To: <slrnah7lum.27i.huaiyu@gauss.almadan.ibm.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEEFPPAA.tim.one@comcast.net>

[Huaiyu Zhu]
> Each time we are having this conversation we go around the same
> circle once or twice.  Let's see if we can find a way out.  I would like
> to divide the issue into three (still somewhat arbitrary) sub-issues, and
> see if any one of them can be resolved in easier/simpler/clearer ways:
>
> 1. How much 754 support is provided by the underlying platforms.
> 2. How do such support vary across the platforms.
> 3. What are the Python designers' wishes.

[on platform variations]
> ...
> I think you are talking about sub-issue 2 here.  Is there some
> place we can find out what these variations are, and how severe they are?

Not that I know of, although several projects have tried and given up.  In
addition to hardware, it varies according to all of compiler, specific
compiler release, libraries, specific library releases, and configuration
options specific to all of the preceding.  If and when vendors chose to
support the optional C99 754 gimmicks, then there will be a uniform way to
proceed.  gcc is making some progress in that direction.

> I can certainly contribute "the patch to enable inf and nan in Python" :-)
> (attached), but I'm sure that's not the reason you are against it.

Because it's a crock <0.1 wink>.  Making x**y different than pow(x, y)
sucks, and even worse if it's in ways that can vary across platforms.  It's
also the case that we believe a majority of Python users *want* overflow
exceptions, and are well served by them.  Unless and until there's a
coherent and reasonably x-platform scheme allowing users to enable and
disable specific numeric exceptions at will, we'll favor griping on overflow
(and divide-by-0, and invalid operation) when it doesn't take heroic effort
to do so (we leave * + - and / alone, except for /-by-0, because it's too
hard to check them -- fpectlmodule.c makes heroic efforts in that direction,
but doesn't work on all platforms, or, over time, even on all platforms it
used to work on).

> ...
> I think you are talking about sub-issue 1 here.  Do you actually believe
> that subset-compliance is worse than none?

For a sufficiently broken subset, yes.  "No possibility ever to raise an
exception" is in fact technically conforming to the standard, as the
possibility for trap handlers of any kind are only encouraged ("should", not
"shall").  That's an area where I want something stronger than the std
requires.

> I'd think that would depend on what subset we are talking about.

For example, fine by me if you can't get at directed rounding modes, despite
that they're mandatory.  Others would disagree with that, btw.

> Certainly "can't provide full support" does not imply "avoid as much as
> possible".  There must be some concrete criteria that are omitted here.

The lack of the possibility to enable and disable exceptions on overflow,
divide-by-0, and invalid operation are hangups to me.  Others may include
underflow in that mix.  Others will include sticky status flags, etc.  You
need to hammer this out in the NumPy community first, though.

[Tim]
>> If Python were 100% 754-conforming, Guido and I would still
>> favor enabling exceptions on overflow, divide-by-zero, and invalid
>> operation, by default.  As things are, you've got no choice at all,
>> unless you want such exceptions enabled and are on a platform where
>> the fpectl module happens to work.

> I think you are talking about sub-issue 3 here.  Could you provide a
> rationale in the context of numerical computation?

Sorry, it's a waste of time to do that again.  There are examples on both
sides, and what they show in aggregate is that different apps need different
exceptional-case behavior, and somtimes even across phases within a single
app, not that NonStop Rulz.  There are even cases that can benefit greatly
just from being able to examine the inexact flag.  754 has lots of good
stuff.

> I can offer an example in support of the opposite side.

Of course you can.  If you tried 3x harder, you could give twenty <wink>.

> ...
> How can this be done by vectorized computation if individual elements
> may throw exceptions?

It's amazing how much numeric computation got done on old Cray boxes, which
didn't give you a choice about blowing up on overflow either.  I don't know
that vectorized computations have any relevance to core Python, but if you
really want to know, in your exp(-x**2) example people used vector merge to
replace "too big" values of x before feeding the X vector into the rest of
the computation.  In the case of Python on a 754 box, you can almost
certainly get away with writing it as exp(-x*x) instead -- and it will
likely run faster, too (x**2 calls the platform pow(); x*x is a straight
multiply).

> Certainly you and Guido have the final say on sub-issue 3.  On
> the other two issues, however, I suspect that
>
> 1. Most platforms these days support 754 to a remarkably good degree, far
>    better than that is allowed in today's Python.

The support code required at the C level is platform-specific, and needs a
phalanx of cooperative platform experts to contribute it.  Slash and burn
will not lead to a good and maintainable result.

> 2. Most number crunchers would be more concerned that it works as well as
>    possible on the platforms available to them, rather than  whether it
>    also works on other platforms.

As I said above, that's why the NumPy community has to agree first.  The
overwhelming majority of plain-core Python users are not number crunchers,
and don't know a NaN from an umlaut.  Giving them either 1j or a NaN from
sqrt(-1) is a Bad Idea.  I would often like 1j or a NaN in my own apps --
but I often want the ValueError Python tries to deliver in that case too.

> 3. The variations across major platforms are minor.

I take it you haven't looked at Python's fpectlmodule.c <wink>.





From kurugant at denshi.ece.utk.edu  Fri Jun 28 09:52:28 2002
From: kurugant at denshi.ece.utk.edu (Teja Sastry)
Date: Fri, 28 Jun 2002 09:52:28 -0400
Subject: Python objects
Message-ID: <3D1C6A1C.2060407@denshi.ece.utk.edu>



Could anybody share their experience on how to call the C++ variable to 
python,  I'm not quite familiar with python . I see there is a way to 
declare that as a python object in C++ program and export it to python. 
I would really appreciate if anybody can clarify me on how do this 
prgramming.

Thanks,
Teja Sastry





From mjais at web.de  Wed Jun  5 09:47:07 2002
From: mjais at web.de (Markus Jais)
Date: Wed, 05 Jun 2002 15:47:07 +0200
Subject: python cookbook at oreilly's web site
References: <dCnL8.5$566.1250@news.ecrc.de> <adl2nq$3hrm$1@ID-41653.news.dfncis.de> <adl3ii$ofq$1@newsreaderg1.core.theplanet.net>
Message-ID: <7DoL8.6$566.2353@news.ecrc.de>

F. GEIGER wrote:

> Yes, and I'd really like to know the difference.

when you click on "full description" on the oreilly site
there is an overview how it differs from the ASPN cookbook

but there is not text that desrcibes the changes in detail
at least I have not found one.

markus


> 
> Franz
> 
> 
> "Tobias Klausmann" <klausman-un040602 at tuts.net> schrieb im Newsbeitrag
> news:adl2nq$3hrm$1 at ID-41653.news.dfncis.de...
>> Markus Jais <mjais at web.de> wrote:
>> > just found
>> >
>> > http://www.ora.com/catalog/pythoncook/
>>
>> Does this have any relation to ASPNs Python cookbook?
>> http://aspn.activestate.com/ASPN/Cookbook/Python
>>
>> Wondering,
>> Tobias
>>



From janto_d at hotmail.com  Sat Jun 29 15:18:37 2002
From: janto_d at hotmail.com (Janto Dreijer)
Date: Sat, 29 Jun 2002 21:18:37 +0200
Subject: doctest for java
References: <afhv70$15mk$1@news.adamastor.ac.za> <afisev$1f6e$1@news.adamastor.ac.za>
Message-ID: <afl16t$2bv1$1@news.adamastor.ac.za>

"Janto Dreijer" <janto_d at hotmail.com> wrote in message
news:afisev$1f6e$1 at news.adamastor.ac.za...
> Tim Peters <tim.one at comcast.net> wrote in message
> news:<mailman.1025281419.16432.python-list at python.org>...
> > [Janto Dreijer]
> > > Are there any attempts to port the "doctest" module to Java? (All
> > > I can find using google is some testing class for the JavaDoc program
> > > calling itself "DocTest".)
> > >
> > > At a glance it looks like a worthwhile effort...extracting Jython
> > > "examples"
> > > from /**...*/ comments also seems simple to implement.
> >
> > Last I heard, Python's std doctest.py worked fine under Jython.  If
> anything
> > else simple is needed, you're elected <wink>.
>
> I in no way meant your code is trivial. Although I'm starting to think you
> did that "import __future__" trick with the sole reason to confuse :-)
>
> I'm actually thinking of running doctests on *.java files and not on *.py
> files, so the current doctest.py wouldn't be 100% usable.
>
> class Obj {
>     /**
>         "Three" is the number it should print. And the number of the
> printing should be "three":
>         This is a usage example written in Jython...maybe.
>
>         >>> Obj.f()
>         3
>
>     */
>     public static void f() {
>         System.out.println(3);
>     }
> }
>
> That's it. It can't be helped. I'm spending my weekend on this :-)
>

sigh. ok, I give up.

I wanted to use Tim's implementation since most of the code would be the
same, so:
After extracting all the comments from a Java source file with a simple
regular expression I simply instantiated the Tester class included with
doctest.py and passed the comments to the runstring() method. That means
about 10 lines from Jython.

The problem, however, is that this can only work for public and friendly
Java functions. There's no way I can access a class' properties that are by
definition private. So that rules out solutions involving outside classes.
Unfortunately Java objects also have limited introspection in that it can't
grab a listing of it's own methods (ala __dict__) and thus also rules out a
hack that involves initiating doctests from within the object itself.

A possible way is to go deeper into the JVM by using either a custom made VM
(...yeah, right) or using the Java debugger tool (jdb.exe) which I assume
has access to private attributes. This seems the way to go if someone has
enough enthusiasm.

Bottom line: I'm not _that_ a big a fan of Java. I might work on it a bit
more during the next few weeks, but don't get your hopes up.

ps. I'll mail my code if anyone asks, but I'm not associating myself with
this half-working code by posting it on Usenet. :-)





From aahz at pythoncraft.com  Mon Jun 24 23:12:45 2002
From: aahz at pythoncraft.com (Aahz)
Date: 24 Jun 2002 23:12:45 -0400
Subject: Names and bindings (was Re: Suggestions for good programming practices?)
References: <WEuR8.2074$5l4.75657@ozemail.com.au> <bko7fa.9t.ln@grey.aerojockey.localdomain> <af7sch$dcp$1@panix1.panix.com> <3D17C0D0.439C5269@engcorp.com>
Message-ID: <af8n3d$2b8$1@panix1.panix.com>

In article <3D17C0D0.439C5269 at engcorp.com>,
Peter Hansen  <peter at engcorp.com> wrote:
>Aahz wrote:
>> In article <bko7fa.9t.ln at grey.aerojockey.localdomain>,
>> Carl Banks  <imbosol at vt.edu> wrote:
>>>
>>>* Learn how references and the object system work as soon as possible.
>> 
>> ObBrokenRecord: "bindings", not "references".
>
>I haven't had a record player for a while :-) so would you please
>expand on that?  I've adopted "references" without giving it any
>thought.  Why is that not as good as "bindings"?

Because unless you're using the C API, you can never access the
references directly.  Python always hides the references from you:

>>> a = []
>>> b = a
>>> a.append(1)
>>> b
[1]

It's pretty clear that the assignment creates a reference to an object,
rather than having a variable contain the value.  It should be equally
clear that the dereferencing is completely automatic -- so automatic, in
fact, that you *can't* get at the reference.

Saying "binding" instead of "reference" helps to remember the critical
semantic differences in Python's object model.  It also helps when one
is trying to explain things like this:

def f(a, x=[]):
    x.append(a)
    print x

f(1)
f('foo')
L=[5]
f(9, L)
f(23)
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From peter at engcorp.com  Mon Jun 17 21:05:55 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 21:05:55 -0400
Subject: NEWbie asks "is Python a good tool for automated file/folder 
 generation?"
References: <mailman.1024073951.24787.python-list@python.org>
Message-ID: <3D0E8773.A1BA95ED@engcorp.com>

Mark McEahern wrote:
> 
> > I've never used Python, and I'm wondering if you, collectively, think
> > that Python is a good language to do this.
> 
> Sure.  I'd recommend starting with simple things:
> 
> Fire up Python and do this:
> 
> >>> import os
> >>> path = "c:\type your path here"
> >>> print os.listdir(path)

I realize this was just a pseudo-example, but I think we should
point out for the pre-newbie :-) that the above has a bug.

Using the backslash in a string like that can cause mysterious
problems when it produces special characters by combining with
the following letters as escape codes.  The above string, for 
example, has no backslash in it, though it does have a TAB 
character which has somehow appeared after the "c:" and before 
the word "ype" ...  :-)

Better to learn to use raw-strings, with an "r" in front,
as in path = r"c:\this is a safe path" or better yet, 
either forward slashes always: path = "c:/this works too".

-Peter


From e9226414 at nospam.stud4.tuwien.ac.at  Mon Jun 17 09:45:05 2002
From: e9226414 at nospam.stud4.tuwien.ac.at (Zsolt Marx)
Date: Mon, 17 Jun 2002 13:45:05 GMT
Subject: non blocking socket.accept()
Message-ID: <qeokea.b6q.ln@neo.fdns.net>

I far as I know socket.accept() waits for a client
to ask for a connection with the server.

This command seems to be blocking, though I would
like to limit the wait time onto a predefined interval
and to throw an error, if there is no connection.

How can I achieve this?

Zsolt Marx




From max at alcyone.com  Wed Jun 19 04:30:58 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 19 Jun 2002 01:30:58 -0700
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com> <mailman.1024460666.24801.python-list@python.org> <B8XP8.47925$GY.15259265@twister.nyroc.rr.com>
Message-ID: <3D104142.8948754@alcyone.com>

George Hester wrote:

> I don't want to learn the language.  At least not right now.  This
> started
> becuase I read there was a new version of Python.  Then I saw an ASP
> at
> Microsoft that used Python.  So I thought I would run an ASP to
> determine my
> version of Python.  And then I thought I would ask how to do that
> here.

If you don't want to learn the language, what possible use could knowing
what version you're running do you?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Who'd ever think it / Such a squalid little ending
\__/ The American and Florence, _Chess_
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From lljy at students.cs.mu.oz.au  Fri Jun 28 00:26:06 2002
From: lljy at students.cs.mu.oz.au (Lars Yencken)
Date: Fri, 28 Jun 2002 14:26:06 +1000
Subject: Plugins in Python
References: <afgeol$sgd$1@mulga.cs.mu.OZ.AU> <3D1BD350.8F0C74FF@alcyone.com>
Message-ID: <afgogu$das$1@mulga.cs.mu.OZ.AU>

Erik Max Francis wrote:

> Lars Yencken wrote:
> 
>> I'm trying to do plugins for a program I'm writing in Python, and I'm
>> not
>> sure how exactly to do it. Basically what I want it to be able to
>> write a
>> module which loads every python file in a directory and stores an
>> instance
>> of each one that loaded successfully.
> 
> The most obvious implementation is simply to have them each imported as
> modules with __import__, and define some kind of simple
> initialization/shutdown API that will be called upon import.
> 

Thanks, it worked well to just use __import__ and have some standard module 
items.

Lars


From david at dataovation.com  Fri Jun 28 09:08:52 2002
From: david at dataovation.com (David McInnis)
Date: Fri, 28 Jun 2002 06:08:52 -0700
Subject: Dialing out on MODEM (Partially Resolved)
In-Reply-To: <3D1C5B65.6CBA96A3@engcorp.com>
Message-ID: <007901c21ea4$f3b60f60$221d7041@bigdaddy>

Peter,

Thanks for the tip.  I forgot that I am on Windows and needed \r\n
rather than just \n

I get the modem to dial, now how do I get it to dial a number when the
phone is picked up.  I am dialing a pager.  When it picks up I need to
outdial a numeric message.

David McInnis

-----Original Message-----
From: python-list-admin at python.org [mailto:python-list-admin at python.org]
On Behalf Of Peter Hansen
Sent: Friday, June 28, 2002 5:50 AM
To: python-list at python.org
Subject: Re: Dialing out on MODEM

David McInnis wrote:
> 
> I need to write a small script that opens com3 and writes an ATDT to
> dial out, answer the line and send a numeric message.  Basically, I
> paging utility.
> 
> I am using pyserial and cannot seem to get my modem to dial with
> 
> Modem.write("ATDT5555555")

Is Modem an object from pyserial, or one of yours?  And does it
automatically append the \n or CR or do you have to do that yourself?

-Peter
-- 
http://mail.python.org/mailman/listinfo/python-list





From phawkins at connact.com  Wed Jun 19 23:26:19 2002
From: phawkins at connact.com (Patricia J. Hawkins)
Date: 19 Jun 2002 23:26:19 -0400
Subject: How can I make a screenshot using BitBlt or something else???
References: <6ac1ecad.0202151338.15d3c4b@posting.google.com> <3C6D93F0.8F83FA9E@olen.to> <6ac1ecad.0202180323.e24dcfe@posting.google.com>
Message-ID: <wkhejyockk.fsf@mail.connact.com>

>>>>> "IC" == I C <theivo at abv.bg> writes:

IC> Joonas Paalasmaa <joonas at olen.to> wrote in message news:<3C6D93F0.8F83FA9E at olen.to>...
>> "I.C." wrote:
>> > 
>> > I am new in PythonWin programming so I have a simple question -
>> > How can I make a screenshot of the whole screen and then get the colors
>> > of some pixels?
>> 
>> Use Fredrik Lundh's grabscreen utility for PIL.
>> http://effbot.org/pil/index.htm

IC> Thank you very much, this seems to work, but do you know how can I
IC> make the
IC> screenshot using win32ui - on the system, where I want to run the
IC> script I won't be able to install the PIL-Module(SysAdmin-Rules).

Can't you just put the PIL directory in the same location as you put
your script?  I don't think it needs any more setup than just being
somewhere on your PYTHONPATH -- and the local directory always is.

-- 
Patricia J. Hawkins
Hawkins Internet Applications, LLC





From aahz at pythoncraft.com  Mon Jun 24 22:58:05 2002
From: aahz at pythoncraft.com (Aahz)
Date: 24 Jun 2002 22:58:05 -0400
Subject: Suggestions for good programming practices?
References: <mailman.1024957832.5633.python-list@python.org>
Message-ID: <af8m7t$rv5$1@panix1.panix.com>

In article <mailman.1024957832.5633.python-list at python.org>,
David LeBlanc <whisper at oz.net> wrote:
>
>Does code run at global scope pay a global scope lookup cost?

Yup.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From bbeck13 at excite.com  Tue Jun 11 23:11:48 2002
From: bbeck13 at excite.com (Brandon Beck)
Date: Wed, 12 Jun 2002 03:11:48 GMT
Subject: Sorting list (maybe stupid)
References: <mailman.1023845235.5315.python-list@python.org>
Message-ID: <3D06BBC4.3050901@excite.com>

> In fact it's almost certain <wink>.  Small lists are handled by binary
> insertion sort, for peak speed, and the way that was coded just so happens
> to be stable.  That's why small examples will always lead you to believe
> .sort() is stable.  The full-blown samplesort used for non-trivial lists
> isn't stable and can't be made stable efficiently.
> 

I personally like Alex Martelli's version of a stable sort using the 
decorate-sort-undecorate idiom.  His recipe for it can be found at:

	http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52234


The meat of his recipe is something along the lines of:

	def stable_sort_copy(alist):
		decorated = zip(alist, xrange(len(alist)))
		decorated.sort()
		return [ item for item, index in decorated ]

	def stable_sort_inplace(alist):
		alist[:] = stable_sort_copy(alist)


The efficiency of this implementation should be on the order of 
O(2*n+n*lg(n)) which we all know is O(n*lg(n)).  It's definitely not the 
most efficient way in the world to perform a stable sort, especially in 
terms of space, but it's short and sweet and lets you worry about more 
important things.  If it turns out that sorting is truly a bottleneck in 
your program, then revisit the sort for a more optimal version.


From Aristotle_00 at yahoo.com  Thu Jun 27 11:46:25 2002
From: Aristotle_00 at yahoo.com (Mark)
Date: Thu, 27 Jun 2002 15:46:25 GMT
Subject: "multi" zip
Message-ID: <lrGS8.13892$bt.307@nwrddc04.gnilink.net>

So,

Let line be a list of lists where the sublists are the same length.

For example:

line = [ [0,1,2], [3,4,5], [6,7,8] ]

Now, I can do:

zip(line[0], line[1], line[2])

to "invert" the array.  Now, I'd like to do two different things:

1) instead of producing tuples, I'd like to produce lists.

2) I'd like to extend this to an arbitrary number of lists.  If it were 
possible, I'd do zip(line[0], line[1], ..., line[n]) but this doesn't seem 
to compile :).

Is there a way to make zip() do this, or should I stop trying to hammer with 
a wrench?

Regards,
Mark
           


From marklists at mceahern.com  Mon Jun 17 19:21:44 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 17 Jun 2002 18:21:44 -0500
Subject: inherrited private class varriables
In-Reply-To: <MPG.17788662b0a1b67a98968c@news.rz.uni-karlsruhe.de>
Message-ID: <NCBBLFCOHHDIKCAFGCFBAEBJLFAA.marklists@mceahern.com>

> why can't i access a class private varriable from within a predecessor?

You can, but you have to use the mangled name:

  mangled_name = "_%s%s" % (self.__class__.__name__, "__length")
  l = getattr(self, mangled_name, self.__length)
  print l

Of course, if you don't want to torture yourself, you could just use _length
instead.  ;-)

Cheers,

// mark

-





From fredrik at pythonware.com  Thu Jun 20 13:48:20 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 20 Jun 2002 17:48:20 GMT
Subject: do...while
References: <slrnah40l1.12v.msoulier@pmerd071.ca.nortel.com>
Message-ID: <EzoQ8.12922$p56.4028292@newsb.telia.net>

Michael P. Soulier wrote:
>     How do you people handle the lack of a do...while in Python? I find that
> at times, I must initialize variables for a loop with identical code to the
> loop, which feels like a waste to me.

the standard pydiom is:

    while 1:
        do stuff
        if condition:
            break
        do stuff

</F>




From whisper at oz.net  Mon Jun 24 04:12:35 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 24 Jun 2002 01:12:35 -0700
Subject: Newbie question (anyone programming with curses?)
In-Reply-To: <004901c21b2a$31254e20$0101010a@local>
Message-ID: <GCEDKONBLEFPPADDJCOECELJDLAA.whisper@oz.net>

Forgive me if i'm incorrect - it's been awhile since I mucked with curses.

Isn't the purpose of curses to convert ANSI standard codes into whatever a
particular terminal understands? If so, consoles on NT/2k/XP should all do
cursor positioning , blinking, color control etc. directly IIRC. Even
Win95/98/ME will do it if ansi.sys is loaded.

Another alternative is S-Lan and PySlan, which is platform independent and
unencumbered I believe. I have built and gotten both running on Win 2k.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Chris Gonnerman
> Sent: Sunday, June 23, 2002 19:52
> To: python-list at python.org
> Subject: Re: Newbie question (anyone programming with curses?)
>
>
> ----- Original Message -----
> From: "Emile van Sebille" <emile at fenx.com>
>
>
> > hellboy
> > > I'm quite new to Python, but I'm positive that version 2.2 (for
> > Windows)
> > > comes without curses module, though documentation states it is
> > included.
> > > import curses gives me "Import Error: no module named curses" message.
> > > Anyone knows where to download curses module (for Windows)?
> > >
> >
> > from http://py-howto.sourceforge.net/curses/node2.html
> >
> > No one has made a Windows port of the curses module. On a Windows
> > platform, try the Console module written by Fredrik Lundh. The Console
> > module provides cursor-addressable text output, plus full support for
> > mouse and keyboard input, and is available from
> > http://effbot.org/efflib/console.
> >
> > Hope this is current and helps,
>
> I found the effbot's version a bit unstable on Windows 9x (a lot actually)
> and I wound up writing WConio, based on the TCCONIO.C sources for MinGW32.
>
>     http://newcenturycomputers.net/projects/wconio.html
>
> but it's still not curses.  I have considered many times creating a GConio
> (Generic Console I/O) module wrapping both WConio on Win32 and curses on
> Unixoid OS's.  Haven't done much other than think about it though.
>
>
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From martin at v.loewis.de  Fri Jun  7 02:11:09 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 07 Jun 2002 08:11:09 +0200
Subject: How do you read unicode files?
References: <adpd7h$6sv$1@slb3.atl.mindspring.net>
Message-ID: <m3vg8vaa9u.fsf@mira.informatik.hu-berlin.de>

"Matt Gerrans" <mgerrans at mindspring.com> writes:

> How do you read in a unicode file and convert it to a standard string?

Depends on what you mean by "unicode file". A file always uses a
byte-oriented encoding; the term "unicode file" does not imply byte
orientation. People usually associated UTF-8 and UTF-16 (with and
without BOM) with that term.

> It seems that when you open a file and read it, what you get is a
> string of single-byte characters.  I've tried all kinds of
> permutations of calls to unicode(), decode(), encode(), etc. with
> different flavors of encoding ('utf-8', 'utf-16' and so on).

The best way to read such a file is to use codecs.open. If you already
have byte data, _decoding_ them to Unicode, with the unicode builtin,
is the right thing to do.

> I could parse the data myself (skipping the initial two bytes and then every
> other one -- I'm only working with ASCII in double byte format, so the high
> order byte is always 0), but I imagine there must be a way to get the
> existing tools to work.

Ah, so there are two initial bytes. That most certainly means that you
have a file in UTF-16 with BOM; so you should use "utf-16" as the
encoding name (without BOM, you need to know the byte order, and
specify utf-16be or utf-16le respectively).

> What I want to be able to do is write a search and replace tool that
> will work equally well on ANSI and Unicode (or double-byte) text
> files (without changing the file type, of course)...

Then codecs.open is the right solution; for "ANSI" files, the encoding
is "mbcs" (neither the term "ANSI" nor the term "mbcs" are
particularly well-chosen to describe what they do).

Regards,
Martin


From sam at webslingerZ.com  Tue Jun 18 15:59:38 2002
From: sam at webslingerZ.com (Sam Brauer)
Date: Tue, 18 Jun 2002 15:59:38 -0400 (EDT)
Subject: cannot unmarshal code objects in restricted execution mode
Message-ID: <Pine.LNX.4.44.0206181541180.1527-100000@localhost.localdomain>

In the context of a long-running Python process (specifically mod_python) 
I occasionally get the following error:

exceptions.RuntimeError: cannot unmarshal code objects in restricted 
execution mode
File "/usr/local/lib/python2.2/encodings/__init__.py", line 51, in 
search_function
mod = __import__(modname,globals(),locals(),'*')

This happens when I call code like this:

utf8string = unicode(val, enc).encode('utf-8')

where 'val' is a string in an encoding identified by the string 'enc'.

What's strange is that the vast majority of the time the error doesn't 
occur, but after it happens once, it seems to happen more frequently 
(although not consistently).

Does anyone know what might be the cause of this error?

-- 
Sam Brauer
Systems Programmer
sam at webslingerZ.com






From martin at v.loewis.de  Thu Jun  6 16:34:06 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 06 Jun 2002 22:34:06 +0200
Subject: creating an instance of a C extension type from another C extension type  (HOW?)
References: <3CFFB567.435254@cpsc.ucalgary.ca>
Message-ID: <m34rggw3i9.fsf@mira.informatik.hu-berlin.de>

Maxwell Sayles <sayles at cpsc.ucalgary.ca> writes:

> is there anything wrong with my logic?  code?

Most likely, there is something wrong with your code. Notice that you
cannot normally share symbols between extension modules. Instead, if
you want two modules to share symbols, you need to generate a shared
library that offers the shared symbols, and link both extension
modules against that shared library.

Regards,
Martin


From burkhard.kloss at kbcfp.com  Thu Jun 27 12:07:25 2002
From: burkhard.kloss at kbcfp.com (news)
Date: Thu, 27 Jun 2002 17:07:25 +0100
Subject: Integrals with Python
References: <3D188551.6BA00DED@lanl.gov>
Message-ID: <1025194045.882109@master.nyc.kbcfp.com>

http://pylab.sourceforge.net/packages/quadrature.py seems to do exactly what
you want.




From whisper at oz.net  Sat Jun 22 15:17:08 2002
From: whisper at oz.net (David LeBlanc)
Date: Sat, 22 Jun 2002 12:17:08 -0700
Subject: Python hits the spot
In-Reply-To: <3D1482AE.8FA299D8@kfunigraz.ac.at>
Message-ID: <GCEDKONBLEFPPADDJCOEIEILDLAA.whisper@oz.net>

<snip>
> I cannot believe that Linux is responsible for Python's ill behavior
> under Linux. Before you start ranting about note the following: In a
> first step it has nothing to do with my programming style (admittedly
> not very good and famous) it has to do that the same calculation cannot
> be performed on Linux and Python 2.2!
>
>
> S. Gonzi

Problems of such magnitude on known good systems are generally corrected by
tightening the wingnut between the keyboard and the seat. (Note: your nut
may not have wings.)

Dave LeBlanc
Seattle, WA USA





From gerhard.haering at gmx.de  Wed Jun 19 18:55:32 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Thu, 20 Jun 2002 00:55:32 +0200
Subject: Windows versions of Python---pros and cons?
In-Reply-To: <3D110B55.4D8AD7FA@astro.cornell.edu>
References: <3D110B55.4D8AD7FA@astro.cornell.edu>
Message-ID: <20020619225532.GA12907@lilith.my-fqdn.de>

* Tom Loredo <loredo at astro.cornell.edu> [2002-06-19 18:53 -0400]:
> [Pythonlabs and ActiveState Python]  Are their substantial differences
> between the versions that I will bump into down the line?

No, none.

> In particular, I write many C extensions; can one use distutils and
> the Borland or gcc free compilers to build C extensions in both
> versions?

Yes, they're the same Python, ActiveState has just packaged a few more
goodies, and have a nicer installer. You can, however, manually get all
the extras that ActiveState packages from several places at the net.
ActiveState has just the advantage that they've already done that job
for you.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From occeanlinux at linuxmail.org  Sun Jun  2 15:20:38 2002
From: occeanlinux at linuxmail.org (Gold Fish)
Date: Sun, 02 Jun 2002 19:20:38 GMT
Subject: List
Message-ID: <3cfa6ed8_1@news.iprimus.com.au>

Assume we got 2 list 
list 1 = ['ABC','CDG','DAV','FE','FECA','FEAFA','FEA']
list 2 = []
how can we copy element from list 1 to list 2 say i just want to copy the 3 
element 'CDG','DAV','FE' to list2 only.


From jjbbox2002 at yahoo.com  Sun Jun 23 20:27:52 2002
From: jjbbox2002 at yahoo.com (Jay)
Date: 23 Jun 2002 17:27:52 -0700
Subject: File IO read returning a string -> but I want a byte -> Question: Is there a faster more efficient way?
Message-ID: <3f8faafc.0206231627.570b8602@posting.google.com>

#Question: Is there a faster more efficient way?

from time import clock

len_data = 20*1024*1024
data = fp.read(len_data)

start = clock()
for i in range(len_data): 
   norm[i] = struct.unpack("B",data[i])
finish = clock()


From tripie at cpan.org  Sat Jun 15 00:01:01 2002
From: tripie at cpan.org (Tomas Styblo)
Date: Sat, 15 Jun 2002 04:01:01 +0000 (UTC)
Subject: Using Python to script a game?
References: <cdc2589b.0206141718.7dc2ae13@posting.google.com> <mailman.1024106376.21850.python-list@python.org>
Message-ID: <slrnagleld.d2l.tripie@tripie.tripie>

In article <mailman.1024106376.21850.python-list at python.org>, 
Gerhard H?ring wrote:

> Yes. And it would be a pretty good match, too. There's even been a
> commercial game that uses Python for internal scripting, its name
> escapes me, though.

Name of the game is Blade. Maybe there exist more games using Python
now, though.

-- 
Tomas Styblo
tripie at cpan.org


From 98barret at utsc.utoronto.ca  Sun Jun 23 23:55:17 2002
From: 98barret at utsc.utoronto.ca (Paul Barrett)
Date: Sun, 23 Jun 2002 23:55:17 -0400
Subject: Python HTML Data Collection
Message-ID: <Pine.GSO.4.33.0206232352240.18910-100000@fissure.scar.utoronto.ca>

Hi,

   I'm fairly new to Python but I do have a little bit of experience from
a class I took which taught some Python. I'm working at a job where I
query a CGI script, collect the data and then post it to another CGI
script. My main question is whether I could write a python script which
would automate this process. Both of these scripts are on external
websites and I only need a select amount of the data. I thought that if
any language would do this job, then it would be Python, but I haven't
been able to find any information for something like this. Do you folks
have any suggestions?

Thanks,

Paul Barrett



From wurmy at earthlink.net  Sun Jun 30 00:07:30 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Sun, 30 Jun 2002 04:07:30 GMT
Subject: newbie Q: delete list few lines in file
References: <7fe97cc4.0206291519.1c95ce31@posting.google.com> <3D1E5123.FDEFADDF@engcorp.com>
Message-ID: <3D1E83C0.8060700@earthlink.net>

Peter Hansen wrote:

>>* how to write to file inplace? (or, do i write to a new file, then
>>delete original, and rename the new file?)
> 
> You cannot write files in place.  

<nitpick>

Sure you can:

   f = open("foobar.txt", "w")
   f.write("foo!"*5)
   f.close()

   g = open("foobar.txt", "r+")
   g.seek(0, 0)
   g.write("bar!")
   g.close()

   print open("foobar.txt", "r").read()

prints

   bar!foo!foo!foo!foo!

</nitpick>

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/



From fperez528 at yahoo.com  Tue Jun 18 20:58:41 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 18 Jun 2002 18:58:41 -0600
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <yXBP8.44417$GY.13865423@twister.nyroc.rr.com>
Message-ID: <aeol0b$9p2$1@peabody.colorado.edu>

George Hester wrote:

> Listen you aren't my daddy.  If you do not want to help constructively then
> please get off my boat.
> 

By responding in this way to what was a perfectly polite, informative and 
correct post from someone who's well known for useful contributions in this 
group, you've no doubt earned yourself a place of honor in a lot of people's 
killfiles (mine included). Good luck in getting any further help from this 
group.

cheers,

f.


From blah at calico-consulting.com  Sat Jun  8 01:45:58 2002
From: blah at calico-consulting.com (Mr Blah)
Date: 7 Jun 2002 22:45:58 -0700
Subject: no module named windowsc?
Message-ID: <3b738180.0206072145.1c4a655b@posting.google.com>

I installed python 2.2 for windows on my windows 2000 professional
machine and also installed fxpy for python 2.2.

When I try to run any of the example programs that come with fxpy I
get the same error - a complaint about windowsc:

C:\Python22\FXPy>python windows.py
Traceback (most recent call last):
  File "windows.py", line 2, in ?
    import windowsc
ImportError: No module named windowsc

Searching on windowsc I can't seem to come up with much of anything.
Can anyone give me a clue about what might be going on?

Mr Blah
blah at calico-consulting.com


From tjreedy at udel.edu  Sun Jun  2 16:22:45 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 02 Jun 2002 20:22:45 GMT
Subject: What is a classmethod for?
References: <addgv30f4d@enews4.newsguy.com>
Message-ID: <p8vK8.102095$%o.9916157@bin3.nnrp.aus1.giganews.com>

"Dennis Peterson" <denpeterson at yahoo.com> wrote in message
news:addgv30f4d at enews4.newsguy.com...
> class C:
>     def foo(cls):
>         print cls
>     foo = classmethod(foo)
...
> That's nice, but I can't think of a use for it, given that we have
> inheritance and isinstance(). Can anyone enlighten me?

Suppose, more realistically, that .__init__() adds each instance to a
list of class instances.  Then a reasonable classmethod would be one
that does something with that class attribute.

TJR





From mwh at python.net  Thu Jun 13 05:42:10 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 13 Jun 2002 09:42:10 GMT
Subject: py2exe 4 linux?
References: <3D08563C.33D041E3@gol.ge>
Message-ID: <lkwut3wm5m.fsf@pc150.maths.bris.ac.uk>

Giorgi Lekishvili <gleki at gol.ge> writes:

> Hi all!
> 
> Is there anything like py2exe for Linux?

I think a there is a version of Gordon McMillan's installer for linux.

Cheers,
M.

-- 
  I'm a keen cyclist and I stop at red lights.  Those who don't need
  hitting with a great big slapping machine.
                                           -- Colin Davidson, cam.misc


From dbasch at yahoo.com  Sat Jun 29 15:45:43 2002
From: dbasch at yahoo.com (Derek Basch)
Date: Sat, 29 Jun 2002 12:45:43 -0700 (PDT)
Subject: Win32/Python serial port event character problem
Message-ID: <20020629194543.25476.qmail@web20804.mail.yahoo.com>

Sorry to post this again but it didn't get any
responses. Maybe better luck this time.

I use the following code to read data from the serial
port when a certain event character ('\n') is
received:
http://www.geocities.com/dbasch/temp.txt

It seems to work fine except for that when I send the
following data via the serial port:
http://www.geocities.com/dbasch/testdata.txt

the output is:

Comport Thread:test 1
Comport Thread:
test 2
te
Comport Thread:st 3
tes
Comport Thread:t 4
test 5

It seems to me that in the time between the event
character being received and the ReadFile operation
occurring additional characters are being placed into
the windows OS serial buffer.
Has anyone ever dealt with this? Is using the event
character not a reliable method in windows?
I know the event character is being triggered because
I can change the event character to 'g' and nothing
happens unless a 'g' is in the received data.
I have obviously left out some of the code in the
example but I hope everyone will understand it ok.
This is driving me batty so any help is greatly
appreciated.
Thanks,
Derek Basch
1 very unskilled win32 programmer.


__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com




From mgerrans at mindspring.com  Sun Jun 16 16:37:22 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Sun, 16 Jun 2002 13:37:22 -0700
Subject: Newbie Help Please.(OO and Tkinter)
References: <B93209DC.7281%sarmstrong13@mac.com>
Message-ID: <aeit4p$h0s$1@slb5.atl.mindspring.net>

Did you intend to be defining the method apply() inside the method savein()
or did accidentally get that extra indentation?

>     def savin(self, question):
>         Label(question, text="Directory:").grid(row=0)
>         Label(question, text="Filename:").grid(row=1)
>         self.e1 = Entry(question)
>         self.e2 = Entry(question)
>         self.e1.grid(row=0, column=1)
>         self.e2.grid(row=1, column=1)
>         def apply(self):
>             self.dir = self.e1.get()
>             self.fn = self.e2.get()
>             self.sav = self.dir + self.fn
>             self.savfile = open(self.sav, "w")
>             for line in self.t1.readlines():
>                 self.savefile.write(line)
>             self.savefile.close()





From juszczak at ptd.net  Sun Jun  9 15:10:50 2002
From: juszczak at ptd.net (Stan)
Date: Sun, 09 Jun 2002 19:10:50 GMT
Subject: Newbie -  Python21.dll not found 
Message-ID: <_KNM8.11799$Q43.303520@nnrp1.ptd.net>

Hi,
        When I try to start up python a message "The dynamic link library
python2.dll could not be found in the specified path". The app was installed
along with other apps when  I have installed the Foxserv 2.1 on W2k system.

Thanks in advance,

Stan




From phd at phd.pp.ru  Wed Jun  5 15:11:31 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 5 Jun 2002 23:11:31 +0400
Subject: Urllib thread safe?
In-Reply-To: <m3it4xobkb.fsf@mira.informatik.hu-berlin.de>; from martin@v.loewis.de on Wed, Jun 05, 2002 at 07:55:00PM +0200
References: <mailman.1023280808.19305.python-list@python.org> <m3it4xobkb.fsf@mira.informatik.hu-berlin.de>
Message-ID: <20020605231131.A30904@phd.pp.ru>

On Wed, Jun 05, 2002 at 07:55:00PM +0200, Martin v. Loewis wrote:
> Yes, that is thread-safe. urllib does not maintain any per-url state,

   Hah? It maintains global _urlopener...

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From spammers.do.not.bother at void.bogus  Mon Jun  3 11:25:37 2002
From: spammers.do.not.bother at void.bogus (Magnus)
Date: Mon, 03 Jun 2002 15:25:37 GMT
Subject: Automated Email
References: <adg1do$10nmr5$1@ID-49758.news.dfncis.de>
Message-ID: <RTLK8.41545$n4.9322089@newsc.telia.net>

Ken wrote:

> How to do automated emailing when results being processed?
> 
> Thanks


1.) enter the python interpreter by e.g. typing "python"
2.) type "import smtplib"
3.) type "help(smtplib)"

Good luck!



From kragen at pobox.com  Sun Jun  2 22:51:48 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 02 Jun 2002 22:51:48 -0400
Subject: HTML links in a GUI tkinter
References: <mailman.1022986323.16196.python-list@python.org>
Message-ID: <83heklrs57.fsf@panacea.canonical.org>

aurel-le-gnou <aurelienviolet at noos.fr> writes:
> How can I put in a frame or in any widget (?) an html link so that the
> browser opens at the url indicated when the user clicks ?
> Just simply as if it were an HTML link in an HTML document but in a
> tkinter GUI ?
> 
> What are the packetages to import ?
> In what king of widget can I put such a link ?
> What kind of lines of code should i put in this widget the link to be
> active ?
> 
> 
> Thanks a lot for your help
> 
> 
> Aur?lien Violet
> aurelienviolet at noos.fr



From emile at fenx.com  Sat Jun  1 10:53:28 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 01 Jun 2002 14:53:28 GMT
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <06da01c208f4$d69003c0$6601a8c0@boostconsulting.com> <200206010121.g511LxX19223@pcp742651pcs.reston01.va.comcast.net> <073501c20960$23f367e0$6601a8c0@boostconsulting.com> <200206011334.g51DY1D21669@pcp742651pcs.reston01.va.comcast.net> <07d801c20970$cdebe050$6601a8c0@boostconsulting.com> <mailman.1022940122.3421.python-list@python.org>
Message-ID: <Id5K8.171932$Po6.340972@rwcrnsc52.ops.asp.att.net>

<jepler at unpythonic.net> wrote:
> IMO the following is an odder behavior:
> >>> t
> ([1], [2], [3])
> >>> t[1] += [3]
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> TypeError: object doesn't support item assignment
> >>> t
> ([1], [2, 3], [3])
> 

Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
>>> class myTuple(tuple):
...     def __setitem__(self, attr, value):
...             print 'in setitem:', attr
...
>>> a = myTuple(([1],[2],[3]))
>>> a
([1], [2], [3])
>>> a[1] += [4]
in setitem: 1
>>> a
([1], [2, 4], [3])
>>>


Here's another that's differently wrong:
>>> a = ('aaa',)
>>> a[0]+='bbb'
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: object doesn't support item assignment
>>> a
('aaa',)
>>>




-- 

Emile van Sebille
emile at fenx.com

--------- 



From erict at millfilm.co.uk  Fri Jun 28 07:59:44 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Fri, 28 Jun 2002 12:59:44 +0100
Subject: Be the first on your block to register your computer.
References: <mailman.1025217276.12118.python-list@python.org>
Message-ID: <3D1C4FB0.A443ACF6@millfilm.co.uk>

Now, I have a good excuse to keep stacking my old
computers in the garage.

Gustavo Cordova wrote:

> >
> > "David LeBlanc" <whisper at oz.net> wrote in message
> > news:mailman.1025211444.21440.python-list at python.org...
> > > Yes! exciting times are coming again as Intel, AMD and
> > > Microsoft (among quite a few others) are banding together
> > > to once more put ID numbers (and now basic encryption)
> > > onto next generation silicon.
> >
> > Not that I _want_ to sound crabby, but: this has relevance to
> > Python _how_ exactly?
> >
> > James
> >
>
> Well, for one, Python is one of those evil Open Sores
> software viruses, unstable, dangerous to the health of
> your machine.
>
> You can bet your snickers that Python will not be one of
> those "M$ Approved" software packages that you may freely
> install, once this comes through.
>
> -gus



From noone at here.invalid  Tue Jun 25 14:34:09 2002
From: noone at here.invalid (JohnJacob)
Date: Tue, 25 Jun 2002 13:34:09 -0500
Subject: Python & zipfiles
References: <uhgfcopr269e0@corp.supernews.com>
Message-ID: <afad1s02hbs@enews1.newsguy.com>

"Patrick Bielen" <bielen at stafa.nl> wrote in message
news:uhgfcopr269e0 at corp.supernews.com...
> Hello all,
>
> I use a module to work with zipfiles.
> This one is only for unpacking files
> from the zip. But i only want to know
> if i have to much code to do the trick.
>
> I'm barely new to python, that's why i ask.
>
> --- begin script
>
> import os
> from zipfile import ZipFile
>
> zipdir = '/as400/Beheer/Databases/'
> os.chdir('/tmp')
> zipfiles = ZipFile(zipdir + 'dbs.zip').namelist()
> ZipFile(zipdir + 'dbs.zip','r')
> for file in zipfiles:
>     print file
>     source = ZipFile(zipdir + 'dbs.zip').read(file)
>     open(file, 'w').write(source)
> ZipFile(zipdir + 'dbs.zip').close()
>
> --- end of script ---
>
> The problem is, that i don't know if i use the
> ZipFile-command too much... because the manual
> told me i have to close it afterwards, but it's
> not clear to me if i have to expecitely open it.
>
> Any help or code cleaning would be really appre-
> ciated.
>
> Greetings
>
> Patrick

This is how I would rewrite it:

import os
from zipfile import ZipFile

zipdir = '/as400/Beheer/Databases/'
os.chdir('/tmp')

zipfile = ZipFile(os.path.join(zipdir, 'dbs.zip'))
for file in zipfiles.namelist():
    print file
    source = zipfile.read(file)
    open(file, 'w').write(source)

zipfile.close()



greg




From bokr at oz.net  Sun Jun 23 02:08:37 2002
From: bokr at oz.net (Bengt Richter)
Date: 23 Jun 2002 06:08:37 GMT
Subject: ConfigParser & converting strings to lists
References: <3D152A28.2177B054@tds.net>
Message-ID: <af3ol5$rn9$0@216.39.172.122>

On Sun, 23 Jun 2002 01:53:44 GMT, "Edward K. Ream" <edream at tds.net> wrote:

>Hi all,
>
>I have a question concerning lists in the ConfigParser module.  What I
>would like to do is something like:
>
>config = ConfigParser.ConfigParser()
>config.set("recent files", "recentFiles, files)
>
>where files is a _list_ of file names.  Actually, this works just fine,
Are you sure? I wish you'd just copy and paste from a real interactive session.
It's not really that much more typing ;-)
For 2.2 on windows, I get

 >>> import ConfigParser
 >>> config = ConfigParser.ConfigParser()

Suppose this is your file name list:
 >>> files = ['file1','file2','file3']

You need a section, or you'll get an exception:
 >>> config.add_section('recent files')

Plain files didn't work for me, but the string representation from `files` did:
 >>> config.set("recent files", "recentFiles", `files`)

Which came back as expected:
 >>> config.get("recent files", "recentFiles")
 "['file1', 'file2', 'file3']"

And they try hard to make x == eval(`x`), so to get the actual list, you can:
 >>> eval(config.get("recent files", "recentFiles"))
 ['file1', 'file2', 'file3']
 
Unless the security risk is unacceptable, which it probably is, and you might want
to do something safer. Knowing the repr format for a list of strings, and assumung
the config.set got `alist` as the value, you could, e.g.,
 >>> theFileList = config.get("recent files", "recentFiles")[2:-2].split("', '")
 >>> theFileList
 ['file1', 'file2', 'file3']
But that's pretty brittle, and won't tolerate hand editing very well, so maybe
after checking for '[' to decide it's a list, get the names something like
 >>> [x.strip()[1:-1] for x in theFileList.strip()[1:-1].split(',')]
 ['file1', 'file2', 'file3']


>and the config file contains:
>
>[recent files]
>recentFiles = [file1, file2, ...]
If it really does, it wasn't generated using backquotes like `files`.

>
>as expected.  However,
>
>files = config.get("recent files", "recentFiles")
>
>sets files to the _string_ "[file1, file2, ...]".  I suppose this must
>be how get should work; after all, settings could be arbitrary
>strings...
Again, that wouldn't have come from `files`
>
>Anyway, I wonder whether there is a clever way of converting the string
>representation of a list containing strings into a true list of
>strings.  In any event, wouldn't a getlist convenience method be nice?
>
Ok, here's a go at it (always returns a list):

 >>> class X(ConfigParser.ConfigParser):
 ...     def getlist(self,section,option):
 ...         s = self.get(section, option)
 ...         if '[' in s:
 ...             s = [x.strip()[1:-1] for x in s.strip()[1:-1].split(',')]
 ...         else:
 ...             s = [s.strip()]
 ...         return s
 ...
 >>> configx = X()
 >>> configx.add_section('recent files')
 >>> configx.set("recent files", "recentFiles", `files`)
 >>> configx.get("recent files", "recentFiles")
 "['file1', 'file2', 'file3']"
 >>> configx.getlist("recent files", "recentFiles")
 ['file1', 'file2', 'file3']
 >>> configx.set("recent files", "recentSingleFile", 'a_single_file_name')
 >>> configx.get("recent files", "recentSingleFile")
 'a_single_file_name'
 >>> configx.getlist("recent files", "recentSingleFile")
 ['a_single_file_name']

You could override set so it would automatically do backquotes if the value arg wasn't
already a string, and then call the base class with the resulting string.

And whatever ...

Regards,
Bengt Richter


From aahz at pythoncraft.com  Sat Jun 15 00:20:35 2002
From: aahz at pythoncraft.com (Aahz)
Date: 15 Jun 2002 00:20:35 -0400
Subject: Using Python to script a game?
References: <cdc2589b.0206141718.7dc2ae13@posting.google.com> <PTyO8.226065$%y.20663204@bin4.nnrp.aus1.giganews.com>
Message-ID: <aeefaj$6kb$1@panix1.panix.com>

In article <PTyO8.226065$%y.20663204 at bin4.nnrp.aus1.giganews.com>,
Terry Reedy <tjreedy at udel.edu> wrote:
>
>Blade of Darkness, a commercial real-time 3d medieval action game, is
>scripted in Python.  The demo is probably still available (check
>google); it comes with readable .py files.

On behalf of my medievalist friends: smile when you say that, and put
medieval in quotes.

It should probably be mentioned that one of the more interesting bits of
Blade of Darkness is the realistic object physics.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From mgilfix at eecs.tufts.edu  Mon Jun 17 20:28:22 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Mon, 17 Jun 2002 20:28:22 -0400
Subject: simple socket redirector app question
In-Reply-To: <5DrP8.25$OT1.34057@newsfeed.slurp.net>; from bedge@troikanetworks.com on Mon, Jun 17, 2002 at 01:27:44PM -0700
References: <5DrP8.25$OT1.34057@newsfeed.slurp.net>
Message-ID: <20020617202822.E17188@eecs.tufts.edu>

  I think SocketServer is a good choice for this. SocketServer is
aimed at providing a easy to drop-in infrastructure for apps that
don't require any complicated server setup. SocketServer would make
this quick and compact. Essentially you just want to read from one
socket and write to the other. Via inhertiance you could construct
a bridge class that contains two socket servers and talks between
them...

                   -- Mike

On Mon, Jun 17 @ 13:27, Bruce Edge wrote:
> I have an app which talks via sockets to local host clients only.
> I'd like to get at it remotely.
> 
> Basically it would accept remote connections from a remote client, 
> relay them the the local app, and forward the replies back to the 
> remote client.
> 
> What's the simplest way to write this type of a socket redirector?
> 
> Use SocketServer, or is that overkill?
> 
> Thanks, Bruce.
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (bedge)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From python at rcn.com  Wed Jun  5 22:50:55 2002
From: python at rcn.com (Raymond Hettinger)
Date: Wed, 5 Jun 2002 22:50:55 -0400
Subject: alignment excel written in Python
References: <6bd9f01b.0206050736.f469082@posting.google.com> <3CFE9861.5090200@skippinet.com.au> <Pine.GSO.4.31.0206051810190.22089-100000@staff1.cso.uiuc.edu>
Message-ID: <admips$7a7$1@bob.news.rcn.net>

> I have a questions now: How can I deal with the alignment of Excel from
> Python? for example, put the value in a cell with alignment right, or
> center.

xlLeft, xlRight, xlCenter = -4131, -4152, -4108
from win32com.client.dynamic import Dispatch
xl = Dispatch('Excel.Application')
xl.Range('A1').HorizontalAlignment = xlRight


Raymond Hettinger




From skip at pobox.com  Fri Jun  7 14:29:43 2002
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 7 Jun 2002 13:29:43 -0500
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <200206071815.g57IFjg01303@europa.research.att.com>
References: <yu99vg8v6orb.fsf@europa.research.att.com>
        <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
        <200206071747.g57HlbH01057@europa.research.att.com>
        <15616.62634.635544.905055@12-248-41-177.client.attbi.com>
        <200206071804.g57I4fp01212@europa.research.att.com>
        <15616.63342.559215.644086@12-248-41-177.client.attbi.com>
        <200206071815.g57IFjg01303@europa.research.att.com>
Message-ID: <15616.64407.863059.729351@12-248-41-177.client.attbi.com>

    $ ./python
    Python 2.2.1 (#4, Jun  7 2002, 13:24:38) 
    [GCC 3.1] on sunos5
    Type "help", "copyright", "credits" or "license" for more information.
    >>> import struct
    Traceback (most recent call last):
    File "<stdin>", line 1, in ?
    ImportError: No module named struct

    Andrew> To what would it have renamed struct.so?

Dunno.  Try "find . -name 'struct*'.  It should be in your build/lib*
directory (where "*" matches your configuration name).

S




From BPettersen at NAREX.com  Sun Jun 30 17:04:52 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Sun, 30 Jun 2002 15:04:52 -0600
Subject: How to get rid the new line
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158F14@admin56.narex.com>

> From: Frank Tobin [mailto:ftobin at neverending.org] 
> 
> Fredrik Lundh, on 2002-06-29, wrote:
> 
> > or add the single character needed to make chomp a
> > bit more robust:
> > 
> >     def chomp(line):
> >         if line[-1:] == '\n':
> >             line = line[:-1]
> >         return line
> 
> I think the fact that the first few tries here were *wrong* 
> about how to implement a chomp indicates that it really 
> should part of Python.  Just because it's easy to implement 
> doesn't mean it can't be part of the, say, string methods.  I 
> mean, look at str.{upper,lower,isdigit}, etc.  A chomp method 
> would be popular; rstrip or implementing your own doesn't cut it.

I submitted a patch for this (that handled mac, win, and unix line
endings) but it was declined by the BDFL. I suspect one of the reasons
were that I named it chomp...

-- bjorn




From nhodgson at bigpond.net.au  Fri Jun 28 06:33:12 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Fri, 28 Jun 2002 10:33:12 GMT
Subject: Be the first on your block to register your computer.
References: <mailman.1025211444.21440.python-list@python.org> <afg1ks$e6lj5$1@ID-71831.news.dfncis.de> <3D1B6E69.1090204@earthlink.net>
Message-ID: <IXWS8.368185$o66.975748@news-server.bigpond.net.au>

Ron Stephens:

> Well, Python just might not be one of the Microsoft approved programs
> allowed to run. In that case, you would not be able to use Python  at all.

   Then we'll just write Python as an Excel macro. Microsoft has over its
entire history stuck programming languages in things *even when they are a
security liability*. They show no signs of deviating from this path with
ever more code-hosting ecosystems with every Windows release.

   I suspect that a Python interpreter could be built out of a set of
Outlook message processing rules although I wouldn't be the one to volunteer
to do it whereas I will volunteer to do the Excel port if it should ever be
needed.

   Neil





From mwarden22 at hotmail.com  Thu Jun 13 14:35:25 2002
From: mwarden22 at hotmail.com (nealj)
Date: 13 Jun 2002 11:35:25 -0700
Subject: telling handled exceptions from unhandled ones
Message-ID: <d3768569.0206131035.6f2305c8@posting.google.com>

An outstanding problem with the python debugger I'm working on is
being able to tell the difference between an exception that is handled
and one that isn't.  When Hap gets PyTrace_EXCEPTION sent to its trace
function it can't reliably tell when an exception isn't handled and
stop at the correct location.  Anyone have any suggestions?


From chris.jackson at ps.ge.com  Tue Jun 25 15:29:17 2002
From: chris.jackson at ps.ge.com (Chris Jackson)
Date: 25 Jun 2002 12:29:17 -0700
Subject: File "in3sun34.py", line 33, in AttributeError: HTTPConnection
References: <mailman.1024982313.3269.python-list@python.org>
Message-ID: <f74c2878.0206251129.3381e049@posting.google.com>

Guillermo Fernandez <guillermo.fernandez at epfl.ch> wrote in message news:<mailman.1024982313.3269.python-list at python.org>...
> Hello!
> 
> Please, if you answer me, include my e-mail in the answer as I am not
> abonned to this mailing list.
> 
> I've written a little program to test python (I'm a beginner). Here it
> is:
> 
> import httplib
> 
> conn = httplib.HTTPConnection("www.python.org")
> print "CONNECTED"
>   conn.request("GET", "/index.html    
> response=conn.getresponse()
> answer=response.read()
> conn.close()
> print answer
> 
> When I execute it, I obtain an error:
> [112]host-cfernand% python test.py
> Traceback (innermost last):
>   File "test.py", line 3, in ?
>     conn = httplib.HTTPConnection("www.python  .org")
> AttributeError: HTTPConnection
> 
> Line number 3 is:
> conn = httplib.HTTPConnection("www.python.org")
> 
> I have searched in the internet and in the documentation and I have not
> been able to solve it.
> 
> On the other hand, this program it works, so it's not a network problem:
> import urllib
> sock = urllib.urlopen("http://www.python.org/index.html")
> htmlSource = sock.read()
> sock.close()
> print htmlSource
> 
> Thank you for your help!
> 
> Guille

I have tried your first block of code and get the following error
message:

Traceback (most recent call last):
  File "<string>", line 1, in ?
  File "C:\Documents and Settings\jacksoc4\Desktop\http.py", line 5,
in ?
    conn.request("GET", "/index.html")
  File "C:\Python22\lib\httplib.py", line 537, in request
    self._send_request(method, url, body, headers)
  File "C:\Python22\lib\httplib.py", line 553, in _send_request
    self.putrequest(method, url)
  File "C:\Python22\lib\httplib.py", line 453, in putrequest
    self.send(str)
  File "C:\Python22\lib\httplib.py", line 395, in send
    self.connect()
  File "C:\Python22\lib\httplib.py", line 363, in connect
    for res in socket.getaddrinfo(self.host, self.port, 0,
socket.SOCK_STREAM):
gaierror: (7, 'getaddrinfo failed')

The second code block using urllib works fine.

Perhaps the httplib.py file is buggy or the problem is something that
has to do the client/server protocol.  socket.getaddrinfo is supposed
to return a list of (family, socktype, proto, canonname, sockaddr)


I'll continue looking to see what I can find.


From a.clarke11 at pop.ntlworld.com  Fri Jun 28 03:38:51 2002
From: a.clarke11 at pop.ntlworld.com (a.clarke11)
Date: Fri, 28 Jun 2002 08:38:51 +0100
Subject: Most elegant python program to date.
References: <mailman.1025198615.10254.python-list@python.org>
Message-ID: <3D1C128C.856E6B80@pop.ntlworld.com>

What about the pizzashop class in Lutz and Ascher, Learning Python O Reilly p
173?m Lucid, readable, concise in its output and educatiuonal as well.

PoulsenL at capecon.com wrote:

> Just curious.  I find it much easier to learn a language by reading lucid
> examples of its prose.  Are there any nominations for the most elegant
> example of a python program to date.  Not sure under what criteria, I leave
> that up to you.  I will say that I use python for matrix manipulation more
> than anything, but that shouldn't prevent you from nominating your favorite
> regardless of purpose.
>
> Loren



From communityservices at racingknowledge.com  Tue Jun  4 22:10:02 2002
From: communityservices at racingknowledge.com (communityservices at racingknowledge.com)
Date: Tue, 04 Jun 2002 22:10:02 -0400
Subject: WE NEED YOUR HELP
Message-ID: <200206042301876.SM01128@127.0.0.1>

		Hello,

Thank you for taking the time for reading this e-mail, my name is Michael Armstrong. I am founding member of
Racing Knowledge (dot) Org [http://www.racingknowledge.org]. As some of you may now our former web site,
although filled with information, was rather archaic. over the past three months the other founding members and
myself have been working with a local development company to produce a new web site, this new web site will
provide the following services for our members:

Online Auctions
Online Forums
Racing News
Racing Articles
How To's
Newsletters
Track Locator
Shop Locator
Web Search (Specific to automotive needs)
Post You Ride / Rate Your Ride

We will also provide a place for shops, clubs and users to post a single page in our web directory of whatever
they want (within reason).

Racing Knowledge has always been and always will remain a free service. We do not and never intend to make any
profit here, the only thing we ask is that our users click on our sponsors banners if they find a banner that
interests them.

HERE IS WHY WE NEED YOUR HELP

Our former content was lost when our contract with our host expired this leaves us in a very sad position. What
I ask of you is to help us get back to good with our new site that will be online within the next month or so,
and even with the generic forum software in place of the old site now [http://www.racingknowledge.org], go
there sign up post a question, get an answer. Browse our online store (we make no profit here, we make a
commission of the sales and that helps pay the bills, this new site is costing us near 30,000 dollars that
we are going to have to pay back over the next 3 months).

Racing Knowledge T-shirts and Accessories: http://www.cafepress.com/cp/store/store.aspx?storeid=rko

Racing Knowledge Store: http://racingknowledge.vstoreauto.com/

I am also looking for ad sponsors, forum moderators, and content. If you would like to write an article and
have it posted on our site with your name and perhaps a banner of your choice to your web site, or anywhere you
wish, please send it to content at racingknowledge.org

Thank You For Your Time

Michael Armstrong
Founding Member
Racing Knowledge
marmstrong at racingknowledge.org
http://www.racingknowledge.org
		




From peter at engcorp.com  Sat Jun 29 21:51:17 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 29 Jun 2002 21:51:17 -0400
Subject: possibly a dumb question
References: <3d1e5eff$1_7@nopics.sjc>
Message-ID: <3D1E6415.8128F3F8@engcorp.com>

Adonis wrote:
> 
> can a class return a value? i have treid to google this, but have found
> myself short of an answer, i know i could place it in a member and retrieve
> it or through a function, but im trying to simplify my life.. just a
> little..

You probably don't actually mean a class.  I think you mean an object,
which is an instance of a class.  (If I've got that wrong, I think
we need clarification of the question.)

If that's the case, there's still one more problem left. :-)
"Returning a value" is a concept related to calling a method
or function.  It doesn't have any particular meaning if you
ask "can an object return a value?" unless you mean something
like "can a method of an object return a value?".  That is
certainly possible, although I'm not this is what you wanted,
either:

>>> class MyClass:
...    def myMethod(self):
...        return 'spam'
...
>>> c = MyClass()
>>> print 'Green eggs and %s.' % c.myMethod()
Green eggs and spam.

Can you ask the question in a different way, or show an example
of what you were hoping to do?

-Peter


From whisper at oz.net  Sun Jun  9 14:29:48 2002
From: whisper at oz.net (David LeBlanc)
Date: Sun, 9 Jun 2002 11:29:48 -0700
Subject: (no subject)
In-Reply-To: <Pine.LNX.4.44.0206091224010.1320-100000@adams.patriot.net>
Message-ID: <GCEDKONBLEFPPADDJCOECELCDHAA.whisper@oz.net>

www.python.org and http://www.vex.net/parnassus/ are good starting places
for virtually everything pythonic.

"Python Web Programming" by Steve Holden is a good book on the subject.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Attila Horvath
> Sent: Sunday, June 09, 2002 9:30
> To: python-list at python.org
> Subject: (no subject)
>
>
> All,
>
> Am new to PYTHON and am looking for simple/sample PY-code to:
>
> 	- provide [generic] CGI functionality
> 	and if possible...
> 	- generate dynamic (DHTML) responses to CGI input parameters
>
> Any suggestions?
>
> Thx
> Attila
>
> PS: Have already looked at 'Dive Into Python'.
>
>
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From sarmstrong13 at mac.com  Sat Jun  8 16:28:53 2002
From: sarmstrong13 at mac.com (SA)
Date: Sat, 08 Jun 2002 15:28:53 -0500
Subject: Tkinter Help Please
Message-ID: <B927D335.6B54%sarmstrong13@mac.com>

Can someone explain what _tkinter in the Tkinter.py refers to?

Specifically from line:35 of Tkinter.py

import _tkinter # If this fails your Python may not be configured for Tk

Ok how do I compile Python to be configured for Tk and where is the _tkinter
module located?

Thanks.
SA



From jarvin24 at lehtori.cc.tut.fi  Tue Jun 11 05:37:58 2002
From: jarvin24 at lehtori.cc.tut.fi (=?iso-8859-1?Q?J=E4rvinen?= Petri)
Date: Tue, 11 Jun 2002 09:37:58 +0000 (UTC)
Subject: Sorting list (maybe stupid)
Message-ID: <slrnagbh7m.h03.jarvin24@lehtori.cc.tut.fi>

Hey,

I have a problem and can't solve it:

There is lines of text like:

2. Header2
1. Header1
102. Header102

and so on...

And I need to sort them according to the number (1,2,3,4...)

Am I missing something.. I have tried to split the lines by whitespace and 
create lists. But how can i sort the list by first element...

Thanks, Petri


From vjovanov at stevens-tech.edu  Mon Jun 10 12:49:40 2002
From: vjovanov at stevens-tech.edu (Vojin Jovanovic)
Date: Mon, 10 Jun 2002 12:49:40 -0400
Subject: self
References: <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu> <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com> <mailman.1023294728.2322.python-list@python.org> <sWsL8.82$v_7.4260962@newnews.cc.stevens-tech.edu> <mailman.1023323409.9676.python-list@python.org> <jpKL8.83$HV1.4324388@newnews.cc.stevens-tech.edu> <mailman.1023409885.8528.python-list@python.org> <rL3M8.85$Cf4.4420736@newnews.cc.stevens-tech.edu> <mailman.1023464131.21384.python-list@python.org>
Message-ID: <cM4N8.86$Bx6.4459580@newnews.cc.stevens-tech.edu>

Yep, this time you did it. You practically implemented a delayed evaluator
in Python.  I added the stuff to my own code and added some other things
like access to global variables etc.  Everything works fine.  I have to
admit that it is a clever trick to collect the names from an expression by
evaluating the expression in an empty dictionary.  Exactly what one needs to
get rid of self.  Now as Fernando Perez suggested this can be used to add
':=' to the language, but I am not familiar with that stuff.

Thank you for your help and for not giving up on the problem

V.J.




From chris_mk at hotmail.com  Mon Jun  3 18:45:29 2002
From: chris_mk at hotmail.com (Christopher)
Date: 3 Jun 2002 15:45:29 -0700
Subject: Problem with urllib.urlopen()
Message-ID: <cc14131d.0206031445.5154de31@posting.google.com>

Hi all,

I'm having a bit of a problem and I'm not sure what I should do.  I've
tried a couple of things and I'm running out of ideas.

Basically, I want to be able to access, directly, the XML versions of
the NCBI results of a search.  There is a default display of 'Summary'
and then there are numberous other displays, one of which is 'XML' You
can view the following link in XML (sort of) merely by changing the
display option (that was just some backgroud info, nothing spectacular
that really matters).
 Here is a link to an example page (do a search, then click details,
then the URL button and PubMed gives the actual url it uses (otherwise
it is just a cgi url; got that from O'Reilly's "Developing
Bioinformatics Computer Skills": 
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=PureSearch&db=PubMed&details_term=%28%28%28%22organization%20and%20administration%22%5BMeSH%20Subheading%5D%20OR%20organ%5BText%20Word%5D%29%20AND%20%28%22diabetes%20mellitus%22%5BMeSH%20Terms%5D%20OR%20diabetes%5BText%20Word%5D%29%29%20AND%20%28%28%22glucagon%22%5BMeSH%20Terms%5D%20OR%20glucagon%5BText%20Word%5D%29%20AND%20%28%22insulin%22%5BMeSH%
0Terms%5D%20OR%20insulin%5BText%20Word%5D%29%29%29


Now, try this url in your webbrowser and see what you get (you should
see the actual xml version of the page above):
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Display&db=PubMed&details_term=%28%28%28%22organization%20and%20administration%22%5BMeSH%20Subheading%5D%20OR%20organ%5BText%20Word%5D%29%20AND%20%28%22diabetes%20mellitus%22%5BMeSH%20Terms%5D%20OR%20diabetes%5BText%20Word%5D%29%29%20AND%20%28%28%22glucagon%22%5BMeSH%20Terms%5D%20OR%20glucagon%5BText%20Word%5D%29%20AND%20%28%22insulin%22%5BMeSH%20T
rms%5D%20OR%20insulin%5BText%20Word%5D%29%29%29&dopt=XML&query_key=1

You should, like I said, get the XML version of the original links
(with some extra information, which is why I want to interface with
the XML site, btw).  Now, if you do an interface with Python instead
of your browser you get a response of an error page instead.  Why is
that and what can I do to change it?  If you have any information, I
would appreciate it ver much.  Thanks.

Chris


From tjreedy at udel.edu  Tue Jun  4 22:40:00 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 05 Jun 2002 02:40:00 GMT
Subject: Thought on PEP 204 and 276
References: <ckp4fuc4fskjcpilnn2mrt3665fqvld92g@4ax.com> <mailman.1022528744.19859.python-list@python.org> <gm17fu40rh6r675g3a6d1ovl49tb9pd822@4ax.com> <mailman.1022594566.27265.python-list@python.org> <slrnaf7mo4.dfe.philh@comuno.freeserve.co.uk> <6w4L8.229234$%u2.120502@atlpnn01.usenetserver.com> <j36qfu4r84j8320iegrujjf8dpvdvs20bm@4ax.com> <YU9L8.240282$%u2.115403@atlpnn01.usenetserver.com> <88hqfu8ks2sfq8hkk63jn02dr9tp7mvhdh@4ax.com>
Message-ID: <4SeL8.120676$Kp.12239366@bin7.nnrp.aus1.giganews.com>

"Stephen Horne" <steve at lurking.demon.co.uk> wrote in message
> The special behaviour is  the [] cannot create a new item in a list

Actually ...

>>> a = range(3)
>>> a[2:2] = [9]
>>> a
[0, 1, 9, 2]
>>> a[4:4] = [8] # to not disturb existing mapping
>>> a
[0, 1, 9, 2, 8]

To me, this supports your main point ...

> The origin of this thread was the idea of supporting the items()
> method for lists in the same way as it is for dictionaries instead
of
> adding an enumerate function. There is no relative performance hit
> whichever notation is used. Being consistent in how key-to-value
> mapping types represent their functionality has no performance cost
at
> all.
>
> If there was a performance hit in what I was suggesting, I'd say
that
> pragmatics were in your favor. Without that performance hit, perhaps
> pragmatics may have more to do with the learning curve.
>
> Someone who has written something like...
>
>   for key, value in l_Dict.items () :
>
> might well try to do the same for a list. Why have two ways to do
the
> same thing?

I agree completely that the 'same' operation should have the same name
for all types it can apply to.

Terry J. Reedy





From mike.christie at athensgroup.com  Fri Jun 21 12:37:18 2002
From: mike.christie at athensgroup.com (Mike Christie)
Date: Fri, 21 Jun 2002 16:37:18 GMT
Subject: [newbie] Overwriting class definitions
Message-ID: <3D13563D.AB80D324@athensgroup.com>

I'm a complete newcomer to Python; I'm teaching myself Python and this is just
about the only Python code I've ever typed in, so my apologies for posting a
dumb beginner question.  Here's my problem.

I tried a slight variation on the Circle class example from "The Quick Python
Book" by Harms and McDonald.  I figured I'd put the class definitions in
script files so I could edit and reload them quickly, rather than retyping.  I
entered this in a file called circle.py:

class Circle:
    def __init__(self):
        print "Hello, version 1"

Then I did this:

import circle
circle.Circle()

and got

Hello, version 1

as expected.  Then I edited the file to say "version 2" and saved it, and then
typed in the import statement and circle.Circle() statement again.  It says
version 1 again.

So what's happening?  I assumed that when I imported the file again, it would
read in the class definition again and overwrite the old definition.  But that
doesn't appear to be happening.  When I exit Pythonwin and reenter, it runs
just fine.

Do I need to destroy the existing class def first?

Mike


From emile at fenx.com  Mon Jun  3 12:38:16 2002
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 03 Jun 2002 16:38:16 GMT
Subject: UnboundLocalError: local variable 'blah' referenced before assignment
References: <40a8aa92.0206030828.3b96333d@posting.google.com>
Message-ID: <YXMK8.157803$L76.243898@rwcrnsc53>

Daniel
> I have a strange problem where a global variable is no longer
> referenced.
> Stripped down, it's something to the effect of:
>
> def func1():
>    blah='bcd'
>
> def func2():

insert global blah here

>    print blah
>    blah='cde'

this assignment causes blah to be interpreted as a local within func2,
so the print (local) blah above occurs before blah has a value.  Adding
the global blah identifies that blah is not local, allowing the global
value to be used.


--

Emile van Sebille
emile at fenx.com

---------



From hst at empolis.co.uk  Fri Jun  7 12:12:21 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Fri, 7 Jun 2002 17:12:21 +0100
Subject: Newby: How do I strip HTML tags?
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB04F3CD@hendrix.empolisuk.com>

netvegetable wrote

> I'm mucking around with cgi, and I'm trying to work out a way 
> to strip the 
> html tags of a string. e.g, I want to convert this...
> 
> ><font size = 12><b><big>Really Big String</big></b></font>
> 
> to this this ...
> 
> >Really Big String
> 
> ... and store it as a value.
> 
> I worked out a crude, but effective way of doing it (see code 
> below), but I 
> can't escape the feeling there must be a built in way of 
> doing it more. If 
> nothing else, I'm sure somebody who knows their regular 
> expressions could 
> neaten it up (please?).
> 
> def strip_html_tags(it):
> 	left = it[:(len(it)/2)]
> 	right = it[(len(it)/2):]
> 	final = left[left.rfind('>')+1:] + right[:right.find('<')]
> 	return final
> 

If your HTML is reasonably legal, then you can use something along the lines of the following very quick and very dirty program:

import re
import sys

s = open(sys.argv[1]).read()
o = open('tmp.tmp', 'w')
r = re.compile('(<!--.*?-->)|(<[^>]*>)([^<]+)', re.DOTALL)
for x, y, z in r.findall(s):
    if z and not z.isspace():   #don't use comments tags and white-space only content
        print >>o, z

Note that you have to test first for HTML comments as a comment can contain a '>' character.

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.




From loewis at informatik.hu-berlin.de  Thu Jun  6 04:31:52 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 06 Jun 2002 10:31:52 +0200
Subject: Urllib thread safe?
References: <mailman.1023280808.19305.python-list@python.org> <m3it4xobkb.fsf@mira.informatik.hu-berlin.de> <mailman.1023304328.9010.python-list@python.org> <just-9B7B3B.21511805062002@news1.xs4all.nl> <mailman.1023308888.6405.python-list@python.org>
Message-ID: <j43cw07qpz.fsf@informatik.hu-berlin.de>

Oleg Broytmann <phd at phd.pp.ru> writes:

>    I am afraid you would have problems mixing calls to the global
> _urlopener from different threads.

Can you be more specific as to how precisely such problems could
arise?

Regards,
Martin



From tanzer at swing.co.at  Thu Jun 13 02:21:34 2002
From: tanzer at swing.co.at (Christian Tanzer)
Date: Thu, 13 Jun 2002 08:21:34 +0200
Subject: get a list of classes in this module
In-Reply-To: Your message of "Wed, 12 Jun 2002 16:19:19 CDT."
             <NCBBLFCOHHDIKCAFGCFBCEFBLEAA.marklists@mceahern.com>
Message-ID: <E17INzC-000057-00@tswings.swing.cluster>

"Mark McEahern" <marklists at mceahern.com> wrote:

> Yeah, it's not so much that I don't trust myself--there's that.  ;-)  But
> it's also the simple DRY principle at work.  Why should I have to list the
> classes in that factory-like dictionary when Python can do that for me?
>
> Thankfully, I didn't have to resort to globals()--but took advantage of
> Python 2.2's __subclasses__ (which requires subclassing the base class from
> object).  That worked smoov.

I'd recommend to use a custom metaclass to put the classes in question
into the dictionary -- this way some other module can add credit card
classes and it should still work.

YMMV,

-- 
Christian Tanzer                                         tanzer at swing.co.at
Glasauergasse 32                                       Tel: +43 1 876 62 36
A-1130 Vienna, Austria                                 Fax: +43 1 877 66 92





From ark at research.att.com  Fri Jun  7 12:24:08 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 7 Jun 2002 16:24:08 GMT
Subject: Python 2.2.1 vs. gcc 3.1?
Message-ID: <yu99vg8v6orb.fsf@europa.research.att.com>

I just installed gcc 3.1 on my Solaris 8 machine
and am trying to rebuild Python 2.2.1

Installation is running into some trouble:

running build
running build_ext
building 'struct' extension
creating build
creating build/temp.solaris-2.8-sun4u-2.2
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/tmp/build-gnu913/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /tmp/build-gnu913/Python-2.2.1/Modules/structmodule.c -o build/temp.solaris-2.8-sun4u-2.2/structmodule.o
creating build/lib.solaris-2.8-sun4u-2.2
gcc -shared build/temp.solaris-2.8-sun4u-2.2/structmodule.o -L/usr/local/lib -o build/lib.solaris-2.8-sun4u-2.2/struct.so
WARNING: removing "struct" since importing it failed
building 'regex' extension
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/tmp/build-gnu913/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /tmp/build-gnu913/Python-2.2.1/Modules/regexmodule.c -o build/temp.solaris-2.8-sun4u-2.2/regexmodule.o
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/tmp/build-gnu913/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /tmp/build-gnu913/Python-2.2.1/Modules/regexpr.c -o build/temp.solaris-2.8-sun4u-2.2/regexpr.o
gcc -shared build/temp.solaris-2.8-sun4u-2.2/regexmodule.o build/temp.solaris-2.8-sun4u-2.2/regexpr.o -L/usr/local/lib -o build/lib.solaris-2.8-sun4u-2.2/regex.so
WARNING: removing "regex" since importing it failed

and so on.

I do not believe that these problems happened with gcc 3.0

Any suggestions as to how to go about troubleshooting it?


-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From erict at millfilm.co.uk  Fri Jun  7 11:08:44 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Fri, 07 Jun 2002 16:08:44 +0100
Subject: self
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu> <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com> <mailman.1023294728.2322.python-list@python.org> <sWsL8.82$v_7.4260962@newnews.cc.stevens-tech.edu> <mailman.1023323409.9676.python-list@python.org> <jpKL8.83$HV1.4324388@newnews.cc.stevens-tech.edu> <mailman.1023409885.8528.python-list@python.org> <rL3M8.85$Cf4.4420736@newnews.cc.stevens-tech.edu>
Message-ID: <3D00CC7C.DB89BFC0@millfilm.co.uk>

It works for me with 2.2 but not with 1.5.2 both on Linux

Vojin Jovanovic wrote:

> > Why such a
> > harsh reaction without bothering to correct the obvious 'typo'?
>
> No, it is not the 'typo'.  I corrected that immediately.
> Also you had in your message >>> e=Evaluator({'a':'3']), a wrong bracket.
> Go back to your post and check it out.  That is why I thought that you
> didn't test it.
> However,  all that aside, the problems is that your code just din't work
> after I corrected those 2 'typos'.  Take a look yourself ....
>
> Python 2.1.2 (#31, Jan 15 2002, 17:28:11) [MSC 32 bit (Intel)] on win32
> Type "copyright", "credits" or "license" for more information.
> IDLE 0.8 -- press F1 for help
> >>> class Evaluator:
>     def __init__(self, eqs={}):
>         self.__dict__['equations']=eqs
>     def __getattr__(self, name):
>         value = self.equations[name]
>         if type(value)!=str:
>             return value   # if it's not a string, just give it back
>         scope={}
>         while 1:
>             try:
>                 return eval(value, scope)
>             except NameError, n:
>                 var=str(n).split("'")[1]
>                 scope[var]=getattr(self, var)
>     def __setattr__(self, name, value):
>         self.equations[name]=value
>
> >>> class Berta:
>     a=5
>
> >>> e=Evaluator({'a':3})
> >>> e.berta=Berta()
> >>> e.g='berta.a*a'
> >>> e.g
> 'berta.a*a'
> >>>
>
> V.J.



From hsbirdi at creatus.com  Thu Jun 13 20:50:09 2002
From: hsbirdi at creatus.com (Harminder S. Virdi)
Date: Thu, 13 Jun 2002 17:50:09 -0700
Subject: python milter module compilation problem
Message-ID: <Pine.SUN.4.30.0206131727550.4718-100000@thunder>

Hello,

I have the latest version of Python installed on my machine with the
libmilter.a distributed with the sendmail 8.12.4, but when I try to
compile the python milter module, it fails with the following error.

Do you have any idea what is going on? I would appreciate if you
could tell me. I dont know why I am getting this error.

[/raid/hsbirdi/milter-0.4.4] % python setup.py build
running build
running build_py
not copying Milter.py (output up-to-date)
not copying mime.py (output up-to-date)
running build_ext
building 'milter' extension
creating build/temp.solaris-2.6-sun4u-2.2
gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC
-I/usr/local/Python-2.2.1/include/python2.2 -c miltermodule.c -o
build/temp.solaris-2.6-sun4u-2.2/miltermodule.o
In file included from /usr/local/Python-2.2.1/include/python2.2/Python.h:24,
                 from miltermodule.c:101:
/usr/local/Python-2.2.1/include/python2.2/pyconfig.h:104: warning:
`_FILE_OFFSET_BITS' redefined
/usr/include/sys/feature_tests.h:80: warning: this is the location of the
previous definition
gcc -shared build/temp.solaris-2.6-sun4u-2.2/miltermodule.o -lmilter -lsm -o
build/lib.solaris-2.6-sun4u-2.2/milter.so
Text relocation remains                         referenced
    against symbol                  offset      in file
free                                0x12c       /usr/lib/libmilter.a(main.o)
free                                0x6f8
/usr/lib/libmilter.a(listener.o)
free                                0xbb0       /usr/lib/libmilter.a(engine.o)
.....................
....................
bunch of similar lines
.....................
.....................
<unknown>                           0x5d8
/usr/lib/libmilter.a(errstring.o)
<unknown>                           0x5e0
/usr/lib/libmilter.a(errstring.o)
<unknown>                           0x208
/usr/lib/libmilter.a(errstring.o)
ld: fatal: relocations remain against allocatable but non-writable sections
collect2: ld returned 1 exit status
error: command 'gcc' failed with exit status 1
[/raid/hsbirdi/milter-0.4.4] %

TIA,
-- Harminder



From huaiyu at gauss.almadan.ibm.com  Tue Jun  4 12:51:57 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Tue, 4 Jun 2002 16:51:57 +0000 (UTC)
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com>
Message-ID: <slrnafps1c.4ea.huaiyu@gauss.almadan.ibm.com>

John Roth <johnroth at ameritech.net> wrote:
>
>"Huaiyu Zhu" <huaiyu at gauss.almadan.ibm.com> wrote in message
>news:slrnafo7qv.37c.huaiyu at gauss.almadan.ibm.com...
>> Gustavo Cordova <gcordova at hebmex.com> wrote:
>
>>
>> Great explanation!  The problem stems from the fact that the current
>+=
>> operator implements two semantically distinct operations, chosen
>according
>> to some properties of the operands.  This is not unlike the situation
>of the
>> division operator with distinct behaviors on floats and ints.
>
>This is not an issue with '+='. It's an issue with assignment,
>regardless
>of whether it is standalone or combined with some other operation.

The point is that += does not have to contain any assignment at all.  In
most cases for mutable objects, the result is the same object (mutated).
This was the motivation for introducing += into Python in the first place.
The need for assignment only comes from immutable objects, but its
implementational implication extends to mutable objects as well, producing
this spurious exception.  It is spurious because the last assignment step is
not necessary.

>
>And I disagree that assignment implements two distinct operations.
>Assignment always rebinds. Always. In the examples cited, it
>attempted to rebind into a tuple, because that's what was asked
>for when the left hand side had a subscript.

This missed the point.  In-place operations should not contain assigments.

>
>The reason that the examples given at the top of the thread fail
>in such strange ways is simply that '+=' does not check whether
>the assignment is possible before doing the addition (or concatination,
>in the specific cases cited.)
>
>To see this, you could try the first example, using explicit '+' and
>'='. I suspect it would fail in exactly the same way, with exactly
>the same side effect.

Huaiyu


From BPettersen at NAREX.com  Tue Jun 18 20:36:28 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 18 Jun 2002 18:36:28 -0600
Subject: tomorrow in yyyymmdd format
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158EE1@admin56.narex.com>

> From: Gerhard H?ring [mailto:gerhard at bigfoot.de] 
> 
> Giulio Cespuglio wrote in comp.lang.python:
> > Hello everybody,
> > 
> > We all know that if I want to get the present GMT date in yyyymmdd 
> > format I should say
> > 
> > 	print time.strftime("%Y%m%d", time.gmtime() )
> > 
> > Now, what about tomorrow? Note that a simple string manipulation is 
> > not the solution I'm looking for, as tomorrow might be next month, 
> > next year, the 29th of February and such. Do you confirm 
> that there's 
> > nothing better than the following silly-looking expression?
> > 
> > 	time.strftime("%Y%m%d", time.gmtime(time.time() + 60*60*24))
> 
> No, there's nothing better. Unless you're using the DateTime 
> type from the eGenix mxExtensions:

Sure there is :-) time.mktime() can take a "malformed" time tuple and do something sensible with it. You can therefore get a time tuple from gmtime, add one to the day position and call time.mktime() on the result:

>>> x = (2002, 2, 29, 0, 31, 42, 2, 170, 0)
>>> time.gmtime(time.mktime(x))
(2002, 3, 1, 7, 31, 42, 4, 60, 0)
>>>

-- bjorn




From emile at fenx.com  Fri Jun  7 08:22:03 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 07 Jun 2002 12:22:03 GMT
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com>
Message-ID: <Lz1M8.123774$352.6153@sccrnsc02>

Peter Hansen
> You forgot the even more important first thing for a beginner:
> get it correct!  The only good way to do this is to write unit
> tests that prove it.  If you don't have tests, you don't have
> working code (or how can you prove it?).

We always used absence of bug reports as a gauge... and always
questioned if that indicated usage or correctness.  ;-)

--

Emile van Sebille
emile at fenx.com

---------



From marklists at mceahern.com  Wed Jun 26 13:22:30 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 26 Jun 2002 12:22:30 -0500
Subject: Problems using samples
In-Reply-To: <337619fa.0206260913.41e08940@posting.google.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEKDLGAA.marklists@mceahern.com>

> 6 if x = y:

This statement assigns (binds) the value of y to the name x.  I assume what
you want to do is test for equality.  If so, you should use the equality
operator, not the assignment operator:

  if x == y:

By the way, if you want to make it easier to help you in the future, please
post an interactive session that demonstrates the error you're experiencing;
i.e.,

  >>> x = 8
  >>> y = 4
  >>> v = x - y
  >>> if x = y:
    File "<string>", line 1
      if x = y:
           ^
  SyntaxError: invalid syntax

This is one of the beauties of Python--you should take advantage of it.

Admittedly, Python's error is cryptic, but the ^ clues you in to the fact
that it has something to do with the equals sign.

// m

-





From edream at tds.net  Thu Jun 20 15:23:32 2002
From: edream at tds.net (Edward K. Ream)
Date: Thu, 20 Jun 2002 19:23:32 GMT
Subject: tkinter font picker?
Message-ID: <3D122BB6.92593026@tds.net>

Does anyone know of tkinter code that will put up a font chooser
dialog?  There are a number of such packages written in tk/tcl.  I'm
wondering if anyone knows of a port to tkinter.

Thanks,

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


From skip at pobox.com  Mon Jun  3 15:22:14 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 3 Jun 2002 14:22:14 -0500
Subject: Created a "file" like command in python (see man 4 magic) and face a showstoper bug
In-Reply-To: <359e75aa.0206031113.6a8bfed1@posting.google.com>
References: <359e75aa.0206031113.6a8bfed1@posting.google.com>
Message-ID: <15611.49638.585990.766462@beluga.mojam.com>

    Thomas> The code causing problem is :

    Thomas> if replace:
    Thomas>   try:
    Thomas>     mime = mime % replace
    Thomas>   except:
    Thomas>     pass

    Thomas> In both case mime is containing the string "%d." and replace is
    Thomas> in both case of type int and valued 1 and 0.

    Thomas> Any idea ?

My guess is that either mime or replace aren't what you think they are and
your catch-all except clause is masking an error in your code.  I suggest
you make the except clause more explicit (what problems are you prepared to
handle?) and/or print out mime and replace before executing the % operator:

    try:
        print (mime, replace)
        mime = mime % replace
    except StuffIKnowHowToRecoverFrom:
        recover_from_nasty_stuff()

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html





From tim.one at comcast.net  Fri Jun  7 14:26:18 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 07 Jun 2002 14:26:18 -0400
Subject: array, list, performance...
In-Reply-To: <3D0108B7.3C206C2E@accessforall.nl>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEELPLAA.tim.one@comcast.net>

[Michael Chermside]
>     li = [x0, x1, x2, ...]
> ...
>     li.append(x)   appending is O(1)
>                    (actually, it SOMETIMES takes O(len(li)) as it
>                     resizes the list. But if you grow a list by
>                     continually appending, the amortized time is
>                     linear (ie, constant amortized time per element)

[Ype Kingma]
> Append is often enough linear in the length of the list
> that growing by appending is O(len(li) * len(li)).

[Tim]
> For Python 2.2 Michael is correct.  For Python 2.1 and earlier, you're
> correct "in theory", although in practice it was often
> linear-time anyway,> depending on internal details of the platform
> realloc.

[back to Ype]
> Good news.
> Multiplying the list's size by a constant factor when necessary would
> give O(log(len(li)) * len(li)) behaviour, so how was this made
> O(len(li)) in 2.2?

By increasing a list's size (when needed) by an amount proportional to its
current size.

> I checked 'What's new in 2.2' but found nothing about this.

Probably not.  I don't write up a news item when I optimize away a useless
branch in the code either <wink>.  Seriously, it's an internal
implementation detail, and, like I said, it often *acted* linear-time
anyway.  The only platforms I know of where it appeared to make a real
difference were assorted flavors of Windows, and some flavor of BSD Unix (I
don't recall which).  On those it helped a lot in cases of extreme list
abuse.





From benjl at cse.unsw.edu.au  Sun Jun  2 19:30:46 2002
From: benjl at cse.unsw.edu.au (Benno)
Date: 2 Jun 2002 16:30:46 -0700
Subject: What is a classmethod for?
References: <addgv30f4d@enews4.newsguy.com> <addlm6$g44$1@panix1.panix.com>
Message-ID: <332c7c11.0206021530.2ada6f5c@posting.google.com>

aahz at pythoncraft.com (Aahz) wrote in message news:<addlm6$g44$1 at panix1.panix.com>...
> In article <addgv30f4d at enews4.newsguy.com>,
> Dennis Peterson <denpeterson at yahoo.com> wrote:
> >
> >
> >That's nice, but I can't think of a use for it, given that we have
> >inheritance and isinstance(). Can anyone enlighten me?
> 
> The point is that you can call class methods *without* creating an
> instance:
> 
> C.foo()
> D.foo()
> 
> This allows you to manipulate class variables through methods, rather
> than using functions or needing to create an instance.

Which is really useful if you are overriding __new__(). (Which is itself a
class method iirc).

Cheers,

Benno


From quinn at baht.ugcs.caltech.edu  Sun Jun 23 17:42:54 2002
From: quinn at baht.ugcs.caltech.edu (Quinn Dunkan)
Date: 23 Jun 2002 21:42:54 GMT
Subject: re.rsearch?
Message-ID: <slrnahcg6u.382.quinn@baht.ugcs.caltech.edu>

Occaisionally I find myself wanting an re.search that will find the rightmost
match.  In those occaisions, I use:

def rsearch(reg, s):
    'like re.search only find last match, return start, end instead of match'
    last_m = None
    start = end = 0
    while 1:
        m = re.search(reg, s[end:])
        if m is None:
            if last_m is None:
                return None, None
            else:
                return start, end
        last_m = m
        end += m.end()
        start = end - (m.end() - m.start())

... of course, it would be nicer to return an actual match object, but I
haven't figured out how to do that short of faking up my own, and I've only
needed (start, end) so far.  Unfortunately, AFAIK, re.search also lacks a
'start_searching_at' type option.

Has anyone else solved this probem in a more graceful way?


From benjl at cse.unsw.edu.au  Sat Jun  8 20:10:20 2002
From: benjl at cse.unsw.edu.au (Benno)
Date: 8 Jun 2002 17:10:20 -0700
Subject: Counting lines
References: <mailman.1023405789.30972.python-list@python.org> <aJTL8.21666$ZF5.2505752733@newssvr10.news.prodigy.com> <3D000B36.5C7BCEBF@engcorp.com> <mailman.1023468709.32328.python-list@python.org> <mailman.1023557203.18547.python-list@python.org>
Message-ID: <332c7c11.0206081610.32d292b2@posting.google.com>

Lulu of the Lotus-Eaters <mertz at gnosis.cx> wrote in message news:<mailman.1023557203.18547.python-list at python.org>...
> Since this thread has drifted into line counts, I think it worth sharing
> a resource that a fellow-developer shared with me:
> 
>     SLOCCount:
>     http://www.dwheeler.com/sloccount/
> 
> Take a look.  It's not written in Python (I don't think), but it seems
> to do a large superset of what pycount.py does.  SLOCCount counts the
> lines of code in a large number of different programming languages,
> using rather clever rules and heuristics both to identify what language
> a file actually is, and to discern which lines are code versus comments
> (including, e.g. Python docstrings).

Nice program. Although lines of count has to be about the most abused metric 
in existence IMHO. It should surely be the aim of any project to *minimise*
the number of lines of code. (Fewer lines genreally means fewer opporunities
for bugs, easier to find bugs, easier to maintain, and less chance of 
duplication.) In my own code I find refactoring (or even rewriting in some 
cases), can reduce the lines of code by a factor of 10.

Cheers,

Benno


From zopestoller at thomas-guettler.de  Fri Jun 28 08:18:13 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Fri, 28 Jun 2002 14:18:13 +0200
Subject: [NEWBIE]- Syntax error
References: <a7YS8.727$yY2.23825@ozemail.com.au>
Message-ID: <3D1C5405.1030302@thomas-guettler.de>

Alistair Campbell wrote:

> Hi,
> I imagine the answer to this post is very simple ... but I'm not proud.
> 
> Using Python2.2 on Windows OS
> 
> Anyway, when I try to run a script fom the command line in IDLE or Python
> 
> 
>>>>python name_of_script.py
>>>>
> 
> the interpreter complains that there is a syntax error and highlights the
> "name_of _script" part.
> 
> I imagine that I have not initialised some PATH variable or other. I guess
> that i need to put something in the autoexec.bat. Can anyone suggest what?


One solution would be that you start a MS-DOS prompt and start the script
from there. (Python.exe must be in your %PATH%)

But better:
the extension .py should be connected to the interpreter. Double-click
on your script from the windows-explorer should work. The window closes
after the script is run. This will help:

import sys

#.. your code

print "end. Press return"
sys.stdin.readline()


 thomas


BTW:
  python foo.py
does not work since you are in the python interpreter. "python" and "foo.py"
are no known functions.

print "foo" will work






From bokr at oz.net  Thu Jun 13 21:58:14 2002
From: bokr at oz.net (Bengt Richter)
Date: 14 Jun 2002 01:58:14 GMT
Subject: Type subclassing: bug or feature
References: <aeau48$111$1@panix1.panix.com> <Xns922CEE783C10Dcliechtigmxnet@62.2.16.82>
Message-ID: <aebijm$aq0$0@216.39.172.122>

On 13 Jun 2002 23:24:09 +0200, Chris Liechti <cliechti at gmx.net> wrote:

>aahz at pythoncraft.com (Aahz) wrote in news:aeau48$111$1 at panix1.panix.com:
>
>> Consider the following code:
>> 
>> class MyStr(str):
>>     def contains(self, value):
>>         return self.find(value) >= 0
>> 
>> s = MyStr("hello, world!")
>> s = s.capitalize()
>> if s.contains('Hello'):
>>     print "Found it!"
>> 
>> It fails with an AttributeError when it calls s.contains(), because
>> s.capitalize() returned a str instead of a MyStr.  Anyone want to take a
>> whack at defending this as the correct behavior?
>
>solving this at the C level would require patching many stingobject factory 
>functions... e.g. PyString_FromStringAndSize for capitalize(). but it would 
>indeed be desireable.
>
>i think it would like to have a even more genral mechanism. i would like to 
>have a possibility to replace the builtin type. e.g. i could replace the 
>builtin "str" and all strings from now on would be instaces of my class. 
>but that's clearly impossible in the current implementation.
>
><hacker warning>
>like Bengt already sugested just automated ;-) ...
>
>>>> class mstr(str):
>... 	for name,m in str.__dict__.items():
>... 		if type(m) == type(str.find):
>... 			exec """def %s(*args, **kwargs):
>... 					q = str.%s(*args, **kwargs)
>... 					if type(q) is str: return mstr(q)
>... 					else: return q\n""" % (name, name)
>... 	def __contains__(self, other):
>... 		return str.find(self, other) >= 0
>... 	
>>>> s = mstr("hello")
>>>> type(s.capitalize())
><class '__main__.mstr'>
>>>> 'll' in s.capitalize()
>1
></hacker warning>
>
Interesting way to go if you can't know except dynamically at run time
what types base class methods will return.

I don't know how you'd fully generalize this process though, and it
seems a shame to have to wrap everything in an extra layer. It would
be nice to be able to specify coercions in the subclass in a way that
base class methods could access quickly to check whether to call a coercion
routine on a result, e.g.,

    __coercions__ = (str, mstr), (bool, No_Yes)  # just as example with several coercions

Also str.split could even access this,  and produce a list of MyStr substrings.
And also deeper bases' methods could check on self.__coercions__

Just OTTOMH, so not very thought through ;-)

BTW, I just thought of a 'where' keyword for 'for' e.g., the above for would become

    for name,m in str.__dict__.items() where type(m) == type(str.find):

sort of like sql select ... where ... or the list comprehension if.

Regards,
Bengt Richter


From andy at dustman.net  Sat Jun 15 19:32:22 2002
From: andy at dustman.net (Andy Dustman)
Date: 15 Jun 2002 19:32:22 -0400
Subject: executemany & UPDATE (MySQL & Python)
In-Reply-To: <20020614191751.GA1520@lilith.my-fqdn.de>
References: <slrnagk7tu.114.gerhard@lilith.my-fqdn.de>
	<NDBBLAEMOMEHKALAMKKJKEJJKPAA.narnett@mccmedia.com> 
	<20020614191751.GA1520@lilith.my-fqdn.de>
Message-ID: <1024183942.3841.6.camel@localhost.localdomain>

On Fri, 2002-06-14 at 15:17, Gerhard H?ring wrote:
> (Cc-ing Andy Dustman)
> 
> * Nick Arnett <narnett at mccmedia.com> [2002-06-14 11:02 -0700]:
> > > Gerhard wrote:
> > > > Nick Arnett wrote:
> > > > > Can anyone confirm that it is *not* possible to pass a list to
> > > > > the Python MySQLdb module in UPDATE operations?
> 
> ...
> 
> > > For database modules that don't support prepared statements, the db module
> > > implements executemany with multiple calls to execute, anyway.
> > 
> > Well, since MySQLdb is much faster at inserting that way, I'm assuming that
> > if it'll do it on an UPDATE, that will also be much faster.
> 
> I mistakenly assumed that MySQLdb's executemany was implemented the same
> way as pyPgSQL's. That assumption was wrong. Looking into the source
> reveals that MySQLdb's implementation is really clever, but
> unfortunately hardcoded to only work with INSERT statements.

Yes, MySQL has syntax for a multi-row INSERT, but apparently no such
similar syntax for UPDATE, so they have to be done as seperate queries.

REPLACE has the same syntax as INSERT (a MySQL extension that INSERTS
but overwrites any existing rows), and that should also work fine with
executemany(), though I haven't used it.

I will change MySQLdb though to simply perform the query multiple times
if it does not seem to match the VALUES (...), (...), ... syntax; I
consider it a bug that it doesn't do this already, but nobody seems to
care (until now).

-- 
Andy Dustman         PGP: 0x930B8AB6
    @       .net     http://dustman.net/andy
"Cogito, ergo sum." -- Rene Descartes
"I yam what I yam and that's all what I yam." -- Popeye




From thor__00 at yahoo.com  Fri Jun 14 06:10:28 2002
From: thor__00 at yahoo.com (Thor)
Date: Fri, 14 Jun 2002 12:10:28 +0200
Subject: system command (thanks to all)
References: <aeajuh$5b5eo$1@ID-108351.news.dfncis.de>
Message-ID: <aecli9$5qfp9$1@ID-108351.news.dfncis.de>

Hej, the problem is solved and it was something non-realed with teh issues 
here, a few lines up the ones I suppouse that were problematic. But having 
tried th suggestons here (working alone, but not working in may prgram) 
they me to locate teh problem. THanks to all!!
-- 
Thor -- Stockholm -- Sverige


From rjones at ekit-inc.com  Thu Jun 27 19:20:28 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Fri, 28 Jun 2002 09:20:28 +1000
Subject: Pythoniac: Thoughts on a hardware Python processor
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F06233015DEC3B@mail.hebmex.com>
References: <F7DB8D13DB61D511B6FF00B0D0F06233015DEC3B@mail.hebmex.com>
Message-ID: <200206280920.28424.rjones@ekit-inc.com>

On Fri, 28 Jun 2002 08:44, Gustavo Cordova wrote:
> > Whoever asked if the Intel/AMD FP stack counted: nah - it's
> > not the primary mode of the CPU like it was on the HP 3000
> > and the Forth chip. BTW, speaking of the Forth chip, I think
> > Forth, Inc. is still offering the IP for that if anyone wants
> > to burn a few hunderd k$ on new silicon.
>
> If I recall correctly, Chuck Moore's still churning out
> forth processors, with on-chip fast memory and hardware
> stacks, really really cool stuff. I think there's more info
> in the ColorForth site, or at least it points you in the
> right direction.

Ooh... from there I found:

  http://www-2.cs.cmu.edu/~koopman/stack_computers/index.html

  "The ENTIRE BOOK is available to read on-line or to download for
   off-line reading!"

Oh dear, there goes my spare time (again) :)


    Richard





From anthony at interlink.com.au  Fri Jun 14 01:19:59 2002
From: anthony at interlink.com.au (Anthony Baxter)
Date: Fri, 14 Jun 2002 15:19:59 +1000
Subject: filter2 
In-Reply-To: Message from Oleg Broytmann <phd@phd.pp.ru> 
   of "Thu, 13 Jun 2002 17:33:13 +0400." <20020613173313.D19424@phd.pp.ru> 
Message-ID: <200206140519.g5E5Jxv24853@localhost.localdomain>

>>> Oleg Broytmann wrote
> Hello!
> 
>    I want to have new python builtin - filter2. It is exactly like filter,
> but it returns also a list of items that didn't pass the test.
 [...]
>    Opinions?

Horrible.

Not a good idea, and a horrible horrible name. Completely non-pythonic.

(Well, you did ask for opinions)

-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From K.Rdt at TU-Berlin.DE  Sun Jun 16 16:11:53 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Sun, 16 Jun 2002 22:11:53 +0200
Subject: HT telnet interact with auto login?
In-Reply-To: <20020616210629.GA20626@soniq.net>
Message-ID: <872UFDQNUSB841A0KHF0LJVQ21LGRMVP.3d0cf109@lap>

Am 16.06.02 23:06:29, schrieb Paul Boehm <paul at soniq.net>:

>is this what you want?

No. Thanks anyway - Klaus
>
># On Sun, Jun 16, 2002 at 06:07:30PM +0000, Klaus Reinhardt wrote:
>import getpass
>import sys
>import telnetlib
> 
>HOST = "...de"
>tn = telnetlib.Telnet(HOST)
> 
>user = '...'
>password = '...'
>
># 2.)
>tn.read_until("login: ")	# this seams to work
>			# but without connection
>tn.write(user + "\n")
>if password:
>    tn.read_until("Password: ")
>    tn.write(password + "\n")
># 1.)
>tn.interact()	# this is working but with manually ..
>
>
----------------- mailto:K.Rdt at TU-Berlin.DE start -------------------
Hi

		K at Rdt
----------------- mailto:K.Rdt at TU-Berlin.DE !end! -------------------







From siegfried.gonzi at kfunigraz.ac.at  Sat Jun 22 15:27:17 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Sat, 22 Jun 2002 21:27:17 +0200
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122>
Message-ID: <3D14CF95.4030501@kfunigraz.ac.at>

Bengt Richter wrote:

> 
> You apparently have a lot of different stuff glued together. Have you tested the
> separate pieces separately? Do you have some test problems that you can use to validate
> numerical results *exactly* (trailing decimal differences need an explanation. They
> can be benign/acceptable or not).

The Mie code (C function) works, because on Windows I wrote a program 
under DISLIN in order to depict the phase function, and I can use the 
same program under Linux without problems. The Fortran function works 
too (the Fortran function is really a tiny piece and straightforward in 
contrast to the C function which is a third party insane pointer hell). 
The C code itself is essentially the same:  on Windows I compiled it 
with SWIG and Microsoft Visual C++ and on Linux I used gcc and SWIG; but 
the name of the function is actually the same. The same goes for all the 
other functions and directories.


> I hate to see Python (or anything) bashed without proof. So far I haven't seen the proof,
> nor an explanation.
> 
> If your code is open source, maybe someone will be interested in discovering why
> it bogs Python down (or Python bogs it down ;-)

The code is open source but not very well documented. But I can send it 
to you if you like. You can also get the data if you like.

First, I thought there is something wrong with my data because I have it 
converted  from DOS to UNIX with: dos2unix.

There is an interesting thing to observe: On Windows I am likely to hit 
the 100MB threshold, but on Unix I only hit 10MB (and sometimes 20MB). 
At least "top" shows me that.


You can get all the functions if you like. I would only have to get it 
from my laptop back to my stationary machine.

Drop me a note if you want the functions and data. I can prepare it that 
way that it will be ready to use for you.

What I further noticed. On Windows the Task manager tells me that in the 
first 20 minutes Python uses constantly 99% of the processor power. The 
same goes under Linux, but after that 20 minutes it sucks down to 87% on 
Linux and on Windows! Note: same functions and that 8 times!

S. Gonzi



From jwbaxter at spamcop.net  Sat Jun 29 21:44:12 2002
From: jwbaxter at spamcop.net (John Baxter)
Date: Sat, 29 Jun 2002 18:44:12 -0700
Subject: Python needs better error reporting
References: <mailman.1025350057.14486.python-list@python.org> <U2iT8.45587$n4.10825231@newsc.telia.net>
Message-ID: <jwbaxter-BBD3FF.18441229062002@corp.supernews.com>

In article <U2iT8.45587$n4.10825231 at newsc.telia.net>,
 "Fredrik Lundh" <fredrik at pythonware.com> wrote:

> Tim Peters wrote:
> 
> > > I really DO have a point!
> >
> > Error-reporting in parsers has been the subject of intense
> > research.  Do a little digging and come back when you can
> > make a case that your point can be addressed effectively.
> 
> fwiw, I've seen Lisp advocates argue that this may be
> one of Lisp's greatest advantages...

It probably is now, but it was not so in the early days, when the LISP 
users walked around MIT with a bunch of cards full of right parens.  
They stuck several such cards on the back of the decks they submitted 
with LISP programs.

The then-current batch-procesing LISP interpreter gave no output at all 
in the absence of enough )s. It gave one *something*--often useful--in 
the presence of too many (as it just ignored the excess).  (And 
turn-around was overnight, although the center often did better.) 

  --John (who never understood anything of significance which John 
McCarthy said in my presence)


From hesterloli at hotmail.com  Wed Jun 19 04:21:43 2002
From: hesterloli at hotmail.com (George Hester)
Date: Wed, 19 Jun 2002 08:21:43 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com> <3D1010C3.C321F8B0@engcorp.com>
Message-ID: <raXP8.47926$GY.15260255@twister.nyroc.rr.com>

Yes my take on it was that you said "I needed to crawl before I walk."  Please
don't hold my inference against me.

--
George Hester
_________________________________
"Peter Hansen" <peter at engcorp.com> wrote in message
news:3D1010C3.C321F8B0 at engcorp.com...
> George Hester wrote:
> >
> > The whole issue started from here.  I did not deserve to be told to "crawl
> > before I walk."  I may have responded to that inappropriately I confess.
But
> > that does not excuse the implied derogatory.  Since that the lambasting
> > crecscendoed.
>
> Perhaps "inferred derogatory" is more appropriate.  "Implied"
> suggests that I intended it to be so, which is not true.  "Inferred"
> refers to how you interpreted it.
>
> Anyway, maybe we'd best move on... no long-term harm done
> to any party, I think.
>
> -Peter




From dianne at dogmac.com  Sun Jun 23 21:32:38 2002
From: dianne at dogmac.com (Dianne van Dulken)
Date: Mon, 24 Jun 2002 11:32:38 +1000
Subject: Suggestions for good programming practices?
Message-ID: <WEuR8.2074$5l4.75657@ozemail.com.au>

Hi,

I'm fairly new to python, coming from a perl background, and was wondering
if anyone has a list of things that they consider as good programming
practice in Python (I thought this might be an interesting topic for
discussion, anyway)

For example, in perl you obviously are always encouraged to use strict, and
we always use eval around our main code to pick up any unexpected errors.  I
presume the equivalent in python is the try statement.  Would you suggest
that this should be used, or would that be redundant here?  Any other tips
that you would suggest a new python user should be always doing as a good
programming practice?

Thanks

Di




From radix at twistedmatrix.com  Mon Jun 10 13:17:37 2002
From: radix at twistedmatrix.com (Christopher Armstrong)
Date: 10 Jun 2002 13:17:37 -0400
Subject: Source formatting for long expressions.
In-Reply-To: <15620.50614.922005.965796@12-248-41-177.client.attbi.com>
References: <87hekdxypd.fsf@twistedmatrix.com>
	<15620.50614.922005.965796@12-248-41-177.client.attbi.com>
Message-ID: <87r8jfvyry.fsf@twistedmatrix.com>


>>>>> "Skip" == Skip Montanaro <skip at pobox.com> writes:

    Chris> Does anyone know of a tool that will beak up a very, very long
    Chris> expression like this into multiple lines with sane indentation?

    Skip> How about the pprint module?

I've looked at the source for pprint (on Python 2.1), and I saw nothing
that allows me to extend it with my own classes. If you check my other
replies to this thread, you can see my (stupid, but effective) solution.

-- 
                                Chris Armstrong
                         << radix at twistedmatrix.com >>
                http://twistedmatrix.com/users/carmstro.twistd/





From skip at pobox.com  Mon Jun  3 16:12:02 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 3 Jun 2002 15:12:02 -0500
Subject: mySQL on Solaris .8
In-Reply-To: <79d1869.0206031202.228c4387@posting.google.com>
References: <79d1869.0206030646.512af52c@posting.google.com>
        <79d1869.0206031202.228c4387@posting.google.com>
Message-ID: <15611.52626.927671.555007@beluga.mojam.com>

    >> I am trying to install a mySQL database in solaris 2.8. I have a copy
    >> of the distro for mySQL-python-0.9.1. When I try to run setup.py, I
    >> am told that I am missing distutils.core

    >> Where do I get this? What package is it part of?

Joe,

It's part of the core Python distribution (since 2.0 or 2.1 I think).
Sounds like either you're running an older version of Python or the one you
have is misinstalled in some way.  Assuming you have MySQL's client libs and
a recent enough version of Python correctly installed, Installing
mysql-python should take less than 30 seconds from the time you execute

    python setup.py install

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html





From garth at deadlybloodyserious.com  Mon Jun 17 01:23:02 2002
From: garth at deadlybloodyserious.com (Garth T Kidd)
Date: 16 Jun 2002 22:23:02 -0700
Subject: MS Word -- finding text
References: <mailman.1024247060.12127.python-list@python.org> <4c877253.0206161457.a2289c8@posting.google.com>
Message-ID: <4c877253.0206162123.5c8de8a3@posting.google.com>

> > conversion str() first. When necessary, it slowly goes through 
> > each character, handling the exceptions that are raised. 
> 
> Great technique! Now, I don't suppose you could identify the
> characters str() isn't handling, so that someone can fix str()? :)

Mark's come up with a much better method. How bloody *typical*. :)

Regards,
Garth.


From CousinStanley at HotMail.Com  Tue Jun 11 19:29:18 2002
From: CousinStanley at HotMail.Com (Cousin Stanley)
Date: 11 Jun 2002 23:29:18 GMT
Subject: PIL _imaging
Message-ID: <Xns922AA829A8C5FHotMailcs@130.133.1.4>

Greetings PythonIstas ... 

I'm a dinosaur-age programmer trying to get started 
using Python 2.2.1 under Win98 1st ed ... 

I've installed the  wxPython  and  PMW  packages
and the  samples/demos  from both seem to run OK ... 

However, I'm having some problems trying to run demos
from the Python Imaging Library ... 

  python Samples/viewer.py ButtonD.jpg

  Leads to the bottom-line on the trace below
    ImportError: The _imaging C module is not installed ...

  There are 2 files in K:\Python\Lib\PIL\DLLs
  but apparently these are not the ones that are needed ...

    _imaging.pyd
    _imagingtk.pyd 


PYTHONPATH is set in the  autoexec.bat  file
  set PYTHONPATH=K:\Python\Lib\PIL
  set PYTHONPATH=%PYTHONPATH%;K:\Python\Lib\PIL\DLLs
  set PYTHONPATH=%PYTHONPATH%;K:\Python\Lib\PIL\PIL
  set PYTHONPATH=%PYTHONPATH%;K:\Python\Lib\PIL\Samples 

Have I blown the set_Up / search_Rules 
or do I need to look for an additional module
to install ???

Cousin Stanley

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

< K:\Python\Lib\PIL >
Tue 06-11-2002 15:24:35.35 > python Samples/viewer.py ButtonD.jpg
Traceback (most recent call last):
  File "Samples/viewer.py", line 46, in ?
    UI(root, im).pack()
  File "Samples/viewer.py", line 25, in __init__
    self.image = ImageTk.PhotoImage(im)
  File "K:\Python\Lib\PIL\PIL\ImageTk.py", line 86, in __init__
    self.paste(image)
  File "K:\Python\Lib\PIL\PIL\ImageTk.py", line 108, in paste
    im.load()
  File "K:\Python\Lib\PIL\PIL\ImageFile.py", line 126, in load
    self.load_prepare()
  File "K:\Python\Lib\PIL\PIL\ImageFile.py", line 180, in load_prepare
    self.im = Image.core.new(self.mode, self.size)
  File "K:\Python\Lib\PIL\PIL\Image.py", line 37, in __getattr__
    raise ImportError, "The _imaging C module is not installed"
ImportError: The _imaging C module is not installed

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



From fredrik at pythonware.com  Sat Jun 29 08:06:24 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 29 Jun 2002 12:06:24 GMT
Subject: I'm an idiot
References: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16> <afj27n023m7@enews2.newsguy.com> <3D1D5C5E.3A5EB81D@earthlink.net>
Message-ID: <4phT8.45579$n4.10823008@newsc.telia.net>

Joseph A Knapka wrote:

> Now, personally, I'd like to have any read after EOF throw
> an exception.

I suggest subclassing the file object, using your custom
subclass for a while, and throwing it out when you get
bored adding all those extra try/except clauses around
your code...

> But I'm sure there are Very Good Reasons not to do this,
> and that some wise person will presently educate me :-)

To many people, this is a religous topic.  Google can probably
tell you all about the different religions.

The most common view is that as the name implies, exceptions
should be used to deal with exceptional, unexpected conditions.
That you'll reach the end of a file if you keep reading data from
it should hardly come as a surprise...

</F>




From mjais at web.de  Tue Jun 18 03:27:37 2002
From: mjais at web.de (Markus Jais)
Date: Tue, 18 Jun 2002 09:27:37 +0200
Subject: why len(list) instead of list.len()  ?
Message-ID: <EhBP8.1$Qn1.1124@news.ecrc.de>

hello

I way wondering why one has to write

list = [2, 3, 5]
print len(list)

instead of list.len()

in Ruby I can write 

array = [2, 3, 4]
puts array.length()


which feels more intuitive to me

please, I do not want to start a flame war
Ruby <-> Python
I like and use both !!

maybe there is something like that for python
and I have just not found it.

markus


From claird at starbase.neosoft.com  Fri Jun 28 08:26:09 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 28 Jun 2002 07:26:09 -0500
Subject: Be the first on your block to register your computer.
References: <mailman.1025211444.21440.python-list@python.org> <afg1ks$e6lj5$1@ID-71831.news.dfncis.de> <3D1B6E69.1090204@earthlink.net> <IXWS8.368185$o66.975748@news-server.bigpond.net.au>
Message-ID: <0F2CD32D80F7F061.3D8D80DBD1BF70E9.B80FB476A9BA57A6@lp.airnews.net>

In article <IXWS8.368185$o66.975748 at news-server.bigpond.net.au>,
Neil Hodgson <nhodgson at bigpond.net.au> wrote:
			.
			.
			.
>   Then we'll just write Python as an Excel macro. Microsoft has over its
>entire history stuck programming languages in things *even when they are a
>security liability*. They show no signs of deviating from this path with
			.
			.
			.
'Guess you didn't get the memo.  Microsoft doesn't do this
in spite of the security implications; Microsoft does this
*because* of the security implications <URL: http://
www.pbs.org/cringely/pulpit/pulpit20020627.html >.  Micro-
soft's marketing machine is sufficiently powerful to 
induce people to believe the people who create problems
should be trusted to solve them.

'Sort of like all the other scandals in the headlines.
-- 

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


From sadams123 at optushome.com.au  Wed Jun 26 00:05:22 2002
From: sadams123 at optushome.com.au (Steven Adams)
Date: Wed, 26 Jun 2002 14:05:22 +1000
Subject: Python and Windows XP
References: <20020625233842.29966.00000386@mb-ma.aol.com>
Message-ID: <3d193d85$0$20999$afc38c87@news.optusnet.com.au>

>    I apologise for this newbie question, but I use Windows XP and wonder
how I
> can open a shell session to use Python?  There used to be the DOS command
> option, but now...

I just open up python.exe in my Python directory and it brings up the Python
interpreter, the standard 2.2 install setup shortcuts to it as

Start->Programs->Python -->Python (command line)

There is a command prompt that behaves like a DOS prompt - though I can't
recall where it sits in the default menu.

I'm using XP Pro so YMMV

Steven




From printers at sendme.cz  Thu Jun 20 04:12:20 2002
From: printers at sendme.cz (A)
Date: Thu, 20 Jun 2002 10:12:20 +0200
Subject: How to find all emails (on the same line)
Message-ID: <3D11AA84.4222.A8556@localhost>

Hi,
I need to find ALL emails in a file. I can use re module but I do not 
know how to find all emails if there are more emails on the same 
line because re module ( as far as I know )finds only one/first 
occurrence on that line.
Thanks for help.
Ladislav






From ianb at colorstudy.com  Tue Jun  4 01:23:29 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 04 Jun 2002 00:23:29 -0500
Subject: Login Authentication with cookies...
In-Reply-To: <e271e93c.0206031905.73796d31@posting.google.com>
References: <e271e93c.0206031905.73796d31@posting.google.com>
Message-ID: <1023168209.16910.4.camel@localhost>

That's a weird little error -- a missing import at the top of
ClientCookie.py ?  Anyway, you could also use HTTPSession, included with
webunit.sf.net.

On Mon, 2002-06-03 at 22:05, chandan wrote:
> Hi.., 
> 
> Problem: - write a python script to access pages that are password
> protected, but DONT return a 401 forbidden code. i.e. Login to a
> website like hotmail and check you mail message or in my particular
> case, http://www.fool.com/login.asp
> and then download the discussion board messages for a particular
> stock.
> 
> Attempts so far: - The login and password protection is based on
> cookies, so I used ClientCookie.py as follows (although, i am not sure
> how to use it properly)
> 
> import ClientCookie
> import urllib2
> import urllib
> params = urllib.urlencode({'login':'login','password':'password'})
> request = urllib2.Request("http://www.fool.com/login.asp", params)
> # note we're using the urlopen from ClientCookie, not urllib2
> result = ClientCookie.urlopen(request)
> 
> Error message: 
> Traceback (innermost last)
>   File "c:\python22\programs\demos\trial", line 7, in ?
>     result = ClientCookie.urlopen(request)
> NameError: global name 'types' is not defined
> 
> Program disconnected.
> 
> I dunno what to do about the types in ClientCookie.py... even the
> example code given in the README file for this library returns the
> same error.., i am using windows XP, python 2.2
> 
> ClientCookie can be obtained from here -
> http://wwwsearch.sourceforge.net/ClientCookie/
> 
> Thanx for the help
> Chandan
> 
> P.S : Newbie...
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 






From jdhunter at nitace.bsd.uchicago.edu  Mon Jun 24 15:03:57 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Mon, 24 Jun 2002 14:03:57 -0500
Subject: ? and %s placeholders, help?
References: <af5oq7$sj1$1@news8.svr.pol.co.uk>
Message-ID: <m2k7oolcrm.fsf@video.paradise.lost>

>>>>> "Duncan" == Duncan Smith <buzzard at urubu.freeserve.co.uk> writes:

    Duncan> Can anyone please explain the following.  Basically I need
    Duncan> to know the difference between the ? placeholder and the
    Duncan> %s placeholder.  I have only been able to find examples of
    Duncan> the use of '?' in situations where it clearly works
    Duncan> (integers?).  But I have no good idea why I cannot use it
    Duncan> as below.  Or am I stuck with having the statement parsed
    Duncan> each time I execute it?

As far as I know, the '?' syntax is used in the perl DBI but nowhere
in the python MySQLdb.  So forget about that one.

The %s in python is string interpolation, but there is a special way
to do it in MySQLdb.  You can do

>>> query = "INSERT INTO %s (%s) VALUES %s"
>>> tbl = 'AA_temp'
>>> vars = 'var1, var2, var3, var4, var5, var6, var7'
>>> vals = '("level1", "level1", "level1", "level1", "level1", "level1",
"level1")'
>>> curs.execute(query % (tbl, vars, vals))

but it's not the right way.  When converting the value arguments, you
want to use MySQLdb's built in conversion functions, otherwise strings
and other things won't be properly quoted.

To start with a simple example from the docs:

c.execute("""SELECT spam, eggs, sausage FROM breakfast
            WHERE price < %s""", (max_price,))

Note that this *is not* the same as

c.execute("""SELECT spam, eggs, sausage FROM breakfast
            WHERE price < %s""" % max_price)

The former uses the MySQLdb built in conversions (which are
extensible), the latter uses python's string format conversions.

Your example is more involved, but something like this should work:

import string
(val1, val2, val3) = (1, 'John', 6.0)
tbl = 'mytable'
vars = ('var1', 'var2', 'var3')
vals = (val1, val2, val3)
valsfmt = len(vals) * '%s,'
valsfmt = valsfmt[:-1]  # valsfmt == '%s,%s,%s'

# use python string format to build the mysqldb query string
# query == 'INSERT INTO mytable (var1,var2,var3) VALUES %s,%s,%s'
query =  "INSERT INTO %s (%s) VALUES %s" % \
	(tbl, string.join(vars, ','), valsfmt) 


# use mysqldb to format the values
curs.execute(query, vals)

Don't know if this will help you with your efficiency problem, but at
least you'll be doing the conversions properly.

John Hunter







From hsauer at marnie.teuto.de  Sun Jun 23 13:50:53 2002
From: hsauer at marnie.teuto.de (Henning Sauer)
Date: 23 Jun 2002 19:50:53 +0200
Subject: Trouble with urllib, auth and logging into Netgear RT314
Message-ID: <m3u1ntyjcy.fsf@marnie.henning-sauer.de>

Hello,

when I try to get a information from my Netgear RT 314 Router through
it's Webinterface with Python, I have to authenticate with user and
password. I do this with passing a URL like this:

result = urllib.urlopen("http://user:password at 192.168.0.1")

With this I get prompted one the console for user and password. 
When I retry this immediately I don't get asked for user and password.
This problem seems to be specific to Python and the Netgear RT 314,
because when I try to connect with browser with the above URL it works,
and also when try another URL with user password URL from Python it also
works.
How can I solve this problem ?

Bye,
Henning



From achim.domma at syynx.de  Mon Jun 17 18:15:38 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Tue, 18 Jun 2002 00:15:38 +0200
Subject: ImageMagick for Python incl. Source
Message-ID: <aelmhc$35m$03$1@news.t-online.com>

Hi,

I have just updated http://www.procoders.net/imagemagick with some
documentation and small examples. There is also source code to compile
ImageMagick for Python on non-windows plattforms, but this might be a bit
tricky. Feel free to ask, if you have problems.

greetings
Achim




From morton at dennisinter.com  Sat Jun  8 13:41:28 2002
From: morton at dennisinter.com (damien  morton)
Date: 8 Jun 2002 10:41:28 -0700
Subject: Comparison: HTTP with Java, Perl or Python
References: <3D009911.9000904@gmx.de>
Message-ID: <4abd9ce7.0206080941.564ef212@posting.google.com>

Ingo Linkweiler <i.linkweiler at gmx.de> wrote in message news:<3D009911.9000904 at gmx.de>...
> Hello,
> 
> for a comparison of languages I need a complete "translation" of this short skript into Perl, Tcl and Java:
> 
>  myurl = "http://www.uni-dortmund.de"
>  import urllib
>  file  = urllib.urlopen(myurl)
>  text  = file.read()
>  file.close()
>  print text
> 
> Can you send me some solutions?
> The skripts must be able to run, please do not post "partial" solutions.
> 
> Ingo


Your python code could be a whole lot shorter:

import urllib
print urllib.open("http://www.uni-dortmund.de").read()


From fredrik at pythonware.com  Wed Jun 26 04:10:22 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 26 Jun 2002 08:10:22 GMT
Subject: Parsing strings (\n and \\)
References: <3D188006.7050202@thomas-guettler.de> <mailman.1025017712.20580.python-list@python.org> <lt0S8.45025$n4.10623651@newsc.telia.net> <slrnahijff.m4g.look@pikkukorppi.cs.tut.fi>
Message-ID: <OFeS8.45132$n4.10657162@newsc.telia.net>

Simo Salminen wrote:

> safe(?) version:
>
> >>> PARSED = eval(UNPARSED, {"__builtins__":None}, {})
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
>   File "<string>", line 0, in ?
> NameError: name '__import__' is not defined

now try this:

UNPARSED = "'*'*10000000*2*2*2*2*2*2*2*2*2"

</F>




From deathtospam43423 at altavista.com  Mon Jun 10 09:04:18 2002
From: deathtospam43423 at altavista.com (fishfinger)
Date: 10 Jun 2002 13:04:18 GMT
Subject: Newby: What's the problem with this string?
Message-ID: <Xns9229E662C1E0AAusYourStandingInIt@130.133.1.4>

For some reason I can't read this string from the command line ...

......

Khaksar, a founding member of the Taliban, said he contacted the Americans 
because he feared the Islamic movement had been hijacked ??? first by 
Pakistan's powerful intelligence agency and then by bin Laden and his al-
Qaida group.

Khaksar said he and others in the Taliban wanted to ''keep Afghanistan for 
Afghans'' but found themselves marginalized because of bin Laden's 
influence over Mullah Omar. Bin Laden donated suitcases full of money to 
finance the Taliban's war-effort against the northern-based alliance led by 
the late guerrilla leader, Ahmed Shah Massood

......

Why is that? And how can I make it work properly? The method I'm using is 
.......

ttt = raw_input("a string:")


-- 
netvegetable at excite.com


From johnroth at ameritech.net  Thu Jun 20 13:29:48 2002
From: johnroth at ameritech.net (John Roth)
Date: Thu, 20 Jun 2002 13:29:48 -0400
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com> <lkznxqjqx2.fsf@pc150.maths.bris.ac.uk> <3D11A21D.9010809@mxm.dk> <uh3e1usajle42c@news.supernews.com> <3D11DA9B.163A7BD0@info.unicaen.fr>
Message-ID: <uh446elev7af68@news.supernews.com>

"Jerzy Karczmarczuk" <karczma at info.unicaen.fr> wrote in message
news:3D11DA9B.163A7BD0 at info.unicaen.fr...
> John Roth after Michael Hudson after Max M.:
>
> > > Goto's are simple ;-)
> > > def goto(lable):
> > >      lable()
> > > def start():
> > >
> > >      print 'starting'
> > >      goto(lable2)
> > >
> > > def lable1():
> > >
> > >      print 'at lable 1'
> > >      goto(stop)
>
> etc.
>
> > Those aren't goto's in the classical sense. For a real goto,
> > you need to think assembler: if it's a statement anywhere,
> > you can branch to it, and who cares about the execution
> > history! Having the compiler manage your flow control is
> > for sissys!
>
> Gentlemen, what are you talking about?
> (I confess I didn't follow this thread, so forgive me if I
> try to open a non-existent door)
>
> Before, John Roth suggested using structured exceptions as GOTOs,

Nope. I did no such thing. I pointed out that structured
exceptions were what drove the last goto's out of more
modern languages. Structured programming is what got
rid of most of them.

> and Michael Hudson wrote something like that:
>
> > There are also technical difficulties implementing it (the
> > blockstack), or I'd have probably done goto functionality in
> > bytecodehacks :)
>
>
> Well...
>
> 1. The "solution" of Max M. is erroneous for two reasons.
>    A. A procedure call stacks the return address, so a few hundreds
>       such gotos, and the system blows up.
>       Unfortunately Python does not optimize tail calls, I WONDER WHY;
>       for a member of the Functional Programming Church this is
>       a mortal sin, and I am serious.
>    B. The code chunks linked in such a way provide different variable
>       scopes, so there is little sense in calling it a "branching".
>
> 2. "Structured exception" is something quite heavy, it may be heavier
>    than the setjump/longjump contraption, and may also be useless to
>    those who simply would love to do some Python programming ? la
>    Fortran.
>
> 3. No, in order to speak about genuine gotos *you don't need to think
>    assembler*.
>    What you need is to understand correctly the notion of continuation
>    (something quite standard - again - in the Functional Programming
>    world). This is a high-level model of "flat" branching, but may be
>    used to more exquisite constructions, such as backtracking.

That's still not a goto in the classical sense. In the classical sense,
a goto allows you to go to any statement. The point you're missing is
that the goto statement is a denial of structure - since it can go
anywhere, you can't make any assumptions about the program
structure. The piece of code you're looking at can be mysteriously
activated from somewhere else at any time. The technique is called
"spaghetti coding," and is not recommended by anyone sane (except
for job security, of course.)

Continuations, on the other hand, have been discussed, and may
very well wind up in Python some day. See the discussion around
a variant called 'stackless.'

>    I don't believe that there are any technical difficulties in the
>    implementation of gotos at the low level. At the high level this
>    is then a question of the language structure, as seen by Van
Rossum,
>    the lack of gotos was a conscious decision.

At the time he designed the language, I suspect that if anyone had
brought it up, he'd have been laughed out of the room. We're talking
late 80's, early 90s here, not the '60s or '70s.

John Roth
>
>
> Jerzy Karczmarczuk
> Caen, France




From marklists at mceahern.com  Wed Jun 12 15:24:04 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 12 Jun 2002 14:24:04 -0500
Subject: get a list of classes in this module
In-Reply-To: <NCBBLFCOHHDIKCAFGCFBCEEKLEAA.mark@mceahern.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBEEEMLEAA.marklists@mceahern.com>

> I know I can use inspect.getmembers() to do this, but what if I
> want to put
> the helper method in the same file as the classes (i.e., validation.py)?
> How do I refer to *this* module?  Why do I want the helper in the same
> module?  Well, heck, why not?  ;-)

What I'm doing currently is using the base class, which I know will be
there, and using it to get the module and from there the subclasses:

class CreditCardValidation:
  ...

def validation_factory():
  c = CreditCardValidation
  m = inspect.getmodule(c)
  all_classes = inspect.getmembers(m, inspect.isclass)
  d = [k, v for k, v in all_classes.items() if issubclass(v, c)]

Then I doh! realized I can just use $baseclass.__subclasses_() method!

def validation_factory():
  classes = CreditCardValidation.__subclasses__()
  d = {}
  for c in classes:
    d[c.__name__] = c
  return d

Cheers,

// mark

p.s.  If I'm misusing the "factory" name, I'm all ears for other suggestions
for the name of this method.

-





From rpm1deletethis at nospamfrontiernet.net  Sun Jun 23 05:35:15 2002
From: rpm1deletethis at nospamfrontiernet.net (RPM1)
Date: Sun, 23 Jun 2002 05:35:15 -0400
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at>
Message-ID: <uhb5ii5b16dle8@corp.supernews.com>

"Siegfried Gonzi" <siegfried.gonzi at kfunigraz.ac.at> wrote
> I am not any longer willing to try the calculation on my Linux box and
> Python 2.2 (20 Linux power off  shut downs are enough in a day; should I
> destroy my laptop?).


Just curious, is your laptop a Hewlitt Packard?

Patrick





From andreia.31jesus at bol.com.br  Tue Jun 11 09:45:04 2002
From: andreia.31jesus at bol.com.br (emagrecersemfome.cjb.net)
Date: Tue, 11 Jun 2002 10:45:04 -0300
Subject: Emagreça 1,5Kg por semana!!  Sem dietas e sem fome !!!
Message-ID: <20020611134445.8100DC94BB@srv11-sao.terra.com.br>


DESEJA EMAGRECER 1,5Kg POR SEMANA??? SEM DIETAS, 
E SEM FOME!!!
Veja em http://www.emagrecersemfome.cjb.net .

Pedidos somente para:  andreia.31jesus at bol.com.br

Emagre?a 1,5Kg por semana, tira a fome sem dietas e ? 90% 
natural.
Psor?ase? Queda de cabelo? Asma? ?lcera e gastrite? Rinite?

http://www.emagrecersemfome.cjb.net

Tel.:(021-31) 3487-3702  /  (021-31) 9666-0580.

--------------------------------------------------------------------
Este e-mail promocional ser? enviado somente uma vez.



























































-----------
NEVER SEND SPAM. IT IS BAD.




From tim.one at comcast.net  Tue Jun 18 13:51:53 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 18 Jun 2002 13:51:53 -0400
Subject: Pop a list from beginning ? and memory saving...
In-Reply-To: <aem8e8$4kq$1@news.idiom.com>
Message-ID: <BIEJKCLHCIOIHAGOKOLHAELEDEAA.tim.one@comcast.net>

[James T. Dennis]
> ...
>  So, the critical question is:
>
> 	Is the list's reverse() method done in place

Yes.

> and does it run in linear time

Yes.

> (or better?

No.  It does len(list)/2 swaps.




From chris.gonnerman at newcenturycomputers.net  Mon Jun  3 08:28:37 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Mon, 3 Jun 2002 07:28:37 -0500
Subject: List
References: <20020602205610.31575.qmail@linuxmail.org> <009b01c20a7e$12a5ad80$0101010a@local> <mailman.1023062343.28782.python-list@python.org> <3cfb07cf_1@news.iprimus.com.au>
Message-ID: <002601c20afa$3430a460$0101010a@local>

----- Original Message -----
From: "Gold Fish" <occeanlinux at linuxmail.org>


> Mike C. Fletcher wrote:
>
> > I don't know if you'll consider this a more elegant way to divide up the
> > lists, but I think it's a little more beautiful than the index-mucking
> > approaches IMO:
> >
> > def into( source, count=3 ):
> >      source = source[:]
> >      return intoDestructive( source, count )
> >
> > def intoDestructive( source, count=3 ):
> >      if source and not count:
> >          raise ValueError( """Cannot divide into 0-length
sub-elements"""
> >          )
> >      elif not source:
> >          return []
> >      results = []
> >      if count > 0:
> >          while source:
> >              results.append( source[:count])
> >              del source[:count]
> >      else:
> >          while source:
> >              results.append( source[count:])
> >              del source[count:]
> >      return results
> >
> > It will be very slow for very large lists with positive slice values
> > (something the index-mucking approach doesn't suffer from), but for most
> > reasonably sized lists, should be fairly good.  It handles negative
> > steps as well, so you can slice up the list from the right _or_ the left
> > side :) .
> >
> > Enjoy all,
> > Mike
> >
> > Chris Gonnerman wrote:
> > ...
> >>
> >>     Lists = []
> >>     N = 2
> >>
> >>     adj = 0
> >>
> >>     if len(BigList)%N:
> >>         adj = 1
> >>
> >>     for i in range(len(BigList)/N+adj):
> >>         Lists.append(BigList[i*N:(i+1)*N])
> >>
> >> (If anyone knows a more elegant solution please post it,
> >> as I am getting headaches trying to think of one.)
> > ...
> > _______________________________________
> >    Mike C. Fletcher
> >    http://members.rogers.com/mcfletch/
>
> Sorry for the very stupid question but i still don't get my goal, that's
is
> using Mike C. Fletcher method i will generate the list of elements again
> and don't know what size it is. For example :
> [['p1','p2','p3'],['p4','p5','p6'],['p7','p8','p9'],['p10']]

Eh?  What size it is?  len(Lists) answers that question.

> I want to seperate this list again in order to using them later .It look
> like you go to the realestate agent website and find the list of property
> then you just want to display 3 properties per pages. And you can go to
> other pages which also contain 3 properties as well until the last pages
> which has only 1 items. For example:
> Page 1: ['p1','p2','p3']
> Page 2: ['p4','p5','p6']
> Page 3: ['p7','p8','p9']
> Page 4: ['p10']
> I didn't sleep well due to this problem Can any help me.

Maybe if we understood the problem better.  The solutions given by both Mike
and myself break the main list down into sublists; we stored them in an
"outer" list because, well, they have to be stored SOMEWHERE.

Are you envisioning:

    list1 = ['p1','p2','p3']
    list2 = ['p4','p5','p6']
    list3 = ['p7','p8','p9']
    list4 = ['p10']

that is, each sublist has its own variable name?  It can be done but it is
almost always the wrong way.  Using the methods we gave above, the result is
(in effect):

    List[1] = ['p1','p2','p3']
    List[2] = ['p4','p5','p6']
    List[3] = ['p7','p8','p9']
    List[4] = ['p10']

which is basically equivalent to what I think you are trying to do.

Could you maybe explain the *entire* project?  Or do you, like, work for the
CIA?  (I'd rather you not tell me if you then have to kill me.)

Chris Gonnerman chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net






From tim_one at email.msn.com  Mon Jun  3 00:18:53 2002
From: tim_one at email.msn.com (Tim Peters)
Date: Mon, 3 Jun 2002 00:18:53 -0400
Subject: Which is fastest dict or list.index() ?
In-Reply-To: <83n0udrs8x.fsf@panacea.canonical.org>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEPIPJAA.tim_one@email.msn.com>

[Kragen Sitaker, to Raymond Hettinger]
> I'm curious how your skiplist version compares with the built-in
> Python dict type on performance; I'd been thinking about building a
> skiplist ordered dictionary extension type myself, but if you've
> already done it, maybe I should just use yours.

One of the better-kept secrets of the StandaloneZODB project

    <http://www.zope.org/Products/StandaloneZODB>

is that it comes with several flavors of C-coded B-Tree mappings and B-Tree
based sets, where B-Tree nodes and bucket leaf nodes are Python objects in
their own right, and you get persistence "for free".  Operations include
range search, intersection, union, difference, and weighted flavors of
intersection and union that compute linear combinations of integer-valued
maps (think of that as a tiny subset of NumPy <wink>).

In practice, they're sluggish compared to Python dicts, and especially for
string-keyed B-Trees (string-keyed Python dicts use all sorts of
implementation tricks for speed).  OTOH, B-Trees support O(log N) insertion
and deletion while preserving sorted order, can spill pieces to disk
gracefully in contexts dicts can't, have good locality of reference during
traversal, and are more memory-efficient than Python dicts, especially for
int-keyed and int-valued B-Trees (for Zope Reasons, those are special-cased
to store C ints directly as C ints rather than as pointers to fat Python int
objects).  One Big Surprise is that len(BTree) can take time proportional to
the number of elements in the tree:  to reduce the load on the persistence
machinery during mutation, B-Tree nodes don't save a count of the number of
key+value pairs reachable from them, and counting them requires traversing
all the leaf buckets.

They're powerful data structures and should be better known.  They should
also be better documented <wink>.  They're fresh on my mind because I'm
working on optimizing the snot out of some common B-Tree operations.

if-you-can-figure-out-how-to-use-'em-you'll-like-'em-ly y'rs  - tim





From jan.eric-enlund at multi.fi  Sun Jun 16 08:51:45 2002
From: jan.eric-enlund at multi.fi (Jan-Eric)
Date: Sun, 16 Jun 2002 15:51:45 +0300
Subject: Snort alert tail...
Message-ID: <AB0P8.107$W91.886@news1.global-one.fi>

HEllo !
I'm trying to write a dynamic firewall script in python that scans the Snort 
alert file like 'tail -f' and takes action based on the infomation it gets 
from that file. But I can't get the 'tail' function to work.It reads the 
file, but any new information that Snort is writing to the file doesn't 
show up to the script.

ex.
file = open('/var/log/snort', 'r')
while 1:
        file = file.read()
        print file
....

Any suggestions???
Jan-Eric / Finland


From justin at iago.org  Sat Jun 15 17:13:18 2002
From: justin at iago.org (Justin Sheehy)
Date: Sat, 15 Jun 2002 17:13:18 -0400
Subject: python install through fink, mac os x
In-Reply-To: <7fe97cc4.0206142251.164d9ec5@posting.google.com> (xah@xahlee.org's
 message of "14 Jun 2002 23:51:36 -0700")
References: <7fe97cc4.0206142251.164d9ec5@posting.google.com>
Message-ID: <x7n0twjle9.fsf@ra.iago.org>

xah at xahlee.org (Xah Lee) writes:

> Python installation through fink failed with the following message.

> can anyone help?

I don't know much about fink, but a quick scan of the output made it
obvious that the problem was with the download of BerkeleyDB, not a
problem with installing Python itself.

-Justin

 


> [xah at localhost ~][Fri Jun 14,23:48:52]
> fink install python
> sudo /sw/bin/fink install python
> Password:
> Reading package info...
> Information about 435 packages read in 5 seconds.
> The following 9 additional packages will be installed:
>  db3 db3-shlibs expat gdbm gdbm-shlibs gmp gmp-shlibs manconf tcltk
> Do you want to continue? [Y/n] y
> curl -f -L -s -S -O http://www.sleepycat.com/update/3.3.11/db-3.3.11.tar.gz
> curl: (22) The requested file was not found
> ### curl failed, exit code 22
> Downloading the file "db-3.3.11.tar.gz" failed.
>
> (1)  Give up
> (2)  Retry
>
> How do you want to proceed? [2]





From amckay at merlintechnologies.com  Fri Jun  7 13:07:19 2002
From: amckay at merlintechnologies.com (Andy McKay)
Date: Fri, 7 Jun 2002 10:07:19 -0700
Subject: Newby: How do I strip HTML tags?
In-Reply-To: <3d00e1b1.47853937@news.t-online.de>
References: <Xns92278CC566FEAusYourStandingInIt@130.133.1.4> <3d00e1b1.47853937@news.t-online.de>
Message-ID: <20020607170719.GB11618@mckay.merlintechnologies.com>

> This is a quite straight forward function: 

This of course assumes that there are no "<" or ">" anywhere else in
attributes etc...

I would recommend using HTML Parser to get an accurate representation, or
even stripogram which is pretty good (note it doesnt need Zope at all),
although the html2text isnt perfect:

>>> from stripogram import html2text
>>> html2text("<b>This is a <i>test</i></b>")
'This is a test'
-- 
  Andy McKay
  Merlin Technologies




From jepler at unpythonic.net  Thu Jun 20 15:50:36 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Thu, 20 Jun 2002 14:50:36 -0500
Subject: Pop a list from beginning ? and memory saving...
In-Reply-To: <aep4jb$1q8h$1@news.idiom.com>
References: <aem8e8$4kq$1@news.idiom.com> <BIEJKCLHCIOIHAGOKOLHAELEDEAA.tim.one@comcast.net> <mailman.1024424930.10523.python-list@python.org> <aep4jb$1q8h$1@news.idiom.com>
Message-ID: <20020620195034.GA17857@unpythonic.net>

I seem to recall a discussion, probably a few years ago by now, for a
list implementation that was efficient to grow or shrink from either
end.  I don't recall why it wasn't adopted.  I also can't find the
discussion. :(

I suspect that having unallocated space at both the front and the back
of the list, growing according to the same exponential curve when the
end is reached, would give this kind of behavior.  

The following isn't finished, but append, insert and pop seem to work
properly, and on my machine already beat the builtin list for at-front
operations at 10000 items (test 3/4) but lag in at-end operations (test
1/2):
    10000   test1   test2   test3   test4
    list    0.042   0.075   1.23    1.28
    delist  0.051   0.23    0.22    0.47

    100000  test1   test2   test3   test4
    list    0.48    0.92    151.2   127.6
    delist  0.56    2.46    5.45    9.31
(I didn't let a 1M-item test finish)

Jeff


def roundupsize(n):
    nbits = 0
    n2 = n >> 5
    while 1:
	n2 >>= 3
	nbits += 3
	if not n2: break
    return ((n>>nbits)+1)<<nbits

class delist(list):
    def __init__(self, data=[]):
	super(delist, self).__init__(data)
	self.start = 0
	self.__super = super(delist, self)

    def __growstart(self):
	l = len(self)
	nl = roundupsize(l)
	self[0:0] = [None] * (nl-l)
	self.start = self.start + nl-l

    def __len__(self):
	return self.__super.__len__() - self.start

    def pop(self, idx=None):
	if idx is None:
	    return self.__super.pop()
	if idx == 0:
	    ret = self[idx]
	    start = self.start
	    self[start] = None
	    self.start = start + 1
	    return ret
	else:
	    if idx > 0: idx = idx + len(self)
	    return self.__super.pop(idx + self.start)

    def __getitem__(self, idx):
	self.__super.__getitem__(idx + self.start)

    def insert(self, idx, item):
	if idx == 0:
	    if not self.start:
		self.__growstart()
	    self.start = self.start - 1
	    self[self.start] = item
	    return
	self.__super.insert(idx+self.start, item)

    def __repr__(self):
	return "<dlist +%d %s>" % (self.start, list(self[self.start:]))

def test():
    import time

    def test1(x):
	# Test speed of append
	assert not x
	for i in r:
	    x.append(i)
	assert len(x) == len(r)
	return x

    def test2(x):
	# Test speed of pop-from-end
	assert len(x) == len(r)
	for i in r:
	    x.pop()
	assert not x

    def test3(x):
	# test speed of insert-at-start
	assert not x
	for i in r:
	    x.insert(0, i)
#	    print x
	assert len(x) == len(r)
	return x

    def test4(x):
	# test speed of pop-from-start
	assert len(x) == len(r)
	for i in r:
	    x.pop(0)
	assert not x

    def run(f, arg):
	t0 = time.time()
	ret = f(arg)
	dt = time.time() - t0
	print "%s/%s/%d: %f" % (f.__name__, type(arg), len(r), dt)
	return ret

    for rr in (10000, 100000, 1000000):
	r = range(rr)
	for ll in list(), delist():
	    ll = run(test1, ll)
	    run(test2, ll)
	for ll in list(), delist():
	    ll = run(test3, ll)
	    run(test4, ll)

test()




From giulio.agostini.remove.this at libero.it  Mon Jun  3 11:14:54 2002
From: giulio.agostini.remove.this at libero.it (Giulio Cespuglio)
Date: Mon, 3 Jun 2002 15:14:54 +0000 (UTC)
Subject: htmllib examples?
Message-ID: <ut1nfuo5skj0d5g0ff6dngo1icelpdfmg1@4ax.com>

Hi there,

I'm trying to use htmllib, but I can't find any examples in the
documentation
(http://www.python.org/doc/current/lib/module-htmllib.html).
Is there a (very well hidden) example codebase in www.python.org?
Where else can I look?

Thanks a lot guys,
Giulio


From djc at object-craft.com.au  Sat Jun  8 08:24:00 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 08 Jun 2002 22:24:00 +1000
Subject: Creating Dynamic Web Pages (ZOPE)
In-Reply-To: <3D01CB91.8B98F2D7@optushome.com.au>
References: <Pine.LNX.4.21.0206081919430.1782-100000@localhost.localdomain>
	<3D01CB91.8B98F2D7@optushome.com.au>
Message-ID: <m37klagdr3.fsf@ferret.object-craft.com.au>

>>>>> "Tim" == Tim Churches <tchur at optushome.com.au> writes:

Tim> Michael Hall wrote:
>>  I notice no-one has suggested good ol' CGI. For a relatively
>> simple job such as this, a few basic Python CGI scripts, some
>> templates and MySQLdb could also be used ... oh, just noticed the
>> MS-SQL database. There is probably a module around to cope with
>> MS-SQL, no?

Tim> Object Craft have a Python wrapper for MS SQL Server. See
Tim> http://www.object-craft.com.au

Depending upon how you are going to access MS SQL Server, the Sybase
module might be a better choice.

If you are compiling against the MS SQL Server libraries, then the
MSSQL server module should be OK.  If you are using FreeTDS then the
Sybase module is probably better (although still not perfect).  If
using Sybase client libraries then either the MSSQL or Sybase modules
can probably be used.

Confused yet?

Tim> And while you are there, have a look at their Albatross CGI-based
Tim> stateful Web application framework, which could be just what is
Tim> required for the rest of the Web site envisaged.  In my
Tim> experience, nearly all Web applications and database-driven Web
Tim> sites end up needing to retain state information sooner or
Tim> later. Might as well start with a framework which can go the
Tim> distance. Albatross is especially nice in this respect because
Tim> you can choose alternative application models, starting with
Tim> "SimpleApp" which is just what the name suggests, and later
Tim> progress to more complex application models through mixins
Tim> without having to start from scratch or throw away your existing
Tim> code.

Thanks for the plug Tim.

One of the goals I had when developing Albatross was to try to keep
the toolkit subservient to your code, not the other way around.
Whether I met that goal or not is a matter of opinion :-).  If you
want to craft an application from a collection of code fragments which
run within an application server then Zope or Webware might be better
for you.

This page:

   http://www.object-craft.com.au/projects/albatross/albatross/tug-simple2.html

shows a very simple CGI application which just uses the templating
parts of Albatross.  If you do not want to use the application
functionality, you don't have to.

The application functionality is there to take most of the drudgery
out of organising your application code, maintaining sessions, and
working out just which page / request has been requested by the
browser.

When building a collection of small web applications (especially for
the intranet) you find yourself doing the same infrastructure and glue
things over and over.  Albatross is designed to let you spend your
effort on the interesting parts of the applications.

Tim is correct about the ability to change application types with
minimal change to your application code.  The toolkit calls your
request processing logic via the same interface regardless of how you
have decided to structure and deploy your application.  This means you
can develop using CGI then deploy using mod_python.  You can even
switch between server and client side sessions without changing much
more than the mainline.

Another important goal in Albatross is to keep the toolkit line count
very small so that people can easily look inside and see how it works.
Code is organised as a collection of mixin classes so that you can
easily substitute your own functionality where the Albatross code does
not suit you.

- Dave

-- 
http://www.object-craft.com.au





From hesterloli at hotmail.com  Wed Jun 19 09:36:57 2002
From: hesterloli at hotmail.com (George Hester)
Date: Wed, 19 Jun 2002 13:36:57 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com> <mailman.1024460666.24801.python-list@python.org> <B8XP8.47925$GY.15259265@twister.nyroc.rr.com> <mailman.1024490732.9209.python-list@python.org>
Message-ID: <ZN%P8.33112$uk2.13703915@twister.nyroc.rr.com>

OK

--
George Hester
_________________________________
"Chris Gonnerman" <chris.gonnerman at newcenturycomputers.net> wrote in message
news:mailman.1024490732.9209.python-list at python.org...
> ----- Original Message -----
> From: "George Hester" <hesterloli at hotmail.com>
>
>
> > I don't want to learn the language.
>
> Ah, so.  Then go away, and come back when you do want to learn.  And don't
> whine.
>
>
> Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
> http://newcenturycomputers.net
>
>
>
>




From gerhard at bigfoot.de  Thu Jun 20 09:19:19 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 20 Jun 2002 13:19:19 GMT
Subject: MySQLdb question
References: <aesi66$cgb$1@sunsite.icm.edu.pl>
Message-ID: <slrnah3leo.18g.gerhard@lilith.my-fqdn.de>

In article <aesi66$cgb$1 at sunsite.icm.edu.pl>, =?Latin-2?Q?Micha=B3?= Kurowski wrote:
> Hi,
> 
> I'm working on a small project involving sql (MySQLdb-0.9.1) and I constantly
> have to "get around" a  quite weird problem:
> 
> I cannot find a way to pass a sql statement a table as a parameter.

These are two different things:

(1) cursor.execute("... %s ...", "foo")
(2) cursor.execute("... %s ..." % foo)

(1) is the DB-API method to get proper quoting of parameters for SQL.
(2) is the normal Python quoting

For table names, you need (2), for about everything else, you need (1).

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From m2 at plusseven.com  Tue Jun 18 12:55:48 2002
From: m2 at plusseven.com (Alex Polite)
Date: Tue, 18 Jun 2002 18:55:48 +0200
Subject: pychecker + XEmacs
In-Reply-To: <aekok8$7p118$1@ID-59885.news.dfncis.de>
References: <3d0de483$0$3881$e4fe514c@dreader4.news.xs4all.nl> <aekok8$7p118$1@ID-59885.news.dfncis.de>
Message-ID: <20020618165548.GB29257@matijek.plusseven.com>

On Mon, Jun 17, 2002 at 03:34:02PM +0200, Thomas Heller wrote:
> python-mode.el from CVS contains pychecker support.
> C-c C-w runs 'pychecker --stdlib <current-python-file>',
> and shows the results in a *compilation* window.
> Then you can M-x next-error or M-x previous-error
> to your hearts content.

Interesting.  Where is the CVS repository for python-mode.el?

-- 
Alex Polite
http://plusseven.com/gpg/




From quinn at lira.ugcs.caltech.edu  Sat Jun  8 23:36:07 2002
From: quinn at lira.ugcs.caltech.edu (Quinn Dunkan)
Date: 9 Jun 2002 03:36:07 GMT
Subject: email: msg.add_payload not synonymous with msg.attach?
References: <slrnafvtu6.7q0.mlh@vier.idi.ntnu.no>
Message-ID: <slrnag5j96.17b.quinn@lira.ugcs.caltech.edu>

On Fri, 7 Jun 2002 00:01:10 +0000 (UTC), Magnus Lie Hetland
<mlh at vier.idi.ntnu.no> wrote:
>I've been fiddling with the email package, and tried to construct a
>Message object, using add_payload to add the email body (It's a
>text/plain message). This worked fine, except for a deprecation
>warning telling me to use attach instead. But when I did, my program

In 1.0 (comes with python 2.2.1), attach is defined as 'attach = add_payload'
so it's synonymous.  It also doesn't have any mention of 'add_payload" being
deprecated.  I hope it's not really, since I use it all the time.

CVS is probably not at 1.0 anymore.

>crashed with the message "TypeError: string payload expected: <type
>'list'>" (from Generator.py). Well... My payload was indeed simply a
>string, so I don't understand the error message. Also, I don't
>understand how this discrepancy could have occurred if, as the docs
>claim, add_payload and attach are synonymous.

Well, at least the 1.0 verson of email.Message.Message has an annoying bug that
relates _payload being either a string or list depending on how many times
add_payload() has been called (not a great design, IMO).  The bug arises
because Message assumes a list _payload in the presence of the multipart/*
type, which leads to trouble because some multipart/* messages only have one
part (I'm not sure if this is legal according to the RFCs, but many messages
out there do that, including Crispin's MIME torture test (which I would
recommend be used in the testing of the package).

What the error msg probably means is that your message somehow has multiple
payloads, but doesn't have a multipart/* type.  Maybe they changed attach
to always make _payload a list or something.

>(The code is very simple -- basically just instantiating Message and
>adding a payload.)
>
>I'm using the email package from CVS (downloaded everything a day or
>two ago).

I wouldn't be surprised if the docs are not up to date with the latest CVS.
You should probably either use 1.0 or check the Message.py source.


From christophe.delord at free.fr  Thu Jun 20 14:27:32 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Thu, 20 Jun 2002 20:27:32 +0200
Subject: How to represent the infinite ?
References: <3d11ee9b$1@news.swissonline.ch>
	<20020620193230.53d58f8f.christophe.delord@free.fr>
	<aet4p6$33kk$1@nntp6.u.washington.edu>
Message-ID: <20020620202732.68e67873.christophe.delord@free.fr>

On 20 Jun 2002 17:52:38 GMT
"Donn Cave" <donn at u.washington.edu> wrote:

With Python 2.2 and Linux it works. The example I gave is a copy-paste from IDLE. It also works with my Python 1.5.2 and 2.3 (I'm under Linux on an Intel processor)

So it seems not to be implemented in every binaries and/or platform and/or compiler and/or ... So this solution may not be portable and it may be safer to make a special class for the infinite like in my first reply.

Here is the proof that I'm not totaly drunk yet ;-)

[christ at localhost py]$ python1.5
Python 1.5.2 (#1, Apr  3 2002, 18:16:26)  [GCC 2.96 20000731 (Red Hat Linux 7.2 2 on linux-i386
Copyright 1991-1995 Stichting Mathematisch Centrum, Amsterdam
>>> x=1e1000
>>> x
inf
>>>
[christ at localhost py]$ python2.2
Python 2.2 (#1, Apr 12 2002, 15:29:57)
[GCC 2.96 20000731 (Red Hat Linux 7.2 2.96-109)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> x=1e1000
>>> x
inf
>>>
[christ at localhost py]$ python
Python 2.3a0 (#2, Jun 20 2002, 19:48:46)
[GCC 2.96 20000731 (Red Hat Linux 7.3 2.96-110)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> x=1e1000
>>> x
inf
>>>
[christ at localhost py]$






> Quoth Christophe Delord <christophe.delord at free.fr>:
> |
> | There is another simple solution. Just use a float that is bigger than any 64-bit float. For example 10^1000. This float has a special encoding meaning +oo !
> | So no need to redefine anything else ;-)
> |
> | >>> x=1e1000
> | >>> x
> | inf
> ...
> | >>> x-x
> | nan
> 
> $ python
> Python 2.0 (#1, Dec 18 2000, 10:19:52) [C] on osf1V4
> Type "copyright", "credits" or "license" for more information.
> >>> x = 1e1000
> >>> x
> 1.7976931348623157e+308
> >>> x > 10
> 1
> >>> x - x
> 0.0
> >>>
> 
> I missed the original post, hope this helps.
> 
> 	Donn Cave, donn at u.washington.edu
> 
> | > Not need to say to me that I can use 99999999999999 or -99999999999999 or I
> | > do not know what.
> | > Nor from a MAX_INT of the machine... bus Python can go higher.
> | > 
> |
> |
> |
> |
> | -- 
> |
> | (o_   Christophe Delord                   _o)
> | //\   http://christophe.delord.free.fr/   /\\
> | V_/_  mailto:christophe.delord at free.fr   _\_V


-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From bokr at oz.net  Sat Jun 29 15:14:46 2002
From: bokr at oz.net (Bengt Richter)
Date: 29 Jun 2002 19:14:46 GMT
Subject: Newbie that don't understand
References: <AIlT8.13914$p56.4478653@newsb.telia.net>
Message-ID: <afl0v6$gvl$0@216.39.172.122>

On Sat, 29 Jun 2002 17:00:16 GMT, "Martin St?hl" <martin.stahl at telia.com> wrote:

>I have never been programming before and now I am reading the
>"Non-Programmers Tutorial For Python"
>http://www.honors.montana.edu/~jjc/easytut/easytut/node9.html#SECTION0091000
>0000000000000
>And don't understand the following
>I quote
>
>"To start off this chapter I am going to give you a example of what you
>could do but shouldn't (so don't type it in):
>
>a = 23
>b = -23
>
>if a < 0:
23 is not less than 0, so the indented statements(only one here) are (is) not executed,
and a remains bound to the value 23
>    a = -a
>
>if b < 0:
-23 _is_ less than 0, so the indented statement(s)(only one here) _is_ executed,
and b becomes bound to the value of the expression -b which is -(-23) which is positive 23
>    b = -b
>
>if a == b:
at this point a is bound to 23 and b is also bound to a number whose value is 23,
so the indented statement(s) below the if statement is(are) executed
and you see the output printed.
>    print "The absolute values of", a,"and",b,"are equal"

>else:
This block of indented statement(s) would only get executed if the preceding
if statement at the same indentation level were false (i.e., if a==b were
false in this example). But it was true, so this statement doesn't get executed.
>    print "The absolute values of a and b are different"
>
>with the output being:
>The absolute values of 23 and 23 are equal"
As you might expect now ;-)
>
>how can 23 and -23 be equal?
>
It's not saying that. The absolute values are the positive magnitudes of the numbers,
ignoring sign. It would be clearer to write

    a = abs(a)

instead of writing the equivalent code

    if a < 0:
        a = - a

But the result is the same, and that is why the printed message is
referring to "The absolute values."

HTH

Regards,
Bengt Richter


From djmitchell at NOSPAM.optushome.com.au  Fri Jun 21 06:09:02 2002
From: djmitchell at NOSPAM.optushome.com.au (David Mitchell)
Date: Fri, 21 Jun 2002 20:09:02 +1000
Subject: I'd give up Perl tomorrow if only...
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au> <3D12F6FC.3000608@wedgetail.com>
Message-ID: <pan.2002.06.21.10.09.01.691351.4527@NOSPAM.optushome.com.au>

> Among other things, it's one of the reasons I'm tinkering with a Perl
> ORB.
> 
> http://sourceforge.net/projects/perlorb
> 
> If you use a Python ORB, such a Fnorb,
> [http://sourceforge.net/projects/fnorb], on the Python side, you should
> be able to write wrappers for your favourite Perl module pretty easily
> and use them from Python.
> 
> Yes, it's stop-gap, but when you already have a non-trivial Perl module,
> either on CPAN or in-house, it can save a lot of rewriting. What we
> *really* need to do is dynamically create CORBA interfaces for Perl
> modules, but that might be impossible given that Perl is like Python in
> that module and class interfaces aren't pre-declared :(
> 
> Anyway, feel free to email me if you want to try this approach and you
> need guidance.

Thanks Derek, I'll check these out over the weekend.  I knew of the
existence of Fnorb, but not perlorb

I've thought several times "why don't I just start porting e.g. DBI and
DBD::Oracle myself, and maybe someone else will do DBD::*", and we'd then
be on our way.  Time, as always, is the big problem - I just don't have
the time on my hands to take on such a big task myself, and commit to
supporting and updating it afterwards, although I'd happily commit small
chunks of time to such a project.

I'm busy playing with Jython at the moment to see if I can deal with the
problem from another angle.  Like Perl, Java's also got some nice
"generic" solutions (e.g. JDBC) to common problems, and Jython looks like
potentially a really nice fit for a bunch of people I know who are struggling
with WebSphere at present.  WebSphere on sites that need to change rapidly is a
particularly ugly mix...

Regards

David Mitchell


From threeseas at earthlink.net  Wed Jun 26 09:33:49 2002
From: threeseas at earthlink.net (Timothy Rue)
Date: Wed, 26 Jun 2002 13:33:49 GMT
Subject: autocoder redux (was Re: Reminder of a python Project.)
References: <mailman.1025076514.1684.python-list@python.org>
Message-ID: <1622.942T1631T5554565threeseas@earthlink.net>

On 26-Jun-02 02:27:12 David LeBlanc <whisper at oz.net> wrote:
><snip>
>> Timothy Rue wrote:
>> >
>> > On 25-Jun-02 22:38:16 Peter Hansen <peter at engcorp.com> wrote:
>> > >I'm afraid there is no longer any need for an autocoder.
>> >
>> > Geee, maybe you should tell that to NIST
>>
>> No need.  They called me the other day and I let them know, but
>> it was too late to get it into the Computerworld article.
>>
>> Seriously, there's really no issue with software development any
>> more.  The problems are solved.  I guess if you've invented the
>> world's first autocoder, it kinda sux to be you then, eh?
>>
>> -Peter
>> --

>Yes, software development techniques have gotten so good, they're now going
>to have Windows XP running atomic Reactors, Air Traffic Control and the
>Space Shuttle. I expect they where encouraged by US Navy tests of a warship
>being run on Windows NT where the ship (conventional fueled carrier) ended
>up laying dead in the water and tugs had to be summoned to bring her into
>port (source: Procedings of the United States Naval Institute). Gives "blue
>screen of death" a whole new meaning.

>Dave LeBlanc
>Seattle, WA USA


Well considering where you are from....

Guess that means the 60 billion is intentional huh?


---
*3 S.E.A.S - Virtual Interaction Configuration (VIC) - VISION OF VISIONS!*
   *~ ~ ~      Advancing How we Perceive and Use the Tool of Computers!*
Timothy Rue      What's *DONE* in all we do?  *AI PK OI IP OP SF IQ ID KE*
Email @ mailto:timrue at mindspring.com      >INPUT->(Processing)->OUTPUT>v
Web @ http://www.mindspring.com/~timrue/  ^<--------<----9----<--------<



From candiazoo at attbi.com  Thu Jun 13 01:00:11 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Thu, 13 Jun 2002 05:00:11 GMT
Subject: Help with ADO and dynamic insert statement...
Message-ID: <3d0826ec.240888125@netnews.attbi.com>

I have a program written in VB that creates an ado COMMAND object, sets the text
to a dynamic sql string (insert into table (?, ?, ?, etc))...  it works in VB...
this is my first foray into doing something similar with Python and it isn't
working.  Could someone take a gander at this and tell me what I might be doing
incorrectly?

Thanks!

Mike

============================================
segment of code below
============================================

    if status == SUCCESS:
        #   Generate our dynamic insert statement, create new
        #   ADODB connection and command objects which we'll
        #   use to connect to our access database (via the
        #   CatalogMDB DSN object)...

        sql = "insert into works (title, wrk_inst, trs_inst, aas_inst) " + \
              "values (?, ?, ?, ?)"
    
        cnnTempDB = win32com.client.Dispatch('ADODB.Connection')
        cmdInsert = win32com.client.Dispatch('ADODB.Command')
        
        cnnTempDB.Open(TEMP_DB)
        cmdInsert.ActiveConnection = cnnTempDB
        cmdInsert.CommandType = 1
        cmdInsert.CommandText = sql
        
        if not rstOracle.EOF:
            #   Let's bind up our dynamic sql statement...
            
            cmdInsert.Parameters.Append(cmdInsert.CreateParameter("title", 200,
1, len(rstOracle.Fields("title").Value), rstOracle.Fields("title").Value))
            cmdInsert.Parameters.Append(cmdInsert.CreateParameter("wrk_inst",
131, 1, 4, rstOracle.Fields("wrk_inst").Value))
            cmdInsert.Parameters.Append(cmdInsert.CreateParameter("trs_inst",
131, 1, 4, rstOracle.Fields("trs_inst").Value))
            cmdInsert.Parameters.Append(cmdInsert.CreateParameter("aas_inst",
131, 1, 4, rstOracle.Fields("aas_inst").Value))
            cmdInsert.Execute()
            rstOracle.MoveNext()


From emile at fenx.com  Thu Jun 20 09:17:58 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 20 Jun 2002 13:17:58 GMT
Subject: can I call operator overloader in superclass?
References: <H2dQ8.1805$CZ4.569945@news1.news.adelphia.net>
Message-ID: <aCkQ8.19693$EP.1305@sccrnsc03>

Jon J. Morin
> classAListSub(AList):
>         repr_calls = 0
>         def __init__(self, list=[]):
>                 AList.__init__(self, list)
>         def __repr__(self):
>                 AListSub.repr_calls = AListSub.repr_calls + 1
>                 print "Calling __repr__ in superclass"
>                 AList.__repr__(self)
>                 print "Method called %d times." %
(AListSub.repr_calls)
> ...
> ...
> ...
> I can create instances on AListSub by passing in a string, or a list
as in
> >>> x = AListSub('sasquatch')
> >>> x
> Calling __repr__ in superclass
> Method called 1 times.
> Traceback (innermost last):
>   File "<stdin>", line 1, in ?
> TypeError: repr not string
>
> For one, I don't understand the error.  For seconds, can I call
__repr__ in
> the superclass from __repr__ in the subclass?

You should return a string from the __repr__ method:

>>> class test:
...     def __repr__(self):
...             print 'in __repr__'
...
>>> t = test()
>>> `t`
in __repr__
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: __repr__ returned non-string (type None)
>>> class test:
...     def __repr__(self):
...             print 'in __repr__'
...             return "string"
...
>>> t = test()
>>> `t`
in __repr__
'string'
>>>

--

Emile van Sebille
emile at fenx.com

---------



From merkosh at hadiko.de  Sat Jun 22 12:28:49 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Sat, 22 Jun 2002 18:28:49 +0200
Subject: shallow copy vs. deep copy
Message-ID: <MPG.177ec42fc19fcbf598968f@news.rz.uni-karlsruhe.de>

hi,

am am wondering where to draw the line between a shallow copy and a deep 
copy of compound objects.

when creating a shallow copy i could most easily create a new object and 
say 

new.__dict__ = old.__dict__

on the other side i could also:

new.__dict__ = old.__dict__.copy()

which creates a shallow copy of the old instance dictionary.
if all attributes of the class are primitive (like int, long, string) 
which one is the correct way of creating a shallow copy?

also: if i create a deep copy and my instance attributes only contain 
immutable objects there's no difference if i create another copy or just 
pass the reference. if one of the objects attribute gets "changes" it'll 
be reassigned a new object and thats it.
what's the "fine" way here?

Ciao
Uwe


From rhymes at myself.com  Sat Jun 29 11:41:52 2002
From: rhymes at myself.com (Rhymes)
Date: Sat, 29 Jun 2002 17:41:52 +0200
Subject: exception handing
Message-ID: <pan.2002.06.29.17.41.28.926443.440@myself.com>

I'm reading "Learning Python" and in the section "gotchas"
of the exception handling chapter the book itself states
that the exception matching is made by identity not equality.
It also write an example such as this:

>>> ex1 = "spam"
>>> ex2 = "spam"
>>>
>>> ex1 == ex2, ex1 is ex2
(1, 0)

<--- here i get (1, 1) ---->

>>> try:
...    raise ex1
... except ex1:
...    print 'got it'
...
got it
>>> try:
...    raise ex1
... except ex2:
...    print 'Got it'
...
Traceback (innermost last):
  File "<stdin>", line 2, in ?
spam

but in my Python 2.2 i don't get back any error from the interpreter
about the second example... why? What has changed (since 1.5 release)
about objects identity?

--
Rhymes
rhymes at myself.com
" ride or die "


From tjreedy at udel.edu  Fri Jun 14 23:54:23 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sat, 15 Jun 2002 03:54:23 GMT
Subject: Using Python to script a game?
References: <cdc2589b.0206141718.7dc2ae13@posting.google.com>
Message-ID: <PTyO8.226065$%y.20663204@bin4.nnrp.aus1.giganews.com>

"Ben Blonder" <ben at sourcecod.com> wrote in message
news:cdc2589b.0206141718.7dc2ae13 at posting.google.com...
> I'm writing a 3D naval tactics in C++ for the Mac. I'm planning on
> having all the object behaviors, movements, and AI controlled
through
> a scripting layer. The scripts would be text files the game would
> parse and then carry out actions. All the scripts would be read at
> runtime, not compile time, so they'd all have to be interpreted.
> In the scripting system, there would have to be some way to create a
> new foo object, define x, and then call print() while the user was
> playing.
..
> Can I use Python to accomplish this, and if so, can you point me to
> some tutorials on integratring Python with my existing code?

Blade of Darkness, a commercial real-time 3d medieval action game, is
scripted in Python.  The demo is probably still available (check
google); it comes with readable .py files.

Terry J. Reedy





From mgilfix at eecs.tufts.edu  Fri Jun  7 00:32:17 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Fri, 7 Jun 2002 00:32:17 -0400
Subject: Creating Dynamic Web Pages (ZOPE)
In-Reply-To: <46ca81a0.0206062016.789d3af6@posting.google.com>; from hwcowan@hotmail.com on Thu, Jun 06, 2002 at 09:16:22PM -0700
References: <46ca81a0.0206062016.789d3af6@posting.google.com>
Message-ID: <20020607003217.B20029@eecs.tufts.edu>

  Hiya!

On Thu, Jun 06 @ 21:16, Hugh Cowan wrote:
> I don't know if ZOPE would fit my situation or not.  I know that it is
> / can be used to create complex E-commerce sites, but I am wondering
> if it is overkill for what I need done.  Would I be better off trying
> to find some sort of RAD Web Tool instead (anyone have any good
> suggestions) ?

  I think Zope is definitely the way to go as it is designed to
address exactly what you need and easily. Zope allows you to add
'drop-in' components to your web service (such as an SQL query,
or called a 'ZSQL query' in Zope-speak), which tends to be a big
time saver, and then reference that query to dynamically produce a
webpage. More so, it allows for easy remote management with delegation
of roles and authority to multiple parties; seems like a good way to
structure your intranet.

                     -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From mhammond at skippinet.com.au  Fri Jun 28 20:10:29 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sat, 29 Jun 2002 00:10:29 GMT
Subject: Python COM problem - accessing properties with optional parameters
References: <1025282888.585264@lave>
Message-ID: <3D1CFBB8.90709@skippinet.com.au>

Dave Kirby wrote:
> I am using Python to access the Microsoft Accessibility interface.  I have
> got as far as getting an IAccessible object, but am having problems calling
> some of its methods.
> 
> The problem is that some of the methods are declared as properties with
> [propget] or [propput] in the IDL, but also have optional parmeters.  For
> example:
> 
>   [hidden, propget, id(DISPID_ACC_NAME)] HRESULT accName([in, optional]
> VARIANT varChild,[out, retval] BSTR* pszName);
> 
> if I have a an object called obj I can access the property accName like
> this:
> 
> 
>>>>obj.accName
>>>
> u'Some Object Name'
> 
> 
> which is all fine and good.  However there are times when I need to call the
> accName method with an int parameter, e.g.
> 
> 
>>>>obj.accName(0)
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
> TypeError: object of type 'unicode' is not callable
> 
> 
> this obviously doesnt work, since it treating it as a property and returning
> a string, then trying to call that string as a function.

If you run makepy over the type library, you *should* find that the 
property has *also* been exposed as a "SetAccName()" method.  Check the 
generated code for details.

Mark.



From samschul at pacbell.net  Sat Jun 29 22:04:39 2002
From: samschul at pacbell.net (Samuel Schulenburg)
Date: 29 Jun 2002 19:04:39 -0700
Subject: Intercepting a process shutdown command under windows?
Message-ID: <d7fcf8b4.0206291804.3146614b@posting.google.com>

Is their a way to keep another process from shutdown your python
application before you can perform some cleanup activities?

Sam Schulenburg


From trentm at ActiveState.com  Tue Jun 25 12:39:19 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Tue, 25 Jun 2002 09:39:19 -0700
Subject: os.kill for win32
In-Reply-To: <pan.2002.06.24.14.40.03.222610.147@no-spam.netscape.net>; from gupsgrotto@no-spam.netscape.net on Tue, Jun 25, 2002 at 02:37:45PM +0000
References: <pan.2002.06.24.14.40.03.222610.147@no-spam.netscape.net>
Message-ID: <20020625093919.A4331@ActiveState.com>

[grant wrote]
> Anyone know the same functionality of the os.kill module for the win32
> platform.  All examples that I've seen are just for Unix.  In the program
> that I've written I'm lauching an external program via a wxCheckbox
> switch and would like to kill the program when the switch is unchecked
> however wxKill only works on Unix, and os.kill is the same.  Is there any
> relatively simple way to do this same thing on win32 (NT platform)

winprocess.py was recently checked into the PyWin32 CVS tree. You should
be able to glean the kill code you need from there:

(attached)


Trent

-- 
Trent Mick
TrentM at ActiveState.com
-------------- next part --------------
"""
Windows Process Control

winprocess.run launches a child process and returns the exit code.
Optionally, it can:
  redirect stdin, stdout & stderr to files
  run the command as another user
  limit the process's running time
  control the process window (location, size, window state, desktop)
Works on Windows NT, 2000 & XP. Requires Mark Hammond's win32
extensions.

This code is free for any purpose, with no warranty of any kind.
-- John B. Dell'Aquila <jbd at alum.mit.edu>
"""

import win32api, win32process, win32security
import win32event, win32con, msvcrt, win32gui


def logonUser(loginString):
    """
    Login as specified user and return handle.
    loginString:  'Domain\nUser\nPassword'; for local
        login use . or empty string as domain 
        e.g. '.\nadministrator\nsecret_password'
    """
    domain, user, passwd = loginString.split('\n')
    return win32security.LogonUser(
        user,
        domain,
        passwd,
        win32con.LOGON32_LOGON_INTERACTIVE,
        win32con.LOGON32_PROVIDER_DEFAULT
        )


class Process:
    """
    A Windows process.
    """

    def __init__(self, cmd, login=None,
                 hStdin=None, hStdout=None, hStderr=None,
                 show=1, xy=None, xySize=None,
                 desktop=None):
        """
        Create a Windows process.
        cmd:     command to run
        login:   run as user 'Domain\nUser\nPassword'
        hStdin, hStdout, hStderr:
                 handles for process I/O; default is caller's stdin,
                 stdout & stderr
        show:    wShowWindow (0=SW_HIDE, 1=SW_NORMAL, ...)
        xy:      window offset (x, y) of upper left corner in pixels
        xySize:  window size (width, height) in pixels
        desktop: lpDesktop - name of desktop e.g. 'winsta0\\default'
                 None = inherit current desktop
                 '' = create new desktop if necessary

        User calling login requires additional privileges:
          Act as part of the operating system [not needed on Windows XP]
          Increase quotas
          Replace a process level token
        Login string must EITHER be an administrator's account
        (ordinary user can't access current desktop - see Microsoft
        Q165194) OR use desktop='' to run another desktop invisibly
        (may be very slow to startup & finalize).
        """
        si = win32process.STARTUPINFO()
        si.dwFlags = (win32con.STARTF_USESTDHANDLES ^
                      win32con.STARTF_USESHOWWINDOW)
        if hStdin is None:
            si.hStdInput = win32api.GetStdHandle(win32api.STD_INPUT_HANDLE)
        else:
            si.hStdInput = hStdin
        if hStdout is None:
            si.hStdOutput = win32api.GetStdHandle(win32api.STD_OUTPUT_HANDLE)
        else:
            si.hStdOutput = hStdout
        if hStderr is None:
            si.hStdError = win32api.GetStdHandle(win32api.STD_ERROR_HANDLE)
        else:
            si.hStdError = hStderr
        si.wShowWindow = show
        if xy is not None:
            si.dwX, si.dwY = xy
            si.dwFlags ^= win32con.STARTF_USEPOSITION
        if xySize is not None:
            si.dwXSize, si.dwYSize = xySize
            si.dwFlags ^= win32con.STARTF_USESIZE
        if desktop is not None:
            si.lpDesktop = desktop
        procArgs = (None,  # appName
                    cmd,  # commandLine
                    None,  # processAttributes
                    None,  # threadAttributes
                    1,  # bInheritHandles
                    win32process.CREATE_NEW_CONSOLE,  # dwCreationFlags
                    None,  # newEnvironment
                    None,  # currentDirectory
                    si)  # startupinfo
        if login is not None:
            hUser = logonUser(login)
            win32security.ImpersonateLoggedOnUser(hUser)
            procHandles = win32process.CreateProcessAsUser(hUser, *procArgs)
            win32security.RevertToSelf()
        else:
            procHandles = win32process.CreateProcess(*procArgs)
        self.hProcess, self.hThread, self.PId, self.TId = procHandles

    def wait(self, mSec=None):
        """
        Wait for process to finish or for specified number of
        milliseconds to elapse.
        """
        if mSec is None:
            mSec = win32event.INFINITE
        return win32event.WaitForSingleObject(self.hProcess, mSec)

    def kill(self, gracePeriod=5000):
        """
        Kill process. Try for an orderly shutdown via WM_CLOSE.  If
        still running after gracePeriod (5 sec. default), terminate.
        """
        win32gui.EnumWindows(self.__close__, 0)
        if self.wait(gracePeriod) != win32event.WAIT_OBJECT_0:
            win32process.TerminateProcess(self.hProcess, 0)
            win32api.Sleep(100) # wait for resources to be released
            
    def __close__(self, hwnd, dummy):
        """
        EnumWindows callback - sends WM_CLOSE to any window
        owned by this process.
        """
        TId, PId = win32process.GetWindowThreadProcessId(hwnd)
        if PId == self.PId:
            win32gui.PostMessage(hwnd, win32con.WM_CLOSE, 0, 0)
        
    def exitCode(self):
        """
        Return process exit code.
        """
        return win32process.GetExitCodeProcess(self.hProcess)
    
    
def run(cmd, mSec=None, stdin=None, stdout=None, stderr=None, **kw):
    """
    Run cmd as a child process and return exit code.
    mSec:  terminate cmd after specified number of milliseconds
    stdin, stdout, stderr:
           file objects for child I/O (use hStdin etc. to attach
           handles instead of files); default is caller's stdin,
           stdout & stderr;
    kw:    see Process.__init__ for more keyword options
    """
    if stdin is not None:
        kw['hStdin'] = msvcrt.get_osfhandle(stdin.fileno())
    if stdout is not None:
        kw['hStdout'] = msvcrt.get_osfhandle(stdout.fileno())
    if stderr is not None:
        kw['hStderr'] = msvcrt.get_osfhandle(stderr.fileno())
    child = Process(cmd, **kw)
    if child.wait(mSec) != win32event.WAIT_OBJECT_0:
        child.kill()
        raise WindowsError, 'process timeout exceeded'
    return child.exitCode()


if __name__ == '__main__':

    # Pipe commands to a shell and display the output in notepad
    print 'Testing winprocess.py...'

    import tempfile

    timeoutSeconds = 15
    cmdString = """\
REM      Test of winprocess.py piping commands to a shell.\r
REM      This window will close in %d seconds.\r
vol\r
net user\r
_this_is_a_test_of_stderr_\r
""" % timeoutSeconds

    cmd, out = tempfile.TemporaryFile(), tempfile.TemporaryFile()
    cmd.write(cmdString)
    cmd.seek(0)
    print 'CMD.EXE exit code:', run('cmd.exe', show=0, stdin=cmd,
                                    stdout=out, stderr=out)
    cmd.close()
    print 'NOTEPAD exit code:', run('notepad.exe %s' % out.file.name,
                                    show=win32con.SW_MAXIMIZE,
                                    mSec=timeoutSeconds*1000)
    out.close()



From fredrik at pythonware.com  Sat Jun  8 12:30:41 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 08 Jun 2002 16:30:41 GMT
Subject: Comparison: HTTP with Java, Perl or Python
References: <3D009911.9000904@gmx.de>
Message-ID: <RiqM8.11236$p56.3411617@newsb.telia.net>

Ingo Linkweiler wrote:
> for a comparison of languages I need a complete "translation"
> of this short skript into Perl, Tcl and Java:
>
>  myurl = "http://www.uni-dortmund.de"
>  import urllib
>  file  = urllib.urlopen(myurl)
>  text  = file.read()
>  file.close()
>  print text

here's a Python version:

    import urllib
    MYURL = "http://www.uni-dortmund.de"
    print urllib.urlopen(MYURL).read()

</F>




From jonathan at onegoodidea.com  Sun Jun 30 16:02:34 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Sun, 30 Jun 2002 21:02:34 +0100
Subject: Newbie that don't understand
References: <AIlT8.13914$p56.4478653@newsb.telia.net> <3d1e262f.3755054@news.clara.net> <3D1E4F80.D8EFBB3E@engcorp.com> <3D1E56EC.7080203@SPAMnwinternet.com>
Message-ID: <B945226A.D35D%jonathan@onegoodidea.com>

On 30/6/2002 1:55, in article 3D1E56EC.7080203 at SPAMnwinternet.com, "Jim"
<jbublitzNO at SPAMnwinternet.com> wrote:

> Some British
> sports cars from the 50's and 60's have the positive
> terminal connected to the chassis (ground for a car),
> so they have a -12v (-6v?) electrical system (or "positive
> ground"). There is no difference, however, between the
> battery in an American car and these British cars - just
> a difference in reference point. As Einstein said, it's
> all relative..

[Still being off-topic:]

Wouldn't a positive ground car rust faster? I seem to vaguely remember
something from school physics about negative ground helping to slow
oxidation by replacing electrons in the chassis.

If I remember rightly, that's also why iron ships have (or used to have?)
zinc plates bolted to the hull on the inside. When the zinc oxidises it
gains a negative charge which replaces electrons being stripped from the
iron on the outside.

-face-screwed-up-in-desperate-recollection-ly y'rs,

Jonathan



From jmichelz at mail.com  Wed Jun 26 18:14:02 2002
From: jmichelz at mail.com (John Michelsen)
Date: 26 Jun 2002 15:14:02 -0700
Subject: Wanted:  Potential Python Authors
References: <ae7gke$mhb$1@ccis12.baylor.edu> <3D08C96B.1020509@linuxmail.org>
Message-ID: <df6ea3a2.0206261414.26fcc03d@posting.google.com>

Bob X <bobx at linuxmail.org> wrote in message news:<3D08C96B.1020509 at linuxmail.org>...
> Sheesh...and all he wanted was some potential authors and we debate it 
> to death!  : )
> 
> Jeff Donahoo wrote:
> > I'm looking for somebody to author a short book on Python in my programming
> > technology series.  Who do you think would make a great author that hasn't
> > already written a book on Python?  Candidates include:
> > 
> > - Competent and frequent posters to this newsgroup
> > - Author of articles in the various trade magazines
> > - Expert programmer

I'm thinking of taking up the challenge.  What python related subjects
would you all like to see a new book on?  I have my list, but am
keeping it quiet for
now. =)

John


From whisper at oz.net  Thu Jun 13 23:52:57 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 13 Jun 2002 20:52:57 -0700
Subject: More Tkinter Help please...
In-Reply-To: <GCEDKONBLEFPPADDJCOEEEANDJAA.whisper@oz.net>
Message-ID: <GCEDKONBLEFPPADDJCOEAEAODJAA.whisper@oz.net>

Oops, my bad - PyCrust uses WxPython. The Idle app in tools/idle would be a
better role model for you.


David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: David LeBlanc [mailto:whisper at oz.net]
> Sent: Thursday, June 13, 2002 20:50
> To: SA; python-list at python.org
> Subject: RE: More Tkinter Help please...
>
>
> > -----Original Message-----
> > From: python-list-admin at python.org
> > [mailto:python-list-admin at python.org]On Behalf Of SA
> > Sent: Thursday, June 13, 2002 20:04
> > To: python-list at python.org
> > Subject: More Tkinter Help please...
> >
> >
> > Hi Everyone-
> >
> >     I have another Tkinter question for you. I'm still trying
> to translate
> > "that previously mentioned" Tcl/Tk program into Python/Tkinter. The
> > following is all of the code I've managed so far:
> >
> > from Tkinter import *
> > import sys
> > import os
> >
> > class PyShell:
> >     def __init__(self, top):
> >
> >         f = Frame(top)
> >         f.pack()
> >
> >         self.t1 = Text(f, height="12", width="84", font="Courier 12")
> >         self.t1.pack(side=TOP, pady=2)
> >
> >         self.t2 = Text(f, height="12", width="84", bg="lightblue",
> > font="Courier 12")
> >         self.t2.pack(side=TOP, pady=2)
> >
> >         self.b1 = Button(f, text="Execute", command=self.expyth)
> >         self.b1.pack(side=LEFT)
> >
> >         self.b2 = Button(f, text="Clear Input", command=self.clearin)
> >         self.b2.pack(side=LEFT)
> >
> >         self.b3 = Button(f, text="Clear Output", command=self.clearout)
> >         self.b3.pack(side=LEFT)
> >
> >         self.b4 = Button(f, text="Save Input", command=self.savin)
> >         self.b4.pack(side=LEFT)
> >
> >     def clearin(self):
> >         self.t1.delete(0,END)
> >
> >     def clearout(self):
> >         self.t2.delete(0,END)
> >
> >     def expyth(self):
> >         output = os.popen("python -c").t1()
> >         self.t2.delete(0,END)
> >         sys.stdout.write.t2(output)
> >
> >     def savin(self):
> >         pass
> >
> > root = Tk()
> >
> > app = PyShell(root)
> >
> > root.mainloop()
> >
> > When I run this I get the following error:
> >
> > Traceback (most recent call last):
> >   File "PyShell.py", line 45, in ?
> >     app = PyShell(root)
> >   File "PyShell.py", line 17, in __init__
> >     self.b1 = Button(f, text="Execute", command=self.expyth)
> > AttributeError: PyShell instance has no attribute 'expyth'
> >
> >
> > Any ideas on what I'm doing wrong here?
> >
> > Thanks.
> > SA
> >
> > --
> > http://mail.python.org/mailman/listinfo/python-list
>
>
> Code comes up for me, but has other bugs I'm not able to diagnose.
>
> FWIW, if you're trying to create a python shell, are you aware of
> PyCrust on Sourceforge? If nothing else, you might get some
> pointers out of it :)
>
> Dave LeBlanc
> Seattle, WA USA





From cliechti at gmx.net  Fri Jun 21 17:01:55 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 21 Jun 2002 23:01:55 +0200
Subject: why extending copy.*
References: <MPG.177d438123ae240598968d@news.rz.uni-karlsruhe.de>
Message-ID: <Xns9234EAB9AC746cliechtigmxnet@62.2.16.82>

Uwe Mayer <merkosh at hadiko.de> wrote in 
news:MPG.177d438123ae240598968d at news.rz.uni-karlsruhe.de:

> hi,
> 
> i need shallow and deep copies of objects and I found the copy module in 
> the reference library which does quite a good job.
> however, they write that one could implement the copy protocol by 
> extending/ overwriting the __copy__() and __deepcopy__() method.
> 
> since __copy__() and __deepcopy__() already duplicate the instance 
> dictionary, __deepcopy__() even recursively. For what reasons would I 
> want to extend /overwrite the capabilities of __copy__() or __deepcopy__
> ()?

its important when you use external resources. e.g. a file descriptor can 
be duplicated so that you can call close() on each file object (in a copied 
object and in the original) without infuence on the other (i think, but 
you'll get the point...)

or a connection to a DB could be opened on the copied object so that the 
copy and the original do not interfere.

chris

-- 
Chris <cliechti at gmx.net>



From jdhunter at nitace.bsd.uchicago.edu  Sat Jun 22 13:32:05 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Sat, 22 Jun 2002 12:32:05 -0500
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <m21yazxtgs.fsf@mother.paradise.lost>
 <3D148E71.2070201@kfunigraz.ac.at>
Message-ID: <m2wusrw76y.fsf@mother.paradise.lost>

>>>>> "Siegfried" == Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> writes:


    Siegfried> Note: the first 20 minutes everything went on right
    Siegfried> (including the two external functions). On Windows
    Siegfried> after the first wavelength the system sucks down too,
    Siegfried> but it keeps up.

It is very difficult to test and debug code when your system freezes,
because it takes so long to reboot.  One trick I used when I was
encountering system freezes (ultimately it was a pointer problem) was
to set up an 'at' process to kill the job for me in case my system
froze.  Suppose your program is called 'myprog' and it takes roughly 5
minutes to get to the freeze point.  Try:

unix> at now +6min
at> killall myprog
at> CTRL-d                # Hit Control and D together
unix> ./myprog

With luck, the 'at' process will kill the job after 6 minutes and you
can try something else.  Much less painful than a reboot.

    Siegfried> I can send you the code (without) data, but that won't
    Siegfried> cure the situation.

Send it to the group.  Especially any C code involving memory
allocation and pointer dereferencing.  Before you start the 'myprog'
job, you may want to open a separate X term and run

unix> top

Then type 'M' to sort by memory usage.  Watch the % memory usage as
the program runs and see if it is increasing with time.  If it is, and
you aren't intentionally allocating increasing memory, then it is
likely that there is a free or delete statement somewhere in the C
code that runs fine that isn't being wrapped or called properly in the
python code.

JDH


From sarmstrong13 at mac.com  Tue Jun  4 11:59:18 2002
From: sarmstrong13 at mac.com (SA)
Date: Tue, 04 Jun 2002 10:59:18 -0500
Subject: Python command line?
Message-ID: <B9224E06.628B%sarmstrong13@mac.com>

Hi Everyone-

I would like to run python code line by line using:

python -c some line of code

Will this work?

I have a gui app that lets me type in code in one fiels and then executes
the code typed into that field line by line and prints the output to another
field for evaluation.

Thanks for your help.
SA



From pierpaolo.glave at ciaolab.com  Mon Jun 24 12:03:03 2002
From: pierpaolo.glave at ciaolab.com (Pier Paolo Glave)
Date: 24 Jun 2002 09:03:03 -0700
Subject: Embedding in Tkinter (reprise)
Message-ID: <1907a5e4.0206240803.6ca0a4c@posting.google.com>

Hi everybody,

I wanted to ask you if it is possible to embed an external graphical
program in
a Tkinter Window: e.g. run xclock in a Tkinter frame or canvas.

The same question was posted by someone else almost one year ago, but
witout
answers. (see http://groups.google.com/groups?q=window+embed+canvas+group:*python*&hl=en&lr=&ie=UTF-8&oe=UTF8&selm=3BA69295.4240BDEB%40bioeng.ucsd.edu&rnum=4)

Sorry for the re-post and thank you in advance.

--
Pier Paolo Glave


From mgerrans at mindspring.com  Mon Jun 10 03:26:01 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Mon, 10 Jun 2002 00:26:01 -0700
Subject: Newbie Question (executing Program from within)
References: <B92826C3.6BF9%sarmstrong13@mac.com>
Message-ID: <ae1kae$1h4$1@slb0.atl.mindspring.net>

# This will do the trick:

def runScript(filename):
   namespace = { '__name__' : '__main__' }
   execfile(filename, namespace)
   return namespace

# - mfg




From observer at NOSPAM.space.pl  Wed Jun 12 07:07:12 2002
From: observer at NOSPAM.space.pl (Johann)
Date: Wed, 12 Jun 2002 13:07:12 +0200
Subject: PythonWorks and non-English characters...
Message-ID: <jq9eguc8jkc35diss43s5etm5gcmcrqfrd@4ax.com>

I would like to change default settings of PythonWorks
(http://www.pythonware.com/downloads/index.htm) to non-English
character set. When I work with layout resource I can get  iso-8859-1
encodings only for text in my elements whatever kind of font I choose.
All xml files for layout elements have also iso-8859-1 encoding. :(

I would like to change it to utf-8, iso8859-2 or cp1250. I found in
manual that should make "editpref" for it, but I do not know where I
should create such file. 

--
Johann


From fredrik at pythonware.com  Sat Jun 29 08:06:21 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 29 Jun 2002 12:06:21 GMT
Subject: Problem installing PIL 1.1.3
References: <Xns923C386473E29AusYourStandingInIt@130.133.1.4>
Message-ID: <1phT8.45576$n4.10822862@newsc.telia.net>

"netvegetable" wrote:

> Keep getting this error everytime I try to run setup.py
>
> File "D:\PYTHON22\lib\site-
> packages\Pythonwin\pywin\framework\scriptutils.py", line 301, in RunScript
>     exec codeObject in __main__.__dict__
>   File "D:\Python22\Imaging-1.1.3\setup.py", line 52, in ?
>     for line in open(os.path.join("libImaging", "ImConfig.h")).readlines():
> IOError: [Errno 2] No such file or directory: 'libImaging\\ImConfig.h'
>
> Anyone seen this happen before? Or even better, know how I can get it to
> work?

the setup.py script assumes that 1) you have the full source
kit installed, and 2) you're running the script from the source
directory.

I suggest running the configuration/build files from a Windows
console, not from within a Python IDE.

</F>




From jadestar at idiom.com  Wed Jun 12 04:17:47 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 12 Jun 2002 08:17:47 GMT
Subject: mx library
References: <3CF338E5.F3F0931C@katamail.com> <mailman.1022574753.24571.python-list@python.org>
Message-ID: <ae703b$on0$1@news.idiom.com>

Gerhard H?ring <gh_pythonlist at gmx.de> wrote:

> * Yallowman <yallowman at katamail.com> [2002-05-28 09:59 +0200]:
>> I need us mx library on RedHat 6.2. Where can i find this rpm ?

> In the dist/ directory. Don't believe me?

> Download the egenix-mx Extensions from Marc-Andr?'s homepage, unpack
> them, enter in the directory and issue:

> $ python setup.py build
> $ python setup.py bdist_rpm

> You could skip the rpm creation step and just install the mx-Extensions
> with:

> $ python setup.py install

> Using the RPM has the advantage of easier uninstalling.

> Gerhard

 It also has the considerable advantages of maintaining a database
 of installed packages, file "ownership" (associating files to their
 packages), permissions and meta-data (--setperms, etc) and MD5
 checksums.  It makes the packages files easy to identify, verify and
 manage.  (I'm particularly fond of rpm -qf /full/path/to/file
 command --- which I use with rpm -ql or -qld to find the associated
 documentation for many commands, config files, etc).  Actually I prefer
 Debian apt to that; and the dpkg -S command to find the "source" of a 
 given file.  Also Debian's apt-get build-dep && apt-get -b source ROCKS!
 --- you can find, fetch, and install al of the requisite packages
 that you need to find, fetch and BUILD a package from source with
 just one command!  That's great for programming students, tech support
 (sometimes the answers can't only be found in the sources) and 
 open source consultants. [I'm in all three categories].






From quinn at retch.ugcs.caltech.edu  Thu Jun 20 20:28:48 2002
From: quinn at retch.ugcs.caltech.edu (Quinn Dunkan)
Date: 21 Jun 2002 00:28:48 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com> <3D1010C3.C321F8B0@engcorp.com> <raXP8.47926$GY.15260255@twister.nyroc.rr.com> <3D107948.B3FEE4BB@engcorp.com>
Message-ID: <slrnah4spv.557.quinn@retch.ugcs.caltech.edu>

On Wed, 19 Jun 2002 08:30:00 -0400, Peter Hansen <peter at engcorp.com> wrote:
>George Hester wrote:
>> 
>> Yes my take on it was that you said "I needed to crawl before I walk."
Please
>> don't hold my inference against me.
>
>That's precisely what I said.  Nothing derogatory about it, except

... maybe if he's Balinese?

The name doesn't sound like it though.


From jadestar at idiom.com  Mon Jun 10 02:04:35 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 10 Jun 2002 06:04:35 GMT
Subject: Backwards emulation rather than backwards compatibility?
References: <3CEF8E04.6040909@ix.netcom.com> <m3it5cp8hj.fsf@mira.informatik.hu-berlin.de> <uf06o3f6tu8ud3@news.supernews.com>
Message-ID: <ae1fhj$12ab$1@news.idiom.com>

John Roth <johnroth at ameritech.net> wrote:


> "Martin v. Loewis" <martin at v.loewis.de> wrote in message
> news:m3it5cp8hj.fsf at mira.informatik.hu-berlin.de...
>> Antaeus Feldspar <feldspar at ix.netcom.com> writes:
>>
>>> Could the python interpreter be equipped with a mode that would
> allow it to emulate past interpreters?

 Why emulate, why not exec the old interpreter with the open filehandle
 of the script that's to be interpreted?

>> No, that won't be possible. It is easy enough to install multiple
>> versions of the interpreter, though.

 Ever since I saw "from __future__ import" I've thought that Python
 was crying out for a "from __past__ import"

 ... though I have no real idea what that would mean.

 (The big issue would be when you want to mix modules; run a 1.5.2 
 module in a 2.x interpreter, etc.  If there was a good "conduit" to
 pass data between the two interpreters it might work (though we'd
 have two processes, and the "newer" version would be responsible for
 pickling up objects and passing them through some IPC to be unpickled
 on the other side; and some objects (generators) would have to be 
 passed as proxies (the other side would call on them; pass the arguments
 back to the other intpreter and get the results back.

 This is probably technical possible, but I'm not sure the idea is worth
 any effort; it probably would have inevitable limitations that would make
 it less than transparent --- at which point upgrading the programs and 
 modules to be compatible with the new version looks more attractive.



From peter at engcorp.com  Wed Jun 26 00:18:07 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 26 Jun 2002 00:18:07 -0400
Subject: Python and Windows XP
References: <20020625233842.29966.00000386@mb-ma.aol.com>
Message-ID: <3D19407F.B7529B42@engcorp.com>

Marc wrote:
> 
> Hello all,
>    I apologise for this newbie question, but I use Windows XP and wonder how I
> can open a shell session to use Python?  There used to be the DOS command
> option, but now...

Assuming Windows XP is just Windows NT with a couple of letters shifted
around randomly, and assuming you can find the equivalent of the Start Menu
"Run..." selection, you can probably type "CMD" there to get a prompt.

That's a lot of assuming...

-Peter


From stochastic at eresmas.net  Sat Jun 15 14:33:22 2002
From: stochastic at eresmas.net (Jean Pierre DENYS)
Date: Sat, 15 Jun 2002 20:33:22 +0200
Subject: SWIG and Callbacks
In-Reply-To: <3d0a8b6e@nntp.server.uni-frankfurt.de>
References: <3d0a8b6e@nntp.server.uni-frankfurt.de>
Message-ID: <p04320403b930c7f699ea@[62.174.12.219]>

Hi,

Do you know where I could find the GRAD system?

regards,

Jean Pierre DENYS


>Hi,
>
>I have to bind a C-library which does lots of low-level
>ioctl and uses c-callback functions. Using SWIG has given
>me the possibility to call most of the wrapped functions
>without too much hassle, but I just learned that SWIG can't
>call Python-functions as callbacks.
>
>I had a quick look at the other wrappers (SILOON, GRAD, CXX, etc.)
>but most look either outdated or not appropriate for this task.
>Do you know an (automatic) wrapper generator which has this feature?
>
>I know sip can do this for c++ libraries - can sip also be used
>to wrap plain c-libs without "handwriting" code?
>
>Yours,
>
>:M:





From huaiyu at gauss.almadan.ibm.com  Sun Jun 30 16:01:12 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Sun, 30 Jun 2002 20:01:12 +0000 (UTC)
Subject: possibly a dumb question
References: <3d1e5eff$1_7@nopics.sjc> <3D1E6415.8128F3F8@engcorp.com> <3d1e6251$1_3@nopics.sjc>
Message-ID: <slrnahuos8.hjr.huaiyu@gauss.almadan.ibm.com>

Adonis <deltapigz at telocity.com> wrote:
>sure:
>
>class foo:
>    def __init__(self, value):
>        return value
>x = foo(0)
>print x ;yeilds 0
>
>i know the code provided is wrong, but its the general idea.
>

Is this what you want?

>>> class foo:
...     def __init__(self, value):
...         self.value = value
...     def __call__(self):
...         return self.value
... 
>>> x = foo(0)
>>> print x()+3
3

If you only want string representations instead of the original value
object, you can use __str__ as others have pointed out.

Huaiyu


From tismer at tismer.com  Sat Jun 29 11:24:25 2002
From: tismer at tismer.com (Christian Tismer)
Date: Sat, 29 Jun 2002 17:24:25 +0200
Subject: I'm an idiot
References: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16> <afj27n023m7@enews2.newsguy.com> <3D1D5C5E.3A5EB81D@earthlink.net> <4phT8.45579$n4.10823008@newsc.telia.net>
Message-ID: <3D1DD129.9090507@tismer.com>

Fredrik Lundh wrote:
> Joseph A Knapka wrote:

...

>>But I'm sure there are Very Good Reasons not to do this,
>>and that some wise person will presently educate me :-)
> 
> 
> To many people, this is a religous topic.  Google can probably
> tell you all about the different religions.
> 
> The most common view is that as the name implies, exceptions
> should be used to deal with exceptional, unexpected conditions.
> That you'll reach the end of a file if you keep reading data from
> it should hardly come as a surprise...

Well, there *might* be one other reason, while maybe
not too remarkable, but the following examples show
that using exceptions just for the sake of speed
may be a consideration.

Test 1 iterates over realines() output (well, kidding,
I just used a cont string) and asks if the string is
empty.
Test 2 does it different, by wrapping the whole loop
into a try...except and does a test by indexing the
string, which will raise an IndexError on failure.
Test 3 is just an empty loop for comparison.

 >>> def test1(n):
... 	s='abc'
... 	for i in xrange(n):
... 		if not s:
... 			break
... 		
 >>> def test2(n):
... 	s="abc"
... 	try:
... 		for i in xrange(n):
... 			s[0]
... 	except IndexError:
... 		pass
... 	
 >>> def test3(n):
... 	for i in xrange(n):
... 		pass
... 	
 >>> timing(test1, 1000000)
(0.99099999999999999, None)
 >>> timing(test2, 1000000)
(0.90100000000000002, None)
 >>> timing(test3, 1000000)
(0.5, None)
 >>>

Comparing the results suggests, that an exception
based solution is 20 percent faster, or even better.
Since the string must be touched anyway if we want
to do something with it, it can be managed to write
the string processing in a way, that the above
exception raising indexing is performed as a side
effect of the processing, if written carefully.

I have to admit that this all doesn't increase
readability, it's just a trick that came to mind.

For the general case, I really recommend to follow
your advice!

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From whisper at oz.net  Tue Jun 25 17:22:00 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 25 Jun 2002 14:22:00 -0700
Subject: How to ImpersonateLoggedOnUser on Windows NT/2000?
In-Reply-To: <200206251034.35021.gherron@islandtraining.com>
Message-ID: <GCEDKONBLEFPPADDJCOEAECLDMAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Gary Herron
> Sent: Tuesday, June 25, 2002 10:35
> To: python-list at python.org
> Subject: How to ImpersonateLoggedOnUser on Windows NT/2000?
>
>
> Hi,
>
> I'm trying to write a script for both Windows NT/2000 and Linux which
> needs to perform actions on the behalf of several different users.
>
> On Linux I use "os.seteuid" and friends and all is fine.
>
> On NT the comparable thing seems to be LogonUser,
> ImpersonateLoggedOnUser, and RevertToSelf.  However for these calls to
> be successful, one needs various privileges, such as SE_TCB_NAME, and
> SE_CHANGE_NOTIFY_NAME..
>
> At this point I start getting lost, so I hope someone who has been
> through this can help me.  How can a Python program running as local
> administrator on Windows NT/2000 set those privileges, and then once
> set, how does one actually *do* the impersonation?
>
> Thank you,
> Dr. Gary Herron
>

After a quick look at MSDN, it looks like it goes like this on Win2000
Pro(psuedo C code):
1. Use Start | Programs | Administrative tools | Computer Management to add
a user.
2. Use Start | Programs | Administrative tools | Local Security Policy to
add the user to the desired access control list (which is backwards from the
way NT did it - you added priviledges to users).
3. This code:
PHandle userHandle;
LogonUser(user, domain, password, LOGON32_LOGON_BATCH,
LOGON32_PROVIDER_DEFAULT, userHandle);
// LOGON32_LOGON_BATCH can also be LOGON32_LOGON_INTERACTIVE
ImpersonateLoggedOnUser(userHandle);
// do whatever
RevertToSelf();	// or just exit to revert to self.

Presumably you've seen the discussion of this in the Python Win32 extensions
help?

After a bunch more reading in MSDN, it looks like the simplest way to have a
user with sufficient priveledges is to use the GUI tools to create such a
user and then impersonate it. I suspect the programmatic way is to create
Access Control Lists and Security Descriptors - and MSDN doesn't connect the
dots for doing that. "Sufficient Priviledges" include both the user's
permissions and the permissions of the object the user is trying to
manipulate.

You might find this of interest: http://www.codeguru.com/system/Logon.shtml

I wish I could be more help - a way to do this 100% programmatically
interests me too.

Dave LeBlanc
Seattle, WA USA





From chris_mk at hotmail.com  Wed Jun  5 13:23:09 2002
From: chris_mk at hotmail.com (Christopher)
Date: 5 Jun 2002 10:23:09 -0700
Subject: Problem with urllib.urlopen()
References: <cc14131d.0206031445.5154de31@posting.google.com> <cc14131d.0206032128.5b2efcf6@posting.google.com> <cc14131d.0206041631.157f11d1@posting.google.com> <WSnL8.268879$%u2.175369@atlpnn01.usenetserver.com>
Message-ID: <cc14131d.0206050923.3c1da687@posting.google.com>

Actually, the link in the browser (IE and Netscape on a Win2K machine
with a LAN connection (work) and IE on Win98 and IE on WinXP on a
56K(both home)) should show a page containing XML versions of the
journal entries as returned in the first link.  Python, instead,
returns the "no current query" reponse.  Thanks.

Chris

(some of the code is attached(the current format doesn't matter, I
have a script that will translate it into 'real' XML...)

&lt;PubMedArticle&gt;
&lt;MedlineCitation Status=&quot;In-Process&quot;&gt;
&lt;MedlineID&gt;21999273&lt;/MedlineID&gt;
&lt;PMID&gt;12003952&lt;/PMID&gt;
&lt;DateCreated&gt;
&lt;Year&gt;2002&lt;/Year&gt;
&lt;Month&gt;05&lt;/Month&gt;
&lt;Day&gt;10&lt;/Day&gt;
&lt;/DateCreated&gt;
&lt;Article&gt;
&lt;Journal&gt;
&lt;ISSN&gt;0021-9193&lt;/ISSN&gt;
&lt;JournalIssue&gt;
&lt;Volume&gt;184&lt;/Volume&gt;
&lt;Issue&gt;11&lt;/Issue&gt;
&lt;PubDate&gt;
&lt;Year&gt;2002&lt;/Year&gt;
&lt;Month&gt;Jun&lt;/Month&gt;
&lt;Day&gt;1&lt;/Day&gt;
&lt;/PubDate&gt;
&lt;/JournalIssue&gt;
&lt;/Journal&gt;
&lt;ArticleTitle&gt;The ner Gene of Photorhabdus: Effects on
Primary-Form-Specific Phenotypes and Outer Membrane Protein
Composition.&lt;/ArticleTitle&gt;
&lt;Pagination&gt;
&lt;MedlinePgn&gt;3096-105&lt;/MedlinePgn&gt;
&lt;/Pagination&gt;
&lt;Abstract&gt;
&lt;AbstractText&gt;The nematode-bacterium complex of
Heterorhabditis-Photorhabdus is pathogenic to insect larvae. The
bacteria undergo a form of phenotypic switching whereby the primary
form, at the stationary phase of the growth cycle, makes a range of
products and has the capacity to support nematode growth, whereas the
secondary form does not express these phenotypes. The work described
here investigated the mechanism regulating phenotypic variation by
transforming the primary cells with secondary-form DNA on a
low-copy-number vector and screening for colonies which did not
produce the yellow pigment characteristic of primaries. Four
transformants all carrying the same gene were found to loose
primary-form-specific characteristics, and the gene was sequenced and
identified as ner, a regulatory gene in gram-negative bacteria and
their phages. Unexpectedly, inactivation of the endogenous gene in the
secondaries did not cause them to revert to the primary phenotype, and
the gene was expressed in the primary form as well as the secondary
form during exponential but not stationary phase and deregulated in
the plasmid-bearing primary form. These and other pieces of evidence
indicate that the endogenous ner gene is not responsible for the
secondary phenotype, but that ner, when overexpressed, can repress
expression of primary phenotypes at stationary phase. Inactivation of
the endogenous ner gene in the primary form affected the outer
membrane protein profile. A number of outer membrane proteins
displayed differential accumulation in the primary and secondary forms
at stationary phase, and two of the primary-form-specific proteins
were absent from the ner primary strain.&lt;/AbstractText&gt;
&lt;/Abstract&gt;
&lt;Affiliation&gt;Department of Biology, National University of
Ireland, Maynooth, Ireland.&lt;/Affiliation&gt;
&lt;AuthorList&gt;
&lt;Author&gt;
&lt;LastName&gt;O'Neill&lt;/LastName&gt;
&lt;ForeName&gt;Keith H&lt;/ForeName&gt;
&lt;Initials&gt;KH&lt;/Initials&gt;
&lt;/Author&gt;
&lt;Author&gt;
&lt;LastName&gt;Roche&lt;/LastName&gt;
&lt;ForeName&gt;Declan M&lt;/ForeName&gt;
&lt;Initials&gt;DM&lt;/Initials&gt;
&lt;/Author&gt;
&lt;Author&gt;
&lt;LastName&gt;Clarke&lt;/LastName&gt;
&lt;ForeName&gt;David J&lt;/ForeName&gt;
&lt;Initials&gt;DJ&lt;/Initials&gt;
&lt;/Author&gt;
&lt;Author&gt;
&lt;LastName&gt;Dowds&lt;/LastName&gt;
&lt;ForeName&gt;Barbara C A&lt;/ForeName&gt;
&lt;Initials&gt;BC&lt;/Initials&gt;
&lt;/Author&gt;
&lt;/AuthorList&gt;
&lt;Language&gt;eng&lt;/Language&gt;
&lt;PublicationTypeList&gt;
&lt;PublicationType&gt;Journal Article&lt;/PublicationType&gt;
&lt;/PublicationTypeList&gt;
&lt;/Article&gt;
&lt;MedlineJournalInfo&gt;
&lt;Country&gt;United States&lt;/Country&gt;
&lt;MedlineTA&gt;J Bacteriol&lt;/MedlineTA&gt;
&lt;NlmUniqueID&gt;2985120R&lt;/NlmUniqueID&gt;
&lt;/MedlineJournalInfo&gt;
&lt;CitationSubset&gt;IM&lt;/CitationSubset&gt;
&lt;/MedlineCitation&gt;
&lt;PubmedData&gt;


"Steve Holden" <sholden at holdenweb.com> wrote in message news:<WSnL8.268879$%u2.175369 at atlpnn01.usenetserver.com>...
> 
> Just to be sure, do you expect the second link (in a browser) to show "Error
> handling request: no current query"? This seems to work w/Python 2.2 under
> cygwin, at least. There was so much output I omitted quite a lot, replaced
> by ellipses below:
> 
> >>> import urllib
> >>> f =
> urllib.urlopen("""http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Display&
> db=PubMed&details_term=%28%22glucagon%22%5BMeSH%20Terms%5D%20OR%20glucagon%5
> BText%20Word%5D%29&dopt=XML&query_key=1""")
> >>> ll  = f.read()
> >>> ll
> '<html>\n  <head> ...
> <table
> cellpadding="0" cellspacing="0" width="100%"><tr><td bgcolor="#ffcccc">Error
> handling request: no current query<BR></td></tr></table> ...
> </body>\n</html>\n'
> >>>
> 
> regards
>  Steve


From narnett at mccmedia.com  Fri Jun 14 17:19:53 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Fri, 14 Jun 2002 14:19:53 -0700
Subject: executemany & UPDATE (MySQL & Python)
In-Reply-To: <20020614191751.GA1520@lilith.my-fqdn.de>
Message-ID: <NDBBLAEMOMEHKALAMKKJMEKIKPAA.narnett@mccmedia.com>


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Gerhard H?ring
> Sent: Friday, June 14, 2002 12:18 PM

[snip]

> I mistakenly assumed that MySQLdb's executemany was implemented the same
> way as pyPgSQL's. That assumption was wrong. Looking into the source
> reveals that MySQLdb's implementation is really clever, but
> unfortunately hardcoded to only work with INSERT statements.

Thanks for looking at the source -- I was trying to avoid doing that
myself...  I guess that settles it.

Nick





From emile at fenx.com  Thu Jun 27 11:39:24 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 27 Jun 2002 15:39:24 GMT
Subject: Var substitution
References: <c07d0edc.0206270557.28180893@posting.google.com>
Message-ID: <MkGS8.336559$cQ3.20337@sccrnsc01>

Guy
> >>> guy="name"
> >>> import os
> >>> os.guy
> 
> This is what should happen :
> 
> >>> import os
> >>> os.name
> 'nt'

Python 2.2.1 (#34, Apr 15 2002, 09:51:39) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> guy='name'
>>> import os

preferred is:
>>> getattr(os, guy)
'nt'

although:
>>> os.__dict__[guy]
'nt'

--

Emile van Sebille
emile at fenx.com

---------



From sarmstrong13 at mac.com  Wed Jun  5 00:28:50 2002
From: sarmstrong13 at mac.com (SA)
Date: Tue, 04 Jun 2002 23:28:50 -0500
Subject: A simple tcl/tk program to run Python.
References: <B922CA56.638E%sarmstrong13@mac.com>
Message-ID: <B922FDB2.661E%sarmstrong13@mac.com>

Found Typo. Here is corrected script:

#!/usr/bin/wish

text .t1 -height 12 -width 84 -font {Times 12}
frame .f
text .t2 -height 8 -width 84 -bg lightblue -font {Courier 9}
button .f.b1 -text Execute -command python
button .f.b2 -text "Clear input" -command ".t1 delete 0.0 end"
button .f.b3 -text "Clear output" -command ".t2 delete 0.0 end"
pack .f.b1 .f.b2 .f.b3 -side left -padx 2
pack .t1 .f .t2 -side top -pady 2

proc python {} {
        catch {exec python -c [.t1 get 0.0 end]} output
    .t2 delete 0.0 end
    .t2 insert end $output
}



From jb at cascade-sys.com  Wed Jun  5 19:15:46 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Wed, 05 Jun 2002 16:15:46 -0700
Subject: Interfacing to 1-Wire LAN, especially Dallas temperature sensors?
References: <a7iefuk650noamrqcmpvroqj60mtgbpf9v@4ax.com>
Message-ID: <3CFE9BA2.3CBEAF7D@cascade-sys.com>

It's possible you'll have better results via the parallel port:

    http://www.google.com/search?hl=en&q=i2c+parallel+port

"I2C" is similar to the two-wire protocol (data plus clock) used on the
dallas parts.

Good luck

--jb


Christian Aastorp wrote:

> I'm considering doing some temperature control and logging, both
> enviromental and inside computers. Did a google search, and found
> references to the range of temperature sensors from Dallas
> Semiconductors.
>
> I also found schematics for interfacing with the serial port, and
> timing diagrams for implementing 1-Wire Lan using direct control of
> the port.
>
> As I'm using  Python for my programming, these days, I would like some
> pointers to modules or techniques to use Python to read the
> temperature sensors.
>
> Christian Aastorp
> --
> http://mail.python.org/mailman/listinfo/python-list

--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From erict at millfilm.co.uk  Thu Jun  6 07:48:36 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Thu, 06 Jun 2002 12:48:36 +0100
Subject: HttpServer help!
Message-ID: <3CFF4C14.450850D8@millfilm.co.uk>

I am trying to get some simple example running locally on my
machine of http(server/client)  code from the book 'Python & Xml'
and I am having trouble.

if someone can look at the following script and tell me if
the problem can be because I don't have the privilege and/or
I didn't setup my system properly:
Thanks.

I am running on Linux Red Hat with python2.2.


The client code to POST:
"""
post.py
"""
from httplib import HTTP
from urllib  import quote

# establish POST data
myquote = 'This is my quote: "I think therefore I am."'

# be sure not to quote the key= sequence...
postdata = "favquote=" + quote(myquote)

print "Will POST ", len(postdata), "bytes:"
print postdata

# begin HTTP request

#req = HTTP("127.0.0.1")  # change to your IP or localhost
#req = HTTP("localhost")  # change to your IP or localhost

##### MY IP address
req = HTTP("10.16.3.132")  # change to your IP or localhost


#req.putrequest("POST", "/c8/favquote.cgi")
req.putrequest("POST",
"file:/usr/people/erict/python/srcToolskit/pythonXmlBook/c8/favquote.cgi")

.
.
.

##### That the Error after the ruprequest->"#####
Traceback (most recent call last):
  File "post.py", line 25, in ?
    req.putrequest("POST",
"file:/usr/people/erict/python/srcToolskit/pythonXmlBook/c8/favquote.cgi")

  File "/usr/people/erict/bin//lib/python2.2/httplib.py", line 453, in
putrequest
    self.send(str)
  File "/usr/people/erict/bin//lib/python2.2/httplib.py", line 395, in
send
    self.connect()
  File "/usr/people/erict/bin//lib/python2.2/httplib.py", line 379, in
connect
    raise socket.error, msg
socket.error: (111, 'Connection refused')


The server seems to work fine: Here is the code:  HTTPServer.py
##########################################################
"""
HTTPServer.py - an simple implementation
of the BaseHTTPServer module
"""

from BaseHTTPServer import HTTPServer
from BaseHTTPServer import BaseHTTPRequestHandler
import string

"""
class myRequestHandler - Handles any and all request
    coming in, regardless of path, file, or request
    method (GET/POST)
"""
class myRequestHandler(BaseHTTPRequestHandler):
  """
  do_GET will be called if the browser does a GET
  request.
  """
  def do_GET(self):
    # give them back a 200 OK every time.
    self.printCustomHTTPResponse(200)

    # start HTML output
    self.wfile.write("<html><body>")
    self.wfile.write("<p>Hello, I am a web server, sort of.</p>")
    self.wfile.write("<p>GET string: " + self.path + "</p>")

    # show browser headers
    self.printBrowserHeaders()

    # finish off HTML
    self.wfile.write("</html></body>")

  """
  do_POST is called if the browser is POSTing data
  from a form
  """
  def do_POST(self):
    # send back a 200 OK
    self.printCustomHTTPResponse(200)

    # start HTML and show browser headers again
    self.wfile.write("<html><body>")
    self.printBrowserHeaders()
    self.wfile.write("<b>Post Data:</b><br>")

    # track down length of the post, so that you
    # can read in the correct number of bytes.  The
    # length of the post is in the browser header
    # named 'content-length'.
    if self.headers.dict.has_key("content-length"):
       # convert content-length from string to int
       content_length = string.atoi(self.headers.dict["content-length"])

       # read in the correct number of bytes from the client
       # connection and send it back to the browser
       raw_post_data = self.rfile.read(content_length)
       print("[Post Data:]", raw_post_data)
       self.wfile.write(raw_post_data)

    # finish off HTML
    self.wfile.write("</html></body>")

  """
  printBrowserHeaders - this method prints the HTTP
      headers sent by the client
  """
  def printBrowserHeaders(self):
    # iterate through header dictionary and
    # display the name and value of each pair.
    self.wfile.write("<p>Headers: <br>")
    header_keys = self.headers.dict.keys()
    for key in header_keys:
      self.wfile.write("<b>" + key + "</b>: ")
      self.wfile.write(self.headers.dict[key] + "<br>")

  """
  printCustomHTTPResponse - this method takes a response
      code and sends the code and custom headers
      back to the browser
  """
  def printCustomHTTPResponse(self, respcode):
    # send back appropriate HTTP code
    self.send_response(respcode)

    # tell them we're sending HTML
    self.send_header("Content-type", "text/html")

    # describe the server software in use!
    self.send_header("Server", "myRequestHandler :-)")

    # close off headers
    self.end_headers()


# start the server on port 2112, requires browser URLs
# to be addressed as http://servername:2112/pathtofile
myServer = HTTPServer(('', 2112), myRequestHandler)

# loop to handle 5 requests
for lp in range(5):
  myServer.handle_request()
##########################################################




From sameer_ at email.com  Tue Jun 11 09:43:51 2002
From: sameer_ at email.com (sameer)
Date: 11 Jun 2002 06:43:51 -0700
Subject: Source formatting for long expressions.
References: <mailman.1023570972.27430.python-list@python.org>	<m3it4srhq5.fsf@mira.informatik.hu-berlin.de>	<874rgcxy99.fsf@twistedmatrix.com> <mailman.1023668414.25769.python-list@python.org>
Message-ID: <6bd9f01b.0206110543.168b0f83@posting.google.com>

why not just do it in xml?

Christopher Armstrong <radix at twistedmatrix.com> wrote in message news:<mailman.1023668414.25769.python-list at python.org>...
> >>>>> "ca" == Christopher Armstrong <radix at twistedmatrix.com> writes:
>  
> >>>>> "mvl" == Martin v Loewis <martin at v.loewis.de> writes:
> 
>     mvl> As you can see, it tries to align further arguments to a function
>     mvl> together with the opening parenthesis. I would assume that any other
>     mvl> auto-formatters use the same style, so you won't get any "sane"
>     mvl> indentation until you drop the nesting level (perhaps in favour of
>     mvl> method invocations).
> 
> [snip..]
>     ca> I'm probably going to end up doing this myself, and I thought of
>     ca> something while reading your message. I could have two steps for
>     ca> formatting: template- generation and indentation. the first step would
>     ca> generate something like a repr() of all of my objects, but with \ns and
>     ca> \ts embedded at strategic locations. Then I could replace all \ts with
>     ca> current_indent_level, calculating current_indent_level by counting (s,
>     ca> {s, and [s. I'm going to try this out now.  :-)
> 
> 
> I got this working. It was fairly easy, once I got a templated source 
> representation::
> 
> def indentify(s):
>     out = []
>     stack = []
>     for ch in s:
>         if ch in ['[', '(', '{']:
>             stack.append(ch)
>         elif ch in [']', ')', '}']:
>             stack.pop()
>         if ch == '\t':
>             out.append('  '*len(stack))
>         else:
>             out.append(ch)
>     return string.join(out, '')
> 
> 
> This is a fairly stupid indenter, but it works for my needs::
> 
> >>> print aot.indentify("Instance('twisted.internet.app.Application', \
> \n\tfoo=bar, \n\tbaz={\n\t'quux': 1, \n\t'spam': \n\tInstance('Eggs', \
> \n\tmore='state')})") #this source will really be generated by my __repr__s.
> Instance('twisted.internet.app.Application', 
>   foo=bar, 
>   baz={
>     'quux': 1, 
>     'spam': 
>     Instance('Eggs', 
>       more='state')})


From ellers at iinet.net.au  Fri Jun 21 02:19:10 2002
From: ellers at iinet.net.au (Ellers)
Date: 20 Jun 2002 23:19:10 -0700
Subject: wxPython: wxEvtHandler leak on exit?
Message-ID: <382bfff5.0206202219.e65e5d2@posting.google.com>

Hello all

I'm developing an app using wxPython (2.3.2) on win32, solaris and
linux. This particular problem doesn't appear on solaris/linux but on
win32 when I close all the frames (ie exit the app) I get this:

14:18:34: There were memory leaks.
14:18:34: ----- Memory dump -----
14:18:34: wxEvtHandler at $173F538, size 32
14:18:34: wxPyCallback at $173F478, size 12
14:18:34: wxObject at $173F428, size 28
14:18:34:
14:18:34:
14:18:34: ----- Memory statistics -----
14:18:34: 1 objects of class wxObject, total size 28
14:18:34: 1 objects of class wxPyCallback, total size 12
14:18:34: 1 objects of class wxEvtHandler, total size 32
14:18:34:
14:18:34: Number of object items: 3
14:18:34: Number of non-object items: 0
14:18:34: Total allocated size: 72
14:18:34:
14:18:34:

I have an event handler class that routes commands from a thread and
effectively does this:

	eventHandler = MyEventHandler( )

	class MyEventHandler( wxEvtHandler ):
	
		def __init__( self ):
			wxEvtHandler.__init__( self )
			EVT_MY_EVENT( self, self.OnMyEvent )
		
		def OnMyEvent( self, event ):
			exec event.code

And the event stuff is like this:

	wxEVT_MY_EVENT = wxNewEventType()

	def EVT_MY_EVENT(win, func):
		win.Connect(-1, -1, wxEVT_MY_EVENT, func)

	class MyEvent( wxPyEvent ):

		"""Allow the shell thread to post some code to be executed in the
main thread"""
			
		def __init__(self, code ):
			wxPyEvent.__init__(self)
			self.SetEventType(wxEVT_MY_EVENT)
			self.code = code


This works nicely 99% of the time.

However, when I send this command:

    wxPostEvent( eventHandler, MyEvent( "app.frame.Close( )))

The frame closes correctly but... somewhere in the process of shutting
down the app the above memory leak occurs.

I don't get it.

Any ideas?

Thanks 
Ellers


From greg at cosc.canterbury.ac.nz  Sun Jun 30 03:01:59 2002
From: greg at cosc.canterbury.ac.nz (greg)
Date: Sun, 30 Jun 2002 19:01:59 +1200
Subject: ANN: Pyrex 0.3.2
Message-ID: <mailman.1025533982.26482.clpa-moderators@python.org>

Pyrex 0.3.2 is now available:

http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/

Lots of little bug fixes and improvements,
including:

- Fixed the & operator that got broken in 0.3.1
- Stopped the weird "SystemError: 'finally' pops bad exception"
- Various crashes
- 2-element slice indexing expressions work now
- Most arithmetic operations now allowed in constant
  expressions (so you can have negative enum values now!)

What is Pyrex?
--------------

Pyrex is a new language for writing Python extension modules.
It lets you freely mix operations on Python and C data, with
all Python reference counting and error checking handled
automatically.




From spam at eatshit.net  Mon Jun 17 19:52:42 2002
From: spam at eatshit.net (paul colby)
Date: Mon, 17 Jun 2002 23:52:42 GMT
Subject: Newbie Help Please.(OO and Tkinter)
References: <B93209DC.7281%sarmstrong13@mac.com> <r3fP8.401$uH2.132@newsread1.prod.itd.earthlink.net> <B933AE19.7379%sarmstrong13@mac.com>
Message-ID: <eDuP8.1591$uH2.431@newsread1.prod.itd.earthlink.net>

SA wrote:

> 
> from Tkinter import *
> import os
> import commands
> import tkSimpleDialog
> 
> class PyShell:
> def?clearin(self):
> self.t1.delete(0.0,END)
> def?clearout(self):
> self.t2.delete(0.0,END)
> def?expyth(self):
> self.t2.delete(0.0,?END)
> self.output?=?commands.getoutput(self.t1.get(0.0,?END))
> self.t2.insert(END,?self.output)
> def?saving_callback():
> savin()
> def?__init__(self,?top):
> self.SAVE?=?Saving(self)
> self.t1?=?Text(top,?height="12",?width="84",?font="Courier?12")
> self.t1.pack(side=TOP,?pady=2)
> f?=?Frame(top)
> f.pack()
> self.b1?=?Button(f,?text="Execute",?command=self.expyth)
> self.b1.pack(side=LEFT)
> self.b2?=?Button(f,?text="Clear?Input",?command=self.clearin)
> self.b2.pack(side=LEFT)
> self.b3?=?Button(f,?text="Clear?Output",?command=self.clearout)
> self.b3.pack(side=LEFT)
> self.b4?=?Button(f,?text="Save?Input",?command=self.SAVE)
> self.b4.pack(side=LEFT)
> self.t2?=?Text(top,?height="12",?width="84",?bg="lightblue",
> font="Courier 12")
> self.t2.pack(side=TOP,?pady=2)
> 
> class Saving(tkSimpleDialog.Dialog):
> def?savin(self,?question):
> Label(question,?text="Directory:").grid(row=0)
> Label(question,?text="Filename:").grid(row=1)
> self.e1?=?Entry(question)
> self.e2?=?Entry(question)
> self.e1.grid(row=0,?column=1)
> self.e2.grid(row=1,?column=1)
> def?apply(self):
> self.dir?=?self.e1.get()
> self.fn?=?self.e2.get()
> self.sav?=?self.dir?+?self.fn
> self.savfile?=?open(self.sav,?"w")
> for?line?in?self.t1.readlines():
> self.savefile.write(line)
> self.savefile.close()
> 
> 
> 
> root = Tk()
> app = PyShell(root)
> root.mainloop()
> 

Sure. I got a little further using

   self.SAVE = Saving(top)

which passes the correct parameter at least. 
Beyond this your code needs more work.

Regards
Paul Colby


From m.laranjeira at datacraft.com.br  Fri Jun 21 14:36:14 2002
From: m.laranjeira at datacraft.com.br (Marcus Vinicius Laranjeira)
Date: Fri, 21 Jun 2002 15:36:14 -0300
Subject: [newbie] How class works ?
Message-ID: <5.1.1.6.0.20020621153612.009ea870@imap.datacraft.com.br>


I have a problem. I wrote a file with a class. When I try to instantiate 
this class in another file, I receive an error telling me : "Failed to 
reload - 'module' object is not callable

I've tried many things to solve this problem, but I can't find where is the 
error or how to solve !

I would appreciate any help !

Marcus





From kzaragoza at attbi.com  Sun Jun  9 20:14:02 2002
From: kzaragoza at attbi.com (Kris J. Zaragoza)
Date: 9 Jun 2002 19:14:02 -0500
Subject: python interacting with other programs
References: <64eb9145.0206041454.4f70e86e@posting.google.com> <7ZRM8.9358$bt2.116439@charlie.risq.qc.ca>
Message-ID: <slrnag7rs8.o5g.kzaragoza@localhost.localdomain>

In article <7ZRM8.9358$bt2.116439 at charlie.risq.qc.ca>, Marc Cote wrote:
> "Jakob" <da_woym at hotmail.com> wrote in message
> news:64eb9145.0206041454.4f70e86e at posting.google.com...
>> now I know python can interact with programs like Microsoft wor and
>> access and all but has anyone ever tried to interact with Lotus Notes?
>> Is it possible?
> 
> Jakob,
> 
> Google Search: "Lotus Notes Python" gives:
> 
>     Using Python to create a command line interpreter for Notes
> 
> URL: http://www.dominopower.com/issues/issue200008/command001.html
> 

Using Notes via COM on a Windows box is really the most viable option.
I did a bit of this while working at Lotus, and it really is quite a
nice way to get data in and out of a Notes database.  Thankfully, the
COM API exposed by Lotus is well structured, and easy to work with.

Unfortunately, the COM interface doesn't expose the entirety of the
Notes API.  If you need to perform any low-level functions or access
parts of the API that aren't exposed via COM, you'll have to roll your
own extension module.  If you want to use Python to access Notes
databases on any platform other than Windows, you'll have to roll your
own.

Even with tools like SWIG, you'll have quite a job of building an
extension library, as it's a very big API.  On the bright side, you
generally only need to call a handful of functions, so building a
limited Notes API wrapper shouldn't be too difficult.  Unfortunately,
I never started down this path, so I don't have any code to
contribute.  (I'm also working in a pretty much all-Microsoft shop, so
we're using that POS Exchange and a bunch of folders on a file server
instead of a nice set of Notes-based collaborative apps.  *grumble*)

-Kris


-- 
Kris J. Zaragoza       | On the face of it, Microsoft complaining about
kzaragoza at attbi.com    | the source license used by Linux is like the
                       | event horizon calling the kettle black.
                       | -- Adam Barr, article on kuro5hin.org


-----------== Posted via Newsfeeds.Com - Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Ulimited Fast Downloads - 19 Servers =-----


From rhaper at houston.rr.com  Sun Jun  2 19:57:04 2002
From: rhaper at houston.rr.com (Rod Haper)
Date: Sun, 02 Jun 2002 23:57:04 GMT
Subject: List
References: <3cfa6ed8_1@news.iprimus.com.au> <addrjh$imq$1@news.tudelft.nl> <3cfa7619_1@news.iprimus.com.au> <addths$jb9$1@news.tudelft.nl> <3cfa8ef5_1@news.iprimus.com.au>
Message-ID: <3CFAB0DB.4020309@houston.rr.com>


Gold Fish wrote:

> What i am trying to do is that i got one biglist don't know how big is it 
> say list1 then i like to divided into some small sublists. These sublists 
> will have some elements in the biglist for example 3 elements maximum in 
> each sublist. So if the biglist got 10 elements then it'll be divided into 
> 4 sublists the first 3 will have 3 elements and the last one will have one 
> elements.
> 


[rodh at rodh rodh]$ python
Python 2.2.1 (#1, Apr 20 2002, 20:12:24)
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-85)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
 >>> bigList = ['a','b','c','d','e','f','g','h','i','j']
 >>> subLists = []
 >>> subListSize = 3
 >>> for i in range(0,len(bigList),subListSize):
...     subLists.append(bigList[i:i+subListSize])
...
 >>> subLists
[['a', 'b', 'c'], ['d', 'e', 'f'], ['g', 'h', 'i'], ['j']]
 >>>

Is this what you are looking for?



-- 
Rod

+----------------------------------+
| Rod Haper, rhaper at houston.rr.com |
|     There is a better way ...    |
|   LAP => Linux + Ada95 + Python  |
+----------------------------------+



From paoloinvernizzi at dmsware.com  Wed Jun 26 03:19:58 2002
From: paoloinvernizzi at dmsware.com (Paolo Invernizzi)
Date: Wed, 26 Jun 2002 07:19:58 GMT
Subject: Python and Eclipse
References: <3D17FAF4.35CC9F1B@earthlink.net> <81CC01E58BF2A99B.DFCD5EC6E4654907.32D731380951B9CA@lp.airnews.net> <3D189A8A.577C8186@earthlink.net> <3D18A6D2.5080404@dmsware.com> <3D18D159.2F66A9E8@earthlink.net>
Message-ID: <3D196AA0.2090202@dmsware.com>

Joseph A Knapka wrote:

> I think the latter is the best way. It's not clear, though,
> that these are the only two roads. But in any case, we must
> IMO be able to debug code under any Python version, preferably
> selectable among a number of different installed versions
> on the same machine (Eclipse allows this with the JVM, so
> I don't see why it would be a problem for Python). That
> way we could support CPython and Jython equally well.

This night I remembered another tip... pyro, 
http://pyro.sourceforge.net/ ,a distributed framework, has 
jython/cpython client and cpython server.
I think this is an attractive choice. The first step can be implement a 
native support for jython, endebbing a jython interpreter in the plugin, 
and then use it as a bridge for external cpython server via pyro...
But I remember that when I tried to compile/manage the jython code via 
eclipse, I bounced against ANT wierdness (I haven't used ant at all in 
my previous java works...)

Cya
Paolo Invernizzi





From Kemp100 at attbi.com  Tue Jun 25 13:21:29 2002
From: Kemp100 at attbi.com (Dirgesh Patel)
Date: 25 Jun 2002 10:21:29 -0700
Subject: How do i COPY files from 1 directory to another..
Message-ID: <aec84c2f.0206250921.535f3d83@posting.google.com>

How do i Copy multiple files from 1 directory to another..this is the
code I have so far..but it gives me a PERMISSION DENIED ERROR..


import shutil
import os

shutil.copy2(r"C:\test",r"C:\test1")

I wanna copy everythign from C:\test to C:\test1....


From hst at empolis.co.uk  Fri Jun  7 12:17:44 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Fri, 7 Jun 2002 17:17:44 +0100
Subject: Newby: How do I strip HTML tags?
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220BF4@hendrix.empolisuk.com>

Harvey Thomas wrote

> netvegetable wrote
> 
> > I'm mucking around with cgi, and I'm trying to work out a way 
> > to strip the 
> > html tags of a string. e.g, I want to convert this...
> > 
> > ><font size = 12><b><big>Really Big String</big></b></font>
> > 
> > to this this ...
> > 
> > >Really Big String
> > 
> > ... and store it as a value.
> > 
> > I worked out a crude, but effective way of doing it (see code 
> > below), but I 
> > can't escape the feeling there must be a built in way of 
> > doing it more. If 
> > nothing else, I'm sure somebody who knows their regular 
> > expressions could 
> > neaten it up (please?).
> > 
> > def strip_html_tags(it):
> > 	left = it[:(len(it)/2)]
> > 	right = it[(len(it)/2):]
> > 	final = left[left.rfind('>')+1:] + right[:right.find('<')]
> > 	return final
> > 
> 
> If your HTML is reasonably legal, then you can use something 
> along the lines of the following very quick and very dirty program:
> 
> import re
> import sys
> 
> s = open(sys.argv[1]).read()
> o = open('tmp.tmp', 'w')
> r = re.compile('(<!--.*?-->)|(<[^>]*>)([^<]+)', re.DOTALL)
> for x, y, z in r.findall(s):
>     if z and not z.isspace():   #don't use comments tags and 
> white-space only content
>         print >>o, z
> 
> Note that you have to test first for HTML comments as a 
> comment can contain a '>' character.
> 

Sorry, the RE should read
r = re.compile('(<!--.*?-->)|(<[^>]*>)|([^<]+)', re.DOTALL)

Harvey

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.




From peter at engcorp.com  Mon Jun 17 21:12:54 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 21:12:54 -0400
Subject: you guys needs a logo
References: <7fe97cc4.0206142325.4df3a174@posting.google.com> <3eeda89d.0206170232.675596b8@posting.google.com>
Message-ID: <3D0E8916.65AC2F13@engcorp.com>

Thinkit wrote:
> 
> xah at xahlee.org (Xah Lee) wrote in message news:<7fe97cc4.0206142325.4df3a174 at posting.google.com>...
> > you guys need a logo.
> >
> > the light-bulb is quite lame as far as logo goes.
> >
> > I suggest that the O'Reilly's python book cover a good choice. It
> > should not be something funny and cute.
> >
> > The MacPython's some 16 Ton weight block is also a good logo.
> >
> > Perhaps you guys can run a logo competition.
> >
> >  Xah
> >  xah at xahlee.org
> >  http://xahlee.org/UnixResource_dir/complang.html#python
> 
> How about a penguin masturbating?  To show what else besides python linux males do.

I think a logo with a penguin stroking its python would be _cute_.

Some might misinterpret it, however...

-Peter


From mcfletch at rogers.com  Mon Jun  3 22:14:14 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 03 Jun 2002 22:14:14 -0400
Subject: Time
References: <mailman.1023144005.19450.python-list@python.org> <ZmUK8.205609$%u2.66972@atlpnn01.usenetserver.com>
Message-ID: <3CFC2276.70105@rogers.com>

Um, actually, you probably want:

datePlus5 = day + RelativeDateTime( days = 5 )

(Note the 's'), without it, you find "day = 5", that is, the 5th of the 
month.

For simple stuff like file-date manipulation, GoldFish can probably just 
use the time module and not need the extra package download.

t = t - (60.0*60*24*5)

Not accurate when Daylight Savings Time switches over, or leap-seconds 
occur, but likely gets the job done for a beginner's web-site scripting.

Enjoy,
Mike

Steve Holden wrote:
...
>>>filename.py is Mon Jun 3 03:25:57 2002 ,if i want to set the time
>>
> forward
> 
>>>or backwards in 5 days time, how can i do that.
>>
> 
> The "mx" way to do what GoldFish wants is to create a
> RelativeDateTime(day=5) and subtract it from the current date.
> 
...

_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From johnroth at ameritech.net  Tue Jun 18 07:41:36 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 18 Jun 2002 07:41:36 -0400
Subject: why len(list) instead of list.len()  ?
References: <EhBP8.1$Qn1.1124@news.ecrc.de>
Message-ID: <ugu728h51f7c83@news.supernews.com>

"Markus Jais" <mjais at web.de> wrote in message
news:EhBP8.1$Qn1.1124 at news.ecrc.de...
> hello
>
> I way wondering why one has to write
>
> list = [2, 3, 5]
> print len(list)
>
> instead of list.len()

This has come up before, many, many times. There
are two answers. One is that there is, in fact, a
list method to return a length. It is list.__len__().
The len() global function calls this method.
All objects that implement the len() function
do it by implementing this 'magic method.'

The other is that what feels intuitive to one
person doesn't necessarily feel intuitive to
someone else with a different background, etc.

John Roth





From shalehperry at attbi.com  Fri Jun 28 22:57:50 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 28 Jun 2002 19:57:50 -0700 (PDT)
Subject: I'm an idiot
In-Reply-To: <20020629031556.GB2953@schaduw.felnet>
Message-ID: <XFMail.20020628195750.shalehperry@attbi.com>

> 
> Many people would rely on Python's automatic file closing behaviour
> and write it like this:
> 
>    g = open('c:/temp/temp1.txt', 'w')
>    for line in open('c:/temp/temp.txt'):
>        g.write(line.strip())
> 

no, we wouldn't.




From dfackrell at DELETETHIS.linuxmail.org  Fri Jun 14 16:03:14 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Fri, 14 Jun 2002 14:03:14 -0600
Subject: why not "'in' in 'in'"?
References: <mailman.1024019967.16051.python-list@python.org> <aed9v40e68@drn.newsguy.com>
Message-ID: <3d0a4c01$1_1@hpb10302.boi.hp.com>

I'll try to figure out how to make this change next week sometime.  I've
been looking for a challenge where I'll be able to accomplish something
useful anyway, and this looks like one where I'll learn a lot without
spending too much time.

If I get stuck on the way, I'll be sure to ask for help here.

Tim, your post begs a question about bets, bots, and currency.  <wink>

--
Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.

"Grant Griffin" <not.this at seebelow.org> wrote in message
news:aed9v40e68 at drn.newsguy.com...
> In article <mailman.1024019967.16051.python-list at python.org>, Tim says...
> >
> >[Grant Griffin]
> >> One of the things I like about Python is that it works as expected.
> >> However, I sometimes accidently expect it to allow one to use "in"
> >> to determine if one string is "in" another, e.g.:
> >>
> >> >>> 'in' in 'in'
> >>
> >> which yields a traceback:
> ...
> >So if somebody wants this change enough to submit a patch (code, doc, +
test
> >suite changes), it's likely to be accepted for 2.3.  As code changes go,
> >this one should be particularly easy.  We have an internal bet going as
to
> >whether this invitation is enough to kill the idea <wink>.
>
> Thanks for the info, Tim.  I'm not in a position to do this work myself,
so if
> somebody else wants to do it, I don't mind.  (Besides, I was just asking a
> _question_ <wink>.)
>
> But until this change magically appears, just knowing that there is no
grand
> philosophical reason why something as intuitive and practical as this must
be
> excluded from Python restores my faith.
>
> faithful-ly y'rs,
>
> =g2
>
> _________________________________________________________________________
>
> Grant R. Griffin                                           g2 at dspguru.com
> Publisher of dspGuru                               http://www.dspguru.com
> Iowegian International Corporation                http://www.iowegian.com
>




From mwh at python.net  Wed Jun 19 09:57:48 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 19 Jun 2002 13:57:48 GMT
Subject: pychecker + XEmacs
References: <3d0de483$0$3881$e4fe514c@dreader4.news.xs4all.nl> <aekok8$7p118$1@ID-59885.news.dfncis.de> <20020618165548.GB29257@matijek.plusseven.com> <20020618175005.GA613@lilith.my-fqdn.de> <mailman.1024425924.24614.python-list@python.org>
Message-ID: <lk8z5bl6ce.fsf@pc150.maths.bris.ac.uk>

"Daniel 'eikeon' Krech" <eikeon at eikeon.com> writes:

> Gerhard H?ring <gerhard at bigfoot.de> writes:
> 
> > * Alex Polite <m2 at plusseven.com> [2002-06-18 18:55 +0200]:
> > > Interesting.  Where is the CVS repository for python-mode.el?
> > 
> > It's in the Python CVS repository. Directory dist/src/Misc/
> 
> I had trouble finding it when I was looking around too. When does the
> one on http://www.python.org/emacs/python-mode/ get updated... 

At a guess, when Barry gets round to it.

> Could a link to the CVS version be added to the python-mode page? 

Probably :)

I've added a brief note to the page.

Cheers,
M.

-- 
  Hmmm... its Sunday afternoon: I could do my work, or I could do a
  Fourier analysis of my computer's fan noise.
       -- Amit Muthu, ucam.chat (from Owen Dunn's summary of the year)


From tjreedy at udel.edu  Sun Jun  2 16:58:27 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Sun, 02 Jun 2002 20:58:27 GMT
Subject: string generator
References: <3cfa8039_1@news.iprimus.com.au>
Message-ID: <TFvK8.104315$Gs.10133770@bin5.nnrp.aus1.giganews.com>

"Gold Fish" <occeanlinux at linuxmail.org> wrote in message
news:3cfa8039_1 at news.iprimus.com.au...
> I like to create a list of file automatical based on series. For
example
> user input the number of file say 5. So i want to generate the file
> according to this range , this mean the generated files could be
> filename1
> filename2
> filename3
> filename4
> filename5
> I using the for loop like that
>
> for file in range(5):
>         lis = [file]
>         string.join('filename',lis[1:])
> how can it didn't work. I am confusing how to do it.

Perhaps you are looking for something like (untested):

define fileseries(basename, n):
  for i in range(n):
    yield basename+str(i)

Terry J. Reedy




From amuys at shortech.com.au  Mon Jun 24 00:36:50 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 23 Jun 2002 21:36:50 -0700
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at>	<3d149f19$0$53133$edfadb0f@dspool01.news.tele.dk>	<af28vp$j57$1@panix1.panix.com> <3D14AD7F.3D5D243F@engcorp.com> 	<af2ia8$a92$1@panix1.panix.com> <mailman.1024809149.4737.python-list@python.org>
Message-ID: <7934d084.0206232036.3df77ba4@posting.google.com>

Cliff Wells <logiplexsoftware at earthlink.net> wrote in message news:<mailman.1024809149.4737.python-list at python.org>...
> On Sat, 2002-06-22 at 12:14, Aahz wrote:
> > In article <3D14AD7F.3D5D243F at engcorp.com>,
> > Peter Hansen  <peter at engcorp.com> wrote:
> > >Aahz wrote:
> > >> Ya know, you could have saved yourself a lot of typing by simply saying
> > >> PBCAK.  ;-)
> > >
> > >To save us from those who can't Google:  Problem Between Chair And Keyboard.
> > >(It was new to me, too.)
> > 
> > <grin>  Ya coulda ast me, too.
> 
> I'd always heard of it called "chair-keyboard interface problem".
> 
And I've always known it as PEBKAC as per the Jargon File entry

http://tuxedo.org/~esr/jargon/html/entry/PEBKAC.html

Andrae


From sarmstrong13 at mac.com  Sun Jun 16 19:38:35 2002
From: sarmstrong13 at mac.com (SA)
Date: Sun, 16 Jun 2002 18:38:35 -0500
Subject: Newbie Help Please.(OO and Tkinter)
References: <B93209DC.7281%sarmstrong13@mac.com> <aeit4p$h0s$1@slb5.atl.mindspring.net>
Message-ID: <B9328BAB.72CB%sarmstrong13@mac.com>

On 6/16/02 3:37 PM, in article aeit4p$h0s$1 at slb5.atl.mindspring.net, "Matt
Gerrans" <mgerrans at mindspring.com> wrote:

> Did you intend to be defining the method apply() inside the method savein()
> or did accidentally get that extra indentation?
> 
>>     def savin(self, question):
>>         Label(question, text="Directory:").grid(row=0)
>>         Label(question, text="Filename:").grid(row=1)
>>         self.e1 = Entry(question)
>>         self.e2 = Entry(question)
>>         self.e1.grid(row=0, column=1)
>>         self.e2.grid(row=1, column=1)
>>         def apply(self):
>>             self.dir = self.e1.get()
>>             self.fn = self.e2.get()
>>             self.sav = self.dir + self.fn
>>             self.savfile = open(self.sav, "w")
>>             for line in self.t1.readlines():
>>                 self.savefile.write(line)
>>             self.savefile.close()
> 
> 
> 
Sorry. Accidental indentation.

SA



From emile at fenx.com  Fri Jun  7 08:05:06 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 07 Jun 2002 12:05:06 GMT
Subject: Python as a Game Dev tool on GBA
References: <3cffe769$1_1@nnrp1.news.uk.psi.net>
Message-ID: <Sj1M8.129475$cQ3.3800@sccrnsc01>

"Gianluca Cancelmi" <cancelmi at cli.di.unipi.it> wrote in message
news:3cffe769$1_1 at nnrp1.news.uk.psi.net...
> Hi,
>
> I've been using python for a lot of little tools for Gameboy Advance.
>
> At the moment I'm using it to convert multiple bitmaps into a large
shared
> tile set and multiple maps.
>
> I'm using PIL to load the bitmaps and generate the tiles.
>
> At the moment I'm having a little bit of trouble with memory usage.
>
> I'm cutting the 8x8 pixels tiles using Image.crop().
> I'm creating 3 other versions of the current tile for Horizontal,
Vertical
> and H+V flipping.
> I convert the tiles in strings and use this string as a key for the
tile
> dictionary. The value for that key is an object (TileImage).
>
> That's the source code:
[snip]
> The problem is that is taking too much memory and I get something like
> memory error in python.
> The test images are 9600x512 pixels each (4 images).
>
> My PC has 512M of RAM and I'm running the script in win2000.
>
> Is there any way to debug memory usage?
>
> Anyone has an idea of what I'm doing wrong?
>
> Any help would be really appreciated.
>
> Cheers,
>
> Gianluca
>
>

It looks like 4 images of 9600 by 512 pixels of high color reduced to
16x16 or 2 bytes per pixel?  Then cut it into 8 pixel by 8 pixel pieces,
and save the pieces in objects in a dict keyed by a string of the image.

So, check my math:

9600*512  * 4bytes/pixel (high color)
~= 50 Mb per image * 4 images
~= 200Mb in the source images?

Then save two copies of in altered string forms and you're over 500Mb.
And that's before adding in overhead of the python structures.

You might try restructuring the need for multiple string versions, and
look into changing:

    for img in img_list:

to
    for img in img_list_names:
        img_list = <<get image loaded with PIL>>

Perhaps you can sneak through by saving that 150Mb...

--

Emile van Sebille
emile at fenx.com

---------



From just at xs4all.nl  Wed Jun  5 15:51:18 2002
From: just at xs4all.nl (Just)
Date: Wed, 05 Jun 2002 21:51:18 +0200
Subject: Urllib thread safe?
References: <mailman.1023280808.19305.python-list@python.org> <m3it4xobkb.fsf@mira.informatik.hu-berlin.de> <mailman.1023304328.9010.python-list@python.org>
Message-ID: <just-9B7B3B.21511805062002@news1.xs4all.nl>

In article <mailman.1023304328.9010.python-list at python.org>,
 Oleg Broytmann <phd at phd.pp.ru> wrote:

> On Wed, Jun 05, 2002 at 07:55:00PM +0200, Martin v. Loewis wrote:
> > Yes, that is thread-safe. urllib does not maintain any per-url state,
> 
>    Hah? It maintains global _urlopener...

But that's not per-url state, is it?

Just


From ethanw at rediffmail.com  Sat Jun 15 15:56:44 2002
From: ethanw at rediffmail.com (Ethan)
Date: 15 Jun 2002 12:56:44 -0700
Subject: PythonWin Problem!
References: <1a52c7f1.0206131455.6df50cf8@posting.google.com> <3D0973C5.60407@skippinet.com.au> <1a52c7f1.0206141158.1c03b3f5@posting.google.com> <3D0A7F8E.3090606@skippinet.com.au>
Message-ID: <1a52c7f1.0206151156.464c028e@posting.google.com>

Mark!
 I am sorry that the way I type 'me' was confusing. 'me' referred to
'myself' and NOT 'Windows ME'. I capitalized it to say that the right
click problem is specific to myself because I bungled sometime ago
with regedit and lost the right click menu and so was unable to try
your simpler second solution (I wrote this in my first post and that I
use Win98).


Mark Hammond <mhammond at skippinet.com.au> wrote in message news:<3D0A7F8E.3090606 at skippinet.com.au>...
> Ethan wrote:
> > Thanks Mark. It worked like a charm. One thing though, the key I had
> > to delete was in HKCU rather than in HKLM. The second method would not
> > work with ME since like I said in my earlier post, I don't have a
> > right click menu for items in the taskbar.
> > 
> Yes, I meant HKCU - sorry.
> 
> I haven't used ME, but I know that Win9x, 2k and XP all have a 
> right-click menu on the taskbar.
> 
> Mark.


From whisper at oz.net  Mon Jun 24 22:30:23 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 24 Jun 2002 19:30:23 -0700
Subject: Suggestions for good programming practices?
In-Reply-To: <3D17D51E.7A47162A@engcorp.com>
Message-ID: <GCEDKONBLEFPPADDJCOEMEOODLAA.whisper@oz.net>

Grrrr!

Look at it this way:

	a,b,c = func()

	abc = func()
	a = abc[0]
	b = abc[1]
	c = abc[2]

Which would you rather write? In C or C++ you have to use an analog of the
second way since the first way isn't an option. I _like_ this behavior of
Python.

It's a piddling point as to how Python returns the info - they come to rest
in discrete variables, not as elements of one aggregate variable. The
original point was about the flexibility of Python not some technical
hair-splitting about what Python might put on the stack.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: Peter Hansen [mailto:peter at engcorp.com]
> Sent: Monday, June 24, 2002 19:28
> To: David LeBlanc
> Subject: Re: Suggestions for good programming practices?
>
>
> David LeBlanc wrote:
> >
> > They may be returned in a tuple, but they're PUT in discrete
> variables and
> > not kept in a tuple that I have to programmatically unpack.
> >
> > Language lawyer should be shot, _then_ heard.
>
> I don't understand the above at all.  Sorry. :)  Are you
> unhappy with this behaviour of Python?  And I don't get
> the language lawyer part.  (Pretend I'm really thick...it
> won't be far from the truth.)
>
> -Peter
>
> >
> > David LeBlanc
> > Seattle, WA USA
> >
> > > -----Original Message-----
> > > From: python-list-admin at python.org
> > > [mailto:python-list-admin at python.org]On Behalf Of Peter Hansen
> > > Sent: Monday, June 24, 2002 18:03
> > > To: python-list at python.org
> > > Subject: Re: Suggestions for good programming practices?
> > >
> > >
> > > David LeBlanc wrote:
> > > >
> > > > The one that always gets me is "lat,lon,timezone =
> > > getLocation("seattle")".
> > > > I.E. the ability to return multiple distinct values as opposed
> > > to returning
> > > > multiple values in the form of a struct as in C or C++.
> > >
> > > You _are_ returning them "in a struct".  It's spelled "tuple"
> > > but other than that what's the difference?
> > >
> > > -Peter
> > > --
> > > http://mail.python.org/mailman/listinfo/python-list





From whisper at oz.net  Wed Jun 19 21:42:26 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 19 Jun 2002 18:42:26 -0700
Subject: procmail replacement in Python
In-Reply-To: <20020618235901.GA2197@lilith.my-fqdn.de>
Message-ID: <GCEDKONBLEFPPADDJCOEAEFHDKAA.whisper@oz.net>

Did you have a look at Animail on SF? Various distribution schemes including
forwarding to another mta. Has both allow and deny lists. Uses regex.

A least one of the bundled 3rd party modules (timeoutsocket) is way out of
date, but the software had a recent release-candidate distro.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Gerhard H?ring
> Sent: Tuesday, June 18, 2002 16:59
> To: python-list at python.org
> Subject: procmail replacement in Python
>
>
> Does such a thing exist? I've recently looked into the procmail sources
> and run away screaming. I'd very much prefer a Python replacement to
> such badly written C code.
>
> Does anything exist that I can start from? I've even contemplated not
> inventing yet another filter language, but just using plain Python
> functions.
>
> If I do this, I'd like to replace fetchmail with getmail as well.
>
> What would you suggest as approach?
>
> - writing it as a getmail plug-in
> - writing it as a standalone program to deliver to mailboxes a la
>   procmail
> - let getmail deliver to a temporary folder, then using my procmail
>   replacement to filter into the right folders, delete spam, etc.
>
> The third one looks to be the least likely to lose mail, but the
> temporary folder approach looks like a kludge.
>
> Gerhard
>
> PS: I've seen something _very_ limited written in Python on Sourceforge
> that only does Maildir, but the name is escaping me.
> --
> This sig powered by Python!
> Au?entemperatur in M?nchen: 23.3 ?C      Wind: 0.9 m/s
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From arosado at softhome.net  Mon Jun  3 20:33:41 2002
From: arosado at softhome.net (Andres Rosado)
Date: Mon, 03 Jun 2002 20:33:41 -0400
Subject: HTML links in a GUI tkinter
Message-ID: <5.1.0.14.0.20020603203124.00be6da8@mail.softhome.net>

At 03:41 AM 6/2/2002 +0000, you wrote:
 >If you can also help me to realize this, it would be really fine:
 >
 >
 >How can I put in a frame or in any widget (?) an html link so that the
 >browser opens at the url indicated when the user clicks ?
 >Just simply as if it were an HTML link in an HTML document but in a
 >tkinter GUI ?

In general (it applies to Win32), you create a label capable of obtaining
focus. Change the color to blue and when the label is clicked, open the web
browser.


 >What are the packetages to import ?

Any package that supports clickable labels. I'm currently using Tkinter.

 >In what king of widget can I put such a link ?

Label.

 >What kind of lines of code should i put in this widget the link to be
 >active ?

If it's in Windows, there is a function that you can send the any file to
it. On other systems, you need to know which browser the user is using,
open it and send it to the browser.


-----------------------------------
Andres Rosado
Email: andresr at despammed.com
ICQ: 66750646
Homepage: http://andres980.tripod.com/

If you can survive death, you can probably survive anything.





From jknapka at earthlink.net  Sat Jun 29 03:06:25 2002
From: jknapka at earthlink.net (Joseph A Knapka)
Date: Sat, 29 Jun 2002 07:06:25 GMT
Subject: I'm an idiot
References: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16> <afj27n023m7@enews2.newsguy.com>
Message-ID: <3D1D5C5E.3A5EB81D@earthlink.net>

"Beej J?rgensen" wrote:
> 
> In article <Xns923BB72B1BAF7davidgriswold1netsca at 204.127.202.16>,
> David  <david_griswold1 at yahoo.com> wrote:
> >Any help would be appreciated.
> 
> Ok, I'm no pro and don't have the manuals in front of me, so I hope I
> don't lead you too astray here:
> 
> >f=open('c:\\temp\\temp.txt', 'r')
> >g=open('c:\\temp\\temp1.txt', 'w')
> >while 1:
> >    try:
> >        s=f.readline
>              ^^^^^^^^
> 
> This is a method, so I think it should be "f.readline()"
> 
> >        g.write(s.split())
> 
> Er, did you mean strip()?  And maybe you have to use String.strip(s),
> but I can't remember.
> 
> >    except IOError:
> >        break
> 
> Close--but I don't believe readline() throws an exception on EOF;
> instead it just returns "".

Now, personally, I'd like to have any read after EOF throw
an exception. That's a nice way to break out of a loop, and
it removes the confusion that newbies and occasionally
oldbies (cough) sometimes experience: "If it returns '' at
EOF, what the heck does it return for empty lines???".

But I'm sure there are Very Good Reasons not to do this,
and that some wise person will presently educate me :-)

Cheers,

-- Joe


From gerhard.haering at gmx.de  Wed Jun 19 23:45:01 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Thu, 20 Jun 2002 05:45:01 +0200
Subject: Variable Interpolation - status of PEP 215
In-Reply-To: <ouit4e63fr.fsf@andlx-anamika.cisco.com>
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com>
Message-ID: <20020620034501.GA13974@lilith.my-fqdn.de>

* Ganesan R <rganesan at myrealbox.com> [2002-06-20 08:50 +0530]:
> 
> Hi,
> 
> I am a relative newbie to Python. I have a reasonably good perl background
> (don't worry I am not about to start a language flame war :-), and one of
> the things I really miss is variable interpolation in strings. I searched
> the net and found PEP 215 (http://www.python.org/peps/pep-0215.html). This
> seems to be exactly what I need. The proposal seems to be quite old, so I'd
> like to know if it's still in consideration or has it been dropped
> completely?
 
It's currently being discussed on the python-dev mailing list.  I
personally find the "%(substname)s" % {"substname": "somevalue"} scheme
good enough, and don't see the point of the PEP. Maybe one reason is
that I haven't actually read it. <0.5 wink>

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From NineOfSix at gmx.de  Tue Jun 18 07:22:59 2002
From: NineOfSix at gmx.de (Axel Grune)
Date: Tue, 18 Jun 2002 13:22:59 +0200
Subject: python cgi + IIS
Message-ID: <3D0F1813.3010304@gmx.de>

Hi, I'm trying to get my IIS (Win2k Pro) execute my python scripts.
I added the 'App Mapping' info for '.pycgi' files in the properties of 
IISAdmin (Computermanagement) as I read in the internet. The string
for the executable file is 'D:\Python22\python.exe -u %s %s' (the path 
is valid).
But when i try to call it within my browser, it just displays it like a 
normal text file instead of executing it.


the script (its called test.pycgi):
#!/usr/bin/python

print "Content-Type: text/plain\n\n"
print "Hallo Welt"


has anyone an idea?

tia,
Axel



From a at a.com  Thu Jun 27 09:12:23 2002
From: a at a.com (Adam)
Date: Thu, 27 Jun 2002 14:12:23 +0100
Subject: speech synthesis
References: <idCS8.378$yY2.9108@ozemail.com.au>
Message-ID: <3D1B0F37.44D3232@a.com>


Alistair Campbell wrote:

> Basically, I am wanting to know whether there are any python modules that
> allow you to write programs which include voice synthesis.

> This url was recomended in the Dr Dobbs links it is in respect of Microsoft
> Speach API although this is only a windows solution but i think thats what u
> are looking for... at the moment

there seem to be some good code examples and comprehensive discussion... i
havent managed to try them yet though...

>  http://groups.google.com/groups?threadm=U_6Q8.39803$ks6.8500 at nwrddc02.gnilink.net

hope it can help...

Thanks

Adam



From jonathan at onegoodidea.com  Fri Jun 28 06:08:36 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Fri, 28 Jun 2002 11:08:36 +0100
Subject: self destruction :)
References: <f4KS8.1326$qq1.14832@castor.casema.net> <7x6604ctb1.fsf@ruckus.brouhaha.com> <3d1c2d6a$0$231$4d4ebb8e@news.nl.uu.net>
Message-ID: <B941F434.D1C6%jonathan@onegoodidea.com>

On 28/6/2002 10:37, in article 3d1c2d6a$0$231$4d4ebb8e at news.nl.uu.net,
"Guyon Mor?e" <gumuz at looze.net> wrote:

> yep this is actually what i mean, but the problem is is that this action
> should come from the object within the list. this is object 'knows' when
> it's done, the list doesn't
> 
> see my problem?

I going to assume that the list is in an instance of some class and that it
points to things that are instances of some other class. My solution would
look something like:

-----
class Parent:
    def __init__( self ):
        self._children = []
    def add_child( self, child ):
        assert child not in self._children
        self._children.append( child )
        child.take_parent( self )
    def remove_child( self, child ):
        assert child in self._children
        child.lose_parent()
        self._children.remove( child )
    def items( self ):
        return self._children


class Child:
    def __init__( self, x ):
        self._x = x
        self._parent = None
    def __repr__( self ):
        return '<Child %s>' % `self._x`
    def take_parent( self, parent ):
        if self._parent:
            parent = self._parent
            self._parent = None
            parent.remove_child( self )
        self._parent = parent
    def lose_parent( self ):
        self.take_parent( None )
-----

Which then works like so:

>>> p = Parent()
>>> c1 = Child(1)
>>> c2 = Child(2)
>>> c1
<Child 1>
>>> c2
<Child 2>
>>> p.add_child( c1 )
>>> p.add_child( c2 )
>>> p.items()
[<Child 1>, <Child 2>]
>>> c1.lose_parent()
>>> p.items()
[<Child 2>]
>>> 

You can write something much more clever than this. I'm a fan of abstracting
out this sort of behaviour into classes that you can mixin to any class. I
usually use a "broadcaster/listener" pattern allowing children to notify
their parents of their imminent demise.

Jonathan



From jeremy at jdyallop.freeserve.co.uk  Tue Jun 11 20:14:46 2002
From: jeremy at jdyallop.freeserve.co.uk (Jeremy Yallop)
Date: 12 Jun 2002 00:14:46 GMT
Subject: html parser etc help
References: <7fe97cc4.0206111445.43b196de@posting.google.com>
Message-ID: <ae63pm$47b95$1@ID-114079.news.dfncis.de>

* Xah Lee
| i'm learning python in hoping to do switch web programing from perl.
| can anyone show me code snipets for the following problem to get
| me started?

I don't speak Perl, but I'll see what I can do.

| * snipet to traverse a directory and list files by typer or suffix.
:
| # perl samples
| # traverse dir and print out files ending in .html
| use File::Find;
| find(\&wanted, '/home/xah');
| sub wanted {if($_ =~ m/\.html&/){print "$File::Find::name\n";}}

   import os
   def wanted(arg, dirname, fnames):
       for file in fnames:
           if file.endswith('.html'):
               print file
   os.path.walk('/home/xah')

| * sample snipet that parse a given html file and print out all <a href> links.

[There's probably a simpler way to do this one]

   import sgmllib
   class link_printer(sgmllib.SGMLParser):
       def unknown_starttag(self, tag, attrs):
           if tag.lower() == 'a' and attrs[0][0].lower() == 'href':
               print "link : ", attrs[0][1]
   lp = link_printer()
   lp.feed(open('file.html').read())

| * a sample snipet that fetch and print out the content of a given url.
:
| # fetch a given webpage
| use LWP::Simple;
| $content = get('http://xahlee.org/');
| print $content;

    import urllib2
    content = urllib2.urlopen('http://xahlee.org/')
    print content.read()

Hope this helps,

          Jeremy.


From dsavitsk at e-coli.net  Sat Jun 15 01:01:38 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sat, 15 Jun 2002 05:01:38 GMT
Subject: Using the locale module on Windows under hte IIS
References: <3D09CA19.5050709@mxm.dk>
Message-ID: <SSzO8.123$Kt4.27760@newssrv26.news.prodigy.com>

"Max M" <maxm at mxm.dk> wrote in message news:3D09CA19.5050709 at mxm.dk...

* * *

> Perhaps I should set it in the global.asa ?
---------------------------------^
i believe this is a good place for asp globals.




From mwh at python.net  Thu Jun 13 12:09:30 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 13 Jun 2002 16:09:30 GMT
Subject: email forwarding program
References: <e21498ef.0206130751.2ef784a7@posting.google.com>
Message-ID: <lklm9jma8x.fsf@pc150.maths.bris.ac.uk>

mbell at cs.pitt.edu (Matthew Bell) writes:

> Hi,
> 
> I have a question regarding the email handling
> libraries in Python.  I'm trying to write a
> program that, based on some information it
> extracts from an email, forwards it to another
> person.  I've about figured out imaplib, but
> don't see how I can use it to forward on emails.

That's because you use smtplib to do that.  IMAP is for getting email.

> One possibility I've considered is to write out
> the email to a temp file, then make a call to
> the operating system to use the UNIX program
> mail

That seems a reasonable approach, too.  But you're looking for
smtplib.

Cheers,
M.

-- 
  But since your post didn't lay out your assumptions, your goals,
  or how you view language characteristics as fitting in with 
  either, you're not a *natural* candidate for embracing Design by 
  Contract <0.6 wink>.    -- Tim Peters, giving Eiffel adoption advice


From jeff at sasmor.com  Tue Jun 18 13:44:56 2002
From: jeff at sasmor.com (Jeff Sasmor)
Date: Tue, 18 Jun 2002 17:44:56 GMT
Subject: If you use Python -OO is it a permanent condition??
References: <rXpP8.32471$ks6.25445@nwrddc02.gnilink.net> <aelamd$fo1$1@panix1.panix.com> <BTsP8.57000$uX3.41630@nwrddc01.gnilink.net> <aelu0a$7lc$02$1@news.t-online.com> <aelvgu$4hr$1@panix1.panix.com>
Message-ID: <skKP8.62170$uX3.17122@nwrddc01.gnilink.net>

Done.

--
#--------------------------------
Jeff Sasmor
jeff at sasmor.com
"Aahz" <aahz at pythoncraft.com> wrote in message
news:aelvgu$4hr$1 at panix1.panix.com...
> In article <aelu0a$7lc$02$1 at news.t-online.com>,
> Georg Mischler  <schorsch at schorsch.com> wrote:
> >Jeff Sasmor wrote:
> >>
> >> 'Fascinating'
> >> I'm sure there's a logical reason... somehow. Anyway, I found
> >> that if I renamed all the .pyo files to .pyc everything works fine.
> >> I guess that's because there's little difference between them. It's
> >> too bad that there's no way to switch optimized mode on and
> >> off programmatically - would be useful for apps that use compile().
> >
> >It would be enough to simply look for the respective "other"
> >bytecode file type as well, if nothing else has been found
> >when trying to import a module. The -O setting only has any
> >real relevance when freshly compiling from *.py, after all.
> >Some small amount of added convenience from the side of the
> >interpreter would be highly appreciated in this point.
>
> File a bug on SourceForge.
> --
> Aahz (aahz at pythoncraft.com)           <*>
http://www.pythoncraft.com/
>
> Project Vote Smart: http://www.vote-smart.org/




From beyerlin at magic-garden.de  Fri Jun  7 05:48:45 2002
From: beyerlin at magic-garden.de (Marc Beyerlin)
Date: Fri, 7 Jun 2002 11:48:45 +0200
Subject: syntax error
Message-ID: <19B481CA418DC04CA76EEF2B90B905F102EB03@srv-mail.magic-garden.local>

hello pythons,
im new to python and have a strange error:
if have installed python 2.4 with treads on my openbsd3.0 box
with also installed zope.
zope works fine, but when i try to start an script like

python ipcheck.py

i get an error msg like syntax error at .....

but i m sure there is no error.

what could be the problem?

greetings,
marc



-------------------------------------------------------
magic garden GmbH - Agentur f?r Informationsarchitektur

Hermannstr. 15 - 70178 Stuttgart (Am Feuersee)

www.magic-garden.de ? beyerlin at magic-garden.de

Tel. (07 11) 619 57-42 ? Fax (07 11) 615 01 38




From pyth at devel.trillke.net  Mon Jun  3 16:43:01 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 3 Jun 2002 22:43:01 +0200
Subject: different gui-toolkits pros/cons...
In-Reply-To: <adgj1n$11edg$1@hades.rz.uni-sb.de>; from uwe@rocksport.de on Mon, Jun 03, 2002 at 08:20:07PM +0000
References: <adfsoo$dmik4$1@hades.rz.uni-sb.de> <_rLK8.12$fd2.979@news.ecrc.de> <adg4e0$dq6mb$3@hades.rz.uni-sb.de> <3cfbc67b$0$52024$e4fe514c@dreader3.news.xs4all.nl> <adgj1n$11edg$1@hades.rz.uni-sb.de>
Message-ID: <20020603224301.I30380@prim.han.de>

Uwe Schmitt wrote:
> Boudewijn Rempt <boud at valdyas.org> wrote:
> | Uwe Schmitt wrote:
> 
> |> Markus Jais <mjais at web.de> wrote:
> |> | with pyQT there might be problems with the license
> |> | if you want to make your programm freely available
> |> | (but I am not a lawyer, so I might be wrong)
> |> 
> |> That's a very important point: what do I have to pay if I used pyQt
> |> for development ??? I did not find any explicit prices at the
> |> trolltec-page...
> 
> | Trolltech doesn't sell PyQt: you need to go to http://www.thekompany.com
> | for BlackAdder. 
> 
> O.K. But what if I only want pyQt ??? I suppose I have to pay Qt...

last time i used PyQt i was not aware of any such issues as long as you
plan to do GPLed software yourself. For Windows this is somewhat different. 

    holger




From shalehperry at attbi.com  Thu Jun 20 14:16:03 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 20 Jun 2002 11:16:03 -0700 (PDT)
Subject: do...while
In-Reply-To: <20020620181230.GB2402@nortelnetworks.com>
Message-ID: <XFMail.20020620111603.shalehperry@attbi.com>

> 
>     Right, I've used that in the past. It doesn't seem nearly as clean
> though,
> as making the conditions for terminating the loop plain in the loop
> declaration. Considering the Python community's leaning towards readable
> code,
> necessitating a break for code factoring goes against that philosophy. 
> 

not every language has had a do .. while construct.  You code with the tools
given.




From ftobin at neverending.org  Sun Jun 30 17:35:56 2002
From: ftobin at neverending.org (Frank Tobin)
Date: Sun, 30 Jun 2002 17:35:56 -0400
Subject: How to get rid the new line
In-Reply-To: <mailman.1025471199.28914.python-list@python.org>
References: <mailman.1025471199.28914.python-list@python.org>
Message-ID: <Pine.LNX.4.44.0206301733410.8215-100000@palanthas.neverending.org>

Bjorn Pettersen, on 2002-06-30, wrote:

> I submitted a patch for this (that handled mac, win, and unix line
> endings) but it was declined by the BDFL. I suspect one of the reasons
> were that I named it chomp...

sed s/chomp/bite/ < patch | mail python-bugs-list at python.org

-- 
Frank Tobin			http://www.neverending.org/~ftobin/



From robin at jessikat.fsnet.co.uk  Sun Jun 30 04:58:26 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 30 Jun 2002 09:58:26 +0100
Subject: yield equivalent in C/JavaScript?
References: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk>
 <7xy9cxhlvx.fsf@ruckus.brouhaha.com> <w3xdKUAuljH9EwgH@jessikat.fsnet.co.uk>
 <7xit41i7i9.fsf@ruckus.brouhaha.com>
Message-ID: <14ms5VAygsH9Eww2@jessikat.fsnet.co.uk>

In article <7xit41i7i9.fsf at ruckus.brouhaha.com>, Paul Rubin <phr-
n2002b at NOSPAMnightsong.com> writes
>Robin Becker <robin at jessikat.fsnet.co.uk> writes:
>> the generators here are things  associated with compression or
>> decompression methods. The complexity is such that a severe
>> reformulation will be tough. I understand the concept, but I think it's
>> easier to just try and consume the whole input at one go and pass it on.
>
>If you are trying to implement data compression in javascript with
>large enough inputs that memory consumption is a problem, you probably
>have worse problems than absence of "yield".  Javascript is so slow
>your program will probably be near unuseable.  What's the JS
>environment and the application?  If it's client side JS in a web
>browser, you're probably using the wrong approach to the problem.  If
>it's something like Netscape server side JS, use the LiveConnect stuff
>to implement the compression in C.
>
You're exactly right and OK as I now understand from the JavaScript guys
that it's easy to set the following headers+content in the server
response

Content-Type: text/html
Content-Length:   ddddd
Content-Encoding: x-gzip

content.gz

and have most modern browsers accept and decode properly. Phew, maybe I
don't have to implement UNBWT/UNARI in JavaScript :)

However the content.gz data has to be exactly compatible with gzip. My
experiments with Python-2.1/zlib on freeBSD give me a compressed string
that's slightly too long for some reason. As an example gzip --> 14099
bytes while zlib.compress-->15008 bytes.

Anyone know what I need to do to make zlib work properly in this
context? Or should I really be using the gzip module?
-- 
Robin Becker


From rhymes at myself.com  Sat Jun 29 14:55:57 2002
From: rhymes at myself.com (Rhymes)
Date: Sat, 29 Jun 2002 20:55:57 +0200
Subject: private
References: <mailman.1025365846.8212.python-list@python.org>
Message-ID: <pan.2002.06.29.20.55.56.33709.440@myself.com>

On Sat, 29 Jun 2002 17:44:24 +0200, Mark McEahern wrote:


> What problem are you trying to solve?

It's only a thought, nothing less nothing more

-- 
Rhymes
rhymes at myself.com
" ride or die "


From laz at strakt.com  Fri Jun  7 02:45:08 2002
From: laz at strakt.com (Fredrik Juhlin)
Date: 07 Jun 2002 08:45:08 +0200
Subject: Help with Black Adder...
In-Reply-To: <adpc36$179jq$1@ID-139866.news.dfncis.de>
References: <adpc36$179jq$1@ID-139866.news.dfncis.de>
Message-ID: <1023432309.24549.8.camel@Raven>

On Fri, 2002-06-07 at 06:16, rapskat wrote:
[snip]

I can't help you with your first (and I'm sure, more pressing) problem,
but:

> 2) Are there any good printed books out there for development swith BA and
> PyQT?  I have Deitel's "Python - How To Program" which is really excellent
> and comprehensive, however I was looking for something more specific to
> development within BA itself.
The book you're looking for is "GUI Programming with Python Using the Qt
Toolkit" by Boudewijn Rempt (ISBN# 0-9700330-4-4). It talks mainly about
programming with PyQT in general, but also about development using
BlackAdder. I have a copy, and even though I don't use it every day (I
find that the QT docs usually gives me the information that I need and I
don't use BlackAdder) I've found it quite useful.

//Fredrik





From storminDOTnorm at excite.com  Wed Jun 12 09:05:34 2002
From: storminDOTnorm at excite.com (Stormin Norm)
Date: Wed, 12 Jun 2002 09:05:34 -0400
Subject: filecmp.cmpfiles on directories
References: <3D06491B.302@excite.com> <Xns922B52DB6F313gmcmhypernetcom@199.171.54.215>
Message-ID: <3D07471E.4050804@excite.com>

Already found the documented details under the dircmp class definition. 
 I did not associate this class with being a function of filecmp.
Thanks for your help.
-Stormin.


Gordon McMillan wrote:

>Stormin wrote:
>
>>I am trying to walk a directory structure and compare directories for 
>>duplicates.  For some reason I get the filecmp to work on individual 
>>files, but not on directories.
>>
>
>>example:
>>DOES NOT WORK:
>>dups=filecmp.cmpfiles("C:\comp1","c:\comp2",common)
>>print dups,"||",common
>>result:
>>([],[],[])||
>>
>
>For each filename in common, cmpfiles will compare the files
>with that basename in the two directories. What you want
>is:
> x = filecmp.dircmp("C:\\comp1","c:\\comp2")
> print x.same_files
> print x.diff_files
>
>which will figure out "common" and call cmpfiles for you.
>
>-- Gordon
>http://www.mcmillan-inc.com/
>



From stephen.boulet at motorola.com  Mon Jun 10 10:01:14 2002
From: stephen.boulet at motorola.com (Stephen Boulet)
Date: Mon, 10 Jun 2002 09:01:14 -0500
Subject: Matrix Multiplications
References: <mailman.1023551951.1785.python-list@python.org>
Message-ID: <3D04B12A.51C3CFFA@motorola.com>

Did you try the "outer" function?

>From the NumPy docs:

<<
outer(a, b)
The third ufunc method is outer, which takes two arrays as arguments and
returns the ?outer ufunc? of the two
arguments. Thus the outer method of the multiply ufunc, results in the
outer product. The outer method
is only supported for binary methods.
>>

-- Stephen

Gavin Lusby wrote:
> 
> Dear all,
> 
> I am an engineering student trying to further my research with python
> programming.  Would someone please tell me how I would do a certain
> matrix multiplication in Pythin/NumPy...
> 
> ----------------------------------------------------------------------
> 
> I have two large matricies A[n,n] and B[n,n]
> 
> I would like to get an outer product of:
> 
> C[k+1:,k:] = B[k+1:,k] * A[k,k:]
> 
> where,  k = 0,...,n-1
> 
> It is in effect, an outer product of two vectors of different length.
> 
> ----------------------------------------------------------------------
> 
> I keep getting errors and in the end it's probably something
> straightforward but I'm no Computer Scientist having been spoon-fed
> Fortran 77.
> 
> I would be most grateful if someone could tell me how to do this
> very efficiently in Python.
> 
> Many thaks in advance,
> 
> Gavin Lusby
> 
> *********************************************************
> Gavin Lusby
> University of Edinburgh
> email:    g.h.lusby at sms.ed.ac.uk


From spammers.do.not.bother at void.bogus  Tue Jun  4 11:17:44 2002
From: spammers.do.not.bother at void.bogus (Magnus)
Date: Tue, 04 Jun 2002 15:17:44 GMT
Subject: Automated Email - What did I do wrong here?
References: <adg1do$10nmr5$1@ID-49758.news.dfncis.de> <adifmg$11i5ks$1@ID-49758.news.dfncis.de> <adifsu$11ifbm$1@ID-49758.news.dfncis.de>
Message-ID: <sS4L8.10704$p56.3153084@newsb.telia.net>

Ken wrote:

> 
> "Ken" <ken at hotmail.com> wrote in message
> news:adifmg$11i5ks$1 at ID-49758.news.dfncis.de...
>> "Ken" <ken at hotmail.com> wrote in message
>> news:adg1do$10nmr5$1 at ID-49758.news.dfncis.de...
>> > How to do automated emailing when results being processed?
>> >
>> > Thanks
>> >
>>
>> What did I do wrong? It crash the program. Without these code, the cgi
>> script works fine. Thanks
>>
>>     fromaddr = 'ken at hotmail.com'
>>     toaddrs = 'ken at hotmail.com'
>>     msg = "testing\n"
>>     server = smtplib.SMTP('smtp.aol.com')
>>     server.sendmail(fromaddr, toaddrs, msg)
>>     server.quit()
>>
> oh...and I did import smtplib...

Try following:

import smtplib
import string, sys

HOST = "smtp.bogus.world"
FROM = "fake at bogus.world"
TO = "fool at bogus.world"
SUBJECT = "Test"
BODY = "Hello........"

body = string.join(("From: %s" % FROM, "To: %s" % TO, "Subject: %s" % 
SUBJECT, "", BODY), "\r\n")
server = smtplib.SMTP(HOST)
server.sendmail(FROM, [TO], body)
server.quit()

/Magnus



From achim.domma at syynx.de  Fri Jun 28 13:58:55 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Fri, 28 Jun 2002 19:58:55 +0200
Subject: distutils/boost question
References: <afhp5p$fqf0d$1@hades.rz.uni-sb.de> <slrnahou7h.140.gerhard@lilith.my-fqdn.de>
Message-ID: <afi7jf$5hh$00$1@news.t-online.com>

"Gerhard H?ring" <gerhard at bigfoot.de> wrote in message
news:slrnahou7h.140.gerhard at lilith.my-fqdn.de...
> The easiest solution is to statically link in the Boost.Python library.

With boost.python V2 dynamic linking will be a requirement. Isn't there
another solution?

Achim





From m.hadfield at niwa.co.nz  Thu Jun  6 19:07:26 2002
From: m.hadfield at niwa.co.nz (Mark Hadfield)
Date: Fri, 7 Jun 2002 11:07:26 +1200
Subject: another near-sighted, time-wasting syntax enhancement idea
References: <e395b886.0206051956.5f447e9b@posting.google.com>
Message-ID: <adoqbd$fst$1@newsreader.mailgate.org>

"Carl Banks" <idot at vt.edu> wrote in message
news:e395b886.0206051956.5f447e9b at posting.google.com...

> Two colons after the class (or whatever) definition means the block
> continues until EOF and is not indented.  Good for big, long classes
> that take up a whole module.  Been suggested before, I'm sure.

Yes, but the BDFL used his time machine to erase all traces of the
offender...

...No, it's never been suggested before.

--
Mark Hadfield            "Ka puwaha et tai nei, Hoea tatou"
m.hadfield at niwa.co.nz
National Institute for Water and Atmospheric Research (NIWA)




From ben at sourcecod.com  Fri Jun 14 21:18:47 2002
From: ben at sourcecod.com (Ben Blonder)
Date: 14 Jun 2002 18:18:47 -0700
Subject: Using Python to script a game?
Message-ID: <cdc2589b.0206141718.7dc2ae13@posting.google.com>

I'm writing a 3D naval tactics in C++ for the Mac. I'm planning on
having all the object behaviors, movements, and AI controlled through
a scripting layer. The scripts would be text files the game would
parse and then carry out actions. All the scripts would be read at
runtime, not compile time, so they'd all have to be interpreted.

For example, I might have the C++ code

class foo { 
public:
int x;
void print() {
cout << x << endl;   
}
};

In the scripting system, there would have to be some way to create a
new foo object, define x, and then call print() while the user was
playing.

Can I use Python to accomplish this, and if so, can you point me to
some tutorials on integratring Python with my existing code?


From mgilfix at eecs.tufts.edu  Fri Jun 14 10:03:19 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Fri, 14 Jun 2002 10:03:19 -0400
Subject: Python GUI
In-Reply-To: <20020614153640.A1863@bork.demon.nl>; from egbert@bork.demon.nl on Fri, Jun 14, 2002 at 03:36:40PM +0200
References: <9ddd154e.0206122021.1b06f89a@posting.google.com> <20020614153640.A1863@bork.demon.nl>
Message-ID: <20020614100319.C4109@eecs.tufts.edu>

  Yes. Currently no support for threading on the windows port. That
was supposed to be fixed for GTK 2.x but I don't think it's available
yet.

                  -- Mike

On Fri, Jun 14 @ 15:36, Egbert Bouwman wrote:
> On Wed, Jun 12, 2002 at 09:21:09PM -0700, H. Safarzadeh wrote:
> 
> > -pyGTK and pyQt: I do not know enough about them. But I know that I do
> > not like GTK look on Windows [...]
> 
> Does PtGTK have any other drawbacks ?
> egbert
> -- 
> Egbert Bouwman - Keizersgracht 197 II - 1016 DS  Amsterdam - 020 6257991
> ========================================================================
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (egbert)




From david.abrahams at rcn.com  Sun Jun 16 16:57:04 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Sun, 16 Jun 2002 16:57:04 -0400
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de> <aeftmj$m29$1@bob.news.rcn.net> <3d0c6d67@nntp.server.uni-frankfurt.de> <3D0CE698.1050806@SPAMnwinternet.com>
Message-ID: <aeiu3a$4ff$1@bob.news.rcn.net>

"Jim" <jbublitzNO at SPAMnwinternet.com> wrote in message
news:3D0CE698.1050806 at SPAMnwinternet.com...

> On the other hand, I maintain PyKDE (I should point out I'm not the sip
> author however), and for an actively maintained project with nearly a
> dozen libraries and several hundred files, sip seems preferable.

I'd really like to know what you find preferable about sip -- aside your own
familiarity with it, of course, which for me usually counts more than any
other single factor ;-). I'm actively developing v2 of Boost.Python and
would really like to make it the best tool possible for this job, so it
would be useful to know more.

-Dave







From sarmstrong13 at mac.com  Tue Jun  4 20:49:42 2002
From: sarmstrong13 at mac.com (SA)
Date: Tue, 04 Jun 2002 19:49:42 -0500
Subject: A simple tcl/tk program to run Python.
Message-ID: <B922CA56.638E%sarmstrong13@mac.com>

Hi Everyone-

    I found this script at :
http://mini.net/tcl/1805.html

It was written by Hugh Dunne as kind of an Idle for Perl on a windows
machine. I modified it to run Python scripts instead. The difference between
this and Idle is that it does not immediately execute your code. And keeps
the code visible so you can go back and edit the mistakes. I think it's cool
because after about twenty lines of code in Idle I tend to forget whether or
not I "import"ed the correct module and then I start losing my spot. So it
is kind of good as a simple code debugger. I want to try and code the same
thing in Python because I find it more useful than Idle. (anyone willing to
help, please jump in).

Let me know what you think, if there are any adjustments you would like to
make, and if it works properly for you. So far I've run it with simple
functions and calling them in the program seems to work OK.

Thanks.
SA

Here's the code:

#!/usr/bin/wish

text .t1 -height 12 -width 84 -font {Courier 12}
frame .f
text .t2 -height 8 -width 84 -bg lightblue -font {Courier 9}
button .f.b1 -text Execute -command perl
button .f.b2 -text "Clear input" -command ".t1 delete 0.0 end"
button .f.b3 -text "Clear output" -command ".t2 delete 0.0 end"
pack .f.b1 .f.b2 .f.b3 -side left -padx 2
pack .t1 .f .t2 -side top -pady 2

proc perl {} {
        catch {exec python -c [.t1 get 0.0 end]} output
    .t2 delete 0.0 end
    .t2 insert end $output
}



From kragen at pobox.com  Mon Jun  3 13:40:08 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 3 Jun 2002 12:40:08 -0500
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Jun  3)
Message-ID: <69E93E7F6CC2D954.1DDB339C5F117AF9.783808701A71D731@lp.airnews.net>

QOTW:  "Imagine two painters.  One spends half his time at arts supplies
stores, checking out the tiniest differences between different models of
brushes, different brands of paint, canvases prepared with minutely
distinct processes yielding minute roughness and absorption differences,
and so on.  The other painter has settled on good, reliable lines of
brushes, paint and canvases, had put these minutiae basically out of her
mind, and spends her time and energy actually painting, working out
different lighting and composition techniques, and so on."  Alex
Martelli, on the perfectibility of computer languages


* Code:
    Pycaml, which interfaces Python to OCaml, a higher-level
    alternative to C for speed-critical code, released version 0.8 on
    2002-05-28.
        http://pycaml.sf.net/

    Twisted, a framework for networked applications that includes
    implementations of a large number of common protocols as well as a
    remote object protocol with bindings in Python, Java, and elisp,
    released version 0.18.0 on 2002-06-01.
        http://twistedmatrix.com/

    Pygame, a Python binding to the popular SDL games-programming
    library, released version 1.5 on 2002-05-30.
        http://groups.google.com/groups?selm=3CF79DA4.7090503%40attbi.com

    Sheila King posted some "liberal" email parsing code that handles
    malformed mail headers better than the current standard email
    module, which doesn't parse Klez emails (among others) very well.
        http://groups.google.com/groups?selm=acvmf3.t8.1%40kserver.org

    Russell Owen posted a GenericCallback class that curries callback
    functions to ease GUI programming.
        http://groups.google.com/groups?selm=adc8kj$1ovk$1%40nntp6.u.washington.edu

    PySQLite, a Python wrapper for the SQLite embedded RDBMS, released
    version 0.2.0 on 2002-06-02.
        http://groups.google.com/groups?selm=mailman.1023065045.19962.python-list%40python.org

    libxml, the GNOME XML library, released version 2.4.22 on
    2002-05-27.
        http://xmlsoft.org/news.html

    PdfSearch, a search engine for archives of PDF files with
    command-line and GUI interfaces, released version 0.1 on
    2002-06-02.
        http://groups.google.com/groups?selm=mailman.1022767921.8082.clpa-moderators%40python.org

* Discussion on Features of Python:
    Jython doesn't run finalizers promptly (or, sometimes, at all),
    with the result that open('foo.txt', 'w').write(something) may
    never write anything to foo.txt.  Some think relying on CPython's
    prompt finalization is a bug, but Francois Pinard explained why he
    thinks it's Jython's behavior that's the bug.
        http://groups.google.com/groups?selm=mailman.1022711561.21409.python-list%40python.org

    There was some discussion of 'for any' and 'for every' boolean
    expressions similar to list comprehensions; Tim Peters pointed out
    that ABC had them and they were nice.
        http://groups.google.com/groups?selm=mailman.1022602963.3818.python-list%40python.org

    Grant Griffin explained why the implicit string concatenation done
    by the Python parser is un-Pythonic.
        http://groups.google.com/groups?selm=ad8dp701452%40drn.newsguy.com

    Aahz advocated using the Queue module for any and all inter-thread
    communication; Mark Hammond thought that position too extreme.
        http://groups.google.com/groups?selm=3CF43BA6.5020106%40skippinet.com.au

    Berthold H?llmann complained that distutils breaks
    platform-independence of scripts when Python executables are
    installed in platform-specific directories.
        http://groups.google.com/groups?selm=m2k7pmlj3l.fsf%40pchoel.privat.uni-hamburg.de

* Problems and Solutions:
    Tim Churches recommended RPy for easy visualization of large
    arrays of numbers, such as Numeric arrays.
        http://groups.google.com/groups?selm=mailman.1022712528.1085.python-list%40python.org

    Andrei Kulakov praised the startup speed of pygtk applications in
    a thread about wxPython startup speed; several other posters on
    the thread had switched from wxPython to pygtk or even Tkinter to
    get better performance.
        http://groups.google.com/groups?selm=slrnafj3eo.86f.ak%40ak.silmarill.org

    Carsten Gaebler noticed that popen would cause your Python program
    to hang forever on exit if you forgot to specify 'w'.  Donn Cave
    explained why.
        http://groups.google.com/groups?selm=ad0bkd$2608$1%40nntp6.u.washington.edu
    There's also a problem with Python's SIGPIPE handling and
    popen()ed programs that don't exit on their own --- they don't
    die.
        http://groups.google.com/groups?selm=ad2vun$1eua$1%40nntp6.u.washington.edu

    Someone had trouble getting McMillan Installer to work on a
    program that used Gadfly; Gordon McMillan thought it was because
    of a particular unusual technique Gadfly used to load its grammar,
    but Richard Jones said it no longer used that technique.
        http://groups.google.com/groups?selm=mailman.1022623853.18650.python-list%40python.org

    Steindl Fritz wanted a variadic Cartesian product function; his
    problem got many solutions, the tersest of which was Steven
    Majewski's.
        http://groups.google.com/groups?selm=mailman.1022695262.8171.python-list%40python.org

* Miscellaneous:
    "gb", from UNC-CH, posted their Python-code-folding setup for
    Emacs.
        http://groups.google.com/groups?selm=uelftqprw.fsf%40cs.unc.edu

    Syver Enstad has a hack to allow HTML links in pydoc output to
    open your preferred editor with the right source file, but only on
    Microsoft Windows.
        http://groups.google.com/groups?selm=un0ugnxa3.fsf%40online.no

    Emile van Sebille and Tim Delaney posted references to good online
    search pages for the Python documentation.
        http://groups.google.com/groups?selm=mailman.1022801931.24850.python-list%40python.org

    EuroPython early-bird registration has been extended until
    2002-06-09 because of a rush of last-minute early-bird
    registrations and Internet problems.
        http://groups.google.com/groups?selm=mailman.1022925562.29795.python-list%40python.org

    The EuroPython web site published an interview with Paul
    F. Dubois, the primary maintainer of NumPy.
        http://europython.zope.nl/interviews/entries/paul_dubois

    David Mertz has reviewed more Python books.
        http://www-106.ibm.com/developerworks/linux/library/l-pbook3.html

    Michal Wallace posted a reference to some code-duplication-finding
    software he'd written; Tim Peters posted a link to some research
    on the subject and mentioned he wished there was a fast
    suffix-tree extension module for Python.
        http://groups.google.com/groups?selm=mailman.1022725195.17768.python-list%40python.org
    It turns out that Danny Yoo has written a suffix-tree extension
    module for Python.
        http://www-hkn.eecs.berkeley.edu/~dyoo/python/suffix_trees/

    Quinn Dunkan extolled the merits of customized development
    environments when Oren Tirosh claimed not to have a personal
    utility library.
        http://groups.google.com/groups?selm=slrnafafkp.eom.quinn%40regurgitate.ugcs.caltech.edu

    PEP 272, the API for block encryption algorithms, has been
    finalized.
        http://groups.google.com/groups?selm=mailman.1022760847.17663.clpa-moderators%40python.org

    There is now a page on python.org about Python-friendly hosting
    providers.
        http://www.python.org/hosting/


========================================================================

Everything you want is probably one or two clicks away in these pages:

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
	 http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Michael Hudson continued Andrew Kuchling's marvelous tradition
    of summarizing action on the python-dev mailing list once every
    other week, into July 2001.  Any volunteers to re-start this
    valuable series?
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

    Much of Python's real work takes place on Special-Interest Group
    mailing lists
        http://www.python.org/sigs/

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
	http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
	http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
	http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

    Links2Go is a new semi-automated link collection; it's impressive
    what AI can generate
        http://www.links2go.com/search?search=python

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning
(approximately), ask <claird at neosoft.com> to subscribe.  Mention
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

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


From whisper at oz.net  Fri Jun 21 18:29:08 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 21 Jun 2002 15:29:08 -0700
Subject: I'd give up Perl tomorrow if only...
In-Reply-To: <3D12F6FC.3000608@wedgetail.com>
Message-ID: <GCEDKONBLEFPPADDJCOECEHDDLAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Derek Thomson
> Sent: Friday, June 21, 2002 2:51
> To: python-list at python.org
> Subject: Re: I'd give up Perl tomorrow if only...
>
>
> David Mitchell wrote:
> > .. there was something equivalent in scope and breadth to CPAN.
>
> I (almost) agree, and sympathise :)
>
> >
> > Is ANYONE working on this, or thinking about working on it, or knows
> > someone who's thinking about working on it, or dreaming about
> it, or ...?
>
> Among other things, it's one of the reasons I'm tinkering with a Perl ORB.
>
> http://sourceforge.net/projects/perlorb
>
> If you use a Python ORB, such a Fnorb,
> [http://sourceforge.net/projects/fnorb], on the Python side, you should
> be able to write wrappers for your favourite Perl module pretty easily
> and use them from Python.
>
> Yes, it's stop-gap, but when you already have a non-trivial Perl module,
> either on CPAN or in-house, it can save a lot of rewriting. What we
> *really* need to do is dynamically create CORBA interfaces for Perl
> modules, but that might be impossible given that Perl is like Python in
> that module and class interfaces aren't pre-declared :(

Even better: a Perl to Python converter :>

> Anyway, feel free to email me if you want to try this approach and you
> need guidance.
>
> As for Parrot, maybe someone will port Python to it one day, and then we
> would have something like Jython, but for Perl!

I've heard that there is a trial implementation of Python on Parrot already.
Don't know where it is though.

> Regards,
> Derek.

It's too bad that ActiveState is keeping PPN more or less proprietary. There
has been some other work done towards a standard for describing archived
packages though.

Dave LeBlanc
Seattle, WA USA





From nico at tekNico.net  Fri Jun  7 04:15:31 2002
From: nico at tekNico.net (Nicola Larosa)
Date: Fri, 07 Jun 2002 10:15:31 +0200
Subject: Creating Dynamic Web Pages (ZOPE)
References: <46ca81a0.0206062016.789d3af6@posting.google.com>
Message-ID: <3D006BA3.1030507@tekNico.net>

> I am not exactly sure where to post this question -- the only
> reference that I could find for ZOPE in the Newsgroups is here.  I
> appologize if this is not the correct place.

There are many mailing lists on the Zope site, ask on the main one for 
further advice, and also search the ZopeDB archives for details on the best 
way of interfacing with MS SqlServer, it was a recent discussion.


> I have the situation where I need to create a simple Intranet for my
> company to display dynamic / data driven web-pages.  We are running a
> Windows NT network with MS-SQL Server and IIS.

Zope runs well on Windows, on its own or behind IIS.


> I had originally thought of using ASP (Server Side Scripting) to
> generate the dynamic web-pages.  While I know how to create the
> Intranet using ASP, I have yet to find any RAD type Web-Tools that
> will allow me to quickly and easily generate the ASP pages required,
> and I don't fancy using the Notepad approach either.

You will have to do some editing, but while not exactly RAD, there are tools 
that will help you, and you will end with a much more maintainable solution.


> My scenario is typical where I need to put together a company Intranet
> with a few front end Web-Pages where users can select information,
> various links, and enter search criteria.  The information is then
> retrieved from the MS-SQL Server and the results are sent back to the
> user.  It's nothing fancy, no shopping carts, or remembering sessions
> states, etc.. just really a simple front-end interface for
> non-technical people to easily retrieve information.

Use the ZDataQueryKit Product:

http://www.zope.org/Members/haqa/ZDataQueryKit

It contains a number of wizards that will help with SQL queries creation and 
report generation.


> I don't know if ZOPE would fit my situation or not.

I think so, and nicely too.


 > I know that it is / can be used to create complex E-commerce sites,
 > but I am wondering if it is overkill for what I need done.

No, it's not. It contains lots of tools, but you will only use what you 
need. The important thing is that the infrastructure is solid, and the 
synchronization with external RDBMS transactions is outstanding. The 
extensive authentication and authorization model has already been mentioned.


> Would I be better off trying to find some sort of RAD Web Tool
 > instead (anyone have any good suggestions) ?

Nothing more RAD than Zope that I know of, in the open source world.

Only, just give a thorough look at the ZopeBook, you will get a much better 
start:

http://www.zope.org/Documentation/ZopeBook/

And lastly, just jump over DTML and go straight to Page Templates. :^)


-- 
"Too much cleverness in the parser can turn against you."
   Guido Van Rossum

Nicola Larosa - nico at tekNico.net





From kragen at pobox.com  Mon Jun 10 18:06:47 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 18:06:47 -0400
Subject: Guru advice needed for mxTextTools
References: <3CFBCC68.D35A03D8@vip.fi>
Message-ID: <833cvupz48.fsf@panacea.canonical.org>

Pekka Niiranen <krissepu at vip.fi> writes:
> I am trying to optimize a function that searches nested strings from
> set of (allmost) flat files (about 2 MB each) . If I use regular
> expressions, I must
> fix the amount of nesting:

It sounds like you want to build a parser; I haven't done this in
Python, but if I did, my first thought would be to look at SPARK to
see if it's any good.



From phd at phd.pp.ru  Thu Jun 13 09:33:13 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 13 Jun 2002 17:33:13 +0400
Subject: filter2
Message-ID: <20020613173313.D19424@phd.pp.ru>

Hello!

   I want to have new python builtin - filter2. It is exactly like filter,
but it returns also a list of items that didn't pass the test.

   Of course I can do it with just "for" loop:

l1 = []
l2 = []
for item in alist:
   if test(item):
      l1.append(item)
   else:
      l2.append(item)

but this way I am missing filter's loop written in C.

   Running filter two times (with test and inverted test) is not an option,
as test could be a costly function.

   Opinions?

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From sam at webslingerZ.com  Thu Jun 20 14:18:44 2002
From: sam at webslingerZ.com (Sam Brauer)
Date: Thu, 20 Jun 2002 14:18:44 -0400 (EDT)
Subject: cannot unmarshal code objects in restricted execution mode
In-Reply-To: <Pine.LNX.4.44.0206181541180.1527-100000@localhost.localdomain>
Message-ID: <Pine.LNX.4.44.0206201409490.1527-100000@localhost.localdomain>

At the risk of annoying folks, I thought I'd re-post a question I asked a 
couple of days ago that no one responded to.  
(Yes, I'm really that desparate.)

I promise I won't post about this problem again on this list.

Any help would be much appreciated....

Thanks!!!! (and sorry for double-posting),
Sam

On Tue, 18 Jun 2002, Sam Brauer wrote:

> In the context of a long-running Python process (specifically mod_python) 
> I occasionally get the following error:
> 
> exceptions.RuntimeError: cannot unmarshal code objects in restricted 
> execution mode
> File "/usr/local/lib/python2.2/encodings/__init__.py", line 51, in 
> search_function
> mod = __import__(modname,globals(),locals(),'*')
> 
> This happens when I call code like this:
> 
> utf8string = unicode(val, enc).encode('utf-8')
> 
> where 'val' is a string in an encoding identified by the string 'enc'.
> 
> What's strange is that the vast majority of the time the error doesn't 
> occur, but after it happens once, it seems to happen more frequently 
> (although not consistently).
> 
> Does anyone know what might be the cause of this error?
> 
> 

-- 
Sam Brauer





From logiplexsoftware at earthlink.net  Sun Jun 16 13:20:54 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: 16 Jun 2002 10:20:54 -0700
Subject: Snort alert tail...
In-Reply-To: <AB0P8.107$W91.886@news1.global-one.fi>
References: <AB0P8.107$W91.886@news1.global-one.fi>
Message-ID: <1024248055.1899.3.camel@localhost.localdomain>

On Sun, 2002-06-16 at 05:51, Jan-Eric wrote:
> HEllo !
> I'm trying to write a dynamic firewall script in python that scans the Snort 
> alert file like 'tail -f' and takes action based on the infomation it gets 
> from that file. But I can't get the 'tail' function to work.It reads the 
> file, but any new information that Snort is writing to the file doesn't 
> show up to the script.
> 
> ex.
> file = open('/var/log/snort', 'r')
> while 1:
>         file = file.read()
>         print file
> ....


import os, stat
import time

def snort(pathname):
    size = os.stat(pathname)[stat.ST_SIZE]
    
    while 1:
        lastsize = size
        size = os.stat(pathname)[stat.ST_SIZE]
        if size > lastsize: # there's new data
            f = open(pathname, 'r')
            f.seek(lastsize)
            print f.readlines()
            f.close()
        else:
            time.sleep(2)

snort("/var/log/snort")







From bokr at oz.net  Sat Jun 15 11:48:21 2002
From: bokr at oz.net (Bengt Richter)
Date: 15 Jun 2002 15:48:21 GMT
Subject: filter2
References: <mailman.1023975618.6474.python-list@python.org> <aeaava$56l8p$1@ID-59885.news.dfncis.de> <mailman.1023979895.13512.python-list@python.org> <slrnagllbc.5mj.quinn@groat.ugcs.caltech.edu> <aefm8m$3ng$0@216.39.172.122>
Message-ID: <aefnk5$5ol$0@216.39.172.122>

On 15 Jun 2002 15:25:10 GMT, bokr at oz.net (Bengt Richter) wrote:

>On 15 Jun 2002 05:49:33 GMT, quinn at groat.ugcs.caltech.edu (Quinn Dunkan) wrote:
>
>>On Thu, 13 Jun 2002 18:50:23 +0400, Oleg Broytmann <phd at phd.pp.ru> wrote:
>>>On Thu, Jun 13, 2002 at 04:39:51PM +0200, Thomas Heller wrote:
>>>> 
>>>> "Oleg Broytmann" <phd at phd.pp.ru> wrote in message
>>news:mailman.1023975618.6474.python-list at python.org...
>>>> >    I want to have new python builtin - filter2. It is exactly like filter,
>>>> > but it returns also a list of items that didn't pass the test.
>>>> 
>>>> IIRC, in Smalltalk they were named 'select' (like filter), and 'reject' (like your filter2).
>>>> IMO these are much better names then filter and filter2.
>>>
>>>   It is too late to rename builtin filter, so I modelled fiter2 after it.
>>
>>In haskell they call it "partition" which I think is better since the
>>emphasis is that it returns both selected and rejeted.
>>
>>I wrote mine in python with a plain for loop.  I'm not sure C would really get
>>you that much efficiency, but it would be easy enough to find out.
>>
>>Not sure it's worth being in the core (maybe there should be a module for
>>list utils or something), but I like partition better than filter2.
>
>I wonder how this version whould do:
> >>> def partition(test, alist):
> ...     l2=[]
> ...     return filter(lambda x: test(x) or l2.append(x), alist), l2
> ...
> >>> partition(lambda x: x%3==0, range(16))
> ([0, 3, 6, 9, 12, 15], [1, 2, 4, 5, 7, 8, 10, 11, 13, 14])
>
Sorry, that should have been

 >>> def partition(test, alist):
 ...     l2=[]; app=l2.append
 ...     return filter(lambda x: test(x) or app(x), alist), l2
 ...
 >>> partition(lambda x: x%3==0, range(16))
 ([0, 3, 6, 9, 12, 15], [1, 2, 4, 5, 7, 8, 10, 11, 13, 14])

>
>I imagine it's faster than my other one (clipped from other post)
>
> >>> def filter3(test, alist): # not to use keyword
> ...     l2=[]
> ...     return [x for x in alist if test(x) or l2.append(x)], l2
> ...
> >>> print filter3(lambda x: x%3==0,range(16))
> ([0, 3, 6, 9, 12, 15], [1, 2, 4, 5, 7, 8, 10, 11, 13, 14])

Regards,
Bengt Richter


From jhndnn at yahoo.com  Fri Jun 14 16:12:41 2002
From: jhndnn at yahoo.com (John Dunn)
Date: 14 Jun 2002 13:12:41 -0700
Subject: Associating a C++ class with a Python class
Message-ID: <2cc96dc7.0206141212.5b6ca5df@posting.google.com>

I would like to add a new class type to my embedded python engine and
I would like this class to wrap an exisiting C++ class instance. All
of the examples I have found are more 'c implementation' than 'c
wrappers', meaning that the python instance is not associated with an
different c++ class. The methods of the python class will just be
straight calls through to the C++ class.

These classes will only be created from C, so I am basically looking
for a way to stuff a pointer to my C++ class into the python object. I
guess I could have a map of PyObject->C++ class, but I was hoping
there was a cleaner way to accompish this task. Is there a standard
way to do this? Am I missing something obvious?

Thanks-

John
--
John Dunn
Peak Audio, a division of Cirrus Logic, Inc.
http://www.peakaudio.com


From amuys at shortech.com.au  Thu Jun 27 20:28:16 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 27 Jun 2002 17:28:16 -0700
Subject: Help with an algorithm wanted
References: <rowen-CE3C0F.13020226062002@nntp1.u.washington.edu>
Message-ID: <7934d084.0206271628.166c6532@posting.google.com>

"Russell E. Owen" <rowen at cesmail.net> wrote in message news:<rowen-CE3C0F.13020226062002 at nntp1.u.washington.edu>...
> I'd like some help solving a problem in Python.
> 
> The basic idea is I have an item of data that can be represented in any 
> of four ways: A,B,C and D. I also have conversion functions between 
> certain pairs of representations, such that there's always a path 
> between any two data representations. In this case suppose I know A<->C, 
> B<->C and C<->D, or graphically:
> 
> A <--+
>      +--> C <---> D
> B <--+
> 
> I'd like to create objects that have one of these items (A,B,C,D) as a 
> given, then be able to query any such object for any data representation.
> 
> For instance, an object with A as the given would return D by solving 
> A->C->D.
> 
> What I'm trying to figure out is a simple and reasonably elegant way to 
> code such objects.
> 

So effectively you have a conversion graph, where each vertex
represents a known representation, and each edge represents a
conversion (presumably with a conversion function as an annotation to
the edge).

What you are asking for appears to be an algorthm to find the shortest
path between to representations in the graph.  Fortunately for you,
Dijkstra provided a solution to this problem a number of years ago,
and implementing it is a standard algorthms assignment (I do hope
you're not asking for help with a homework assignment?).

You'll find an implementation in the Python Cookbook @
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/119466

The recipie was submitted by David Eppstein.  It dosn't look
optimised, but that's ok because it is legible (much more important)
and the algorithm is efficient.

Of signifigant interest to you is that it represents the graph as a
dictionary of edges, and as you will want to annotate the edges with
conversion functions, this is a good thing.

Also note that Dijkstra's algorthm uses weighted edges, so you get to
consider differing conversion costs for free.

Andrae Muys


From duncan at NOSPAMrcp.co.uk  Thu Jun 13 11:19:11 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Thu, 13 Jun 2002 15:19:11 +0000 (UTC)
Subject: Is the weakref.ref callback is called deterministically?
References: <ba1e306f.0206130451.795085d6@posting.google.com>
Message-ID: <Xns922CA4B7FBD3Fduncanrcpcouk@127.0.0.1>

janeaustine50 at hotmail.com (Jane Austine) wrote in 
news:ba1e306f.0206130451.795085d6 at posting.google.com:

> Even though the index i does not change, it loops till the
> result is not None. Seemingly it depends on the weakref.ref
> callback to remove dead weakrefs.
> 
> However, the while loop wouldn't be needed if the callback
> was called right after the last reference to the object
> was "del"ed.
> 
Have you considered what would happen in a multithreaded program if the 
weak reference became invalid after it was retrieved from the dictionary or 
list, but before it was converted to a strong reference? There may be other 
reasons, but you need loops here to avoid the race condition.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From bedge at troikanetworks.com  Mon Jun 17 16:27:44 2002
From: bedge at troikanetworks.com (Bruce Edge)
Date: Mon, 17 Jun 2002 13:27:44 -0700
Subject: simple socket redirector app question
Message-ID: <5DrP8.25$OT1.34057@newsfeed.slurp.net>

I have an app which talks via sockets to local host clients only.
I'd like to get at it remotely.

Basically it would accept remote connections from a remote client, 
relay them the the local app, and forward the replies back to the 
remote client.

What's the simplest way to write this type of a socket redirector?

Use SocketServer, or is that overkill?

Thanks, Bruce.


From bokr at oz.net  Wed Jun 26 19:06:22 2002
From: bokr at oz.net (Bengt Richter)
Date: 26 Jun 2002 23:06:22 GMT
Subject: question od default args
References: <o3vjhuguen6gfm8vnq7rbk5volmi80shl5@4ax.com> <YynS8.13571$p56.4313990@newsb.telia.net>
Message-ID: <afdhde$bg$0@216.39.172.122>

On Wed, 26 Jun 2002 18:17:28 GMT, "Fredrik Lundh" <fredrik at pythonware.com> wrote:

>Gon?alo Rodrigues wrote:
>
>> When I want default args I usually do
>>
>> def (arg = None):
>>     ...
>>
>> But what if I want to make None a reasonable argument too? That is, I
>> want to know if the user has in fact passed an argument, and if not to
>> do something special, with None (or whatever object) being a reasonable
>> arg to be passed.
>
>hmm.  wasn't this just discussed in some thread on a
>newsgroup near you?
>
>three alternatives:
>
>    __UNDEF__ = [] # guaranteed to have a unique id
>
>    def myfunc(arg=__UNDEF__):
>        if arg is __UNDEF__:
>            print "no argument"
>        else:
>            ...
>
>or, quite ugly:
>
>    def myfunc(*args):
>        if not args:
>            print "no argument"
>        else:
>            (arg,) = args
>            ...
>
>or, preferred:
>
>    come up with a better design; good designs allow people to
>    explicitly pass in a "use the default value" argument value...
>
You could also just use keyword args to override one or more defaults by name, e.g.,

 >>> def myfunc(arg, **kw):
 ...     if len(kw)==1 and kw.has_key('default_name'):
 ...         use_this = kw['default_name']
 ...     elif len(kw): raise TypeError,'myfunc received bad or too many kw args'
 ...     else: use_this = '<default value>'
 ...     print 'Using %s and %s' % (`arg`, `use_this` )
 ...
 >>> myfunc('1st arg')
 Using '1st arg' and '<default value>'
 >>> myfunc('1st arg',default_name='override')
 Using '1st arg' and 'override'
 >>> myfunc('1st arg','override')
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
 TypeError: myfunc() takes exactly 1 argument (2 given)
 >>> myfunc('1st arg',default_name='override',bad=123)
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "<stdin>", line 4, in myfunc
 TypeError: myfunc received bad or too many kw args
 >>> myfunc('1st arg',wrong_name='override')
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "<stdin>", line 4, in myfunc
 TypeError: myfunc received bad or too many kw args

And, just to throw something else into the picture ... ;-)

 >>> def foo(x='arbitrary default'):
 ...     if x is foo.func_defaults[0]: print 'Using default object'
 ...     else: print `x`, 'is not the same object as the default'
 ...     if x == foo.func_defaults[0]: print 'Using object equal to default.'
 ...
 >>> foo()
 Using default object
 Using object equal to default.
 >>> foo('arbitrary default')
 'arbitrary default' is not the same object as the default
 Using object equal to default.
 >>> foo(None)
 None is not the same object as the default

Of course, this doesn't differentiate well between passed and default args if
they are in the magically shared common ints and strings, but the distinction
can be interesting and maybe useful (and a potential source of bugs -- watch
out for default mutables ;-) Here's a contrivance to tag things to you can see effects:

 >>> def baz(d={}):
 ...     baz.c = baz.__dict__.get('c',{})
 ...     if d is baz.func_defaults[0]: p = 'default'
 ...     else:
 ...         idd = id(d)
 ...         p = baz.c.get(idd,None)
 ...         if not p:
 ...             p = 'caller%d' % len(baz.c)
 ...             baz.c[idd] = p
 ...     baz.n = baz.__dict__.get('n',0)+1
 ...     d['%s_%s'%(p,baz.n)] = baz.n
 ...     return d
 ...
 >>> baz()
 {'default_1': 1}
 >>> baz()
 {'default_2': 2, 'default_1': 1}
 >>> my_d={}
 >>> baz(my_d)
 {'caller0_3': 3}
 >>> my_d2={}
 >>> baz(my_d2)
 {'caller1_4': 4}
 >>> baz(my_d2)
 {'caller1_4': 4, 'caller1_5': 5}
 >>> baz(my_d)
 {'caller0_6': 6, 'caller0_3': 3}
 >>> baz()
 {'default_7': 7, 'default_2': 2, 'default_1': 1} 

Regards,
Bengt Richter


From opengeometry at NOSPAM.yahoo.ca  Fri Jun 21 05:15:04 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 21 Jun 2002 09:15:04 GMT
Subject: How to find all emails (on the same line)
References: <mailman.1024644929.25732.python-list@python.org>
Message-ID: <aeuqqn$9tbc8$1@ID-99293.news.dfncis.de>

A <printers at sendme.cz> wrote:
> Hi,
> I need to find ALL emails in a file. I can use re module but I do not 
> know how to find all emails if there are more emails on the same 
> line because re module ( as far as I know )finds only one/first 
> occurrence on that line.
> Thanks for help.
> Ladislav

Re-phrase your question, and post only once.  I think you meant email
address, as in <aaa at bbb.ccc>...

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin


From jepler at unpythonic.net  Tue Jun  4 12:28:25 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 4 Jun 2002 11:28:25 -0500
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
In-Reply-To: <9A5L8.232032$%u2.84870@atlpnn01.usenetserver.com>
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <9A5L8.232032$%u2.84870@atlpnn01.usenetserver.com>
Message-ID: <20020604112824.B24006@unpythonic.net>

On Tue, Jun 04, 2002 at 12:01:56PM -0400, Steve Holden wrote:
> >>> a = b = [1,2,3]
> >>> a += [4]
> >>> a, id(a), b, id(b)
> ([1, 2, 3, 4], 269472088, [1, 2, 3, 4], 269472088)
> >>>
> 
> Where did the rebinding take place? ISTM that "a" is still bound to the same
> list, which has been modified in place.

Sure it rebinds a (just to the same object formerly bound to it) ..
>>> class X:
...     def __setattr__(self, a, v):
...         print "rebinding %s, sucker" % a
...         self.__dict__[a] = v
... 
>>> x = X() 
>>> x.a = b = [1,2,3]
rebinding a, sucker
>>> x.a += [4]
rebinding a, sucker
>>> x.a is b
True

Or are you saying that in
    a = b = [1,2,3]
    a = b
the second statement isn't a rebinding operation?  That seems like an
odd position to take....

> Clearly an explicit assignment to a tuple element is always going to fail,
> so I don't really see what this would tell you.

But you're claiming that there *is* no binding, so it would be
surprising that this augmented assignment wouldn't work
    >>> t = ([1])
    >>> t[0] += [2]
t[0] "isn't" being rebound (according to your above logic, since if the
operation could complete, the "is" relation would hold between old t[0]
and new t[0]) .. but, of course, this actually causes a traceback.

I hate augmented assignment more and more each day, but at least I think
I understand it.  Your claim that there is no rebinding operation makes
me think that you may not.

Jeff




From altis at semi-retired.com  Sat Jun 29 14:33:44 2002
From: altis at semi-retired.com (Kevin Altis)
Date: Sat, 29 Jun 2002 11:33:44 -0700
Subject: Outlook PST and DBX mail file support for standard libraries?
Message-ID: <f1nT8.59$jM6.73893@news.uswest.net>

Jon Udell made a blog post today about converting Outlook PST files to
standard formats like MBOX.

http://radio.weblogs.com/0100887/2002/06/28.html#a321

In this case he used Mozilla. There is also a libPST project

http://freshmeat.net/projects/libpst/?topic_id=31

"LibPST provides functions in library form for accessing Outlook's Personal
Folders. Included with this library is a program that will take a PST file
and convert it to an mbox format."

And a libDBX

"LibDBX contains C and a C++ interfaces for accessing Outlook Express dbx
files (the C++ interface is just a wrapper for the C interface). Included
are 2 programs. One will read a dbx file and output the emails in mbox
format to stdout, the other will read the main "folder.dbx" file and export
all the folders into seperate mbox files."

http://freshmeat.net/projects/libdbx/?topic_id=28

Now, I'm not a candidate for the work being out of shape in the C/C++ SWIG
realm, but I'm wondering whether it might be possible to add PST file
support to the Internet Data Handling group of modules

http://www.python.org/doc/current/lib/netdata.html

Even if all that was supported was import/export type operations that would
be extremely useful, but if it was actually possible to read and write
contacts, mail, etc. from the PST then that opens up a whole slew of
utilities and applications that would normally have to be done using COM.

Of course, it might be that the licenses for libPST and libDBX would prevent
them getting into the Python distribution.

Just an idea for anyone with the skills to pursue it.

ka




From albgrig at tiscalinet.it  Fri Jun 21 15:54:26 2002
From: albgrig at tiscalinet.it (Alberto Griggio)
Date: Fri, 21 Jun 2002 21:54:26 +0200
Subject: python mode for emacs
References: <u660cfwo3.fsf@online.no>
Message-ID: <af008d$ft1$1@newsreader.mailgate.org>

On Fri, 21 Jun 2002 18:02:14 +0200, Syver Enstad wrote:


> I've downloaded the latest version from the python cvs tree, and it works
> fine. But... I am not too happy with the default face for the
> py-pseudo-keywords face (None, self....), and tried to change it by
> customizing the face, which worked fine until I restarted emacs and the face
> got it's ugly default. I looked at the font-lock hook and it seems that the
> face is overwritten each time python-mode is loaded with the keyword face.
> 
> Is there a straight forward way of deciding how the py-pseudo-keyword-face
> should look?
 
(defun my-py-mode-hook ()
  (set-face-foreground 'py-pseudo-keyword-face "CadetBlue")
  ;; ... and anything else you want
  )
(add-hook 'python-mode-hook 'my-py-mode-hook)

Bye,
Alberto


From BPettersen at NAREX.com  Thu Jun 13 13:35:09 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 13 Jun 2002 11:35:09 -0600
Subject: why not "'in' in 'in'"?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158ED1@admin56.narex.com>

> From: Grant Griffin [mailto:not.this at seebelow.org] 
> 
> In article <mailman.1023978378.11084.python-list at python.org>, 
> "Mark says...
> >
> >I haven't followed this thread.  Someone else has probably already 
> >suggested this, but in case they haven't, it's fairly easy to define 
> >your own function that does what you want:

[snip]

> 
> That's not a bad suggestion, Mark, but if I did that I guess 
> I would be forever wondering if I was testing whether the 
> first one was in the second or the second one was in the 
> first.  Again, it doesn't read nearly as well as:
> 
>    if x in subject:
>        ...
> 
> which leaves no doubt.

Ok, how about:

>>> class mystr(str):
...   def __contains__(self, s):
...     return self.find(s) >= 0
...
>>> 'hello' in mystr('hello world')
1
>>> 'foo' in mystr('hello world')
0
>>>

-- bjorn




From dmitri.gouliaev at telkel.net  Fri Jun  7 20:28:51 2002
From: dmitri.gouliaev at telkel.net (Dmitri I GOULIAEV)
Date: Fri, 7 Jun 2002 19:28:51 -0500
Subject: Efficient python programming...
In-Reply-To: <3D013FB1.3D7477EA@engcorp.com>; from "Peter Hansen" on Fri, Jun 07, 2002 at 07:20:17PM
References: <mailman.1023452786.27923.python-list@python.org> <3D013FB1.3D7477EA@engcorp.com>
Message-ID: <20020607192851.A29755@lifebook>

Hi, Peter Hansen !

 On Fri, Jun 07, 2002 at 07:20:17PM -0400, Peter Hansen wrote:

> Roman Suzi wrote:
> > 
> > On Fri, 7 Jun 2002, Peter Hansen wrote:
> > > You forgot the even more important first thing for a beginner:
> > > get it correct!  The only good way to do this is to write unit
> > > tests that prove it.  If you don't have tests, you don't have
> > > working code (or how can you prove it?).
> > 
> > To prove that code is correct is very difficult (practically impossible).
> > Unittests are there to prove that you have errors, not the other way
> > around.
> > 
> > Of course, they can increase your confidence in that optimized code is
> > equivalent to what you had before.
> 
> I disagree, if the tests are written first (and run, proving that
> they will fail when the code is wrong or missing).

That's what Roman said. If you have the results of tests for your code and your optiimazed code, you show only that in regard of THIS tests your code remains the same ("increase your confidence in that optimized code is equivalent to what you had before").

> The tests are executable requirement specifications.  Running
> them "proves" (yes, not 100%... nothing could) that the code
> meets the specs.  The specs could be wrong of course, but that
> doesn't mean the code is broken...

... if you can write such tests (I mean the tests covering the specifications). In most case, all you can "prove" is that for this particular tests you obtain this particular results.


Best regards,

-- 
DIG (Dmitri I GOULIAEV)
 
All below this line is added by my e-mail provider.




From g.szolderits at gmx.at  Wed Jun 12 17:12:37 2002
From: g.szolderits at gmx.at (Guenter Szolderits)
Date: Wed, 12 Jun 2002 23:12:37 +0200
Subject: Post installation problem with Python
References: <9dafguci28vdiv5lr3qbhset2jko78t6lm@4ax.com>
Message-ID: <3d07b91c$0$21420$91cee783@newsreader02.highway.telekom.at>

"Larry Taitt" <lt6789remove at hotmail.com> wrote:
> Installed ActivePython 2.2.1 build 222 on my NT4.0 (sp6a) workstation.
> After the installation I'm unable to execute any .exe, .com and .bat.
> program (e.g. notepad).  It tells me that it can't find the files.
> However, if I add the proper extension (e.g., notepad.exe), the
> programs execute without a problem.  I checked the associations and
> filetypes, and everything appears normal.  Any ideas?

Maybe your PATHEXT environment variable is damaged. It should look similar
as mine:

PATHEXT=.COM;.EXE;.BAT;.CMD;.py;.pyw;.pyc;.pys;.pyo

It should at least include .COM;.EXE;.BAT

G?nter




From walter at livinglogic.de  Tue Jun 25 11:45:49 2002
From: walter at livinglogic.de (=?ISO-8859-1?Q?Walter_D=F6rwald?=)
Date: Tue, 25 Jun 2002 17:45:49 +0200
Subject: Sending string var with apply sends all chars as params!???
References: <002501c21c5a$16f514f0$6f0d10ac@integer>
Message-ID: <3D18902D.6040909@livinglogic.de>

GimsMail wrote:

> Hi all,
>  
> I am having problems with the apply function in Python.  When applying a 
> string variable to a method, the string var gets chopped up into all it 
> chars and each char gets sent as a parameter!  The result is getting a 
> TypeError because I am sending too much parameters to the function (more 
> detail below if you still don't understand)
>  
> Regards
> Etienne
>  
> Detailed explanation:
> When running my program I get the following:
>  
> My print line debugging string output:
> Formatting value '20020625140609' of type <type 'string' with formatter 
> <function TimeFormatter at 00D613F4>
>  
> as you can see, I am sending the string '20020625140609' to the function 
> TimeFormatter.
>  
> The code that does this is:
> print "UCP: Formatting value '%s' of type '%s' with formatter '%s'" 
> %(value, type(value), formatter)
> value = apply(formatter,(str(value)))

This should be
value = apply(formatter,(str(value),))

But you have only one argument, so this can be simply written
as

value = formatter(str(value))

> in this case, formatter is the function TimeFormatter (as defined in a 
> mapping table)
>  
> When this code executes I get a TypeError:
>  
> TypeError: TimeFormatter() takes exactly 1 argument (14 given).
>  
> Why is the string (which is 14 chars long) broken up into 14 values when 
> passed using apply!?

Because apply expects a sequence of arguments as the second parameter.

> The weird thing is this code has been working for MONTHS and have only 
> recently stopped working after some refactoring in a completely 
> different part of code!
>  
> The problem goes away when I don't use apply:
> value = formatter(value)

Bye,
    Walter D?rwald





From gat at jpl.nasa.gov  Wed Jun 12 20:35:06 2002
From: gat at jpl.nasa.gov (Erann Gat)
Date: Wed, 12 Jun 2002 17:35:06 -0700
Subject: Code checksums at compile time?
Message-ID: <gat-1206021735060001@k-137-79-50-101.jpl.nasa.gov>

I am running some code that computes results that are stored in a
database.  The code is modified on occasion, and it's important to keep
track of which version of the code generated which results.  It's
important enough to get this right that I want to do this automatically
and not risk the possibility of manual error.

I'm doing the following: the code in question lives in a module called
'foo'.  When I load foo, I do the following:

import foo
code = []
hash = md5.new()
for line in xreadlines.xreadlines(open(foo.__file__)):
  hash.update(line)
  code.append(line)

I then write the (hash,code) pair into the database, and annotate all the
results with the hash value.  Now I have a foolproof record of which code
generated which results.

This works, but it's expensive, because this code is part of a cgi script,
so it has to recompute the hash every time it starts up.

My question is: is there a way to arrange for the module hash to be
computed at compile time rather than run time?  This would seem to be a
straightforward adaptation of my current code, except that I don't know
how to get a handle on "the module that is currently being compiled" (and
its associated file).

Thanks,
Erann Gat


From hesterloli at hotmail.com  Tue Jun 18 17:56:11 2002
From: hesterloli at hotmail.com (George Hester)
Date: Tue, 18 Jun 2002 21:56:11 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com><aegfck$6tkre$1@ID-114079.news.dfncis.de><wEPO8.16066$uk2.8447171@twister.nyroc.rr.com><3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com><gYgP8.42356$GY.12361719@twister.nyroc.rr.com><aekcug$7js8n$1@ID-114079.news.dfncis.de><9YBP8.44418$GY.13865843@twister.nyroc.rr.com><mailman.1024423525.24359.python-list@python.org> <20020619091900.70ae2b7f.larooy@xtar.co.nz>
Message-ID: <%%NP8.46690$GY.14560922@twister.nyroc.rr.com>

oops:

No George was asking for a script he could put in a ASP to determine his version
of Python.

--
George Hester
_________________________________
"John La Rooy" <larooy at xtar.co.nz> wrote in message
news:20020619091900.70ae2b7f.larooy at xtar.co.nz...
> On Tue, 18 Jun 2002 11:02:25 -0700
> Cliff Wells <logiplexsoftware at earthlink.net> wrote:
>
>
> >
> > Perhaps you'd like someone to come out and type it for you as well.
> >
> > Please remember that the people on this list aren't being paid to help you
> > write your program.  They expect you to do at least /a little/ of the
research
> > yourself.  It's understandable if you don't know where to look, but when
> > someone points you to the relevant portion of the docs, you should at least
> > read it before asking again.  It's clear from your postings that you haven't
> > even read the tutorial.
> > If you can't do that then you should pay someone to write your program for
you.
> >
> > --
> > Cliff Wells, Software Engineer
> > Logiplex Corporation (www.logiplex.net)
> > (503) 978-6726 x308  (800) 735-0555 x308
> >
> >
> I've been utterly astonished reading this thread. I was sure George was
flame-baiting
> but after his last couple of posts I'm not so sure. Well said Cliff
>
> John




From krissepu at vip.fi  Sat Jun  8 14:56:41 2002
From: krissepu at vip.fi (Pekka Niiranen)
Date: Sat, 08 Jun 2002 21:56:41 +0300
Subject: Help needed: Handling the search result of mxTextTool
Message-ID: <3D025369.8EF98B13@vip.fi>

After running mxTextTool + Simpleparse search of nested
parameters to single line I got  result like this:

(1, [('match', 2, 12, [('match', 5, 9, [])])], 14)

which means there has been two nested matches in searched string
(of 14 characters in total) and they are at slices [2:12] and [5:9].

Now I have to traverse the tuple/list -structure recursively to
build the strings again, since I am collecting the matched
strings instead of their slice positions. I am afraid this (post-)
looping
nulls the speed benefits gained by using mxTextTools.

Does anybody know a parameter in mxTextTools or a FAST
recursive function (no for loops please !) to help mxTextTools
to return the matched strings in a tuple or a list ( [a[2:12], a[5:9]) ?



-pekka-




From kragen at pobox.com  Sat Jun  8 21:42:40 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 08 Jun 2002 21:42:40 -0400
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com>
Message-ID: <834rgd5isv.fsf@panacea.canonical.org>

Peter Hansen <peter at engcorp.com> writes:
> You forgot the even more important first thing for a beginner: 
> get it correct!  The only good way to do this is to write unit
> tests that prove it.  If you don't have tests, you don't have 
> working code (or how can you prove it?).

I agree that correctness is important, although I think it is
sometimes overemphasized; buggy programs can still be useful.  (I hate
them, though.)  For example, in the ICFP programming contests, any
entry found to contain a bug, however small, is automatically
disqualified.

I also agree that unit tests are very useful for reducing bugginess.

But I do not think that they are either necessary or sufficient for
writing correct code.  I think (and extensive experimentation has
shown) that careful reasoning and code reviewing is necessary and
sufficient to write correct code.

It is certainly possible to err too far in the direction of believing
your code correct simply because it passes its test suite, and this is
a common beginner's error.  I think it is also possible to err too far
in the direction of writing and running tests instead of desk-checking
code.



From ctgaff at attbi.com  Wed Jun  5 22:28:19 2002
From: ctgaff at attbi.com (Corey G.)
Date: Wed, 5 Jun 2002 21:28:19 -0500
Subject: simple pattern matching question
Message-ID: <20020606022819.GA72435@attbi.com>

Gurus,

How would I match with an item from a list?  I would imagine something
this simple would not require a regular expression.

web_servers=('www01','www02','www03')

        for server_match in web_servers:
           if (trap == "Service:"server_match "State:down"):
	      print "found match"

The comparsion should be: if trap == Service:www01 State:down

server_match prints "as is, but not from the list".  I tried many 
combinations but decided to leave the most obviously wrong one.

-- 
Best Regards,
Corey




From maxm at mxm.dk  Wed Jun 12 02:13:40 2002
From: maxm at mxm.dk (Max M)
Date: Wed, 12 Jun 2002 08:13:40 +0200
Subject: PIL _imaging
References: <Xns922AA829A8C5FHotMailcs@130.133.1.4>
Message-ID: <3D06E694.4040103@mxm.dk>

Cousin Stanley wrote:


>   There are 2 files in K:\Python\Lib\PIL\DLLs
>   but apparently these are not the ones that are needed ...
> 
>     _imaging.pyd
>     _imagingtk.pyd 


Put those two files into:
K:\Python\DLLs

And you should be good to go.

regards Max M



From kzaragoza at attbi.com  Sun Jun 30 21:28:03 2002
From: kzaragoza at attbi.com (Kris J. Zaragoza)
Date: 30 Jun 2002 20:28:03 -0500
Subject: COM server threads/unloading
References: <10d944cb.0206261529.741da082@posting.google.com> <slrnahmos0.eh5.kzaragoza@localhost.localdomain> <10d944cb.0206281302.43de821a@posting.google.com>
Message-ID: <slrnahvc61.ao6.kzaragoza@localhost.localdomain>

In article <10d944cb.0206281302.43de821a at posting.google.com>, Uncle Enzo wrote:
> Thanks Kris, for your reply. I have a follow-up question...

No problem.  Glad I could help. :)

>> As I see it, you have two options.  First, you can build a mechanism
>> to explicitly send the logger the message to shut down.
> 
> How would you do this? Start another thread that periodically checks
> mainthread.GetInterfaceCount(), and if it returns zero put in the quit
> event and then return? I'd rather not have a shutdown or release
> method exposed to users.
> 

If you don't want to use an explicit shutdown call, look into defining
a __del__ method on your COM object.  These get called when the
reference count on Python objects reaches zero, and I think they get
called when the reference count on the COM object goes to zero and the
object is released.  (Mark the man could say more on this.)  You could
send the magic quit command to the background thread in there.

I'd dig into this more for you, but I'm actually on vacation and
borrowing a pathetically slow dialup line from my cousin.  The docs
for the Python COM extensions should be able to tell you more.

Good luck!

Kris

-- 
Kris J. Zaragoza       | On the face of it, Microsoft complaining about
kzaragoza at attbi.com    | the source license used by Linux is like the
                       | event horizon calling the kettle black.
                       | -- Adam Barr, article on kuro5hin.org


From jepler at unpythonic.net  Wed Jun  5 08:34:40 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 5 Jun 2002 07:34:40 -0500
Subject: self
In-Reply-To: <0%hL8.22575$tK.5490038@news02.optonline.net>
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net>
Message-ID: <20020605073435.A25640@unpythonic.net>

Nicola,
If you are going to always store the equation, never the value, what
happens when the user does
    c.y = 37
    c.y = 'y+1'
?  As far as I understand it, this would have to put the parser into an
infinite loop when
    c.y
is later referenced (since it must be determined by taking the value of
y and adding 1, but the value of y must be determined by ...)

You'd need to detect this problem somehow, and presumably not let
    c.y = 'y+1'
execute at all, but (especially if you allow dotted references like q.y)
this will be a bit of a pain to detect... regular expressions *might* be
up to the task (for a particular expression), but quite possibly not.
Then you're left with detecting stuff like
    c.x = "y"
    c.y = "x+1"
where it's a two-step process.  You have to determine that x uses
the value of y, and that y uses the value of x, to notice the cycle and
prevent the second assignment.

You're getting closer to the point where using the parser module to
parse the string as though it were an expression using
    >>> tree = parser.expr("x+1").totuple()
    >>> pprint.pprint(tree)
    (258,
     (313,
      (292,
       (293,
	(294,
	 (295,
	  (297,
	   (298,
	    (299,
	     (300,
	      (301,
	       (302, (303, (304, (305, (1, 'x'))))),
	       (14, '+'),
	       (302, (303, (304, (305, (2, '1'))))))))))))))),
     (4, ''),
     (0, ''))
you can use this form to both find the dependencies of the expression 
and to interpret the expression.  (btw I think the parser module doesn't
exist in jython)

Well, good luck with your project -- and don't spring any more
requirements on us, it's getting harder each time to solve your problem.

Jeff




From garrett at bgb.cc  Mon Jun 10 13:07:46 2002
From: garrett at bgb.cc (Don Garrett)
Date: Mon, 10 Jun 2002 17:07:46 GMT
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet> <3CFE23DB.994F8A6C@engcorp.com> <3CFFC67E.3070609@bgb.cc> <uhekfu9wt.fsf@ctwd0143.fitlinxx.com>
Message-ID: <3D04DC94.1040301@bgb.cc>

David Bolen wrote:
> While not MSI, there are free setup programs (I'm a fan of InnoSetup)
> that will produce a perfectly standard looking Windows installation
> executable for your application.  You can couple that with the
> installer/py2exe output to look like a standard Windows setup.
> 
> If you're already doing this and just specifically want an MSI format,
> then ignore the point. :-)

   I haven't been using anything like that, but was planning to start shortly. 
I was just hoping to have py2exe (or someone) output MSI to make it a single 
idiot proof step. And sometimes I can make for a pretty good idiot.

-- 
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc



From emile at fenx.com  Fri Jun 28 07:46:56 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 28 Jun 2002 11:46:56 GMT
Subject: explain this pickle issue to me...
References: <timo-1418C7.00095228062002@corp.supernews.com> <eOXS8.346960$cQ3.21460@sccrnsc01>
Message-ID: <Q0YS8.347036$cQ3.21354@sccrnsc01>

Emile van Sebille
> If you add to the top of testme:
>
> __name__ = 'testme'
> import testme
>
> and change your assert to:
>
> assert isinstance(s, testme.Simple), "Not an instance of Simple"
>
> that should fake it out.  The first makes the creation of pickles use
> testme as the module name, and the second and third allows the
> isinstance to work.
>

No, it's not that hard... ;-)

You don't need to make any changes to testme.py (except fix the assert
typos) if you initialize d properly:

d = (__import__('testme').__dict__)
execfile("testme.py", d)
q = d["writeSimple"]()
d["readSimple"](q)

 Proving-again-that-sombody-always-knows-a-better-way ly y'rs,

--

Emile van Sebille
emile at fenx.com

---------



From Sascha.Ferley at infineon.net  Sun Jun 23 21:30:38 2002
From: Sascha.Ferley at infineon.net (Sascha Ferley)
Date: Sun, 23 Jun 2002 19:30:38 -0600 (MDT)
Subject: PIL, Python and CGI
In-Reply-To: <200206241000.40596.rjones@ekit-inc.com>
Message-ID: <Pine.GSO.4.33.0206231929060.26987-100000@infinity.infineon.net>

Hi Richard
Thanks for replying back to me, this thing has been driving me nuts for a
while now. Anyways, i removed the try/catch block from the Image.py lib,
and checked the error output:

Traceback (most recent call last):
  File "/export/server/web/CPSC461/cgi-bin/interfce.py", line 15, in ?
    print Image.open(urlretrieve(sourcePath, '/tmp/img1')[0]).histogram()
  File "/usr/local/lib/python2.2/site-packages/PIL/Image.py", line 577, in
histogram
    self.load()
  File "/opt/sfw/lib/python2.2/site-packages/PIL/ImageFile.py", line 126,
in load
    self.load_prepare()
  File "/opt/sfw/lib/python2.2/site-packages/PIL/ImageFile.py", line 180,
in load_prepare
    self.im = Image.core.new(self.mode, self.size)
AttributeError: 'module' object has no attribute 'core'

Any new Ideas?
Sascha




On Mon, 24 Jun 2002, Richard Jones wrote:

> On Mon, 24 Jun 2002 09:37, Sascha Ferley wrote:
> > strange reason it can't find the _imaging extention..
> > It is sort of strange that it wouldn't pick it up then.. any more ideas?
>
> I'd say there's a chance that the _imaging module has linked to a shared
> library and the web server doesn't have the appropriate lib dir in its
> LD_LIBRARY_PATH. That'd cause an ImportError.
>
> Unfortunately, PIL is swallowing the exact error - try removing the try/except
> around the "import _imaging" and see what the real error is.
>
>
>    Richard
>
>







From buzzard at urubu.freeserve.co.uk  Wed Jun 19 19:03:26 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Thu, 20 Jun 2002 00:03:26 +0100
Subject: My Python/MySQL code is too slow
Message-ID: <aer2s4$mp7$1@news7.svr.pol.co.uk>

I have already posted this problem on mailing.database.mysql as I see it as
a problem with my lack of knowledge of SQL.  But no responses so far, and
maybe Python could do more of the work.  I repost the original question
below (SQL easier to follow), and the Python code below that.  Any help
appreciated.  TIA.

---------------------------------------------------------------------
I have two tables (say dbA and dbB) and need to find the number of rows in
dbB which have exactly 2 matching rows in dbA (on the selected columns).
I'm using Python/MySQL, but the relevant SQL is below.  This is slow, and
I'm sure it's possible without creating the temporary table and without the
join.  But everything I try fails (because of my limited knowledge of SQL).
Can anyone show me an efficient way of getting the answer?  All I need is
the number of rows in dbB which have exactly 2 matching rows in dbA.  Thanks
in advance.

CREATE TEMPORARY TABLE tmp SELECT var1, var2, var3 FROM dbA GROUP BY var1,
var2, var3 HAVING COUNT(*)=2;

SELECT dbB.var1, dbB.var2, dbB.var3, tmp.var1, tmp.var2, tmp.var3 FROM dbB,
tmp WHERE dbB.var1=tmp.var1 AND dbB.var2=tmp.var2 AND dbB.var3=tmp.var3;

Cheers.

Duncan Smith

------------------------------------------------------------------
import sys
import string
import MySQLdb

def MyFunc(dbname, tblname, pert_table, variables=()):

    """'dbname' is a MySQL database containing tables
    'tblname' and 'pert_table'.  'variables' is a
    sequence listing the relevant columns"""

    variables = list(variables)

    try:
        conn = MySQLdb.connect()
        curs = conn.cursor()
        curs.execute('USE %s' % (dbname,))
        if not variables:
            curs.execute('DESCRIBE %s' % (tblname,))
            variables = [x[0] for x in list(curs.fetchall())]
        vars = string.join(variables, ', ')

        #produce a temporary table containing all records
        #of tblname which occur in pairs and find P

        curs.execute("""CREATE TEMPORARY TABLE tmp
                    SELECT %s FROM %s GROUP BY %s
                    HAVING COUNT(*)=2""" % (vars, tblname, vars))

        def f(s): return string.join([s + var for var in variables], ', ')

        P = curs.execute("""SELECT %s, %s FROM %s, tmp WHERE %s""" %
(f(pert_table+'.'), f('tmp.'), pert_table, string.join([pert_table + '.' +
variables[i] + '=' 'tmp.' + variables[i] for i in range(len(variables))], '
AND ')))

        curs.execute('DROP TABLE tmp')

        curs.close()
        conn.close()

        return P/2

    except MySQLdb.Error, e:
        print "Error %d: %s" % (e.args[0], e.args[1])
        sys.exit(1)




From candiazoo at attbi.com  Sat Jun 22 09:41:03 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Sat, 22 Jun 2002 13:41:03 GMT
Subject: Help with MySQLdb insert with data extracted from odbc <-- Oracle
References: <3d13d5fb.651829078@netnews.attbi.com> <mailman.1024712368.16609.python-list@python.org>
Message-ID: <3d147e64.694941703@netnews.attbi.com>

Ah!  Thank you!

Mike J.

On Sat, 22 Jun 2002 04:17:58 +0200, Gerhard =?iso-8859-15?Q?H=E4ring?=
<gerhard.haering at gmx.de> wrote:

>* candiazoo at attbi.com <candiazoo at attbi.com> [2002-06-22 01:43 +0000]:
>> Can someone tell me what this means?  I mean, I guess I *know* what it means,
>> but why?
>> 
>> The insert statement into a mysql database is...
>> 
>> 'insert into works values(?, ?, ?, ?, ? ,?)'
>
>Wrong. paramstyle = 'format' in MySQLdb, not 'qmark'. Look into the
>DB-API specs (PEP 0249) for details about the various paramstyles that
>are possible.
>
>> The invocation of the dynamic sql is...
>> 
>>     status = cur_mysql.execute(sql, 
>>         (wrk_inst, ZERO, trs_inst, aas_inst, title, NADA))
>
>Simply use %s instead of ?. And read the fine MySQLdb docs that are
>included :-)
>
>Gerhard
>
>PS: I don't like these sort of spam protection attempts.
>-- 
>mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
>web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
>public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
>reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))
>
>



From peter at engcorp.com  Fri Jun  7 19:17:37 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 07 Jun 2002 19:17:37 -0400
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com> <mailman.1023452349.14713.python-list@python.org>
Message-ID: <3D013F11.F25FFA72@engcorp.com>

holger krekel wrote:
> 
> Peter Hansen wrote:
> > You forgot the even more important first thing for a beginner:
> > get it correct!  The only good way to do this is to write unit
> > tests that prove it.  If you don't have tests, you don't have
> > working code (or how can you prove it?).
> 
> yeah. writing tests is *the way* to go if you want other people use
> your code. Although successfull unittests theoretically don't prove
> anything<wink> they tend to do in reality<doublewink>.

Could you explain why you think unittests don't prove anything 
when they are successful?

Are you pointing out that if you write a test after the code
and the test passes, you have learned almost nothing?  I agree
with that.

With Test-Driven Development, however, you write the test first,
run it immediately, and ensure that it fails.  Only after you
have code that makes it pass are you done.

In this case, making sure the test fails when the code is wrong,
wouldn't you say that the test mostly proves that the code works?

(Not quibbling about the fact you are still not 100% sure...)

-Peter


From tim at zope.com  Tue Jun 11 16:50:15 2002
From: tim at zope.com (Tim Peters)
Date: Tue, 11 Jun 2002 16:50:15 -0400
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
In-Reply-To: <iQnN8.153796$352.7562@sccrnsc02>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEDDPMAA.tim@zope.com>

[John Machin]
> === bot bait ================================
> bignum, long, fast, speed, optimize snot out of

[Emile van Sebille]
> Hook many bots this way?  ;-))

John might have, except for his later

> Is the function elegant enough?

Take it from me:  "elegant" has a weight of about minus gazillibob in a
bot's attraction algorithm.

not-as-bad-by-far-as-"practical"-but-a-loser-all-the-same-ly y'rs  - tim





From deckerben at freenet.de  Sun Jun 23 17:58:34 2002
From: deckerben at freenet.de (deckerben)
Date: Sun, 23 Jun 2002 23:58:34 +0200
Subject: Fork
References: <3d15c022$0$26988$9b622d9e@news.freenet.de> <3D15C638.EE5797F9@phekda.freeserve.co.uk> <3d16405b$0$166$9b622d9e@news.freenet.de>
Message-ID: <3d1643f9$0$136$9b622d9e@news.freenet.de>

For those confused by the copy of my post to the DJGPP guys, it pertains to
the MSDOS port of Python 2.2.1 only.

Ben




From pan-newsreader at thomas-guettler.de  Sun Jun 16 16:32:21 2002
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Sun, 16 Jun 2002 22:32:21 +0200
Subject: pygtk: Hourglass Cursor
Message-ID: <pan.2002.06.16.22.32.20.495635.716@thomas-guettler.de>

Hi!

How can I change the cursor to an hourglass with pygtk?

thomas


From huaiyu at gauss.almadan.ibm.com  Mon Jun 24 22:24:30 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Tue, 25 Jun 2002 02:24:30 +0000 (UTC)
Subject: How to represent the infinite ?
References: <mailman.1024718075.18074.python-list@python.org>
Message-ID: <slrnahfl2u.2s7.huaiyu@gauss.almadan.ibm.com>

Tim Peters <tim.one at comcast.net> wrote:
>> 1. Most platforms these days support 754 to a remarkably good degree, far
>>    better than that is allowed in today's Python.
>
>The support code required at the C level is platform-specific, and needs a
>phalanx of cooperative platform experts to contribute it.  Slash and burn
>will not lead to a good and maintainable result.

OK, so I infer that
- each platform may well support 754 to a good degree
- there is probably a common subset that is good enough and large enough
- but they are not invoked the same way, so that a simple approach may only
  support a small set of features and being incorrect on other features in
  various platforms.

This is quite discouraging to hear.  Maybe someone knowledgeable enough can
compile a list of major features on major platforms.  Maybe for suitably
defined set of major features and major platforms, the cost is not high
enough to overwhelm the benefit.  Once there is a start it might become
easier to enlarge the subset.


>> 2. Most number crunchers would be more concerned that it works as well as
>>    possible on the platforms available to them, rather than  whether it
>>    also works on other platforms.
>
>As I said above, that's why the NumPy community has to agree first.  The
>overwhelming majority of plain-core Python users are not number crunchers,
>and don't know a NaN from an umlaut.  Giving them either 1j or a NaN from
>sqrt(-1) is a Bad Idea.  I would often like 1j or a NaN in my own apps --
>but I often want the ValueError Python tries to deliver in that case too.

I realize that there are (at least) two types of numerical calculations,
those where the calculation determines the error bounds, and those where the
error bounds are calculated separately.  The former is likely to occur in
traditional engineering with quite deterministic data, and the latter is
likely to occur in data mining and statistical applications with lots of
uncertainties.  So I can see the point that a proper implementation would
have to have a choice whether overflow generates an exception.


Before that happens (and I understand that is going to be a long time), I'd
like to see overflow exception raised only once for each vector, with proper
information in the exception denoting which element encountered it first.
This is not likely to upset those who want to see the exceptions, while
making it much easier to ignore the exceptions if one wants.  I'm not sure
whether this is a core Python issue or NumPy issue, though.


Huaiyu


From kragen at pobox.com  Sat Jun  8 21:50:33 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 08 Jun 2002 21:50:33 -0400
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <20020606215349.GA13474@node1.opengeometry.net> <mailman.1023401650.18434.python-list@python.org> <adr50a$1n43o$1@ID-71831.news.dfncis.de>
Message-ID: <831ybh5ifq.fsf@panacea.canonical.org>

"James Kew" <james.kew at btinternet.com> writes:

> "gabor" <gabor at realtime.sk> wrote in message
> news:mailman.1023401650.18434.python-list at python.org...
> > for file in directory:
> >     file.printsize()
> 
> Finally, "printsize" as a method name suggests to me that the loop is more
> likely to be IO-bound than CPU-bound, so attempts to optimise the loop
> overhead may buy very little benefit.

Possible, but not likely --- how much I/O do you think it takes to
read an inode and put a line in an output buffer?

> 
> If it were me, I'd go with the first option for clarity; leave it unless and
> until a performance problem emerges and profiling points at this loop;
> consider algorithmic changes first (possibly merge the multiple IO
> operations into a single larger one?); and only then start sacrificing
> clarity for speed.
> 
> James



From peter at engcorp.com  Thu Jun 20 21:43:47 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 20 Jun 2002 21:43:47 -0400
Subject: do...while
References: <slrnah40l1.12v.msoulier@pmerd071.ca.nortel.com> <XFMail.20020620104256.shalehperry@attbi.com> <mailman.1024596866.20597.python-list@python.org>
Message-ID: <3D1284D3.78507FCF@engcorp.com>

"Michael P. Soulier" wrote:
> 
> On Thu, Jun 20, 2002 at 10:42:56AM -0700, Sean 'Shaleh' Perry wrote:
> > while 1:
> >     line = filehandle.readline()
> >     if len(line) <= 5: break
> >     ....
> >     ....
> >
> > is a typical way to implement it.
> 
>     Right, I've used that in the past. It doesn't seem nearly as clean though,
> as making the conditions for terminating the loop plain in the loop
> declaration. Considering the Python community's leaning towards readable code,
> necessitating a break for code factoring goes against that philosophy.

Avoiding duplicated code always trumps an uneasy feeling that
the code might not be as pretty as it could be in another
language or with new language constructs...

Once you get over that uneasy feeling, and implement the loop
that way a few times, and focus on maintainability, you'll
probably stop worrying about it.  At least, I did. :)

-Peter


From merman at snafu.de  Wed Jun 19 10:45:29 2002
From: merman at snafu.de (T. Kaufmann)
Date: Wed, 19 Jun 2002 16:45:29 +0200
Subject: newbie: help please!
References: <aepltq$gl3$1@knossos.btinternet.com>
Message-ID: <3D109909.10204@snafu.de>

Hi Steve,

I run SuSE but there was the same problem.

 > In file included from ./Modules/_tkinter.c:49:
 > /usr/include/tk.h:83:29: X11/Xlib.h: No such file or directory

I think it could be you need also the 'tk.h'! Search this file on your 
machine.

On SuSE there is a separatly TCL/ TK-Developer package - you must 
install this! It contains all TCL/ TK-Headers.

Hope this helps.

Regards Thomas



From peter at engcorp.com  Thu Jun 20 22:05:44 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 20 Jun 2002 22:05:44 -0400
Subject: application development - separating content from function
References: <yroQ8.260949$i6.17986369@amsnews02.chello.com>
Message-ID: <3D1289F8.58A17BA4@engcorp.com>

Henry Baumgartl wrote:
> 
> For eight months now, I am preparing for the development of an open source
> Python application to support individuals in Life Long Learning. It will be
> a mixture of PIM, extended diary, groupware, project management, and modules
> to interact with centralised knowledge / skill exchange servers.
[...]
> What needs to be done is:  [snip description of lots of work]
> 
> The more I have tried to solve these requirements in an elegant and
> efficient manner, the further I seem to stray from anything remotely
> resembling a viable solution. In other words, I can't see the forest for the
> trees anymore.

Welcome to the club! :-)  Undertaking an overwhelming project is pretty
common, especially with many open source projects.  Check out the number
of idle projects on SourceForge, for example.  I've made many many 
false starts in my life as a programmer, and I still regret much of the
time I spent on some of those hopeless projects.

My suggestion to you would be to investigate the practices and principles
of Extreme Programming, especially YAGNI (you aren't gonna need it), 
DTSTTCPW (do the simplest thing that could possibly work), frequent
releases and small iterations, using automated tests to drive the 
development and let the design emerge on its own, and perhaps most
important of all, given your situation, controlling scope.

You've got a really, /really/ huge pile of work ahead of you if you
think of all those pieces as being essential to the project.
Prune it down.  Think of the single most useful piece and exclude 
everything else.  Now think of the single most useful feature of 
that piece, and think of the single most useful aspect of that 
feature.... drill down until you have something that you think you 
can implement in a few days or a week of work.  Now do it.  Have 
automated acceptance tests for it so you know when you've got it 
done, so you can stop working. Release it to your users, or 
yourself.  Use it.  Get feedback from it.

Now... sit back and think about the next most useful feature that
you could implement.  Add it.  Make sure you have those automated
tests.  They let you be confident that the previous code still
works, that the new code works and that you're done when you're
done.  They let you aggressively modify the structure without
having to worry that you'll break something (you will, but you'll
find out immediately, back out the change, and carry on without
getting bogged down in lengthy debugging sessions).

If you picture your project as a very large number of potential
features being implemented one by one, you can probably think of
some that are obvious "must-haves", with a big bang for the buck.
There are others that maybe could be postponed a while, and probably
a bunch that you might never need to do.  Your goal is to pick
those items with the highest value first and implement them, 
without worrying about the others.  That way, if you have to
stop, you at least have something useful.  If you don't focus on 
doing the highest value items first, you'll probably find that
at some point you're ready to quit, you have nothing that can 
actually be used for anything, and you've spent a year and a
half of your life with nothing to show for it but a bunch of 
code.  (And in most projects, code without good tests, which is
as good as files filled with random bytes.)  With good tests, you
can always hand over the project to another.

I've found Python to be an incredibly effective language to use
in this kind of environment, especially as it makes the automated
testing as close to trivial as it probably can be.  But more than
Python itself, starting to think in XP terms has improved my ability
to produce useful, reliable code in a timely fashion.  If you
visit http://www.extremeprogramming.org/ for a quick intro, or
http://c2.com/cgi/wiki/?ExtremeProgrammingRoadmap for a much
larger set of info, you might gain a least a little inspiration,
even if you decide to stick with more traditional approaches.

-Peter


From mgerrans at mindspring.com  Sat Jun 15 03:35:16 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Sat, 15 Jun 2002 00:35:16 -0700
Subject: VBA to Python
References: <GjAO8.4198$iR2.398848@news.xtra.co.nz>
Message-ID: <aeeqnv$i6n$1@slb6.atl.mindspring.net>

> In moving from using Access front end forms and code connected to an
Access
> back end I am looking at changing to Java front end to SQL back end with a
> business rules engine using Python.

Using the Win32 extensions, in particular win32com.client.Dispatch() is
quite handy for dealing with COM automation objects; I use this quite often
to access the same objects commonly used in JScript and VBScript (and
probably also VBA).

> Is Python a good choice to manage and run the rules ...

Of course!   Knocks the socks off BASIC programming.   You have the temerity
to even ask such a question in *this* newsgroup?   :-)

> ...and is there a way of
> transcribing the VBA code across to Python other than literally doing it
> line by line?

As opposed to figuratively doing it line by line?    I guess you want to
convert existing code to Python, rather than just doing the new code in
Python?   There may be translators, but my guess is that this would open a
big can of worms (figuratively) and be the source (so to speak) of lots of
headaches.    I like the idea of doing new work in Python, while slowly
converting modules that you come across (while bug fixing, adding new
features, etc.) to Python.   If you have a lot of working code (even if it
is encoded in a distasteful syntax like VB), there is no sense in junking it
carte blanche -- if (when) there are any problems there will be naysayers
blaming it on Python, which would be a shame.

- mfg




From mgerrans at mindspring.com  Mon Jun 10 12:23:58 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Mon, 10 Jun 2002 09:23:58 -0700
Subject: (no subject)
References: <mailman.1023640273.18210.python-list@python.org>
Message-ID: <ae2jr4$8so$1@slb6.atl.mindspring.net>

CGI in this day and age?   Why not use Java Servlets and JSPs?   If you toss
in Jython (http://www.jython.org), you get the best of both worlds.


> - provide [generic] CGI functionality
> and if possible...
> - generate dynamic (DHTML) responses to CGI input parameters





From boud at valdyas.org  Sun Jun 16 09:26:24 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Sun, 16 Jun 2002 15:26:24 +0200
Subject: IDE recommendations??
References: <b9veea.l85.ln@majestix.gallien.de> <3d0b06bd$0$52043$e4fe514c@dreader3.news.xs4all.nl> <83lm9gtla9.fsf@panacea.canonical.org>
Message-ID: <3d0c9270$0$52048$e4fe514c@dreader3.news.xs4all.nl>

Kragen Sitaker wrote:

> Boudewijn Rempt <boud at valdyas.org> writes:
>> I've tried to use most IDE's out there -- from Pythonwin to Pythonworks.
>> None was really comfortable, except for Wing IDE, which is really
>> powerful. It doesn't include integration of a GUI designer, though, which
>> makes it little more than a glorified editor/debugger (but _very_
>> glorified). At one point there was a version of KDevelop that supported
>> Python development, but that has disappeared, it seems.
>> 
>> I use bash & XEmacs & Designer nowadays, and that gives me a fair
>> productivity.
> 
> I'd like to improve XEmacs to be a better Python IDE.  What kind of
> integration does it need with Designer?  I haven't used GUI-designer
> IDEs much, so I don't really know what features I'm missing.

I don't need much integration: I go esc-x compile, and tell it to
run the Makefile that generates python code from the designer designs.

-- 
Boudewijn Rempt | http://www.valdyas.org


From vlindberg at verio.net  Fri Jun  7 14:28:47 2002
From: vlindberg at verio.net (VanL)
Date: Fri, 07 Jun 2002 12:28:47 -0600
Subject: Help with Black Adder...
References: <adpc36$179jq$1@ID-139866.news.dfncis.de> <mailman.1023432372.7831.python-list@python.org>
Message-ID: <3D00FB5F.5010101@verio.net>

The Book mentioned above (GUI programming with QT or some 
such) is also available for free on the web.

    http://www.opendocspublishing.com/pyqt/index.lxp

Enjoy!

VanL



From akuchlin at ute.mems-exchange.org  Wed Jun  5 12:18:40 2002
From: akuchlin at ute.mems-exchange.org (A.M. Kuchling)
Date: 05 Jun 2002 16:18:40 GMT
Subject: distutils and distributing "scripts"
References: <adl4e4$19ep$1@news.adamastor.ac.za>
Message-ID: <slrnafsef0.57v.akuchlin@ute.mems-exchange.org>

In article <adl4e4$19ep$1 at news.adamastor.ac.za>,
	Hugo van der Merwe wrote:
> just the py_modules. The docs seem to confirm this. How do I get my .pyw 
> file distributed, do I really have to go and do all that MANIFEST.in stuff, 
> just for this one file?

You could just write a MANIFEST file that lists all of the files to be
included, one per line.  That's less complicated than writing a
MANIFEST.in, but it's also more error-prone if you add or remove files
a lot, because you'll have to edit the MANIFEST file each time.

--amk                                                             (www.amk.ca)
>From even the greatest of horrors irony is seldom absent.
    -- H.P. Lovecraft, "The Shunned House"



From martin at v.loewis.de  Sun Jun  9 16:16:35 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 09 Jun 2002 22:16:35 +0200
Subject: Socket error problem
References: <mailman.1023646813.20804.python-list@python.org>
Message-ID: <m33cvwb42k.fsf@mira.informatik.hu-berlin.de>

"David LeBlanc" <whisper at oz.net> writes:

> The general.* constant references are to colorize the output which is turned
> off but still results in the "31" in the connection error line in the
> program output below. NOTE: I didn't write this code, so I have no idea why
> it's returning the username in the socket.error instead of the host. FWIW,
> this code is running in a thread, but there is only one thread trying to get
> mail.

That deserves some investigation since this, again, sounds impossible
(how can it possibly know what the user name as, when this piece of is
not being passed the user name). The only possible explanation is that
self.error_msg prints unrelated information; I recommend to remove all
except clauses and let the exception pass through.

>                 self.sock = socket.socket(af, socktype, proto)

Please put a print statement before this socket call, to print all
three arguments.

>                 self.sock.connect(sa)
>             except socket.error, msg:

And put one here, printing socket.error.

Regards,
Martin


From jadestar at idiom.com  Wed Jun 19 15:52:41 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 19 Jun 2002 19:52:41 GMT
Subject: Pop a list from beginning ? and memory saving...
References: <aem8e8$4kq$1@news.idiom.com> <BIEJKCLHCIOIHAGOKOLHAELEDEAA.tim.one@comcast.net> <mailman.1024424930.10523.python-list@python.org> <aep4jb$1q8h$1@news.idiom.com> <aeqlgo$gbk$1@panix1.panix.com>
Message-ID: <aeqne9$4h1$1@news.idiom.com>

Aahz <aahz at pythoncraft.com> wrote:

> In article <aep4jb$1q8h$1 at news.idiom.com>,
> James T. Dennis <jadestar at idiom.com> wrote:

>> It might be better to have a hack to the implementation of the
>> list primitives --- so they defer the shift of remaining elements
>> until GC.  The head of the list could point at the first none 
>> deleted item.  Then pop(0) could be implemented in O(1) time and
>> the garbage collector would eventually come along to do it's 
>> job.

> While this might be a worthwhile optimization, it has absolutely nothing
> to do with GC.  Each time an element were popped off the front, its
> object would be decreffed and possibly deleted.  GC runs completely
> independently of individual objects.

 Sorry, I was speaking abstractly about garbage collection, I know
 nothing about Python's GC implementation.  Perhaps it would be an 
 opportunity to augment Python's GC.  Some containers (and even 
 some classes) could implement their own compaction routines that 
 could be called by the GC. 



From SBrunning at trisystems.co.uk  Tue Jun 25 06:12:14 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 25 Jun 2002 11:12:14 +0100
Subject: Python & zipfiles
Message-ID: <31575A892FF6D1118F5800600846864DCBD427@intrepid>

> From:	Patrick Bielen [SMTP:bielen at stafa.nl]
> The problem is, that i don't know if i use the
> ZipFile-command too much... because the manual
> told me i have to close it afterwards, but it's
> not clear to me if i have to expecitely open it.
> 
> Any help or code cleaning would be really appre-
> ciated.
 
As luck would have it, zipfile is demonstrated on the online sample chapter
of the effbot's "Python Standard Library" - see
<http://www.oreilly.com/catalog/pythonsl/chapter/ch05.html#t11>. Worth
getting hold of.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.




From insaney at ufl.edu  Mon Jun  3 12:28:47 2002
From: insaney at ufl.edu (Daniel)
Date: 3 Jun 2002 09:28:47 -0700
Subject: UnboundLocalError: local variable 'blah' referenced before assignment
Message-ID: <40a8aa92.0206030828.3b96333d@posting.google.com>

I have a strange problem where a global variable is no longer
referenced.
Stripped down, it's something to the effect of:

def func1():
   blah='bcd'

def func2():
   print blah
   blah='cde'

def main():
   func1()
   print blah
   try:
      func2()
   except:
      pass
   print blah   

blah='abc'
main()

func2 raises an UnboundLocalError: local variable 'blah' referenced
before assignment and the final print is the 'bcd'.  This only happens
in one script I'm working on, all the others it seems fine.  I'm
working in vi, so I suspect that maybe the problem is a funky
character got thrown in somewhere in that script, and the interpreter
isn't liking that.  I did a quick search and all characters of the
file are in string.whitespace, string.letters, string.digits, or
string.punctuation.  What characters might cause this to happen, or if
you suspect the problem is something else, any ideas?

Python 2.2 (#1, Feb 11 2002, 09:52:08)
[GCC 2.95.3 20010315 (release)] on linux2

Thanks,
Daniel


From emile at fenx.com  Mon Jun 24 06:39:09 2002
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 24 Jun 2002 10:39:09 GMT
Subject: converting strings to numbers
References: <3d16f3e0@clear.net.nz>
Message-ID: <hFCR8.55067$EP.5276@sccrnsc03>

Tony Cradwick
> How do you convert (cast?) a string, eg from a user input entry box,
to a
> number, eg integer?
>

int(raw_input('Enter an integer: '))
float(raw_input('Enter a float: '))

--

Emile van Sebille
emile at fenx.com

---------



From DavidA at ActiveState.com  Fri Jun  7 14:52:44 2002
From: DavidA at ActiveState.com (David Ascher)
Date: Fri, 07 Jun 2002 11:52:44 -0700
Subject: Looking for some Cookbook contributors
Message-ID: <3D0100FC.3080701@ActiveState.com>

Hello everyone --

O'Reilly is looking for the following contributors to the
Python Cookbook, to get their addresses to send them free
copies of the book when it comes out:

    Nick Perkins

    David Shaw

    Matthew Wood

    Horst Hansen

If you are one of these people, please write back to me,
so that I can put the O'Reilly people in touch with you.

If you know how to reach one of these people and could
send me their contact information, that would be helpful too.

--David Ascher

PS: Alex and I are proofing the production ready copy, and
I at least am quite proud of what the Python community has
put together.  We'll show it off at OSCON!







From shalehperry at attbi.com  Wed Jun  5 15:12:06 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 05 Jun 2002 12:12:06 -0700 (PDT)
Subject: Detecting OS and Window Manager
In-Reply-To: <ufsnh8pmmde6cf@corp.supernews.com>
Message-ID: <XFMail.20020605121206.shalehperry@attbi.com>

On 05-Jun-2002 Andrew Wilkinson wrote:
> Hi,
> 
> I'm writing a program that integrates with a number of different OSes and
> Window Managers (under Linux), but I need to check what the user is running.
> I know that...
> 
> import os
> if os.name == 'nt':
> 
> will tell me they're running NT. But under Linux I'd like to know if they
> are running KDE or Gnome, does anyone know a (preferably) simple test for
> this?
> 

the only fool proof way to detect KDE, GNOME, etc is to check for X atoms.  The
bigger question is why do you care?  As long as they can display a GUI you
should not be bothered about which one it is.  Especially under the X Window
System it was designed so that random applications DIDN'T know what they were
running under.

Also remember KDE is not a window manager, it is a desktop environment.  It is
perfectly reasonable to run KDE and use fvwm (or even twm) as the window
manager.  Same applies to GNOME.  Also, what about someone who only runs say
Window Maker and never uses GNOME or KDE?




From marklists at mceahern.com  Wed Jun 19 10:08:53 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 19 Jun 2002 09:08:53 -0500
Subject: procmail replacement in Python
In-Reply-To: <oqelf3s7n4.fsf@carouge.sram.qc.ca>
Message-ID: <NCBBLFCOHHDIKCAFGCFBAEGILFAA.marklists@mceahern.com>

> Would you kindly give me an URL for `getmail'?

I think the top hit here is what you're looking for:

  http://www.google.com/search?q=python+getmail

Namely:

  http://www.qcc.sk.ca/~charlesc/software/getmail-2.0/

It's one of those rare cases where you can use the google I'm feeling lucky
button:

  http://www.google.com/search?q=python+getmail&btnI=I'm%20feeling%20lucky

Cheers,

// m

-





From whisper at oz.net  Fri Jun 21 18:44:42 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 21 Jun 2002 15:44:42 -0700
Subject: Windows versions of Python---pros and cons?
In-Reply-To: <3D138293.85892240@noaa.gov>
Message-ID: <GCEDKONBLEFPPADDJCOEOEHEDLAA.whisper@oz.net>

<snip>
> Mark Hammond wrote:
> > Add the .py extension to your PATHEXT environment variable, and you can
> > execute "foo.py" at the command prompt.  Like much of Windows, this is
> > based on the extension rather than file attributes.

Even cooler, you no longer need the .py extension (if you're on Win2000 at
least) and if you're in the same directory as the script you're trying to
execute. I'm not sure that even with a .py extension that it will do a path
search.

> Which is really annoying, kind of  like the Macintosh file creator
> concept: I sometimes want the file to be edited, and sometime run...
>
> Anyway, it looks like my best bet is to re-associate *.py with the
> python interpreter, rather than pythonwin. I rarely have a use for
> opening pythonwin from the command line, and in the GUI, I can use
> openwith from the right button menu if I want to open a *.py file
> in pythonwin.
>
> Now I have to figure out how to change my PATHEXT environment variable,
> which is pretty OT for this group...

It's in your environment settings right under "path" :-)

> -Chris
>

Greetings fellow latte slurping moss backed emerald city dweller ;-)

Dave LeBlanc
Seattle, WA USA





From look at sig.invalid  Wed Jun 26 06:31:55 2002
From: look at sig.invalid (Simo Salminen)
Date: Wed, 26 Jun 2002 10:31:55 +0000 (UTC)
Subject: Parsing strings (\n and \\)
References: <3D188006.7050202@thomas-guettler.de> <mailman.1025017712.20580.python-list@python.org> <lt0S8.45025$n4.10623651@newsc.telia.net> <slrnahijff.m4g.look@pikkukorppi.cs.tut.fi> <OFeS8.45132$n4.10657162@newsc.telia.net>
Message-ID: <slrnahj60r.rli.look@pikkukorppi.cs.tut.fi>

* Fredrik Lundh [Wed, 26 Jun 2002 08:10:22 GMT]
> Simo Salminen wrote:
> 
>> safe(?) version:
>>
>> >>> PARSED = eval(UNPARSED, {"__builtins__":None}, {})
>> Traceback (most recent call last):
>>   File "<interactive input>", line 1, in ?
>>   File "<string>", line 0, in ?
>> NameError: name '__import__' is not defined
> 
> now try this:
> 
> UNPARSED = "'*'*10000000*2*2*2*2*2*2*2*2*2"
> 

yes, but if you don't mind it consuming lots of cpu time and memory, its
safe. i'm not saying eval should be used, but if you do, restricting it is
good idea.


-- 
simo <dot> salminen <at> iki <dot> fi


From radix at twistedmatrix.com  Sat Jun  8 17:11:42 2002
From: radix at twistedmatrix.com (Christopher Armstrong)
Date: 08 Jun 2002 17:11:42 -0400
Subject: Source formatting for long expressions.
Message-ID: <87hekdxypd.fsf@twistedmatrix.com>

Background:
  I'm writing a source persistence mechanism for Twisted 
(http://twistedmatrix.com) as an alternative to the Pickle and XML persistence
that we currently have. Basically, I have code that turns Python objects into
something that looks like::

Ref(1, Instance('twisted.internet.app.Application', persistenceVersion=7,
services={}, udpPorts=[], gid=1000, connectors=[], uid=1000,
tcpPorts=[(9080, Instance('twisted.web.server.Site', sessions={}, 
resource=Instance('twisted.coil.web.ConfigRoot', children={}, 
modules=[], widgets={'config': 
Instance('twisted.coil.web.AppConfiguratorPage', variables={}, 
dispensers=Instance('twisted.coil.coil.DispenserStorage', dispensers={}}),
app= ...
<bazillion more lines>

  As you can see, this is one really long expression :-) I already have the
code that persists to/unpersists from this format (it was fairly easy with the
'jelly' serialization framework), but it's not very useful unless it's in a
pretty format that's easy for people to edit. Does anyone know of a tool that
will beak up a very, very long expression like this into multiple lines with
sane indentation? I'll probably end up writing the beast myself, but I don't
want to waste my time if there's something out there.

-- 
                                Chris Armstrong
                         << radix at twistedmatrix.com >>
                http://twistedmatrix.com/users/carmstro.twistd/





From gabor at realtime.sk  Fri Jun  7 00:15:52 2002
From: gabor at realtime.sk (gabor)
Date: 07 Jun 2002 00:15:52 -0400
Subject: Efficient python programming...
In-Reply-To: <20020606215349.GA13474@node1.opengeometry.net>
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> 
	<20020606215349.GA13474@node1.opengeometry.net>
Message-ID: <1023423352.7321.82.camel@wintermute.atriaky.sk>

ok,

so you need general programming advices?

then by a book about programming or something like that....

but if you want to know about specific python issues, then you have to
learn things like "dictionary are very well optimized etc.."

those things doesn't make your code obfuscated..

for file in directory:
	file.printsize()


ran = range(len(directory))
for r in ran:
	directory[r].printsize()

[file.printsize() for file in directory]

they're the same, and everyone would recommend you the last one..

it's python specific, but it won't make your code obfuscated..


so to summarize this: what do you want to know/learn?

bye,
gabor


 On Thu, Jun 06, 2002 at 08:40:15PM +0200, Shagshag13 wrote:
 >  (not the kind of thing which made you an obfuscated python nerd,
 > but that "map" should be preferred to "for in loop" if possible, that
 > dictionary are very well optimized, etc.)






From pyth at devel.trillke.net  Sat Jun 15 10:16:23 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 15 Jun 2002 16:16:23 +0200
Subject: Pronouncing '__init__'
In-Reply-To: <20020615165915.A17174@hishome.net>; from oren-py-l@hishome.net on Sat, Jun 15, 2002 at 04:59:15PM +0300
References: <20020615132519.A14452@hishome.net> <20020615134410.A15079@prim.han.de> <20020615165915.A17174@hishome.net>
Message-ID: <20020615161623.D15079@prim.han.de>

Oren Tirosh wrote:
> On Sat, Jun 15, 2002 at 01:44:10PM +0200, holger krekel wrote:
> > for oral language:
> > 
> >     init constructor
> > 
> > and if the context allows
> > 
> >     init
> > 
> > for written languange __init__
> > 
> > Isn't it obvious? :-)
> > 
> >     holger
> 
> I used __init__ just as an example.  I meant how do you pronounce names 
> with leading and trailing double underscore in general.  The magic name 
> __init__ may be by far the one most commonly encountered in everyday 
> programming but __there__ __are__ __many__ __others__.

I guessed that :-) my rough table for common uses:

    a.__getattr__  ->  a's getattr
    a.__setattr__  ->  a's setattr 
    a.__delattr__  ->  a's delattr 
    a.__dict__     ->  a's instance/class dict
    a.__hash__     ->  a's hash function 
    a.__init__     ->  a's init method (thanks to aahz)
    a.__new__      ->  a's meta constructor (or so)
    a.__doc__      ->  a's docstring
    a.__call__     ->  a's call method

if any program dares to use these names without underscores
i handle *that* as the special more verbose case (e.g. non-underscored hash method).
the more esoteric __methods__ i like to spell 'the special xyz method'. 

under_score'ly y'rs holger




From loewis at informatik.hu-berlin.de  Tue Jun 18 07:33:32 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Jun 2002 13:33:32 +0200
Subject: unicode strings and strings mix
References: <mailman.1024393102.14302.python-list@python.org>
Message-ID: <j4vg8gx1mb.fsf@informatik.hu-berlin.de>

Roman Suzi <rnd at onego.ru> writes:

> What I wanted to ask is that it should be considered UNSAFE
> to mix strings and Unicode-strings without explicit {en,de}code
> methods, because the program must not depend on locale setting.

Indeed. Changing the system encoding is evil, for this precise reason.

> For example, a developer is using "ASCII" 128-255 for his national
> alphabet (Oleg pointed siteconfig.py) and writes a program which runs just
> fine for him. However, the same program will fail for almost everyone
> else.

Correct. The I18N people in Python are apparently divided as to how
serious the impact of such a change is. Oleg (and others) point that
out as a solution, and for some usages, this is indeed a
solution. Myself (and others) strongly advise against changing the
system encoding, since it means that you write a program that works
fine on one installation, but fails on another. Explicit is better
than implicit.

> I think, it is a source for very subtle errors in Python programs as it
> easily makes them not 8-bit clean!

The current default for the system encoding ("ascii") was chosen
precisely for the reason to prohibit such errors in the default Python
installation. ASCII only covers the bytes 0-127 (anything above that
is *not* part of the American Standard Code for Information
Interchange). 

ASCII happens to be a subset of most other encodings, so if you have a
string consisting only of characters below 128, you can most likely
convert it to Unicode correctly - that's why "ascii" is a reasonably
safe encoding as a default. There are exceptions, though - iso-2022-jp
could be mistaken for ASCII.

Regards,
Martin


From parker at gol.com  Sat Jun  8 07:45:51 2002
From: parker at gol.com (Ian Parker)
Date: Sat, 08 Jun 2002 11:45:51 GMT
Subject: ANN: PyChecker v0.8.11
References: <mailman.1023396733.8725.python-list@python.org>
Message-ID: <tNmi1rJp2eA9EwoA@hiredata.gol.com>

In article <mailman.1023396733.8725.python-list at python.org>, Neal 
Norwitz <neal at metaslash.com> writes
>A new version of PyChecker is available for your hacking pleasure.
>
>        PyChecker is a tool for finding bugs in Python source code.
>        It finds problems that are typically caught by a compiler for less
>        dynamic languages, like C and C++.  It is similar to lint.
>
>Comments, criticisms, new ideas, and other feedback is welcome.
>
>Changes from 0.8.10 to 0.8.11:
>
>  * Improve error message for syntax errors from user files
>  * Fix pychecker.bat so it should work now
>  * Add a warning for using __coerce__ in new-style classes
>  * Add --deprecated option for using deprecated modules or functions
>  * Add a warning for using functions with security problems (os.t[e]mpnam)
>  * Add a warning for comparing against True/False or defining True/False
>  * Add --badexcept option to warn when using string exceptions or
>        classes not derived from Exception to raise/catch exceptions
>  * Fix spurious warnings from using (test and 'true' or 'false)
>
>Various other bugs were also fixed, including some problems
>using PyChecker in Python 2.2.
>
>
>PyChecker is available on Source Forge:
>    Web page:           http://pychecker.sourceforge.net/
>    Project page:       http://sourceforge.net/projects/pychecker/
>    Mailling List:      pychecker-list at lists.sourceforge.net
>
>Neal
>--
>pychecker-list at lists.sourceforge.net
>
>

I'd just like to throw in that the improved error message format for 
syntax errors is particularly nice.

Regards

-- 
Ian Parker


From akineko at pacbell.net  Sun Jun  2 18:08:47 2002
From: akineko at pacbell.net (Aki Niimura)
Date: 2 Jun 2002 15:08:47 -0700
Subject: My message text won't fit in a line
Message-ID: <b714b8de.0206021408.5398d0ef@posting.google.com>

Hi,

I'm writing a Tkinter program. A dialog box (showwarning) is used
to show my message which is a little longer.

My message in the dialog box is split into two lines which makes my message
very difficult read.

First I thought I could fix this by setting width=?? attribute. However,
it didn't work as the widget I had was the dialog itself not the Label 
object I needed.

Is there any better idea to deal with this problem.
(Such as how to access a child widget of the Tk dialog) 
Changing the geometory seems not a solution as the text won't change
even I resize the dialog size.

Any info, suggestions are highly appreciated.

Best regards,
Aki Niimura


From t.bleicher at tu-bs.de  Wed Jun 26 15:39:36 2002
From: t.bleicher at tu-bs.de (Thomas Bleicher)
Date: Wed, 26 Jun 2002 21:39:36 +0200
Subject: Advice sought on ps module
References: <6e990e29.0206200325.355f0413@posting.google.com>
Message-ID: <slrnahk63n.oe.t.bleicher@w0076.willkommen.daheim>

In article <6e990e29.0206200325.355f0413 at posting.google.com>, 
Hans-Joachim Widmaier wrote:

> Examples:
>   - To calculate the with of a string, you have to read the AFM file,
> add the character widths, aplly kerning, transform it with the CTM and
> the fonts TM.
>   - To know whether a font must be reset or is already the current
> font you have to keep track of all the save/restore, gsave/grestore
> and the like.

piddle (http://piddle.sourceforge.net) can do some of this; i. e. you
can use a "StringWidth()" function on a string set in font XYZ. But you
might want to use the developement version (aka "spinge"); the results
seem to be more accurate (incorrect afm-data in older versions or
something).

piddle has other (simple) drawing commands, too.

Hth,
Thomas



From shagshag13 at yahoo.fr  Thu Jun  6 10:11:04 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Thu, 6 Jun 2002 16:11:04 +0200
Subject: check on numerical value?
References: <3cff56b4.511834453@news.skynet.be> <l5JL8.123663$cQ3.3519@sccrnsc01>
Message-ID: <adnqh7$ri6f$1@ID-146704.news.dfncis.de>

can't you use (after importing types) type(x) == types.FloatType (or type.IntType)

s13.




From michael at damaru.com  Sun Jun 16 10:34:42 2002
From: michael at damaru.com (Michael Davis)
Date: Sun, 16 Jun 2002 10:34:42 -0400
Subject: Snort alert tail...
References: <AB0P8.107$W91.886@news1.global-one.fi>
Message-ID: <Qj1P8.845$TL6.52697@news20.bellglobal.com>

Jan-Eric wrote:

> HEllo !
> I'm trying to write a dynamic firewall script in python that scans the
> Snort alert file like 'tail -f' and takes action based on the infomation
> it gets from that file. But I can't get the 'tail' function to work.It
> reads the file, but any new information that Snort is writing to the file
> doesn't show up to the script.
> 
> ex.
> file = open('/var/log/snort', 'r')
> while 1:
>         file = file.read()
>         print file
> ....
> 
> Any suggestions???
> Jan-Eric / Finland

Hi,

Why don't you try opening a pipe to the output of 'tail -f' ? I haven't 
tried this, but you could probably do:

import os

file = os.popen( 'tail -f /var/log/snort' )

etc.

-- 
Michael Davis
Damaru
Custom Programming - Web Development - Database Design
http://www.damaru.com
416-540-1284


From tjreedy at udel.edu  Mon Jun 10 12:33:58 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 10 Jun 2002 16:33:58 GMT
Subject: __imul__ broken for 'objects' in 2.2.1
References: <yfsd6v3qa86.fsf@black132.ex.ac.uk> <3n79guskvoaeqtr4srg11kaqse7jcqutt4@4ax.com> <yfswut76yat.fsf@black132.ex.ac.uk>
Message-ID: <Wx4N8.54515$Qg.4620223@bin6.nnrp.aus1.giganews.com>

"Alexander Schmolck" <a.schmolck at gmx.net> wrote in message
news:yfswut76yat.fsf at black132.ex.ac.uk...
> > PS It seems that a Breaky instance is not (incorrectly) considered
a
> > sequence in 2.3a0 (20020603).
>
> Nice to hear, but only a limited consolation since it quite some
time will
> pass before 2.3 comes out and is sufficiently widely adopted.

Let's hope (request?) that the fix gets backported to 2.2.2 and that
such appear sooner rather that later.






From a.clarke11 at pop.ntlworld.com  Thu Jun 27 18:34:36 2002
From: a.clarke11 at pop.ntlworld.com (a.clarke11)
Date: Thu, 27 Jun 2002 23:34:36 +0100
Subject: Is Python growing?
Message-ID: <3D1B92FB.59B590A0@pop.ntlworld.com>

I love this language, don't know any other, and use it for all sorts of
things. Thanks Guido and all you guys for the work you do.
My question is, I use 2.1 and by the time I need to update, probably the
version number will have reached Python 3.xx. Will that version have the
inherent simplicity that makes 2.1 so usable? Or will it become more
complex, and lose some of its original ease of use? By more complex, I
mean will it have a steeper learning curve for newbies, will it be
slower, will the download take longer, and will the featurelist be
significantly longer?
Other things develop until they lose their function, eg flowers,
saplings, clouds: can Python avoid increasing complexity, is that the
present trend in development?

Just wondered
Tony Clarke



From peter at engcorp.com  Tue Jun  4 20:29:20 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 04 Jun 2002 20:29:20 -0400
Subject: Pop a list from beginning ? and memory saving...
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de>
		<3CFA0676.68F80DE6@engcorp.com>
		<add289$vud42$1@ID-146704.news.dfncis.de>
		<Xns9221CE2AF76D0cliechtigmxnet@62.2.16.82>
		<adhrsl$118kvn$1@ID-146704.news.dfncis.de> <mailman.1023194347.14669.python-list@python.org>
Message-ID: <3CFD5B60.90831F21@engcorp.com>

Fran?ois Pinard wrote:
> 
> [Shagshag13]
> 
> >  def reset(self):
> >   del self._v
> >   self._v = []
> 
> Just for the record, one could also write, in one line:
> 
>     self._v[:] = []
> 
> Probably unneeded in this case, but it is useful when one wants to keep
> extra references to the list, and empty it nevertheless.

I can't parse that.  What does it do?

-Peter


From peter at engcorp.com  Sun Jun  9 02:59:11 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 09 Jun 2002 02:59:11 -0400
Subject: Efficient python programming...
References: <mailman.1023452786.27923.python-list@python.org> <3D013FB1.3D7477EA@engcorp.com> <83y9dp43d0.fsf@panacea.canonical.org>
Message-ID: <3D02FCBF.EEA954C4@engcorp.com>

Kragen Sitaker wrote:
> 
> Peter Hansen <peter at engcorp.com> writes:
> > Roman Suzi wrote:
> > > To prove that code is correct is very difficult (practically impossible).
> > > Unittests are there to prove that you have errors, not the other way
> > > around.
> >
> > I disagree, if the tests are written first (and run, proving that
> > they will fail when the code is wrong or missing).
> 
> I think you are disagreeing because you are ignorant of formal methods.

No, you're wrong.  I'm familiar with formal methods.  I was actually
disagreeing because I've been writing these way too late at night 
without adequate sleep and my logic centres are deprived and my 
language centres are sluggish.

Suffice to say that in another thread I backed away from "tests
can prove anything" -- I was wrong.

What tests do, however, is vastly increase your confidence in the
correctness of the code, and when written properly (in advance
of each small morsel of code that is written, and having been
shown to fail without that code) they are sufficient to deliver
extremely high quality software in which you can have high 
confidence.  And, I believe, _necessary_ for the same.

-Peter


From marklists at mceahern.com  Tue Jun 11 10:23:21 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 11 Jun 2002 09:23:21 -0500
Subject: [OT] Antispamming the email address
In-Reply-To: <GCEDKONBLEFPPADDJCOEOECHDIAA.whisper@oz.net>
Message-ID: <NCBBLFCOHHDIKCAFGCFBIEAALEAA.marklists@mceahern.com>

And then there's tmda, of course, which is written in Python, iirc:

	http://tmda.sf.net/

// m
-





From Chris.Barker at noaa.gov  Tue Jun  4 13:10:27 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 04 Jun 2002 10:10:27 -0700
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com>
Message-ID: <3CFCF3E7.AE891EBE@noaa.gov>

John Roth wrote:
> And I disagree that assignment implements two distinct operations.
> Assignment always rebinds. Always.

Yes, but += is NOT Always an assignment operator, it is sometimes an
assignment operator.

> In the examples cited, it
> attempted to rebind into a tuple, because that's what was asked
> for when the left hand side had a subscript.

What was asked for was to mutate an element of a tuple, which is
perfectly legal in other cases:

>>> t[1].extend([3])
>>> t
([0], [1, 3], [2])

in fact, list += list is equivalent to list += list, as far as I can tell.
 
> The reason that the examples given at the top of the thread fail
> in such strange ways is simply that '+=' does not check whether
> the assignment is possible before doing the addition (or concatination,
> in the specific cases cited.)

It shouldn't do the assignment at all, as it isn't necessary, but if
that is too difficult, it should certainly check.
 
> To see this, you could try the first example, using explicit '+' and
> '='. I suspect it would fail in exactly the same way, with exactly
> the same side effect.

No need to suspect:
>>> t = ([0],[1],[2])
>>> t[1] = t[1] + [3]
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: object doesn't support item assignment
>>> t
([0], [1], [2])

So, it fails, but with no side effect, as it should.

Huaiyu Zhu wrote:
> Great explanation!  The problem stems from the fact that the current +=
> operator implements two semantically distinct operations, chosen according
> to some properties of the operands.

exactly, and it wa kind of a weird decision. Personally, I feel the only
compelling reason to add The += (and friends) operator was to support
in-place addition, most notably for Numeric arrays. Frankly it ought to
just be disallowed for non-mutable types, but all the folks that want it
for integers would scream. to be honest, I've kind of lost track of why
it is a good idea for numbers to be immutable, but that's another topic!

> This is not unlike the situation of the
> division operator with distinct behaviors on floats and ints.

I suppose that is a pretty good parallel, but not quite the same, as the
behavior of += is much more similar for the different types: most of the
time it doesn't matter if the object is changes in pace or not, the main
reason I want it is for performance with Numeric arrays. Of course,
being able to change a mutable object in place inside a tuple is nice
too...oops, that doesn't work!

> I'd imagine that at some future time there would be a desire to split += to
> two distinct operators, one for true in-place operator and one for the fake
> one currently implemented for immutables.  That would remove the spurious
> exceptions for mutable members of immutable structures.

Do we really need a new notation? is a = a + b so bad? In fact, are the
current semantics of += s bad, or should this bug just be fixed?

> That's not unlike the difficulty involved with the division operator change.
> Maybe it's better to change earlier?

If there is going to be a change, the sooner the better!!

Gustavo Cordova wrote:
> The thing is that += MUST work also for immutable objects,
> as well as mutables.

That's arguable, but it how it is done now...

> So, it does a rebind of the final
> value obtained --which in a list or dictionary, being mutable,
> is the same as the initial reference, but in the case of
> immutable objects, it's a different one than before--.

couldn't the re-bind happen only if required?
 
>From the reference docs:

These methods are called to implement the augmented arithmetic
operations (+=, -=, *=, /=, %=, **=, <<=, >>=, &=, =, |=). These methods
should attempt to do the operation in-place (modifying self) and return
the result (which could be, but does not have to be, self). If a
specific method is not defined, the augmented operation falls back to
the normal methods. For instance, to evaluate the expression x+=y, where
x is an instance of a class that has an __iadd__() method, x.__iadd__(y)
is called. If x is an instance of a class that does not define a
__iadd() method, x.__add__(y) and y.__radd__(x) are considered, as with
the evaluation of x+y. 

So, if you simply do not define a set of __i***__ methods for
non-mutable types, then the interpreter would do "the right thing", and
re-binding would not have to happen for mutable types. I no so little
about Python byte code generating and interpreting, that I'm sure I
missing something, but it certainly looks like it could be done.

> So, it's not actually a bug, but an artifact of the hoops
> Python needs to jump through in order to maintain total
> dynamism in data types.

It certainly is a bug. All bugs are artifacts of something! IF not a
bug, I suppose a MAJOR wart.

> Instead of this:
> 
>    a = ([1],[2],[3])
>    a[0] += [3]
> 
> do this:
> 
>    a = ([1],[2],[3])
>    t = a[0]
>    t += [3]

why not just:
a = ([1],[2],[3])
try:
	a[0] += [3]
except TypeError:
	pass

Because that would work too. And you don't think it's a bug?

> It's not *totally* equivalent, because it's missing a
> single, last step:
> 
>    a[0] = t
> 
> *that* is what's raising the exception, which is what
> the += operator is trying to do in you're original version.

Right, but it shouldn't be, clear and simple. += means "add in place, if
possible" it is possible, there should be no error.

I'm in no position to figure out how this should be solved, but it
should be. I'm dumbfounded that folks seem to think that this is not a problem!

me:     Doctor, it hurts when I do this.
Doctor: Well, then don't do that!

By the way, the += and friends operators also cause similar confusion
with name scopes in functions, += is assumed to be a re-bind, when it
isn't necessarily intended to:

>>> l = [1,2,3]
>>> def test():
...     l += [4]
... 
>>> test()
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "<stdin>", line 2, in test
UnboundLocalError: local variable 'l' referenced before assignment

but:

>>> def test2():
...     l.extend([4])

>>> test2()
>>> l
[1, 2, 3, 5]

Works just fine. Perhaps  Huaiyu is right, we need different operators.
However, then the operator for incrementing a non-mutable would just be
a synonym for a = a + b, which looks like unnecessary syntactical sugar
to me.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From nookie at NOSPA_M.online.no  Sat Jun 22 10:42:28 2002
From: nookie at NOSPA_M.online.no (Erlend J. Leiknes)
Date: Sat, 22 Jun 2002 14:42:28 GMT
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at>
Message-ID: <3D148CD0.4050008@NOSPA_M.online.no>

Siegfried Gonzi wrote:
> A few weeks ago I complaint here that Python on Windows XP sucks. I had
> a problem which would took theoretically about 1.5 hours; but in reality
> it takes 6 hours on Windows XP and Python 2.2. After that 6 hours I must
> reboot my system in order to work on.
> 
> In the following I got the advice to install a real user system on my
> laptop. Okay, I bought Linux SuSE 8.0 and installed it on my laptop (256
> MB RAM, 1000 MHz Celeron). I also installed Python 2.2 (the ActivePython
> version, otherwise I couldn't install idle due to the annoying
> "free_software_garbage_2_weeks_installation_nightmare").

What do you mean with
"free_software_garbage_2_weeks_installation_nightmare",
Ive never had any problem installing python/idle...


> 
> SWIG helped me to make my external Python C function (with gcc); and
> F2PY helped me to create my external Python Fortran 90 function (with
> the Intel Fortran 95 compiler).

Perhaps it wasnt python's fault?
Perhaps it was swig or f2py?


> 
> The directories and all the other programs and data are the same as on
> Windows.
> 
> 
> 1. After starting the process from within idle the complete system
> freezes after the first wavelength (after about 20 minutes). I couldn't
> kill a process or open a different session (via F2) in order to kill
> python. The last resort: turn off the power supply.
> 
> 2. I thought it is better to start the process from the command line and
> in a X terminal: python forcing.py
> The same behavior as aforementioned: last resort: power supply. That is
> really good and fine because a Linux system really is greedy for "shut
> down immediately".
> 
> 3. I thought it has to do with Gnome and I logged in to a KDE session.
> Same behavior as in point 1. and 2.
> 
> 
> My conclusion: I will rewrite my simulation in Fortran 90 and will
> abadon Python. Python is maybe good for some small scripts but not
> appropiate for serious programming.
> 
> Thank you Billy Boy for a forgiving user system Windows XP which in turn
> relieves Python's "wreak havoc" behavior.
> 
> I cannot believe that Linux is responsible for Python's ill behavior
> under Linux. Before you start ranting about note the following: In a
> first step it has nothing to do with my programming style (admittedly
> not very good and famous) it has to do that the same calculation cannot
> be performed on Linux and Python 2.2!
> 
> 
> S. Gonzi




From just at xs4all.nl  Wed Jun 19 16:39:53 2002
From: just at xs4all.nl (Just)
Date: Wed, 19 Jun 2002 22:39:53 +0200
Subject: OS X debugger?
References: <aeqotj$dm$1@panix1.panix.com>
Message-ID: <just-BF33DB.22395319062002@news1.xs4all.nl>

In article <aeqotj$dm$1 at panix1.panix.com>, aahz at pythoncraft.com (Aahz) 
wrote:

> Anyone know a good visual debugger for OS X?  I'd originally been
> planning to just use shell windows for the demos of my OSCON tutorial,
> but it occurs to me that a single-stepping debugger with watch variables
> would be even better.

There's a fairly basic one in in MacPython IDE. This IDE apparently can 
also be made to work with a "macho" Framework build of Python, but 
that's not yet supported out-of-the box. Let me know off line if you 
need help with it (you can blame me for most everything that's wrong 
with it ;-), or CC me if you follow up to the group, as I read c.l.py 
fairly superficially these days... Unfortunately the IDE in the 2.2.1 
MacPython release still wants Mac-style line endings (returns), but you 
could grab the sources from CVS to fix that.

Just


From fredrik at pythonware.com  Sun Jun 23 18:41:40 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 23 Jun 2002 22:41:40 GMT
Subject: Python hits the spot
References: <mailman.1024868430.3289.python-list@python.org>
Message-ID: <E8sR8.44838$n4.10520620@newsc.telia.net>

Tim Peters wrote:

> But when the OS+libc refuse to give any hint of trouble, and themselves go
> nuts later, I don't accept that as being Python's problem.  As noted, more
> recent versions of Linux kill the process (well, kill *some* process(es),
> probably including at least this one) instead of going completely insane.
> That sounds like a misfeature to me too, but it's outside of Python's
> control.

for some reason, the python-dev archives contain an old
post by Linus Torvalds on this topic:

http://mail.python.org/pipermail/python-dev/1995-March/008945.html

    "Take the good with the bad: there is no free lunch.
    And please, don't WHINE."

</F>




From whisper at oz.net  Thu Jun 13 23:50:22 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 13 Jun 2002 20:50:22 -0700
Subject: More Tkinter Help please...
In-Reply-To: <B92EC762.70EC%sarmstrong13@mac.com>
Message-ID: <GCEDKONBLEFPPADDJCOEEEANDJAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of SA
> Sent: Thursday, June 13, 2002 20:04
> To: python-list at python.org
> Subject: More Tkinter Help please...
>
>
> Hi Everyone-
>
>     I have another Tkinter question for you. I'm still trying to translate
> "that previously mentioned" Tcl/Tk program into Python/Tkinter. The
> following is all of the code I've managed so far:
>
> from Tkinter import *
> import sys
> import os
>
> class PyShell:
>     def __init__(self, top):
>
>         f = Frame(top)
>         f.pack()
>
>         self.t1 = Text(f, height="12", width="84", font="Courier 12")
>         self.t1.pack(side=TOP, pady=2)
>
>         self.t2 = Text(f, height="12", width="84", bg="lightblue",
> font="Courier 12")
>         self.t2.pack(side=TOP, pady=2)
>
>         self.b1 = Button(f, text="Execute", command=self.expyth)
>         self.b1.pack(side=LEFT)
>
>         self.b2 = Button(f, text="Clear Input", command=self.clearin)
>         self.b2.pack(side=LEFT)
>
>         self.b3 = Button(f, text="Clear Output", command=self.clearout)
>         self.b3.pack(side=LEFT)
>
>         self.b4 = Button(f, text="Save Input", command=self.savin)
>         self.b4.pack(side=LEFT)
>
>     def clearin(self):
>         self.t1.delete(0,END)
>
>     def clearout(self):
>         self.t2.delete(0,END)
>
>     def expyth(self):
>         output = os.popen("python -c").t1()
>         self.t2.delete(0,END)
>         sys.stdout.write.t2(output)
>
>     def savin(self):
>         pass
>
> root = Tk()
>
> app = PyShell(root)
>
> root.mainloop()
>
> When I run this I get the following error:
>
> Traceback (most recent call last):
>   File "PyShell.py", line 45, in ?
>     app = PyShell(root)
>   File "PyShell.py", line 17, in __init__
>     self.b1 = Button(f, text="Execute", command=self.expyth)
> AttributeError: PyShell instance has no attribute 'expyth'
>
>
> Any ideas on what I'm doing wrong here?
>
> Thanks.
> SA
>
> --
> http://mail.python.org/mailman/listinfo/python-list


Code comes up for me, but has other bugs I'm not able to diagnose.

FWIW, if you're trying to create a python shell, are you aware of PyCrust on
Sourceforge? If nothing else, you might get some pointers out of it :)

Dave LeBlanc
Seattle, WA USA





From pierpaolo.glave at ciaolab.com  Wed Jun 26 04:26:02 2002
From: pierpaolo.glave at ciaolab.com (Pier Paolo Glave)
Date: 26 Jun 2002 01:26:02 -0700
Subject: Embedding in Tkinter (reprise)
References: <1907a5e4.0206240803.6ca0a4c@posting.google.com> <mailman.1024936355.12699.python-list@python.org> <1907a5e4.0206250504.4b417e9a@posting.google.com> <mailman.1025012704.15005.python-list@python.org>
Message-ID: <1907a5e4.0206260026.6a2fd764@posting.google.com>

Thank you Jeff, your code worked at first launch!
(except only for "frame" which had to be capitalized ;-)

Bye
--
Pier Paolo Glave

Jeff Epler <jepler at unpythonic.net> wrote in message news:<mailman.1025012704.15005.python-list at python.org>...
> You first need to make sure your Python's tkinter is linked with blt.
> This can be a bit of a mess ...
> 
> Then, guessing, you could start by trying to define a Container class
> in Python, something like
>     class Container(Widget):
> 	    def __init__(self, master=None, cnf={}, **kw):
> 		    Widget.__init__(self, master, 'blt::container', cnf, kw)
> 
> Now, the program would look something like
>     t = Tk()
>     t.tk.call("package", "require", "BLT")
>     f = frame(t)
>     os.system("xclock &")
>     c = Container(f)
>     f.pack(fill="both", expand="yes")
>     c.pack(fill="both", expand="yes", padx=".1i", pady=".1i")
>     c.configure(relief="raised", bd=2, name="xclock")
>     t.mainloop()
> 
> Jeff


From martin at v.loewis.de  Wed Jun  5 13:55:00 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 05 Jun 2002 19:55:00 +0200
Subject: Urllib thread safe?
References: <mailman.1023280808.19305.python-list@python.org>
Message-ID: <m3it4xobkb.fsf@mira.informatik.hu-berlin.de>

Marcel Opsteegh <Marcel.Opsteegh at nob.nl> writes:

> Does somebody know whether the urllib is thread safe? 
> I use the following call: urllib.URLopener().open(...)

Yes, that is thread-safe. urllib does not maintain any per-url state,
nor does any of the underlying libraries.

Regards,
Martin



From emile at fenx.com  Mon Jun  3 09:48:56 2002
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 03 Jun 2002 13:48:56 GMT
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <06da01c208f4$d69003c0$6601a8c0@boostconsulting.com> <200206010121.g511LxX19223@pcp742651pcs.reston01.va.comcast.net> <073501c20960$23f367e0$6601a8c0@boostconsulting.com> <200206011334.g51DY1D21669@pcp742651pcs.reston01.va.comcast.net> <07d801c20970$cdebe050$6601a8c0@boostconsulting.com> <mailman.1022940122.3421.python-list@python.org> <Id5K8.171932$Po6.340972@rwcrnsc52.ops.asp.att.net> <oWIK8.179611$%u2.31494@atlpnn01.usenetserver.com> <cHJK8.181202$%u2.93890@atlpnn01.usenetserver.com>
Message-ID: <ctKK8.185481$Po6.352104@rwcrnsc52.ops.asp.att.net>

Steve Holden
> Steve Holden
> > Why do you consider either of these examples wrong? The whole point
of
> > tuples is that if d is a dictionary, then d[t1] == d[t2] if, and
only if,
> t1
> > == t2. Both of these examples are trying to make changes that will
modify
> an
> > (immutable) tuple.

No.  They're using a reference to a mutable held in a tuple to modify
the mutable in-place.

> >There's every reason to treat these modifications as
> > illegal, since a before the fiddling is not equal to a after the
fiddling,

This is true in the case of re-assignment, as in t[0]=t[0]+1, but
shouldn't be in the case of in-place using += which is why I'd consider
it wrong.  Moreover, the first example completes the change *and*
produces a traceback!

> > but tuples aren't mutable so such changes are illegal.

The tuple shouldn't be modified.  The object pointed to by a reference
held in the tuple should.

> >
> > regards
> > --
>
> ----------------------------------------------------------------------
-
> > Steve Holden
http://www.holdenweb.com/
> > Python Web Programming
http://pydish.holdenweb.com/pwp/
>
> ----------------------------------------------------------------------
-
> >
> The dictionary example is somewhat misleading (because my assertion is
> wrong). However, I stand by the statement that you are both trying to
modify
> tuples, which is fundamentally against the nature of the type.
>

Why is changing an underlying mutable object using a reference held in a
tuple against the nature of a tuple?  It works just fine like this:
>>> t = ([1],[2],[3])
>>> t[1][0]=4
>>> t
([1], [4], [3])
>>> t[1][0]=t[1][0]+4
>>> t
([1], [8], [3])
>>> t[1][0]+=4
>>> t
([1], [12], [3])

Except the in-place extension of a list:

>>> t[1][0]=t[1][0]+[2]
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: unsupported operand types for +
>>>

But, in-place extensions of lists works without changing the id's:

>>> t1 = [1,2,3]
>>> id(t1)
8282220
>>> t1+=[4]
>>> t1
[1, 2, 3, 4]
>>> id(t1)
8282220
>>>

Why should the source container of a reference change the behavior of an
object?

>>> t = ([1],[2],[3])
>>> t1 = t[0]
>>> t1 += [22]
>>> t
([1, 22], [2], [3])
>>> t[0][0]+=[222]
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: unsupported operand type(s) for +=
>>>

--

Emile van Sebille
emile at fenx.com

---------



From sholden at holdenweb.com  Mon Jun  3 10:29:57 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 3 Jun 2002 10:29:57 -0400
Subject: HTML links in a GUI tkinter
References: <mailman.1022986323.16196.python-list@python.org> <83elfprs1x.fsf@panacea.canonical.org> <C7JK8.180067$%u2.89342@atlpnn01.usenetserver.com> <adfqa4$pnf$1@wanadoo.fr>
Message-ID: <%7LK8.185009$%u2.41439@atlpnn01.usenetserver.com>

"Eric Brunel" <eric.brunel at pragmadev.com> wrote in message
news:adfqa4$pnf$1 at wanadoo.fr...
> Steve Holden wrote:
> > So, all Aurel needs to know now is how to run wxPython widgets inside
> > tkinter programs...
>
> Why? Tkinter has everything to do it:
>
[ ... ]
I cancelled the posting, but of course not quickly enough...

regards
 Steve





From "aglyportat\" at n-o.spa__mmm.yahoo dott com>  Fri Jun 14 13:37:22 2002
From: "aglyportat\" at n-o.spa__mmm.yahoo dott com> (Anton Graph)
Date: Fri, 14 Jun 2002 10:37:22 -0700
Subject: PyArg_Parse() & review for memory leaks
Message-ID: <ugk9amdu63ec36@corp.supernews.com>

int
dumptag(PyObject *layout, char *prndriver, char *extra, char **res, int &len
	) {
	PyObject *args = Py_BuildValue("(Oss)", layout, prndriver, extra);
	if(!args)
		l_fatal("args build failed");

	PyObject *result = PyEval_CallObject(dumptag_callback, args);
	Py_DECREF(args);
	if( !result ) {
		if (PyErr_Occurred()) {
			PyErr_Print();
			l_fatal("error occured while executing dumptag()");
		}
		l_fatal("return value is expected from dumptag()");
		return 0;
	}
	char *pstr;
	if(!PyArg_Parse(result, "s", &pstr)) { // the plug-in just
		// builds a string containing tag
		return 0;
	}
	int st=strlen(pstr);
	if(st>=len) {
		delete [] *res;
		len=st+1;
		*res = new char[len];
	}
	astrncpy(*res, pstr, len);
	Py_DECREF(result);
	return st;
}

do I leak pstr here? Sample code I've googled up seems similar, but I 
don't understand how does PyArg_Parse allocate enough space without 
going to the heap?



From shalehperry at attbi.com  Tue Jun  4 14:49:04 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 04 Jun 2002 11:49:04 -0700 (PDT)
Subject: self
In-Reply-To: <TL7L8.79$4p7.4100181@newnews.cc.stevens-tech.edu>
Message-ID: <XFMail.20020604114904.shalehperry@attbi.com>

> As you can see, equations get added on the fly and the parsing gets done as
> the attribute is being called.  Now, notice that ugly self. thing in the
> equations.  Instead of just typing >>>A.f='a*c' I have to add self in front
> a and c to achieve the desired effect.  I could not find any other way to
> implement the class in order to avoid typing self. which of course it
> doesn't mean that there isn't any. If anybody has an idea how to do this
> please let me know.  Otherwise, here is the situation where the self
> concept is bad and having such a dynamic language like Python I think there
> has to be a way out of this problem.
> 
> Why is this all relevant? Well, I am working on embedding an interpreted
> language in a large engineering application and Python is  one of few
> candidates.  So I have been looking for its pros and cons.  So far it passed
> all the tests except the above.  Because, imagine if we required from the
> user of our system that he needs to type self. or even s. in front of every
> new variable that he has in his equation, it would be a big turn off.
> 

why can't your class be an interpreter for math expressions?

it sees 'a*c' and asks it own dict 'do we know a?' if so it access dict['a']
and returns its value.  Looking at your example you seem to be implementing a
symbolic calculator.  Most implementations I have seen involve writing a small
language with interpreter.




From tdelaney at avaya.com  Thu Jun 13 22:53:03 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 14 Jun 2002 12:53:03 +1000
Subject: filter2
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A38B@natasha.auslabs.avaya.com>

Designed to minimise external factors ...

def filter2 (test, l):
    l1 = filter(test, l)
    l2 = filter(lambda x, test=test: not test(x), l)
    return l1, l2

def filter2t (test, l):
    l1 = []
    l2 = []
    funcs = (l1.append, l2.append,)
    map(lambda x: funcs[not test(x)](x), l)
    return l1, l2

def filter2l (test, l):
    l1 = []
    l2 = []
    funcs = [l1.append, l2.append,]
    map(lambda x: funcs[not test(x)](x), l)
    return l1, l2

def filter3t (test, l):
    l1 = []
    l2 = []
    map(lambda x, test=test, funcs=(l1.append, l2.append,): funcs[not
test(x)](x), l)
    return l1, l2

def filter3l (test, l):
    l1 = []
    l2 = []
    map(lambda x, test=test, funcs=[l1.append, l2.append,]: funcs[not
test(x)](x), l)
    return l1, l2

def test (x):
    return not (x % 3)    

def bench (f, r, l, t=test):
    """"""
    import time
    clock = time.clock

    try:
        start = clock()
        for i in r:
            f(t, l)
        t = clock() - start
        print '%-10s %s' % (func.__name__ + ':', t,)
    except NameError:
        pass

r = range(1000)
l = range(1000)
filters = (filter2, filter2t, filter2l, filter3t, filter3l,)

for func in filters:
    bench(func, r, l)

Python 1.5.2 ...

filter2:   5.0598633013
filter3t:  4.76191299505
filter3l:  4.37944378728

Python 2.2.1 ...

filter2:   5.40253108615
filter2t:  4.38322776899
filter2l:  4.40173851257
filter3t:  4.32153526376
filter3l:  4.26489291133

filter3l is the fastest in both cases. filter3l also has the advantage of
working in any Python version.

Note: in filter3t and filter3l I original defined a nested function, but
this made the function perform worse than filter2t and filter2l - presumably
due to needing to bind the function name, then look it up for the map call.

Tim Delaney




From aoeu at go.com  Mon Jun 17 06:56:16 2002
From: aoeu at go.com (Thinkit)
Date: 17 Jun 2002 03:56:16 -0700
Subject: Getting pixel values under cursor
References: <adrhgi$mh$1@nntp1.jpl.nasa.gov>
Message-ID: <3eeda89d.0206170256.4ba75ec@posting.google.com>

"Bill" <metaliu at yahoo.com> wrote in message news:<adrhgi$mh$1 at nntp1.jpl.nasa.gov>...
> I'm using Tkinter right now for a GUI.  Is there a way for me to get, in
> real time or when the user clicks the mouse button on a pixel, the pixel
> intensity value?  Currently, I have a tkinter label which is showing a large
> grey scale image.  I would like the user to move the mouse cursor over the
> image and show the pixel intensity under the cursor.  Is there a way to do
> this?  Perhaps some kind of event binding with the mouse and the image?
> Would putting the image in a canvas instead provide more flexibility?
> 
> Thanks in advance,
> Bill

Changing the cursor repeatedly will take some considerable background
processing.  Just bind mouse moves to check to see if it's within the
label--if it is, display in a status box, if not, display nothing
there.


From thenault at nerim.net  Thu Jun  6 12:07:50 2002
From: thenault at nerim.net (Sylvain Thenault)
Date: 06 Jun 2002 18:07:50 +0200
Subject: problem with isinstance and relative/absolute import
Message-ID: <87adq8jsq1.fsf@tux.ntw23.fr>

does someone can explain to me the following test case, or should it be
considered as a bug ?

[syt at tux syt]$ ls tests/
__init__.py  module2.py  module.py
[syt at tux syt]$ cat tests/__init__.py 
[syt at tux syt]$ cat tests/module.py   
class A: pass

a = A()

[syt at tux syt]$ cat tests/module2.py 
from module import A

b = A()

[syt at tux syt]$ export PYTHONPATH=$PWD
[syt at tux syt]$ cd tests
[syt at tux tests]$ python
Python 2.1.3 (#1, Apr 20 2002, 10:14:34) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> from module2 import b
>>> from tests.module import A, a
>>> isinstance(a, A)
1
>>> isinstance(b, A)
0
>>> 
[syt at tux tests]$ python2.2
Python 2.2.1 (#1, Apr 21 2002, 08:38:44) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> from module2 import b
>>> from tests.module import A,a
>>> isinstance(a,A)
1
>>> isinstance(b,A)
0
>>> 
[syt at tux tests]$ python
Python 2.1.3 (#1, Apr 20 2002, 10:14:34) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> from tests.module import A, a
>>> from tests.module2 import b   
>>> isinstance(b,A)
1
>>> 

TIA

-- 
Sylvain Th?nault 


From carel.fellinger at iae.nl  Fri Jun 28 22:36:23 2002
From: carel.fellinger at iae.nl (Carel Fellinger)
Date: Sat, 29 Jun 2002 04:36:23 +0200
Subject: Why the colon?
In-Reply-To: <4.3.2.7.2.20020628190712.02450328@mail.gte.net>
References: <XFMail.20020628153929.pedrobados@eresmas.net> <4.3.2.7.2.20020628190712.02450328@mail.gte.net>
Message-ID: <20020629023623.GA2953@schaduw.felnet>

On Fri, Jun 28, 2002 at 07:33:31PM -0400, Erv Young wrote:

...coming from C++ to Python in a mere four days

Welcome here, you sure did the right thing:)

...
> where I have added line numbers for the sake of the discussion.  I 
> understand the colons in the middle of lines 4, 5, and 7 as statement 
> separators.  Makes perfect sense.
> 
> Then what am I to make of the colons at the end of lines 1 and 
> 2?  "Yoo-hoo, interpreter!  I'm about to start indenting now!"  Hmmm....

You're quit right, the interpreter doesn't need those colons, but user
studies for Python's predecessor ABC indicated that we humans do (or
atleast prefer, on average that is:).  So it's there for the most
Pythonic of all Pythonic virtues: readability.


-- 
groetjes, carel





From mbar at sgi.com  Thu Jun 20 17:56:52 2002
From: mbar at sgi.com (Matthew Barker)
Date: Thu, 20 Jun 2002 14:56:52 -0700
Subject: you guys needs a logo
References: <7fe97cc4.0206142325.4df3a174@posting.google.com> <3eeda89d.0206170232.675596b8@posting.google.com> <3D0E8916.65AC2F13@engcorp.com>
Message-ID: <aetj2s$2nc5i$1@fido.engr.sgi.com>

Peter Hansen wrote:

> Thinkit wrote:
>> 
>> xah at xahlee.org (Xah Lee) wrote in message
>> news:<7fe97cc4.0206142325.4df3a174 at posting.google.com>...
>> > you guys need a logo.
>> >
>> > the light-bulb is quite lame as far as logo goes.
>> >
>> > I suggest that the O'Reilly's python book cover a good choice. It
>> > should not be something funny and cute.
>> >
>> > The MacPython's some 16 Ton weight block is also a good logo.
>> >
>> > Perhaps you guys can run a logo competition.
>> >
>> >  Xah
>> >  xah at xahlee.org
>> >  http://xahlee.org/UnixResource_dir/complang.html#python
>> 
>> How about a penguin masturbating?  To show what else besides python linux
>> males do.
> 
> I think a logo with a penguin stroking its python would be _cute_.
> 
> Some might misinterpret it, however...
> 
Especially if the penguin was sitting on the python.  
I wonder if their flippers (flappers? wings? fins?) are long enough to 
reach...?  I guess some things are better left unexplored.  
With that I think I'll go void my pointer.



From ken at hotmail.com  Tue Jun  4 11:02:37 2002
From: ken at hotmail.com (Ken)
Date: Wed, 5 Jun 2002 01:02:37 +1000
Subject: Cookie question
Message-ID: <adikmf$1100j6$1@ID-49758.news.dfncis.de>

Do you set your cookie like this?
import Cookie
c = Cookie.SmartCookie()
c["UserID"] = string.rstrip(tusername)
c["Validated"] = "yes"
print c


And extract cookie like this?
if os.environ.has_key("HTTP_COOKIE"):
        strcookiex = os.environ["HTTP_COOKIE"]
        userid = extractCookieID(strcookiex,"UserID")
        validated = extractCookieID(strcookiex,"Validated")
--
def extractCookieID(strcookie,idx):
        strloc1 = string.find(strcookie,idx )
        strvalue = ""
        if strloc1 > -1:
                strvalue = strcookie[strloc1:]
                strloc1 = string.find(strcookie,"=")
                if strloc1 > 0:
                        strvalue = strvalue[strloc1 + 1:]
                strloc1 = string.find(strvalue,";")
                if strloc1 > 0:
                        strvalue = strvalue[0:strloc1]
        return strvalue


If I have many different cgi scripts checking UserID this way, are they all
refering to the same cookie?

If it is so, how come my other scripts can't read userid correctly? What did
I do wrong?

Also, where does the cookie get saved in my computer (using Win2000 Pro)? I
don't seem to find any cookie inside the Tempory Internet Files folder.

Thanks




From skip at pobox.com  Wed Jun 19 17:22:54 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 19 Jun 2002 16:22:54 -0500
Subject: procmail replacement in Python
In-Reply-To: <oq660froj9.fsf@carouge.sram.qc.ca>
References: <20020618235901.GA2197@lilith.my-fqdn.de>
        <oqit4grq03.fsf@carouge.sram.qc.ca>
        <20020619022110.GA6996@lilith.my-fqdn.de>
        <oq660froj9.fsf@carouge.sram.qc.ca>
Message-ID: <15632.63022.66547.868788@localhost.localdomain>

>>>>> "Fran?ois" == Fran?ois Pinard <pinard at iro.umontreal.ca> writes:

    Fran?ois> [Gerhard H?ring]

    >> I'm currently just using spamassassin (the best spam detector around,
    >> very configurable, but written in Perl ;-).

    Fran?ois> I wanted to peek at Spamassassin, which seems to have a good
    Fran?ois> reputation.  And now, you are trying to discourage me? :-)
    Fran?ois> It's really a strange phenomenon: more I work with Python,
    Fran?ois> more Perl gets unreadable :-)

    >> I'm also investingating Pyzor, as I found the Razor author to be too
    >> clueless (also the razor servers aren't open-source).

I strongly recommend SA.  I just trust Craig Hughes to get the Perl
right. ;-)  As for Razor, it seems fine (and I report spam that slips
through SA to Razor), but I don't use it because the network overhead is to
high.  (SA can automatically ask Razor for a verdict on a message.)

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html




From bounce at iwebrider.com  Wed Jun 26 11:33:17 2002
From: bounce at iwebrider.com (Iwebrider.com)
Date: Wed, 26 Jun 2002 11:33:17 -0400
Subject: Web Design For The Wood Industry
Message-ID: <E17NElW-0001o2-01@server769.dnslive.net>

Hi,
    I would like to invite you over to our new website.  http://www.iwebrider.com. 
We design websites for companies and organizations 
alike. We also do websites for individuals. We work with HTML, CGI, DHTML, JAVASCRIPT, 
JAVA Etc. Packages starting at $199.00. while at our site feel free to post a 
free classified ad. 

If you currently have a site, and would like to update it, or add some Interactivity 
to it, Like a guestbook, or auction or some  other interactive feature, we can 
help. We will help you in all  phases of your design project. If you come to 
some part of your project and need help, We are here. I look forward to answering 
any  questions that you may have.

Best Regards

Cobby
Meridan Design

---------------------------------------------------
This is not an unsolicited email. Your email address has been submitted to us 
via on our website.  If you do not wish to receive any future emails and would 
like to remove yourself from our mailing list, please reply to this e-mail with 
the words "UNSUBSCRIBE" in the subject
field.






From gerhard at bigfoot.de  Tue Jun  4 09:07:50 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 4 Jun 2002 13:07:50 GMT
Subject: Need advice on multithreading problem
Message-ID: <slrnafpeei.p8.gerhard@developer2.opus-gmbh.net>

This morning, I've tried to seriously use multithreading in Python for the
first time. So I'd like to get any comments on the design of my
proof-of-concept that I'll paste below at full length. Hope it's not too long
for usenet, but I don't want to throw my comments/docstrings away.

The problem is this: I have a C library with a function of the following
interface:

    def sqlite_exec(conn, sql, callback, arg1):

The only important parameter here is 'callback', it is a callback function that
you need to give to sqlite_exec. The callback function will be called with the
items sqlite_exec returns. If the callback functions returns a value other than
zero, sqlite_exec will stop its processing.

On the Python side, I wanted to force this C library call into a generator.
Which I did using a thread that's spawned for a new sqlite_exec call and
Queue.Queue.

The proof-of-concept *seems* to work, but there might be ugly errors hidden
still. Any comments are welcome.

Gerhard

--- begin code ---

"""
 This is a proof-of-concept for the implementation of a new alternative
 Cursor class that will exploit Python 2.2 generators and the SQLite
 sqlite_exec function with a custom callback.
 
 The point of this exercise is to provide a means to _not_ have to keep the
 entire result set in memory, but to return row by row to the caller on demand,
 using generators.
 
 Of course it will be slower than the current implementation. If nothing else,
 it's a nice programming exercise for me to learn how to handle multithreading
 reliably.
 """
 
 from __future__ import generators
 from Queue import Queue, Empty, Full
 import threading
 
 # Singleton to mark that the last element of a queue. If this is put into a
 # queue, the consumer will stop reading from the queue.
 END_OF_QUEUE = 1
 
 # SQLite's sqlite_exec function looks similar to this one:
 def sqlite_exec(conn, sql, callback, arg1):
     # Let's assume that SQLite would return 15 results
     for i in range(15):
         # According to the docs, if the callback returns something other
         # than 0, the query will be aborted.
         if callback(arg1, ["a", "b", i]) != 0:
             break
 
 class Producer(threading.Thread):
     """A separate producer thread that wraps sqlite_exec, provides all callback
     function for sqlite_exec and fills a queue with the result rows for the
     consumer to read from.""" 
     def __init__(self, *args, **kwargs):
         self.stopflag = 0
         self.queue = kwargs["queue"]
         self.conn = kwargs["conn"]
         self.sql = kwargs["sql"]
         del kwargs["queue"]
         del kwargs["conn"]
         del kwargs["sql"]
         threading.Thread.__init__(self, *args, **kwargs)
 
     def run(self):
         def callback(arg1, items):
             """Callback function for sqlite_exec."""
             if self.stopflag:
                 self.queue.put(END_OF_QUEUE, 1)
                 # We want to abort the query:
                 return 1
             else:
                 self.queue.put(items, 1)
             return 0
             
         sqlite_exec(self.conn, self.sql, callback, None)
         self.queue.put(END_OF_QUEUE, 1)
             
 class Cursor:
     def __init__(self):
         self.conn = None
         self.queue = None
         self.producer = None
         self.__invalidate()
 
     def __invalidate(self):
         """Invalidate is called when a new query is processed with execute or
         executemany or the cursor is closed or the cursor object gets out of
         scope."""
         self.current_sql = None
         if self.producer is not None:
             self.producer.stopflag = 1
 
             # Try to empty the queue, so the producer can insert its
             # END_OF_QUEUE and won't block forever.
             try:
                 while 1:
                     item = self.queue.get(0)
             except Empty, reason:
                 pass
 
     def execute(self, sql, *parms):
         # The query isn't actually processed at this point. It will only be
         # processed in the fetchone/fetchmany/fetchall methods, using __execute
         self.__invalidate()
         self.current_sql = sql
 
     def __execute(self):
         """Build a queue and start the producer thread."""
 
         # We keep max 10 result rows in memory at any given point in time:
         self.queue = Queue(10)
 
         self.producer = Producer(conn=None, sql=self.current_sql, queue=self.queue)
         self.producer.start()
         self.current_sql = None
 
     def fetchall(self):
         if self.current_sql is not None:
             self.__execute()
 
         while 1:
             item = self.queue.get(1)
             if item is END_OF_QUEUE:
                 raise StopIteration
             yield item
 
     def fetchmany(self, howmany=15):
         if self.current_sql is not None:
             self.__execute()
         for i in range(howmany):
             item = self.queue.get(1)
             if item is END_OF_QUEUE:
                 raise StopIteration
             yield item
 
     def fetchone(self):
         if self.current_sql is not None:
             self.__execute()
         item = self.queue.get(1)
         if item is END_OF_QUEUE:
             return None
         else:
             return item
         
     def close(self):
         self.__invalidate()
 
     def __del__(self):
         self.__invalidate()
 
 c = Cursor()
 c.execute("select * from foo")
 print "--- fetchone() ---"
 print c.fetchone()
 print "--- fetchmany(5) ---"
 for res in c.fetchmany(5):
     print res
 print "--- fetchall() ---"
 for res in c.fetchall():
     print res
 c.close()
 
 



From dfackrell at DELETETHIS.linuxmail.org  Mon Jun 24 14:50:52 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Mon, 24 Jun 2002 12:50:52 -0600
Subject: How to stop a SocketServer?
References: <7xhejv9af2.fsf_-_@ruckus.brouhaha.com> <3d1744d4$1_1@hpb10302.boi.hp.com> <3D1757DF.776A428F@engcorp.com>
Message-ID: <3d176a0c$1_3@hpb10302.boi.hp.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3D1757DF.776A428F at engcorp.com...
> Daniel Fackrell wrote:
> >
> > I faced this problem with a simple chat server I wrote.  I expected that
> > termination of the server with <ctrl-C> should have stopped it
immediately,
> > but it instead waits until some communication is received from each
thread
> > that was stopped at a read().
> >
> > It appears that the current threading implementation does not provide
for
> > termination of threads inside a single python statement (such as a
> > filelike.read()), so I am considering a move to character I/O, checking
to
> > verify that a character is available before each read.
>
> The proper solution to this kind of problem is to use timeouts
> and select.  There's generally no need to check for the data
> being available ahead of time by _polling_, which is inefficient.
>
> -Peter

Any chance I could see a very simple example of this in action?  That would
help my little project out quite a bit.

Would this be something that would allow me to guarantee that the all
threads of the server terminated within, say, five seconds of the server
shutdown request?

--
Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.




From dfackrell at DELETETHIS.linuxmail.org  Thu Jun 13 10:18:19 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Thu, 13 Jun 2002 08:18:19 -0600
Subject: Bits and bytes?
References: <mailman.1023896057.21318.python-list@python.org> <3d08839a$0$226$4d4ebb8e@news.nl.uu.net>
Message-ID: <3d08a9ae$1_1@hpb10302.boi.hp.com>

Guyon,

In a case where the total output from your algorithm does not yield a size
evenly divisible by 8, you'll probably have to pad to the nearest byte
(generally done with 0 bits) before you can write to a file, as files these
days are always (?) stored as a multiple of 8 bits.  In fact, that generally
goes for data in memory as well.

I would avoid, however, any padding inside the bit stream, as you'll lose
effective compression.  Better to treat it as if there were no byte
boundaries at all until you need to write it to disk.

Maybe this, combined with the other ideas presented will help you out.

--
Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.

"Guyon Mor?e" <gumuz at looze.net> wrote in message
news:3d08839a$0$226$4d4ebb8e at news.nl.uu.net...
> ok, really a lot of thanx for your help, this is awesome! :D
>
> but, let me explain what i am actually trying to do here and why i think
> both of your solutions won't work for me.
>
> i am trying to implement my own sort of huffman encoding/compression.
> converting asci-characters to a compressed bitstream requires me to write
it
> at bit-level. Bill's solution might work in some way, but my sequence of
> bits aren't allways divideable by eight. a word like 'ape' (which is
3x8=24
> bits) might become 1001101, which is 7 bits.... catch my drift? strings is
> no option as this would have no positive effect on the 'real size'
;there's
> no compression.
>
> i hope this makes it a bit clearer for you. anyway, thanx a lot.
>
> guyon




From see_reply_address at something.invalid  Thu Jun 27 22:39:26 2002
From: see_reply_address at something.invalid (Greg Ewing)
Date: Fri, 28 Jun 2002 14:39:26 +1200
Subject: Pythoniac: Thoughts on a hardware Python processor
References: <mailman.1025212118.29030.python-list@python.org>
Message-ID: <3D1BCC5E.1040505@something.invalid>

David LeBlanc wrote:

> * Burning Python into silicon will tend to fossilize the language. Is Guido
> done enough with it for that yet? Guido and the gang can whip out a patch
> for the PVM in a matter of hours. How long would that take for a silicon
> change and what would it cost?


Leave it as an FPGA implementation and you can whip
out changes just as fast. (Well, faster than changing
silicon, anyway...)

-- 
Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From llchen223 at hotmail.com  Sun Jun 23 04:53:34 2002
From: llchen223 at hotmail.com (Lei Chen)
Date: 23 Jun 2002 01:53:34 -0700
Subject: The best GUI library in python
References: <8d3f4438.0206220457.b820418@posting.google.com>
Message-ID: <de6c123c.0206230053.2a04395@posting.google.com>

d2002xx at myrealbox.com (d2002xx) wrote in message news:<8d3f4438.0206220457.b820418 at posting.google.com>...
> Hello!!
> 
> I want to learn GUI in python, who would tell me which GUI library is
> the best? (don't tell me Tkinter, it's too slow......)

wxWindows is excellent. 

Go here: http://www.wxpython.org/

Cheers,
Lei


From pyth at devel.trillke.net  Tue Jun 11 07:47:54 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 11 Jun 2002 13:47:54 +0200
Subject: Defining a method final
In-Reply-To: <6glN8.35376$86.884025@twister1.libero.it>; from fcorneti@libero.it on Tue, Jun 11, 2002 at 11:35:30AM +0000
References: <6glN8.35376$86.884025@twister1.libero.it>
Message-ID: <20020611134754.Q6609@prim.han.de>

Fabio Corneti wrote:
> Is there any way to define a method final in Python?
> I've searched through the docs and through the web, but
> it seems there's no way to to such a thing using standard
> language statements. Some suggestions?

not that i know of.  unless you write it in C :-)

would you care to present your reasoning for doing a final method?

regards,

    holger




From mcfletch at rogers.com  Sun Jun 23 22:16:47 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Sun, 23 Jun 2002 22:16:47 -0400
Subject: [PyOpenGL]  Need for a C (lead) developer...
Message-ID: <3D16810F.6020109@rogers.com>

Summary:
	PyOpenGL (the Python binding to the OpenGL graphics library) is in need 
of a C developer (preferably with SWIG experience) to work on 
maintaining and extending the package.  Need is both short-term and 
long-term; we could use an immediate helping hand, but we're also 
looking for someone interested in long term work.

	PyOpenGL is a fairly large project (source-size), which is used by a 
considerable number of people (latest version has been downloaded 
something like 16,000 times).  Developers looking to take on a popular 
project and make a name for themselves would find this a good opportunity.


Current Tasks:
	PyOpenGL is currently generated using an old version of the SWIG wrapper 
generator (originally chosen because of missing features in other 
versions of SWIG with the idea that the missing features would 
eventually be adopted in newer SWIG versions (status of this unknown, 
but some work is definitely required to use the newer versions)).  This 
build process tends to create lots of problems for those porting to new 
architectures, and we'd like to move off the (old, beta) version soon. 
This is a fairly extensive project, and will require familiarising 
yourself with the code base before making the attempt.

	We have a number of cross-platform build issues that need to be 
coordinated and integrated.  In one case, we have already-submitted 
patches/fixes for fixing bug(s), in others, there are patches not yet 
submitted, but available.  We need these integrated and tested. 
Integration is the developer requirement, we can get users with the 
exotic architectures to test the fixed versions.

	We have reports of complete meltdown under OpenGL 1.3 implementations 
under Linux.  This is likely a trivial build or configuration problem, 
but it needs to be tracked down.


Future Tasks:
	Maintenance of the package.

	Creation of new wrappers for common OpenGL modules (optional).

	Simplification of the build system for exotic architectures.

	Insert your PyOpenGL dreams here.


Resources:
	The project has 3 administrators, myself (a designer with minimal (read 
non-existent) C coding skills), Tarn (author of the current code 
(current missing in action)) and Rene (who is/was primarily interested 
in the building and porting work).  I work primarily on the 
OpenGLContext side (which provides a series of tests for the base 
implementation, and tends to uncover errors as I develop it).  I tend to 
do the work on the web-site, and general answering of questions when I can.

	At the moment, we have no active C developers, and just myself working on 
the administrative front.  The PyOpenGL work I'm doing is minimal due to 
other projects and the fact that I'm largely unqualified for the work 
that needs to be done.  I am working on a way to use fonttools for 
rendering 3D text in PyOpenGL+OpenGLContext, but that's not going to 
help with the general coding that needs to get done.


How to Get Involved:
	If you'd be interested in working as a developer, let me know and I'll 
see about adding you to the developer list. People familiar with the 
PyOpenGL codebase and project would be especially welcome.

	http://pyopengl.sourceforge.net/

With thanks,
Mike Fletcher

_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From ljd at nospam.com  Thu Jun  6 10:17:12 2002
From: ljd at nospam.com (LJD)
Date: Thu, 06 Jun 2002 14:17:12 GMT
Subject: Using mxDateTime
Message-ID: <I9KL8.125444$ux5.179714@rwcrnsc51.ops.asp.att.net>

Hi:

I'm new to python and I have a need to use time/date functions.

I am running redhat 7.3 with the default python 1.52.  I just downloaded the 
appropriate mxDateTime RPM and installed it.  

How do I now make those mxDateTime functions available to me in my python 
program?  Since this is an extension module rather than a built in library, 
I assume I can't simply...

import mxDateTime

Your help is most appreciated!

Thanks,
Larry


From oren-py-l at hishome.net  Wed Jun 19 02:51:17 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Wed, 19 Jun 2002 02:51:17 -0400
Subject: __subclasses__
In-Reply-To: <20020618133206.GA19082@parque.homelinux.net>
References: <a9bcagho74.22t.jimrowe@optilink.com> <lk4rg6matg.fsf@pc150.maths.bris.ac.uk> <20020618133206.GA19082@parque.homelinux.net>
Message-ID: <20020619065117.GA22764@hishome.net>

> The __bases__ and __subclasses__ (and others) are listed in the 'type'
> object:
> 
>     >>> class Foo:
>     ...     pass
>     ...
>     >>> dir(type(type(Foo)))

Type is no longer a function, it is the type object itself.  
You can just dir(type)

	Oren




From gregor at hoffleit.de  Thu Jun  6 03:25:01 2002
From: gregor at hoffleit.de (Gregor Hoffleit)
Date: Thu, 6 Jun 2002 09:25:01 +0200
Subject: libpython as a shared library?
In-Reply-To: <Pine.OS2.4.32.0206051931120.1678-100000@tenring.andymac.org>
References: <20020604143402.GB22993@hal.mediasupervision.de> <Pine.OS2.4.32.0206051931120.1678-100000@tenring.andymac.org>
Message-ID: <20020606072501.GA6390@hal.mediasupervision.de>

* Andrew MacIntyre <andymac at bullseye.apana.org.au> [020605 21:15]:
> On Tue, 4 Jun 2002, Gregor Hoffleit wrote:
> 
> > Debian has had libpython as a shared lib since a while. Recently, we got
> > a report that the performance of the shared version is dramatically
> > slower than the statically linked version (e.g. about 30%). I made some
> > pystone tests on different architectures of our GNU/Linux, and tried
> > different flags for building and linking the library (and different
> > Python versions), and in fact there seems a significant speed penalty
> > involved as soon as the interpreter is linked against the shared
> > library. The penalty has some correlation to the architecture, but it is
> > significant on all tested constellations.
> 
> A result of PIC code in the shared library?  I've not been motivated
> enough to try building a shared lib on FreeBSD.

If I compile the sources with -fPIC, but put them in a static library,
there's nearly no speed difference to non-PIC code in a static library.
My best guess is that the speed penalty is caused by the indirection
step that's necessary when calling routines from a shared lib; but then,
that's no real explanation for a loss of 30% speed!

That's why I would be interested in a few datapoints from other systems.

    Gregor




From debl2nononespammyyy at bellatlantic.net  Sun Jun  9 13:59:28 2002
From: debl2nononespammyyy at bellatlantic.net (David Lees)
Date: Sun, 09 Jun 2002 17:59:28 GMT
Subject: Screen resolution again
Message-ID: <3D039794.9549C54@bellatlantic.net>

I am still trying to figure out how to automatically create a canvas
that is sized according to the screen resolution using tkinter on a
win95 box and python 2.2.1.  The hack below sort of does that, but I
would appreciate advice, suggestions or pointers to a url on how to do
it correctly.  Note that I create an extra canvas get the root height
and windth.

david lees
-------------------

from Tkinter import *
root = Tk()

root.title('Drawing')
canvas = Canvas(root,width=20,height=20,bg='white')
canvas.pack()

h = canvas.winfo_vrootheight()
w = canvas.winfo_vrootwidth()


print w,h
canvas2 = Canvas(root,width=w-30,height=h-30,bg='white')
canvas2.pack()

root.mainloop()


From nikolai at micon.no  Wed Jun 19 18:46:44 2002
From: nikolai at micon.no (Nikolai Kirsebom)
Date: Wed, 19 Jun 2002 22:46:44 GMT
Subject: Using calldll with mciSendString
References: <nNOP8.3825$yj7.62301@news4.ulv.nextra.no>
Message-ID: <oR7Q8.4291$yj7.68316@news4.ulv.nextra.no>

Got the first part working.  Problem was the name of function is not
"mciSendString" but "mciSendStringA".  Below part which works.  Still
working on the callback stuff.

---------------------------------------------------------
from dynwin import windll
winmm = windll.module('winmm')

def Send(s, cb=None):
    mciSendString=windll.calldll.get_proc_address(winmm.handle,
"mciSendStringA")
    bufIn=array.array('c', s + '\0')
    bufOut=array.array('c', '\0' * 128)
    ret=windll.calldll.call_foreign_function(mciSendString, 'wwLL', 'l',
                                             (bufIn, bufOut, 128L, 0L))
    s = bufOut.tostring()
    if s.index('\0') > 0:
        s = s[0:s.index('\0')]
    else:
        s = ""
    return (ret, s)






"Nikolai Kirsebom" <nikolai at micon.no> wrote in message
news:nNOP8.3825$yj7.62301 at news4.ulv.nextra.no...
> I'm trying to use the windll/calldll module to access the mciSendString
> (winmm.dll) function.  One of the parameters to the function is a string
> containing returned information.  I'm using wxPython as GUI to show a AVI
/
> MPEG movie (in a wxPanel).  The play command may be given a 'notify' flag,
> which results in the player calling back to notify when movie is finished.
> The last parameter to the mciSendString is a HANDLE (hwndCallback).
>
> Can someone give an example how to retreive the returned information and
> also how I set up the callback.
>
>
> Thanks
> Nikolai Kirsebom
>
>
>




From martapython at yahoo.es  Thu Jun 20 09:31:30 2002
From: martapython at yahoo.es (=?iso-8859-1?q?Marta=20Gutierre?=)
Date: Thu, 20 Jun 2002 15:31:30 +0200 (CEST)
Subject: Access to XML dataBase
Message-ID: <20020620133130.3692.qmail@web21509.mail.yahoo.com>


 

Has anybody tried to use  XML-RPC  in X-indice database   from a python client ???

I 've got some problems :

 

I tried sothing like:

import xmlrpclib

client=xmlrpclib.Server("http://localhost:4080")
try:
     client.listCollections("db")
except xmlrpclib.Fault,faultobj:
    print "ERROR", faultobj

ERROR <Fault 0: 'java.lang.Exception: RPC handler object not found for "listColl
ections": no default handler registered.'>

Any idea of what does it mean ???

I can't find good documentation of XML-RPC in Python

 


---------------------------------
Copa del Mundo de la FIFA 2002
El ?nico lugar de Internet con v?deos de los 64 partidos ?Ap?ntante ya!.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020620/1138859f/attachment.html>

From gerhard at bigfoot.de  Wed Jun  5 14:04:46 2002
From: gerhard at bigfoot.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Wed, 5 Jun 2002 20:04:46 +0200
Subject: [Pypgsql-users] Re: [ANN] pyPgSQL 2.1 released
In-Reply-To: <20020605175554.GB33711@gargamel.hqd-internal>
References: <20020605170847.2631.qmail@www4.nameplanet.com> <20020605175554.GB33711@gargamel.hqd-internal>
Message-ID: <20020605180446.GC33711@gargamel.hqd-internal>

* Gerhard H?ring <gerhard.haering at gmx.de> [2002-06-05 19:55 +0200]:
> >>> from pyPgSQL import PgSQL
> >>> conn = PgSQL.connect(client_encoding="utf-8", unicode_results=1)
> >>> cursor = conn.cursor()

Here I missed something that (still) needs to be done explicitely:

cursor.execute("set client_encoding to UNICODE")

> >>> cursor.execute("insert into test(name) values (%s)", unicode("?sterreich", "latin1"))
> >>> cursor.execute("select * from test")
> >>> res = cursor.fetchone()
> >>> print repr(res.name)
> u'\xd6sterreich'

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From look at sig.invalid  Wed Jun 26 01:15:27 2002
From: look at sig.invalid (Simo Salminen)
Date: Wed, 26 Jun 2002 05:15:27 +0000 (UTC)
Subject: Parsing strings (\n and \\)
References: <3D188006.7050202@thomas-guettler.de> <mailman.1025017712.20580.python-list@python.org> <lt0S8.45025$n4.10623651@newsc.telia.net>
Message-ID: <slrnahijff.m4g.look@pikkukorppi.cs.tut.fi>

* Fredrik Lundh [Tue, 25 Jun 2002 16:01:21 GMT]
> Fran?ois Pinard wrote:
> 
>> A simple avenue is to get Python itself to evaluate the string as a constant
>> (you ensure this by removing evaluation context).  Something like this:
>>
>>     PARSED = eval(UNPARSED, {}, {})
> 
>>>> UNPARSED = "__import__('os').system('echo dream on!')"
>>>> PARSED = eval(UNPARSED, {}, {})
> dream on!
> 

safe(?) version:

>>> PARSED = eval(UNPARSED, {"__builtins__":None}, {})
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
  File "<string>", line 0, in ?
NameError: name '__import__' is not defined


-- 
simo <dot> salminen <at> iki <dot> fi


From python at rcn.com  Thu Jun 13 13:20:47 2002
From: python at rcn.com (Raymond Hettinger)
Date: Thu, 13 Jun 2002 13:20:47 -0400
Subject: why not "'in' in 'in'"?
References: <mailman.1023977529.27368.python-list@python.org> <aeac0l013k@drn.newsguy.com>
Message-ID: <aeakd1$2lg$1@bob.news.rcn.net>

"Grant Griffin" <not.this at seebelow.org> wrote in message
news:aeac0l013k at drn.newsguy.com...
> Yes, guilty as charged!  Here's an example of my poor taste.  Since "find"
is
> sort of a boolean concept ("I found/didn't find my tennis shoes"), and
since the
> "find" function really returns the index (position) of a substring, I
probably
> would have called it "position" or maybe just "pos" (in admiration for
BASIC).

Where did you find your tennis shoes?  When you set out to find your tennis
shoes did you want to know where they were or just whether they were
findable?  IMHO, it would be strange to find your shoes and then not know
where they are.


Raymond Hettinger






From kondwani at cs.mun.ca  Mon Jun 17 13:04:07 2002
From: kondwani at cs.mun.ca (Kondwani Spike Mkandawire)
Date: Mon, 17 Jun 2002 14:34:07 -0230
Subject: Messing with GUIs via Python...
Message-ID: <ael553$3ai$1@coranto.ucs.mun.ca>

Howdy Folks is there a way to create simple GUI apps.
using Python ( I have only used it for 4 months)...  If so
is there a way of doing this without having it as a language
binding of some graphics software (as OpenGL)...

Kondwani Spike Mkandawire
CS Intern. Memorial University, Canada...




From dbasch at yahoo.com  Sun Jun 16 18:01:03 2002
From: dbasch at yahoo.com (Derek Basch)
Date: Sun, 16 Jun 2002 15:01:03 -0700 (PDT)
Subject: xmlrpclib/python and fault codes
Message-ID: <20020616220103.85789.qmail@web20801.mail.yahoo.com>

Hi, 
Can anyone out there tell me how to check for a fault
response from an xml-rpc server using the xmlrpclib
python module? I need to detect the fault message and
then attempt to resend the xmlrpc request. I can find
examples written in Perl and Ruby but not python.
Thanks,
Derek Basch



__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com




From da_woym at hotmail.com  Tue Jun  4 18:54:54 2002
From: da_woym at hotmail.com (Jakob)
Date: 4 Jun 2002 15:54:54 -0700
Subject: python interacting with other programs
Message-ID: <64eb9145.0206041454.4f70e86e@posting.google.com>

now I know python can interact with programs like Microsoft wor and
access and all but has anyone ever tried to interact with Lotus Notes?
Is it possible?


From gustav at morpheus.demon.co.uk  Thu Jun  6 05:32:11 2002
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: Thu, 6 Jun 2002 11:32:11 +0200 
Subject: Installer vs py2exe: Windows DLLs
References: <u1ooxxn8.fsf@morpheus.demon.co.uk> <Xns9220730326E0Fgmcmhypernetcom@199.171.54.215>
Message-ID: <u1ogoiqu.fsf@morpheus.demon.co.uk>

Gordon McMillan <gmcm at hypernet.com> writes:

> Paul Moore wrote:
>
>> When building a distribution for the same script using Gordon
>> McMillan's Installer and py2exe, I notice that Installer includes
>> PyWinTypes22.dll and win32api.pyd, whereas py2exe doesn't. 
>
> [snip]
>
>> More to the point, in some ways, why does Installer think these files
>> are needed? Is there some situation in which py2exe generated
>> executables will fail because these files aren't there?
>
> David's answer was 100% correct.
>
> The specific call here is os.path.abspath(...). Before 2.2 and 
> without win32api, this simply examined its argument. In some
> rare corner cases, it would yield an incorrect answer.

OK, I analyzed this a bit further. This is all for Python 2.2 on
Windows 2000.

The win32api dependency (and hence the PyWinTypes22 dependency) comes
from iu.py, in the RegistryImportDirector. Two questions come to mind
- first is whether it is appropriate for Installould use the
_winreg module, to avoid the dependency in the first place. Ther-built applications
to look in the registry for Python entries in any case, as the whole
point is to run independently of any installed copy of Python. I
imagine that something later in the process disables picking up
registry entries, but that leaves the RegistryImportDirector code as
dead code (which isn't worth worrying about, except for the spurious
dependency).

The second question is whether RegistryImportDirector she
translation is pretty much automatic - I attach a patch. This gives a
dependency on _winreg.pyd (pity - I'd assumed _winreg would be built
in on Windows :-() but that is *much* smaller than win32api +
PyWinTypes22. Minimal testing shows no problems, but a better test
would be worth doing...

The dependency on _sre comes from posixpath (in expandvars()), which
is imported conditionally by os. And os is imported by archive.py, to
get at os.path.basename, and os.path.splitext. Would it not be
possible here to directly use {dos,mac,nt,posix}path?

I know that in the grand scheme of things, removing dependencies like
this is fairly pointless for any "real" application, but it appeals to
my sense of minimalism. At least the _winreg patch is pretty minor,
and so probably reasonable... (and win32api/PyWinTypes22 is the big
hit).

Hope this is useful,
Paul Moore.

[CC'd to the installer list, as that's probably a more appropriate place]

--- Patch to use _winreg in iu.py ---

--- \Applications\Python\Installer\iu.py.orig	Mon Nov 19 17:05:56 2001
+++ \Applications\Python\Installer\iu.py	Thu Jun 06 10:17:52 2002
@@ -111,7 +111,7 @@
         self.path = "WindowsRegistry"
         self.map = {}
         try:
-            import win32api
+            import _winreg
 ##            import win32con
         except ImportError:
             pass
@@ -122,15 +122,15 @@
             subkey = r"Software\Python\PythonCore\%s\Modules" % sys.winver
             for root in (HKEY_CURRENT_USER, HKEY_LOCAL_MACHINE):
                 try:
-                    hkey = win32api.RegOpenKeyEx(root, subkey, 0, KEY_ALL_ACCESS)
+                    hkey = _winreg.OpenKey(root, subkey, 0, KEY_ALL_ACCESS)
                 except:
                     pass
                 else:
-                    numsubkeys, numvalues, lastmodified = win32api.RegQueryInfoKey(hkey)
+                    numsubkeys, numvalues, lastmodified = _winreg.QueryInfoKey(hkey)
                     for i in range(numsubkeys):
-                        subkeyname = win32api.RegEnumKey(hkey, i)
-                        hskey = win32api.RegOpenKeyEx(hkey, subkeyname, 0, KEY_ALL_ACCESS)
-                        val = win32api.RegQueryValueEx(hskey, '')
+                        subkeyname = _winreg.EnumKey(hkey, i)
+                        hskey = _winreg.OpenKey(hkey, subkeyname, 0, KEY_ALL_ACCESS)
+                        val = _winreg.QueryValue(hskey, '')
                         desc = getDescr(val[0])
                         self.map[subkeyname] = (val[0], desc)
                         hskey.Close()



From nico at tekNico.net  Wed Jun  5 09:41:01 2002
From: nico at tekNico.net (Nicola Larosa)
Date: Wed, 05 Jun 2002 15:41:01 +0200
Subject: python cookbook at oreilly's web site
References: <dCnL8.5$566.1250@news.ecrc.de> <adl2nq$3hrm$1@ID-41653.news.dfncis.de>
Message-ID: <3CFE14ED.8090207@tekNico.net>

>> just found
>> http://www.ora.com/catalog/pythoncook/

> Does this have any relation to ASPNs Python cookbook?
> http://aspn.activestate.com/ASPN/Cookbook/Python

Well, yes, the book is a selection of the Recipes on ASPN.

O'Reilly emailed to say they'll send me a copy, since a Recipe of mine 
(PyHeartBeat) will be in the book.

What can I say... wow. :^)))


-- 
"Too much cleverness in the parser can turn against you."
   Guido Van Rossum

Nicola Larosa - nico at tekNico.net





From mkur at poczta.gazeta.pl  Thu Jun 20 08:35:19 2002
From: mkur at poczta.gazeta.pl (=?Latin-2?Q?Micha=B3?= Kurowski)
Date: Thu, 20 Jun 2002 12:35:19 +0000 (UTC)
Subject: MySQLdb question
Message-ID: <aesi66$cgb$1@sunsite.icm.edu.pl>

Hi,

I'm working on a small project involving sql (MySQLdb-0.9.1) and I constantly have to
"get around" a  quite weird problem:

I cannot find a way to pass a sql statement a table as a parameter.

In Zope you can do something like:
"select * from <dtml-var table> where <dtml-sqlvar ... >"

And what about pure MySQLdb ?
Quotes are getting in the way and the client just breaks.

Thanks for your help.


-- 
Michal Kurowski
<michal at poczta.gazeta.pl>


From rowen at cesmail.net  Wed Jun 26 19:59:00 2002
From: rowen at cesmail.net (Russell E. Owen)
Date: Wed, 26 Jun 2002 16:59:00 -0700
Subject: Help with an algorithm wanted
References: <rowen-CE3C0F.13020226062002@nntp1.u.washington.edu> <mailman.1025125654.6891.python-list@python.org>
Message-ID: <rowen-3B9D8A.16590026062002@nntp3.u.washington.edu>

In article <mailman.1025125654.6891.python-list at python.org>,
 William Park <opengeometry at yahoo.ca> wrote:

>On Wed, Jun 26, 2002 at 01:02:02PM -0700, Russell E. Owen wrote:
>> For instance, an object with A as the given would return D by solving 
>> A->C->D.
>
>> The real issue is creating such objects without hand coding each one 
>> (since in the real world I have more representations and the number of 
>> casese proliferates badly).
>
>If pairing (ie. mapping between 2 objects) occurs randomly, then it's
>virtually impossible, because you don't know which direction to go from any
>one object.  It would be travelling from city to city, without knowing
>which state or county.

Fortunately the pairings (conversion functions) are completely fixed. So  
the graph never varies, and it even has a fairly simple shape (basically 
a fork with 3 tines and a long handle, if that makes any sense).

The problem is that the graph has 10 nodes. Rather than hand code 10 
different objects, I'm trying to create a factory that will generate the 
required object based on which node (data representation) is a given 
from which the others are derived.

For example, given conversion functions CtoA(C), AtoC(A), CtoB(C), 
BtoC(B), CtoD(C), DtoC(D), then I'll want to be able to create four 
objects, one with A as a given, one with B, etc. The object for A as a 
given should act like:
class Agiven:
   def getA(self):
      return self.A
  def getB(self):
    return CtoB(self.getC)
  def getC(self):
      return AtoC(self.getA)
  def getD(self):
      return CtoD(self.getC)

I'm not too worried about figuring out what methods to run depending on 
which representation is a given. My main worry is how to generate the 
desired objects programmatically in a way that is fairly understandable.

The obvious thing is probably to create a new object and 
programmatically add methods to it. But I'm not sure how to do it. If 
that's messy, I may have to have generate the created object by passing 
it a set of functions in the init method. Bjorn Pettersen just sent me 
some code that may do the trick. i have to study it. I'm not yet sure if 
it creates objects or pseudo-objects.

Of course I could create a text representation of the object and eval 
it, but that really sounds ugly!

-- Russell


From missive at frontiernet.net  Sat Jun  8 13:52:31 2002
From: missive at frontiernet.net (Lee Harr)
Date: Sat, 08 Jun 2002 17:52:31 -0000
Subject: pyGTK: gtkxpm not found
Message-ID: <ug4h2v2eerfoc4@corp.supernews.com>

I am working throught the pyGTK tutorial at:

http://www.moeraki.com/pygtktutorial/gtk-tut.html

In chapter 10.11, (Toolbar) it is trying to
import gtkxpm, but that module is not found.

I googled for it, but do not see any reference to
this module at all.

I am running py-gtk-0.6.9 and python-2.2_1
(from packages) on FreeBSD 4.5

Any wisdom appreciated.



From whisper at oz.net  Mon Jun 10 04:34:54 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 10 Jun 2002 01:34:54 -0700
Subject: Where is Berkeley DB on your system? Can you live without DB 1.85?
In-Reply-To: <83lm9n4m8t.fsf@panacea.canonical.org>
Message-ID: <GCEDKONBLEFPPADDJCOECEMFDHAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Kragen Sitaker
> Sent: Monday, June 10, 2002 0:38
> To: python-list at python.org
> Subject: Re: Where is Berkeley DB on your system? Can you live without
> DB 1.85?
>
>
> "David LeBlanc" <whisper at oz.net> writes:
> > AFAIK, this is true for any version of Sleepycat > 1.85, and
> has been true
> > for years. Including a later version of Sleepycat BSDDB imposes a
> > constraint, not on Python itself, which is free, but on someone
> who would
> > like to distribute Python along with their
> _commercial_for_sale_ app (think
> > Py2Exe for example).
>
> I don't think anybody has proposed including a modern version of BSD
> DB in Python, only dropping support for an old version, which is also
> not included in Python.

Hmmm... I have Python 2.2.1 binary installed from Sourceforge - what is that
bsddb.pyd file dated 4/9/2002 in the dlls directory?

> "commercial for-sale" is misleading.  The current license of BSD DB
> allows distribution in commercial for-sale apps with no further hassle
> unless said commercial for-sale apps are, approximately, free
> software, and it forbids inclusion in noncommercial not-for-sale apps
> unless said not-for-sale apps are, approximately, free software.

This makes no sense, so I'll just pass on it. I think you got your facts
backwards.

> > If you want to change Python's market dynamics and require
> _all_ Python apps
> > to be free and open source, then you can include any version of
> Sleepycat
> > BSDDB you like. I doubt it will have a positive impact on
> Python's future
> > prospects. It didn't work for Perl in it's early years and
> that's why the
> > Artistic License was change to allow "for fee" distributions
> that included
> > Perl.
>
> I thought the AL allowed this from the beginning.  Certainly
> Sleepycat's license doesn't forbid it.

IIRC, in the early days of Pearl, the "AL" had the same tainting as GPL.
Again, IIRC, some of the Pearl inner circle where the ones who agitated for
the change so they could make some money from all the time they where
pouring into Pearl. Today, the AL is more like the GLL.

Sleepycat's license doesn't prohibit commercial for-fee distribution, it
simply requires that you negotiate a payment scheme with them for the use of
versions after 1.85. They don't charge for redistribution with free open
source apps like Python. However, as I said in my last post (along with
exerpts from Sleepycat's FAQ on the subject), if a Python developer was to
include a Python distribution that included a version of bsddb after 1.85 in
a for-fee commercial program, then they would have to negotiate a payment
scheme with Sleepycat. More to the point, they would have to know that they
would have to do that or risk copyright infringement. (BTW, I was, at one
time, in such discussions with Sleepycat, but the project got dropped. I
found their fees to be extremely reasonable and they where quite flexible -
nice people to work with.)

> In any case, your irresponsible, alarmist accusation is a red herring,
> as you will no doubt agree once you understand the situation properly.
> Dropping support for db 1.85 will not require all Python apps to be
> free software or open source; it won't even require all distributions
> of Python to be free software or open source.  It will only require
> distributions of Python with bsddb support to either link to an
> external BSD DB library (like the one included with glibc) or be free
> software.

I don't think I'm the one not understanding the situation properly, nor
could anything I posted be reasonably construed as irresponsible, alarmest
or an accusation! I never said anthing pro or con wrt to dropping bsddb 1.85
from the distribution. I didn't even make any accusations! I was only
referring to versions of bssdb after 1.85 having commercial licensing
provisions.

> If your application doesn't use BSD DB, it won't be affected by this,
> even if you ship binary distributions of it, made with e.g. McMillan
> Installer.  If your application uses BSD DB and you are shipping db
> 1.85, you are screwing your users, probably without knowing it (unless
> you know enough about db 1.85's bugs to work around them.).  This
> change will allow you to continue to screw your users in the same way,
> but you will be required to know that was what you were doing.

No customers are getting screwed :->

> At present, there are irresponsible vendors shipping proprietary
> distributions of Python including BSD DB 1.85.  This course of action
> subjects the entire Python community to pressure to drop or deprecate
> support for BSD DB, because BSD DB 1.85 is badly broken.  Those of us
> who use sane versions of BSD DB are hurt by this, because BSD DB is by
> far the most featureful, performant, and reliable of the other
> dbm-style databases Python supports.

Something remarkably like bsddb is getting shipped in python.org python
binaries from sourceforge as far as I can tell (Windows binaries).

> I assume these vendors are being irresponsible because they don't know
> they are imperiling their users' data by using this buggy old
> software.  This change will let them know that the problem exists, and
> they can either stop shipping proprietary distributions, stop
> supporting BSD DB, or start maintaining BSD DB 1.x.

<gasp> could Guido be so heartless as to imperil people's data by allowing
the inclusion bsddb in the standard Python distro?!?

> If it's not clear, I'm in favor of this change.  It helps everyone
> everywhere.

Whatever gets done, it would be nice to have some useful db shipped in the
standard Python distro. I have heard a rumor that gadfly is such a
candidate - and also that gadfly has bugs of it's own.

BTW, since you're so keen on the failings of the BSDDB that seems to ship
with standard Python, perhaps you'd like to contribute information on how to
use it and what it's pitfalls are and ways to avoid them to the pythondoc?
Hell, you could even help maintain BSD DB 1.x!

Dave LeBlanc
Seattle, WA USA





From hbl at st-andrews.ac.uk  Thu Jun  6 20:00:47 2002
From: hbl at st-andrews.ac.uk (Hamish Lawson)
Date: Fri, 07 Jun 2002 01:00:47 +0100
Subject: Emulating classmethod in Python 2.1 
Message-ID: <3CFFF7AF.B1B4A6C0@st-andrews.ac.uk>

Franz GEIGER wrote:

> There's a receipe about this in the ASPN, by Alex Martelli. Concise 
> and effective. Strongly recommended.

If you are referring to recipe 52304 [1], then I am in fact using this
to emulate staticmethod() where needed - I agree that it's a concise and
effective implementaion of staticmethod(). But Alex's recipe doesn't
emulate classmethod(), since the class is not passed through to the
method (notwithstanding the fact that Alex uses the terms 'class method'
and 'static method' interchangeably in his accompanying discussion -
although he wrote that before Python 2.2 made the distinction). Hence
the reason I'm still looking for an emulation of classmethod().

[1] http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52304


Hamish




From jjl at pobox.com  Sun Jun  9 17:25:40 2002
From: jjl at pobox.com (John J. Lee)
Date: Sun, 9 Jun 2002 22:25:40 +0100
Subject: Correct code/unit testing (was Re: Efficient python programming...)
In-Reply-To: <3D02FB3B.CF75E925@engcorp.com>
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com> <834rgd5isv.fsf@panacea.canonical.org> <3D02FB3B.CF75E925@engcorp.com>
Message-ID: <Pine.SOL.4.30.0206092211170.15771-100000@mimosa.csv.warwick.ac.uk>

On Sun, 9 Jun 2002, Peter Hansen wrote:

> Kragen Sitaker wrote:
[...]
> > shown) that careful reasoning and code reviewing is necessary and
> > sufficient to write correct code.
[...]
> ego, you'll be devastated by the repeated proof that you will never,
> ever get it right the first time, or even after lots of reviewing.

Not quite never.  It *does* happen, but with the same sort of frequency as
I find dropping from a five club juggle leads to one landing upright on
the floor and staying there: it *has* happened to me, exactly three times
in about a decade of practice.

(what's a club?: http://www.windwizards.com/henrysjuggling.html)

> > It is certainly possible to err too far in the direction of believing
> > your code correct simply because it passes its test suite, and this is
> > a common beginner's error.  I think it is also possible to err too far
> > in the direction of writing and running tests instead of desk-checking
> > code.
[...]

Isn't it true that, however much time you're going to spend, the largest
number of bugs are going to get removed by allocating some of the time to
both activities: testing, and staring at the code?


John



From romany at actimize.com  Thu Jun 27 01:54:22 2002
From: romany at actimize.com (Roman Yakovenko)
Date: Thu, 27 Jun 2002 07:54:22 +0200
Subject: Questiion on list
Message-ID: <7647A9F4B298E74DB6793865DA67285004AD96@exchange.adrembi.com>

import string
lst = ['a','b','c', 'a']
mapCount = {}

for word in lst:
    if not mapCount.has_key(string.lower(word)):
        mapCount[string.lower(word)] = 0
    mapCount[string.lower(word)] += 1
    
print mapCount

-----Original Message-----
From: SiverFish [mailto:occeanlinux at linuxmail.org]
Sent: Thursday, June 27, 2002 4:36 PM
To: python-list at python.org
Subject: Questiion on list


I got the list of words now i want to calculate how many word in that
list which is the same word (insensitive case),can anyone show me how to do
it for example ls = ['abc','cd','abc','adf',abc','dwc','cd']
it will show abc = 3 ,cd = 2, adf = 1, dwc=1
-- 
http://mail.python.org/mailman/listinfo/python-list




From fuf at mageo.cz  Fri Jun  7 10:30:21 2002
From: fuf at mageo.cz (fuf)
Date: Fri, 7 Jun 2002 16:30:21 +0200
Subject: positional modifiers in python?
Message-ID: <20020607143021.GA3038@foof.i3.cz>

 hello everyone,

 does python support positional parameters in the print method? ie.
 something like:
    print "there %2$s something %1$s" % ("there", "is")
 it doesn't work but maybe there's some other way?

    thank you,
-- 
		fuf		(fuf at mageo.cz)




From skip at pobox.com  Fri Jun  7 12:42:19 2002
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 7 Jun 2002 11:42:19 -0500
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <yu99vg8v6orb.fsf@europa.research.att.com>
References: <yu99vg8v6orb.fsf@europa.research.att.com>
Message-ID: <15616.57963.519636.730290@12-248-41-177.client.attbi.com>

    ...
    Andrew> gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/tmp/build-gnu913/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /tmp/build-gnu913/Python-2.2.1/Modules/structmodule.c -o build/temp.solaris-2.8-sun4u-2.2/structmodule.o
    Andrew> creating build/lib.solaris-2.8-sun4u-2.2
    Andrew> gcc -shared build/temp.solaris-2.8-sun4u-2.2/structmodule.o -L/usr/local/lib -o build/lib.solaris-2.8-sun4u-2.2/struct.so
    Andrew> WARNING: removing "struct" since importing it failed
    ...

    Andrew> I do not believe that these problems happened with gcc 3.0
    Andrew> Any suggestions as to how to go about troubleshooting it?

My guess is you have a new libgcc.so file that needs to be introduced to the
system using the ldconfig command.  Try running ldd on the struct.so file
and see if it finds everything the file references.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html





From chris.gonnerman at newcenturycomputers.net  Wed Jun 19 23:50:48 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Wed, 19 Jun 2002 22:50:48 -0500
Subject: Flame fests (was Re: python version?)
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <3D1010C3.C321F8B0@engcorp.com> <raXP8.47926$GY.15260255@twister.nyroc.rr.com> <3D107948.B3FEE4BB@engcorp.com> <aer53s$7is$1@panix1.panix.com> <3D112BA6.4CA5F584@engcorp.com>
Message-ID: <009a01c2180d$abb4f5c0$0101010a@local>

----- Original Message -----
From: "Peter Hansen" <peter at engcorp.com>


> Aahz wrote:
> >
> > Everybody have fun playing "kick the newbie"?  Maybe it's time to quit?
>
> I think we'd all object strongly to the characterization that we were
> "kicking a newbie".
>
> For one thing, any kicking done was entirely unrelated to whether or not
> he was a newbie.
>
> For another thing, he has basically denied being interested in learning
> about Python, which I believe makes the label "newbie" inappropriate.

Indeed.  After careful examination I have decided he is a troll.  It took a
while for me to recognize it... for instance, he apologized a couple of
times
in a backhanded way.

So George Hester joins Timothy Rue in my message ruleset under the "delete
from server" category.  Whatever technology everyone else is using, I
recommend
similar revisions for all.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net






From tdelaney at avaya.com  Thu Jun 27 02:08:51 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 27 Jun 2002 16:08:51 +1000
Subject: background-processes with win-python?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A3CE@natasha.auslabs.avaya.com>

> From: Klaus Reinhardt [mailto:K.Rdt at TU-Berlin.DE]
> 
> Is there a possibility to execute system-processes
> under windows-python in the background? So
> I can do this
> 	system("plink  .. tu-berlin.de")
> without waiting for the real connection.

Look at the 'threading' module.

Tim Delaney




From del_brucedawson_del at del_cygnus-software_del.com  Mon Jun 17 03:45:04 2002
From: del_brucedawson_del at del_cygnus-software_del.com (Bruce Dawson)
Date: Mon, 17 Jun 2002 07:45:04 GMT
Subject: Using Python to script a game?
References: <cdc2589b.0206141718.7dc2ae13@posting.google.com> <mailman.1024106376.21850.python-list@python.org> <aeedci$29e$1@panix1.panix.com>
Message-ID: <3D0D9317.6371F155@del_cygnus-software_del.com>

I did a Game Developers Conference talk on using Python in games.
Our company (Humongous Entertainment's) decided to move to it as
our scripting language for all games (weaning ourselves off of a ten
year dependence on 'SCUMM'). I made a list of all of the games I
could find that were using Python. The list is:

Freedom Force
ToonTown
EveOnline
Blade of Darkness
Frequency (PS2)
Game Blender (game engine)
PyGame (game engine)
MindLathe (AI middleware)
All future Humongous PC/Macintosh games

There are some big games in that list.

Ultima Online 2 was also rumoured to be using Python.

For those who are interested, the article and power point slides can be
found at:

http://www.cygnus-software.com/papers/gamescriptinginpython.html

Aahz wrote:

> In article <mailman.1024106376.21850.python-list at python.org>,
> Gerhard =?iso-8859-15?Q?H=E4ring?=  <gerhard at bigfoot.de> wrote:
> >
> >Yes. And it would be a pretty good match, too. There's even been a
> >commercial game that uses Python for internal scripting, its name
> >escapes me, though.
>
> Blade of Darkness.  It's a first-person shooter, too, so that ought to
> be all the evidence needed to prove that Python is fast enough.
> --
> Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/
>
> Project Vote Smart: http://www.vote-smart.org/



From K.Rdt at TU-Berlin.DE  Fri Jun 28 02:19:11 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Fri, 28 Jun 2002 08:19:11 +0200
Subject: crypt for windows?
In-Reply-To: <afgtj8$us1$07$1@news.t-online.com>
Message-ID: <96ZT08QPTQNHDUP4WYT6175KE4WQKTP.3d1bffdf@FRITZweb>

Am 28.06.02 07:52:38, schrieb Markus <listuser at nicetomeetyou.to>:
>I use a module named fcrypt. It is a pure Python implementation and works 
>fine for Unix *and* Windows.
---------------------------------------------------------------------
Hi
Could you give me an example script?
		K at Rdt
---------------------------------------------------------------------







From david.abrahams at rcn.com  Mon Jun 17 21:14:16 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Mon, 17 Jun 2002 21:14:16 -0400
Subject: ImageMagick for Python incl. Source
References: <aelmhc$35m$03$1@news.t-online.com> <aelo0t$sog$1@peabody.colorado.edu> <fMtP8.57199$uX3.27041@nwrddc01.gnilink.net>
Message-ID: <aem1io$eb8$1@bob.news.rcn.net>

"vze" <vze3myxh at verizon.net> wrote in message
news:fMtP8.57199$uX3.27041 at nwrddc01.gnilink.net...
> >
> > Sorry to display my ignorance here, but what on earth is a "Jamfile"?
>
> Looks like jam is a make -like utility from Perforce
> ( http://public.perforce.com/jam/index.html )

Actually, the one you want is at www.boost.org/tools/build.

-Dave




From timo.spam.linna at iki.me.fi.not.invalid  Thu Jun  6 09:34:45 2002
From: timo.spam.linna at iki.me.fi.not.invalid (Timo Linna)
Date: Thu, 06 Jun 2002 13:34:45 GMT
Subject: minidom and encoding problem
References: <17aafe08.0206051514.4af054d2@posting.google.com>
Message-ID: <VxJL8.6693$ZE1.142129@news1.nokia.com>

"Ehab Teima" <ehab_teima at hotmail.com> wrote in message
news:17aafe08.0206051514.4af054d2 at posting.google.com...
> Hi,
>
> I know it's not possible to add an enconding attribute using writexml,
> so the generated document only has <?xml version="1.0"?>. Is there any
> way to get around this problem. I'd like to be able at least to parse
> the document while reading using the proper encoding, such as,
> encoding="ISO-8859-1". I'm only using minidom, any ideas how?

Thanks to Python, you could redefine minidom's writexml-function:

import xml.dom.minidom

def encoded_writexml(self, writer, indent="", addindent="", newl=""):
    writer.write('<?xml version="1.0" encoding="ISO-8859-1" ?>\n') # My
change
    for node in self.childNodes:
        node.writexml(writer, indent, addindent, newl)

xml.dom.minidom.Document.writexml = encoded_writexml

-- timo





From cliechti at gmx.net  Thu Jun 20 17:34:06 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 20 Jun 2002 23:34:06 +0200
Subject: Inhibiting duplicate apps in Windows?
References: <3D11F9E8.DFFFC8E4@tds.net>
Message-ID: <Xns9233F02DB217Ccliechtigmxnet@62.2.16.82>

"Edward K. Ream" <edream at tds.net> wrote in news:3D11F9E8.DFFFC8E4 at tds.net:

> Hi All,
> 
> I have a pure Python app called Leo.  In Windows, I can open Leo by
> double-clicking a .leo file.  I would like to do more: if Leo is open
> and I double-click a .leo file, I would like the _already open_ copy of
> Leo to open the second .leo file, rather than (permanently) opening a
> second copy of Leo.
> 
> What I would like is for the second copy of Leo to detect that another
> copy of leo.py is already running, send a message to the first copy and
> then exit. I wonder if anyone knows how this might be done?
> 
> The C++ version of Leo works as follows:
> 
> 1. It inhibits multiple copies of leo.py from running using a mutex.  I
> think this part is easy because the Python library supports mutexes. 

but the python mutexes do not work across processes.
you can also use FindWindow of win32all to search a process/window, so you 
don't need a mutex.
 
> 2. It sends a message from the second copy of leo.py (the copy that will
> shortly close), instructing the first copy of leo.py to open a file. 
> This is the part that I'm not sure can be done in pure Python.  I don't
> see anything applicable in section 22 (MS Windows Specific Services) of
> the Python Library.  The C++ version of Leo does this by doing:
> 
> SendMessage(hWnd, WM_COPYDATA, 0, (LPARAM)&cds);
> 
> where hWnd is a handle to the previously open copy of Leo and cds is a
> COPYDATASTRUCT containing the command line used to open the second copy
> of Leo.

that all should be possible with the win32all extensions. those extensions 
are quasi stadard for most windows users, so just use it...

> Perhaps this problem can be solved in Python using processes instead of
> messages, and I don't see how.  Are there any demos that might be of
> use?  Thanks!

dumb solution: create a "lock file" and remove it on programm exit. the 
other problem is that you have to delete it manually when the app crashes 
(i deleted may lock files with older Netscapes ;-)

i would definetly take win32all.

chris

-- 
Chris <cliechti at gmx.net>



From bit_bucket5 at hotmail.com  Tue Jun  4 20:34:55 2002
From: bit_bucket5 at hotmail.com (Chris Stromberger)
Date: Wed, 05 Jun 2002 00:34:55 GMT
Subject: Using AT on XP to run python script
Message-ID: <mchqfu448cnq5q8toh4nq7n1sura1ug72g@4ax.com>

I'm having trouble getting AT on XP to run a python script for me.
Have tried:

at xx:xx "c:\python21\python.exe c:\scripts\script.py"
at xx:xx "start c:\scripts\script.py"
at xx:xx "c:\scripts\script.py"

and get "the system cannot find the file specified" for the first two.
The last version does nothing apparently--no error indication or
anything.

I imagine it is related to user accounts and such.  Anyone have any
pointers?


From opengeometry at NOSPAM.yahoo.ca  Fri Jun 28 22:16:16 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 29 Jun 2002 02:16:16 GMT
Subject: How to find out DNS ?
References: <mailman.1025197535.3041.python-list@python.org> <o2qmhuk964q6ihlv3mpnf63g09dqmg1e3f@4ax.com>
Message-ID: <afj59f$f0958$1@ID-99293.news.dfncis.de>

Wolfgang Strobl <ws at mystrobl.de> wrote:
> Thu, 27 Jun 2002 19:03:34 +0200, "A" <printers at sendme.cz>:
> 
>>Hi,
>>Is there a way how to find out, from Python , what primary or 
>>secondary DNS I use when connecting to internet?
>>Thanks for help
>>Ladislav
> 
> Dear Ladislav (who?),
> 
> didn't you already ask some variants of this question last year? For
> example
> 
> http://groups.google.de/groups?selm=mailman.1006268728.27197.python-list%40python.org
> 
> Weren't you satisfied by the answer given by Michael Str?der in
> 
> http://groups.google.de/groups?selm=3BC16D75.566E9502%40stroeder.com
> for example? If not, why not?

You have good memory.  Usenet posts are not things that I would remember...
Anyways, if it's Unix, then
    - /etc/resolv.conf
    - host -t ns 
And, if it's Windows, then try
    - CGI scripts from <http://www.DNSstuff.com/>

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin


From aahz at pythoncraft.com  Wed Jun 19 21:24:14 2002
From: aahz at pythoncraft.com (Aahz)
Date: 19 Jun 2002 21:24:14 -0400
Subject: Threading question
References: <GCEDKONBLEFPPADDJCOECEFBDKAA.whisper@oz.net> <mailman.1024533445.11872.python-list@python.org>
Message-ID: <aeraru$m3i$1@panix1.panix.com>

In article <mailman.1024533445.11872.python-list at python.org>,
Trent Mick  <trentm at ActiveState.com> wrote:
>[David LeBlanc wrote]
>>
>> with a thread sitting in a loop processing intermittant transactions from a
>> buffer, how does one wait if reading an empty buffer does not block - how do
>> you relinquish control to other threads/processes?
>
>There are probably a bunch of ways to do it but here is one.
>
>If you have access to the code that populates the buffer then given
>everyone a handle to a condition variable and have the reader .wait() on
>that condition and have anyone who adds data to the buffer .notify() or
>.notifyAll() on that condition.

If you're doing that, just have the code populating the buffer change
the buffer to a Queue.  Much simpler.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From dl at bestpriceeu.com  Sun Jun  9 14:30:25 2002
From: dl at bestpriceeu.com (Dan Larsen)
Date: Sun, 9 Jun 2002 20:30:25 +0200
Subject: newbie system() question
References: None <3d035180$0$7017$bc7fd3c@news.sonofon.dk> <1023641774.112216@yasure>
Message-ID: <3d039e18$0$7017$bc7fd3c@news.sonofon.dk>

Hi Donn.
Thanx for your answer.

> try "stty tostop",
I haven't try this, coz I got around the problem with a nasty, but working
hack :o)
I just made a sh script, that started the program, this works fine :o)

> The redirection syntax "program >& /dev/null &" won't help in system(),
> because it isn't valid for the Bourne shell.
Isn't Bourne shell, the same as sh?
It works for me to do it, in the standard shelle (sh)...

Anyways... It works for now, so I'll try it out later :o)
Thank you very much for you suggestions

Dan Larsen




From edream at tds.net  Sat Jun 22 22:03:22 2002
From: edream at tds.net (Edward K. Ream)
Date: Sun, 23 Jun 2002 02:03:22 GMT
Subject: ConfigParser & converting strings to lists
References: <3D152A28.2177B054@tds.net>
Message-ID: <3D152C69.CC41ADB0@tds.net>

> Actually, this works just fine,
> and the config file contains:
> 
> [recent files]
> recentFiles = [file1, file2, ...]
> 
> as expected.  However,

Correction.  Instead of

recentFiles = [file1, file2, ...]

the set method writes:

recentFiles = ['file1' 'file2' ...]

so it is strings of the form "['file1' 'file2' ...]" that must be
converted to a list of strings. The ellipses do not actually appear
anywhere, but I had to indicate "and so on" somehow.  Clear enough?

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


From cliechti at gmx.net  Thu Jun  6 18:31:13 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 7 Jun 2002 00:31:13 +0200
Subject: another near-sighted, time-wasting syntax enhancement idea
References: <e395b886.0206051956.5f447e9b@posting.google.com>
Message-ID: <Xns92265AD72BE0cliechtigmxnet@62.2.16.82>

idot at vt.edu (Carl Banks) wrote in 
news:e395b886.0206051956.5f447e9b at posting.google.com:
> class BigLongClassThatRunsUntilEOF::
> 
> def __init__ (self):
>     pass
> 
> def whatever (self):
>     pass
> 
> ...
> 
> Two colons after the class (or whatever) definition means the block
> continues until EOF and is not indented.  Good for big, long classes
> that take up a whole module.  Been suggested before, I'm sure.

i don't see any advantages for that. it's a very suptile difference almost 
not viewable those "::", it makes code bad readable.
doesn't your editor has auto indentation? don't use notepad an switch to a 
real editor <wink>.

chris

-- 
Chris <cliechti at gmx.net>



From trentm at ActiveState.com  Wed Jun 19 20:35:42 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Wed, 19 Jun 2002 17:35:42 -0700
Subject: Threading question
In-Reply-To: <GCEDKONBLEFPPADDJCOECEFBDKAA.whisper@oz.net>; from whisper@oz.net on Wed, Jun 19, 2002 at 05:22:24PM -0700
References: <GCEDKONBLEFPPADDJCOECEFBDKAA.whisper@oz.net>
Message-ID: <20020619173542.A28839@ActiveState.com>

[David LeBlanc wrote]
> with a thread sitting in a loop processing intermittant transactions from a
> buffer, how does one wait if reading an empty buffer does not block - how do
> you relinquish control to other threads/processes?

There are probably a bunch of ways to do it but here is one.

If you have access to the code that populates the buffer then given
everyone a handle to a condition variable and have the reader .wait() on
that condition and have anyone who adds data to the buffer .notify() or
.notifyAll() on that condition.


import threading
class Reader(threading.Thread):
    def __init__(self, cond):
        self.cond = cond
    def run(self):
        self.cond.acquire()
        self.cond.wait()
        self.cond.release()

        # Get some data from the buffer and process it...
        # (Note that you'll probably want to lock the buffer when
        # modifying it here.)


class Writer(thrading.Thread):
    def __init__(self, cond):
        self.cond = cond
    def run(self):
        # Add some data to the buffer...
        # (Note that you'll probably want to lock the buffer when
        # modifying it here.)

        self.cond.acquire()
        self.cond.notify()  # or .notifyAll()
        self.cond.release()


Trent
        

-- 
Trent Mick
TrentM at ActiveState.com




From tim.one at comcast.net  Mon Jun 10 17:02:49 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 10 Jun 2002 17:02:49 -0400
Subject: False Queue.Empty exception ?
In-Reply-To: <4a53cb75.0206101234.179ba0fd@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEOHPLAA.tim.one@comcast.net>

[Scott Blomquist]
> I use the Queue class to maintain a threadpool of threads which
> currently have no task, rather than creating/reaping threads as they
> are needed. When I have a task, I grab one out of the queue and assign
> the task to it. To add complexity :) I have made the threadpool able
> to have a min and a max number of threads created, and if the queue of
> free threads is Empty, then I (potentially) create more threads (up to
> the max), and drop them into the free queue.
>
> The problem is that the Empty exception appears to be raised even when
> the queue is in fact not empty - causing me to create threads when I
> don't really need to - some sort of race condition in the following
> code from the Queue class?
>
> ... [code snipped] ...
>
> Keeping in mind that I am doing a non-blocking get, it seems as though
> if there is a thread between the self.esema.acquire(0) and the
> self.esema.release(), and it gets switched out and another thread is
> switched in and comes through the get call, it will fail on the
> self.esema.acquire(0), even though the queue is not neccessarily
> emtpy.

Yes.  As the docs say,

    Empty
    Exception raised when non-blocking get() (or get_nowait()) is called
    on a Queue object which is empty or locked.
                                     ^^^^^^^^^

A non-blocking get() can't know for sure whether the queue is empty if
someone else is holding the esema lock, and can't wait for that lock (so
that it *could* know for sure) without blocking.  Hence it does what it's
documented to do, raising Empty either if it knows for sure *or* can't make
progress without blocking.  It doesn't even know which is the case -- that
esema can't be acquired instantly is equally true of both cases, and is the
only clue it has.

This is adequate for most apps, which use an unbounded or fixed-size queue,
and simply loop on Empty or Full, perhaps with a short time.sleep() in
between.

Note that even if you did know for sure that the queue was Empty, you could
still create needless threads:  there's nothing to stop all the threads in
use from returning to the queue between the time you learn it was empty and
the time you get around to creating a new thread.

But threads are cheap.  Stick as many in the pool as you think you can
reasonably use at one time, and save yourself the bother of trying to adjust
that number.

it's-not-going-to-work-right-anyway<wink>-ly y'rs  - tim





From varney at collorado.edu  Mon Jun  3 04:21:05 2002
From: varney at collorado.edu (Mike Varney)
Date: Mon, 3 Jun 2002 02:21:05 -0600
Subject: How to suspend execution for at least x seconds in python ? (ie does "sleep" exist in Python ?)
References: <adf7t3$ki71@news.rd.francetelecom.fr>
Message-ID: <adf8tl$k9t$1@peabody.colorado.edu>

"Nicolas Torzec" <nicolas.torzec at rd.francetelecom.com> wrote in message
news:adf7t3$ki71 at news.rd.francetelecom.fr...
> Dear all,
> I was wondering if an utility like "sleep x" exist in python ?
> Under Unix-like operating systems, the sleep utility suspends execution
for at least x seconds under...
>


from time import sleep # imports sleep timer

sleep(t)   # invokes Sleep timer



From ellisjb at my-deja.com  Mon Jun  3 12:15:15 2002
From: ellisjb at my-deja.com (Jonathan Ellis)
Date: 3 Jun 2002 09:15:15 -0700
Subject: Graph distances, how can I speed it up?
References: <33803989.0206030235.4134d33e@posting.google.com>
Message-ID: <d571ae53.0206030815.2e52aac@posting.google.com>

tebeka at cs.bgu.ac.il (Miki Tebeka) wrote in message news:<33803989.0206030235.4134d33e at posting.google.com>...
> Hello All,
> 
> I'm trying to calculate distances in a graph. (The ultimate goal is 
> to calculate the distances between verbs in WordNet)
> 
> The graph is built edge after edge and each time I recalculate the 
> distances matrix. However this takes ages to complete. (will finish in ~50H).

The obvious idea then is to not calculate anything until all edges
have been added.  Then use dijkstra's algorithm
(http://ciips.ee.uwa.edu.au/~morris/Year2/PLDS210/dijkstra.html)
to compute and store distances as needed.  (You can probably get some
savings over the classic algorithm if you know you are going to NEED
distances for ALL vertices -- otherwise you are better off doing lazy
(as-needed) computation.)

-Jonathan


From claird at starbase.neosoft.com  Mon Jun 17 15:01:44 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 17 Jun 2002 14:01:44 -0500
Subject: Newbie deluxe
References: <hdDO8.29984$Hn4.969427@news1.east.cox.net> <e81be8b2.0206151645.9fa4b30@posting.google.com>
Message-ID: <BDF0B93251A16144.B4B5F5D0C70CF9AF.12114D6B0D7360C7@lp.airnews.net>

In article <e81be8b2.0206151645.9fa4b30 at posting.google.com>,
Dean Goodmanson <ponderor at lycos.com> wrote:
			.
			.
			.
>when you're getting familier with the language/technology. I digress,
>yet that is where I find Python very useful, and wish other languages
>(C#) provided a shell epxerience.  BTW, Jython is on my To-Do list,
			.
			.
			.
While it's not widely known, there are interpretive
shells for many commonly-compiled languages, including
elastiC, which has similarities to C#.  See <URL:
http://www.elasticworld.com/ >, for example.

I don't know one for C# itself, though.
-- 

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


From dmitri.gouliaev at telkel.net  Thu Jun 20 02:40:22 2002
From: dmitri.gouliaev at telkel.net (Dmitri I GOULIAEV)
Date: Thu, 20 Jun 2002 01:40:22 -0500
Subject: procmail replacement in Python
In-Reply-To: <20020619150411.GB11789@lilith.my-fqdn.de>; from "Gerhard Haring" on Wed, Jun 19, 2002 at 05:04:12PM
References: <oqelf3s7n4.fsf@carouge.sram.qc.ca> <NCBBLFCOHHDIKCAFGCFBAEGILFAA.marklists@mceahern.com> <20020619181953.E4127@phd.pp.ru> <20020619150411.GB11789@lilith.my-fqdn.de>
Message-ID: <20020620014021.A4575@lifebook>

Hi, Gerhard Haring !

 On Wed, Jun 19, 2002 at 05:04:12PM +0200, Gerhard Haring wrote:

> * Oleg Broytmann <phdOphd.pp.ru> [2002-06-19 18:19 +0400]:
> > On Wed, Jun 19, 2002 at 09:08:53AM -0500, Mark McEahern wrote:
> > >   http://www.qcc.sk.ca/~charlesc/software/getmail-2.0/
> > 
> >    "A POP3 mail retriever..." Hence, it is not fetchmail replacement, as
> > fetchamil also does IMAP, and does it very well.
> 
> I don't need to retrieve emails from an IMAP server [...]

CMIIamW, I thought Oleg was talking about the description of getmail, not about your particular needs.

[...]

> Gerhard


Best regards,

-- 
DIG (Dmitri I GOULIAEV)




From mcherm at destiny.com  Wed Jun 19 14:45:09 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Wed, 19 Jun 2002 14:45:09 -0400
Subject: Listing all available (not just loaded) modules
Message-ID: <3D10D135.2030206@destiny.com>

 > How do I list all available (not just loaded) modules?

Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
 >>> modules = []
 >>> def addModuleToList( path, moduleName, description ):
...     modules.append( moduleName )
...
 >>> import pydoc
 >>> pydoc.ModuleScanner().run( addModuleToList )
 >>> print modules
    [... LOTS OF OUTPUT HERE ...]

-- Michael Chermside





From paul at boddie.net  Mon Jun 24 11:35:47 2002
From: paul at boddie.net (Paul Boddie)
Date: 24 Jun 2002 08:35:47 -0700
Subject: Web templating/db tool with best designer/coder separation?
References: <af2gh6$2c2$0@216.39.172.122> <slrnah9mj3.rdk.jon+usenet@snowy.squish.net> <3D159696.9010209@mxm.dk> <m3r8iye24k.fsf@ferret.object-craft.com.au>
Message-ID: <23891c90.0206240735.3395d1f1@posting.google.com>

Dave Cole <djc at object-craft.com.au> wrote in message news:<m3r8iye24k.fsf at ferret.object-craft.com.au>...
> >>>>> "Max" == Max M <maxm at mxm.dk> writes:
> 
> Max> Jon Ribbens wrote:
> >> In article <af2gh6$2c2$0 at 216.39.172.122>, Bengt Richter wrote: Take
> >> a look at http://jonpy.sourceforge.net/ . It almost completely
> >> separates code and HTML so that the designer and the coder do not
> >> interfere with each other.
> 
> Max> I do understand that argument, but I almost certainly don't agree
> Max> that it is of much use.
> 
> I agree, I am not sure it is such a good argument myself.

It's a sort-of-incomplete argument, really. In most situations, the
designer and developer need to agree on what is being presented, so
they must "interfere" with each other in some way. However, the nature
of the interference should be focused only on the structure of the
data, not on the way the data is stored, for example.

>                                                           If you are
> working in an environment where HTML editors do not understand the
> templating system then you are going to need at least one person who
> can post process the HTML and integrate it with the templating logic.
> No magic templating system is going to solve that problem.

If the designer and HTML editor together know nothing about the
templating system, the best that one can hope for is a diminished
amount of work doing that integration for the developer. Moreover,
where changes need to be made again by the designer, it really helps
if the developer can avoid doing any work "unintegrating" the
templates from the system. Being able to make a round trip with the
templates is pretty interesting, in my opinion.

> Max> The development of the Cmf and Plone are good examples of this
> Max> wrongfull approach. They make it easy to create one single type
> Max> of application, but makes it very hard to re-use their components
> Max> to build other types of applications.
> 
> Whenever you build any toolkit you have to decide on the set of
> problems that you wish to solve with the toolkit.  The smaller the
> problem domain the more useful you can make the toolkit.

Indeed. Some kinds of applications may be addressed well by a
particular toolkit, and since you might spend most of your time
constructing those applications for customers, for example, it would
definitely be worthwhile abandoning "complete freedom of expression"
for something which gets the job done quickly, effectively and
cheaply. This is arguably where many toolkits/frameworks fall down -
by trying to be everything to everyone, they become as complicated as
the language they're written in and don't always offer much more than
the bare language in doing a particular job. Consequently, people get
the temptation to write "rival frameworks" from scratch.

Paul


From marklists at mceahern.com  Tue Jun 25 10:48:25 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 25 Jun 2002 09:48:25 -0500
Subject: Parsing strings (\n and \\)
In-Reply-To: <3D188006.7050202@thomas-guettler.de>
Message-ID: <NCBBLFCOHHDIKCAFGCFBIEHFLGAA.marklists@mceahern.com>

> Is there a function for parsing c-like
> strings? I need to parse strings containing
> '\n' and '\\' '\"'.

I suppose it's obvious to anyone from a C background, but I find myself
wondering, "What do you mean by parse?"

I might be able to help if you provide a little more detail.

Cheers,

// mark

-





From shagshag13 at yahoo.fr  Thu Jun 13 07:44:14 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Thu, 13 Jun 2002 13:44:14 +0200
Subject: parse tar file with python
References: <ae7498$4jven$1@ID-146704.news.dfncis.de> <ae7o6s$4oh4v$1@ID-50400.news.dfncis.de>
Message-ID: <aea0hs$54qq4$1@ID-146704.news.dfncis.de>

In fact, i don't want to tar/untar files, and especially not in main memory !

I wish i could read the tar-ed file line by line (f.readline) and be able to check when i find the beginning of an "inside file" and
get some info about it like name, how and so on... (that's because my original files are plain text file, and i think that tar will
let them unchanged)

In my tar file there is, for example, a kind of separator like this (but with everything in one long line) with :

shag.py_0100744_0002033_0001750_00000004414_07500237361_0015314_0_ustar_00_shagshag_user_0000040_0000417_beginofmyfilehere

where _ stand for a variable amount of another ascii code that i can't cut/paste...

Do you kwow what each means (for example the first one is undoubtly the file name, but then...) ?
And what are the fixed position of each of theses ?
Have a clue ?

Many thanks,

s13.







From bdesth at nospam.free.fr  Fri Jun 21 15:09:05 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Fri, 21 Jun 2002 15:09:05 -0400
Subject: [newbie] import fails first time, then works :-[
Message-ID: <3D1379D1.5020705@nospam.free.fr>

Hi Everybody

I've got this strange problem :
when importing modules that wrap external modules (like socket and 
_socket etc), the first time, I get an ImportError
(no module named _[module]).
If i try a second time, it works ok ?-o

Exemple :
 >>> import socket
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "/usr/local/lib/python2.2/socket.py", line 41, in ?
     from _socket import *
ImportError: No module named _socket
 >>> import socket
 >>> socket
<module 'socket' from '/usr/local/lib/python2.2/socket.pyc'>
 >>>

A little precision :
the external modules are at the right place (PYTHONPATH/dyn-load)

Help !

laotseu

PS : python 2.2 on a linux-box



From idot at vt.edu  Sat Jun  8 21:35:57 2002
From: idot at vt.edu (Carl Banks)
Date: 8 Jun 2002 18:35:57 -0700
Subject: another near-sighted, time-wasting syntax enhancement idea
References: <e395b886.0206051956.5f447e9b@posting.google.com> <Xns92265AD72BE0cliechtigmxnet@62.2.16.82>
Message-ID: <e395b886.0206081735.1c1d0f4@posting.google.com>

Chris Liechti <cliechti at gmx.net> wrote in message news:<Xns92265AD72BE0cliechtigmxnet at 62.2.16.82>...
> idot at vt.edu (Carl Banks) wrote in 
> news:e395b886.0206051956.5f447e9b at posting.google.com:
> > class BigLongClassThatRunsUntilEOF::
> > 
> > def __init__ (self):
> >     pass
> > 
> > def whatever (self):
> >     pass
> > 
> > ...
> > 
> > Two colons after the class (or whatever) definition means the block
> > continues until EOF and is not indented.  Good for big, long classes
> > that take up a whole module.  Been suggested before, I'm sure.
> 
> i don't see any advantages for that. it's a very suptile difference almost 
> not viewable those "::", it makes code bad readable.


class NewsConverse:
        def write_reply (self, message):
                if self.poster_misses_point (message):
                        self.set_paragraph_reply("Code scrunched up "
                                                 "against the right "
                                                 "margin can be pretty "
                                                 "unreadable itself.")


> doesn't your editor has auto indentation? don't use notepad an switch to a 
> real editor <wink>.

I'm pretty sure if it's a feature, emacs has it.  <wink>.


-- 
CARL BANKS


From jon+usenet at unequivocal.co.uk  Tue Jun 25 11:32:41 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Tue, 25 Jun 2002 15:32:41 -0000
Subject: Web templating/db tool with best designer/coder separation?
References: <af2gh6$2c2$0@216.39.172.122> <af8dtp$5qp$0@216.39.172.122>
Message-ID: <slrnahh38p.sar.jon+usenet@snowy.squish.net>

In article <af8dtp$5qp$0 at 216.39.172.122>, Bengt Richter wrote:
[ZPT]
> and from the description, this would seem to have a real advantage over all of
> the above, in that it uses attributes of html tags as opposed to special tags
> that browsers or design tools won't generally understand and render usefully.

jonpy is also designed to be design-tool-proof, in that the only two
constructs it places into the HTML are simple strings in ordinary
text or attributes (where design tools will not mess with them) and
HTML comments to denote structure (which design tools also will not
generally mess with). So I don't think that ZPT has any advantage over
jonpy here.


From gerhard.haering at gmx.de  Sun Jun 30 22:53:00 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 1 Jul 2002 02:53:00 GMT
Subject: can't import generators
References: <mailman.1025400278.15389.python-list@python.org> <m2wusgvohb.fsf@mother.paradise.lost>
Message-ID: <slrnahvh0b.69e.gerhard.haering@lilith.my-fqdn.de>

John Hunter wrote in comp.lang.python:
>>>>>> "Tim" == Tim Peters <tim.one at comcast.net> writes:
> 
>    Tim> Start Python with the -v switch to see how imports are
>    Tim> getting resolved.  __future__.py is an actual module that
>    Tim> actually gets imported.  If you're picking up a wrong version
>    Tim> of the Python libraries, that would explain it (maybe due to
>    Tim> a bad PYTHONPATH setting, or some local module named
>    Tim> __future__.{py,pyc,pyo}).
> 
> Bingo -- thanks.  Old musty, crusty PYTHONPATH setting was the
> culprit.

There is no reason to set PYTHONPATH to the Python core libraries,
they are found automatically.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From claird at starbase.neosoft.com  Tue Jun 18 11:53:59 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 18 Jun 2002 10:53:59 -0500
Subject: Can python have lib to display PDF ?
References: <a6a2957e.0206180246.511fb932@posting.google.com> <aen3to$2a1$1@norfair.nerim.net>
Message-ID: <44A1879E37A77F29.9928F26C11A55C4C.811205F836B36D79@lp.airnews.net>

In article <aen3to$2a1$1 at norfair.nerim.net>, Gillou <nospam at bigfoot.com> wrote:
			.
			.
			.
>AFAIK, the - excelent - reportlab library does not provide features for
>changing an existing PDF.
			.
			.
			.
It does, but there is indeed a story to the case.

ReportLab has always had the ability to work with
existing .pdf instances, as opposed to generating
them "from scratch".  HOWEVER, this capacity was
for a long time limited to the for-fee license.

Just within the past six months, the no-charge
ReportLab library has begun to receive some of
these functions.  Maybe someone will make time to
explain the specifics.
-- 

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


From rchopra at myrealbox.com  Mon Jun 24 12:09:21 2002
From: rchopra at myrealbox.com (Rajat Chopra)
Date: Mon, 24 Jun 2002 12:09:21 -0400
Subject: File Copying
Message-ID: <1024934961.9bd8aff8rchopra@myrealbox.com>

How do you copy a file from one directory to another. I have looked all over and haven't found any documentation on it. Essentially, I want to copy a file from $SRC_DIRECTORY to $TARGET_DIRECTORY.

Does anyone know which modules are used? Or does anyone have some sample scripts?

Thanks,

-Rajat





From peter at engcorp.com  Mon Jun 17 21:20:08 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 21:20:08 -0400
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com>
Message-ID: <3D0E8AC8.6C5C96EE@engcorp.com>

George Hester wrote:
> 
> "Gerhard H?ring" <gerhard at bigfoot.de> wrote:
> > You did not import the sys module.

> Probably not.  Where do I import that from and how do it?  I am trying to put a
> script in ASP using Python as the script language to give me the version of
> Python that I am running when I call the ASP.  The more the script works out of
> the box the better I'll be able to catch on to this language.  Thanks.

It sounds to me like you should probably crawl before you walk.
I suggest downloading the tutorial and installing Python on
your own machine and using it interactively from the prompt
before you advance to something like ASP scripting.  Your
responses clearly show you aren't ready to write code in an
environment which is doubtless going to mask from you many
of the errors and messages which you will encounter as you learn.

In the above case you would be able to experiment and learn about
these things:

C:\>python
Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
>>> sys.version
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
NameError: name 'sys' is not defined

Here's where you would pause to learn a little more about Python....

>>> import sys
>>> sys.version
'2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)]'

-Peter


From gb at cs.unc.edu  Mon Jun  3 19:41:30 2002
From: gb at cs.unc.edu (gb at cs.unc.edu)
Date: 03 Jun 2002 19:41:30 -0400
Subject: Speeding up the Canvas Widget
References: <ad8jif$dki$1@netnews.upenn.edu> <3CFBA74E.C5EB9213@noaa.gov>
Message-ID: <uu1okexqt.fsf@cs.unc.edu>

PyOpenGL is the way to get really high performance. Virtually all of
the GUI toolkits have OpenGL canvases. OpenGL itself is window system
independent so you just need to establish a context using the
GUI and then use OpenGL calls to do your drawing. People think of it
as exclusively 3D but it does all the usual 2D stuff very well.

gb




From whisper at oz.net  Wed Jun 19 19:50:40 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 19 Jun 2002 16:50:40 -0700
Subject: file .seek() under MS-Windows
In-Reply-To: <4169349c.0206191318.7f0d99d9@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEKEEODKAA.whisper@oz.net>

I believe you'll find the base specificer to be 0,1 or 2 and never a
negative number.

Python doc:
seek(offset[, whence])
Set the file's current position, like stdio's fseek(). The whence argument
is optional and defaults to 0 (absolute file positioning); other values are
1 (seek relative to the current position) and 2 (seek relative to the file's
end). There is no return value. Note that if the file is opened for
appending (mode 'a' or 'a+'), any seek() operations will be undone at the
next write. If the file is only opened for writing in append mode (mode
'a'), this method is essentially a no-op, but it remains useful for files
opened in append mode with reading enabled (mode 'a+').

If by chance, you where trying to seek to EOF, then that's seek(2) (offset
from base optional). (N.B. "whence" can be signed, but can not be of a
magnitude to cause seeking before BOF or after EOF (if I recall correctly).)

HTH,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Frank Schaefer
> Sent: Wednesday, June 19, 2002 14:19
> To: python-list at python.org
> Subject: file .seek() under MS-Windows
>
>
> Hi,
>
> The following code:
>
>
> fh = open("myfile.pot", "r")
>
>
> ...
> fh.seek(-1,1)
> ...
>
> produces an error:
> [IOError 22] Invalid Argument
>
> under MS Windows (XP). It works fine under Unix, though (for years).
>
> I appriciate any help.
>
> Thanks,
>
> Frank.
> --
> http://mail.python.org/mailman/listinfo/python-list





From nothere at nowherre.com  Wed Jun 19 18:20:27 2002
From: nothere at nowherre.com (mark)
Date: Wed, 19 Jun 2002 18:20:27 -0400
Subject: old fashioned COM
Message-ID: <uh20956iiu8q52@corp.supernews.com>

I am a newbie at this, and haven't done a lot of programming in some
years.  The book I have on Python describes a lot about programming
internet and email apps.  I need to just select a COM port, set
parameters like baud rate, parity, etc. and do a file capture in dumb
terminal mode.  This may seem to simple in todays world, but I'm trying
to log some events from a piece of equipment that uses serial I/O.  

Help?

Thanks
Mark


From res04ft9 at gte.net  Sat Jun 29 03:43:36 2002
From: res04ft9 at gte.net (Erv Young)
Date: Sat, 29 Jun 2002 03:43:36 -0400
Subject: Most efficient way to write data out to a text file?
In-Reply-To: <3d1d1de6.425769562@netnews.attbi.com>
References: <3d1a6993.248534625@netnews.attbi.com>
Message-ID: <4.3.2.7.2.20020629034014.025b9570@mail.gte.net>

At 02:42 AM 6/29/2002 +0000, candiazoo at attbi.com wrote:
>Again, thank you for all of your input!  I was able to double the speed with
>only a few, minor changes.  It seems like the biggest change came when (and I
>don't know why) I was selective in importing my os, time and sys routines
>(asctime() in particular?!).  I also localized some class/function calls, used
>format strings instead of appending strings, etc.

Mike, if it would not be too much trouble, for the edifications of newbies 
like me, could you post your complete code in its before and after states, 
so we could see the matters that have been discussed, in context?

Thanks.

--Erv





From look at replyto.address.invalid  Tue Jun  4 23:44:05 2002
From: look at replyto.address.invalid (Greg Ewing)
Date: Wed, 05 Jun 2002 15:44:05 +1200
Subject: self
References: <BI5L8.78$Vv1.4155449@newnews.cc.stevens-tech.edu> <adiq31$e8f$1@wanadoo.fr>
Message-ID: <3CFD8905.872A43AF@replyto.address.invalid>

Eric Brunel wrote:
> 
> I always
> found the C++/Java convention particularly disturbing.

If you want something really disturbing, try the
Apple Object Pascal scope rules:

  type
    Foo = object
      a: integer;
      procedure blarg(a: integer);
    end;

  procedure Foo.blarg(a: integer);
  begin
    writeln(a);
  end;

Quickly, now -- which 'a' does that write out?
.
.
.
(answer below)
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

Believe it or not, it writes the *attribute* a,
not the parameter!

-- 
Greg Ewing, Computer Science Dept, 
University of Canterbury,	  
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg


From seeger at sitewaerts.de  Wed Jun 12 07:15:38 2002
From: seeger at sitewaerts.de (Felix Seeger)
Date: Wed, 12 Jun 2002 13:15:38 +0200
Subject: Reload an imported file
References: <ae76vb$4l6iq$1@ID-79019.news.dfncis.de> <3D071FC6.5000309@mxm.dk> <ae77sk$4bbl7$1@ID-79019.news.dfncis.de>
Message-ID: <ae7agq$4m055$1@ID-79019.news.dfncis.de>

Ok, I have it.
For the people who search the archives or groups:

>> test.py
#! /usr/bin/python -O
import time
import importme
from importme import Tester

while 1:
        print Tester.test
        reload (importme)
        from importme import Tester
        print Tester.test
        time.sleep(3)

>> importme.py
#! /usr/bin/python -O

class Tester:

        test = 'test'

and this part:
        reload (importme)
        from importme import Tester
in the __init__ part of every class.

have fun
Felix


From brian at sweetapp.com  Mon Jun 24 21:17:01 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Mon, 24 Jun 2002 18:17:01 -0700
Subject: Suggestions for good programming practices?
In-Reply-To: <3D17C148.6CC8B3D2@engcorp.com>
Message-ID: <013601c21be6$030a4d30$bd5d4540@Dell2>

Peter Hansen wrote:
> You _are_ returning them "in a struct".  It's spelled "tuple"
> but other than that what's the difference?

struct members are access by name while tuple elements are accessed
positionally. 

Cheers,
Brian





From mbelanger_NOSPAM at NOSPAM_pixelsystems.com  Wed Jun 26 10:17:09 2002
From: mbelanger_NOSPAM at NOSPAM_pixelsystems.com (Mathieu Belanger)
Date: Wed, 26 Jun 2002 10:17:09 -0400
Subject: Switch from 2.1 to 2.2
References: <f29c2586.0206250600.48e0a8e3@posting.google.com> <3D188302.1030109@thomas-guettler.de> <zv0S8.3036$%N4.731526@news20.bellglobal.com> <3D192C3E.6926BE61@engcorp.com>
Message-ID: <O%jS8.3322$ZM3.1139861@news20.bellglobal.com>

> Do you really mean that your stack trace, as print on the stdout of
> your machine, is truncated like that, in the middle of the word
> "site-packages"?  That's a pretty serious problem, I think.  Or was
> it just a cut-and-paste error?

Hahaha!! no, I know how to cut and paste :) It was finally not a problem
with python! My main program is written in C++ and I use Python and XML to
send a script to a server and the server give me back the answer in XML. I
just only increase the buffer length to receive my response and Tadam!!
everything works fine!! Don't ask me why!! I made a bunch of tests to see if
it will work forever and the bug didn't reapear!! So thanks anyway !! :)

Mathieu




From fredrik at pythonware.com  Sun Jun 23 18:10:01 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 23 Jun 2002 22:10:01 GMT
Subject: re.rsearch?
References: <slrnahcg6u.382.quinn@baht.ugcs.caltech.edu>
Message-ID: <ZGrR8.44834$n4.10519185@newsc.telia.net>

Quinn Dunkan wrote:
> Occaisionally I find myself wanting an re.search that will find the rightmost
> match.

just prepend ".*" to the pattern...

> Unfortunately, AFAIK, re.search also lacks a 'start_searching_at'
> type option.

if you use a compiled pattern, you can pass in start
and end offsets:

    p = re.compile(pattern)
    p.search(string, start, end)

</F>




From irmen at NOSPAMREMOVETHISxs4all.nl  Sat Jun 22 06:58:18 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Sat, 22 Jun 2002 12:58:18 +0200
Subject: How to stop a SocketServer?
References: <7xhejv9af2.fsf_-_@ruckus.brouhaha.com>
Message-ID: <3D14584A.9030307@NOSPAMREMOVETHISxs4all.nl>

Paul Rubin wrote:


> Then I instantiate the server and run the serve_forever() method.

Don't use the serve_forever() method.
Use handle_request() which handles a single request.
You can now build your own request loop that
exits cleanly, for instance when a 'stopLoop' flag
has been set.

One problem still with this: the loop only exits after
a request has been processed (handle_request will still block).

Irmen



From jepler at unpythonic.net  Sun Jun  9 23:11:48 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Sun, 9 Jun 2002 22:11:48 -0500
Subject: Screen resolution again
In-Reply-To: <3D039794.9549C54@bellatlantic.net>
References: <3D039794.9549C54@bellatlantic.net>
Message-ID: <20020609221147.B4546@unpythonic.net>

On Windows only (according to the manpage, anyway), you can also use
'wm_state' to maximize your window:
    from Tkinter import *
    root = Tk()
    canvas = Canvas(root)
    root.wm_state("zoomed")
It's unfortuante that this is not supported under Unix.

With modern Unix window managers, you can send a _WIN_STATE message to set
WIN_STATE_MAXIMIZED_VIRT |WIN_STATE_MAXIMIZED_HORIZ .. I don't know how
to do this in pure Tk(inter), though.  If you're into this kind of thing,
here's some C code that is down the right alley...

int maximize( Tk_Window tkw, int flag ) 
{ 
        Window w, root; 
    XClientMessageEvent xev; 
    Status res; 
    static Atom _WIN_STATE = 0; 
 
     
    if( !tkw ) { 
        return TCL_ERROR; 
    } 
     
    if( !_WIN_LAYER ) 
        _WIN_STATE = XInternAtom ( Tk_Display( tkw ), "_WIN_STATE", False); 
     
    root = RootWindow( Tk_Display( tkw ), Tk_ScreenNumber( tkw ) );     
    tkw = TkpGetWrapperWindow( tkw ); 
    w = Tk_WindowId( tkw );     
    xev.type = ClientMessage; 
    xev.window = w; 
    xev.message_type = _WIN_STATE; 
     
    xev.format = 32; 
    xev.data.l[0] = WIN_STATE_MAXIMIZED_VERT | WIN_STATE_MAXIMIZED_HORIZ; 
    if( flag ) 
        xev.data.l[1] = WIN_STATE_MAXIMIZED_VERT | WIN_STATE_MAXIMIZED_HORIZ; 
    else 
        xev.data.l[1] = 0;     
     
    res = XSendEvent ( Tk_Display( tkw ), root, False, SubstructureNotifyMask, 
     
    if( !res ) 
    {    
        Tcl_AppendResult("XSendEvent failed");
	return TCL_ERROR;
    } 
     
    XFlush (display); 
 
    return TCL_OK; 
}




From sholden at holdenweb.com  Tue Jun  4 12:08:24 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 4 Jun 2002 12:08:24 -0400
Subject: Automated Email - What did I do wrong here?
References: <adg1do$10nmr5$1@ID-49758.news.dfncis.de> <adifmg$11i5ks$1@ID-49758.news.dfncis.de> <adifsu$11ifbm$1@ID-49758.news.dfncis.de>
Message-ID: <cG5L8.232294$%u2.139939@atlpnn01.usenetserver.com>

"Ken" <ken at hotmail.com> wrote ...
> "Ken" <ken at hotmail.com> wrote ...
> > "Ken" <ken at hotmail.com> wrote
> > > How to do automated emailing when results being processed?
> > >
> > > Thanks
> > >
> >
> > What did I do wrong? It crash the program. Without these code, the cgi
> > script works fine. Thanks
> >
> >     fromaddr = 'ken at hotmail.com'
> >     toaddrs = 'ken at hotmail.com'
> >     msg = "testing\n"
> >     server = smtplib.SMTP('smtp.aol.com')
> >     server.sendmail(fromaddr, toaddrs, msg)
> >     server.quit()
> >
> oh...and I did import smtplib...

Ken:

You really should stop talking to yourself... <wink>


The "toaddrs" argument to sendmail must be a list, even if there's only one
recipient.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From gavrilov at attbi.com  Wed Jun 26 14:22:33 2002
From: gavrilov at attbi.com (Alexander Gavrilov)
Date: Wed, 26 Jun 2002 18:22:33 GMT
Subject: ASP only runs once
References: <3842735a.0206251112.25609e37@posting.google.com>
Message-ID: <JDnS8.164601$nZ3.75676@rwcrnsc53>

Download newest version of win32all extensions from
http://starship.python.net/crew/mhammond/win32/Downloads.html

Alexander Gavrilov

"Michael Anderson" <m_anderson14 at hotmail.com> wrote in message
news:3842735a.0206251112.25609e37 at posting.google.com...
> Hello,
>
> I've been using python in my asp pages for over a year now with no
> problems, and just ran into a problem I can't figure out. I installed
> it on a new machine and my pages only run the first time you hit them.
> After that they do nothing. Pages in VBScript and JavaScript run fine.
>
> If you edit the file and save it, it will run one time and then
> nothing. Since a regular script gets compiled to a .pyc on the first
> execution, I'm guessing it is somehow related. But I don't know how it
> works in an asp page. Has anyone had this problem? My configuration is
> Win2000 Pro, IIS 5, Python 2.1.3, win32all build 145. I also tried it
> with Python 2.2 and win32all 146 but get the same results.
>
> Thanks.
>
> Mike




From phd at phd.pp.ru  Thu Jun 20 04:01:39 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 20 Jun 2002 12:01:39 +0400
Subject: procmail replacement in Python
In-Reply-To: <oq660froj9.fsf@carouge.sram.qc.ca>; from pinard@iro.umontreal.ca on Wed, Jun 19, 2002 at 04:39:38PM -0400
References: <20020618235901.GA2197@lilith.my-fqdn.de> <oqit4grq03.fsf@carouge.sram.qc.ca> <20020619022110.GA6996@lilith.my-fqdn.de> <oq660froj9.fsf@carouge.sram.qc.ca>
Message-ID: <20020620120139.F22899@phd.pp.ru>

On Wed, Jun 19, 2002 at 04:39:38PM -0400, Fran?ois Pinard wrote:
> I wanted to peek at Spamassassin, which seems to have a good reputation.

   Recently they released version 2.30 which is very buggy. Do not use it,
download previous version, 2.20.

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From andreas at andreas-jung.com  Tue Jun  4 06:15:16 2002
From: andreas at andreas-jung.com (Andreas Jung)
Date: Tue, 4 Jun 2002 06:15:16 -0400
Subject: Which is fastest dict or list.index() ?
References: <mailman.1023078005.25452.python-list@python.org> <3CFC3D6A.5010706@lindbergs.org>
Message-ID: <01ea01c20bb0$b9ac6d70$02010a0a@suxlap>

"VanL" <news at lindbergs.org> wrote in message
news:3CFC3D6A.5010706 at lindbergs.org...

>
> I guess I am asking what you mean by sluggish. Sluggish as
> in "I get 10,000 ops per second on a Btree, but 100,000 ops
> per second on a dict" or sluggish as in "I only get 10 ops
> per second on a Btree because I have to go to disk every
> time I pull a node?"
>

It will be hard to beat Python dictionaries in performance since
they are very heavily optimized. Zope BTrees are optimized for
usage inside the ZODB and they can keep large data without loading
the complete datastructure into memory. See also
http://www.zope.org/Members/ajung/BTrees/

Andreas





From dreed at capital.edu  Sat Jun 29 16:51:11 2002
From: dreed at capital.edu (Dave Reed)
Date: Sat, 29 Jun 2002 16:51:11 -0400 (EDT)
Subject: private
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E201922151BA@admin56.narex.com>
	(BPettersen@narex.com)
References: <60FB8BB7F0EFC7409B75EEEC13E201922151BA@admin56.narex.com>
Message-ID: <200206292051.g5TKpB101434@fourier.capital.edu>

> From: "Bjorn Pettersen" <BPettersen at narex.com>
> 
> > From: Dave Reed [mailto:dreed at capital.edu] 
> > 
> [snip private members]
> > 
> > Kind of off-topic, but this got me curious. The only way I 
> > could think of is to assume (which I think is always true) 
> > that the private data members are in order from the starting 
> > address so based on the offset from the address of the object 
> > you could access each one.
> 
> .. and of course #define private public, or just go in and edit the .h
> file...
> 
> -- bjorn


Ok, I was thinking about doing it w/o recompiling the object or lib
file containing the class.

Dave





From rnd at onego.ru  Thu Jun  6 03:35:06 2002
From: rnd at onego.ru (Roman Suzi)
Date: Thu, 6 Jun 2002 11:35:06 +0400 (MSD)
Subject: prototyping good OOdesign in Python?
In-Reply-To: <e81be8b2.0206022315.3a0bf9a9@posting.google.com>
Message-ID: <Pine.LNX.4.44.0206061124230.15882-100000@suzi.com.onego.ru>

On 3 Jun 2002, Dean Goodmanson wrote:

(I oversaw this post and I feel I need to answer it)

> Ken Seehof <kseehof at neuralintegrator.com> wrote in message \
> > > Maybe this was already discussed, but I'd liked to raise this topic again.
> > >
> > > Python has superb OOProgramming support. Lets not dispute this for a
> > > moment. Now let's consider that we need to make well-designed OOP solution
> > > first prototyped in Python.
> > >
> > > What troubles are there? Python is too dynamic and all that. And all
> > > those nifty features need to be translated properly into, say, C++.
> > >
> 
> Wow! A topic that I have a keen interest in and slight experience in!
> 
> re troubles...
 
> - Implementing the design in Python using idioms that translate into
> stuff only possible through extraneous C++ code and/or STL.
> (range->foreach'isms, function pointers, ... mind is blanking ...
> python's data types and support sure are handy, but it does take a few
> extra brain cycles to write them in a C/C++ shop fashion 

>(note the C,
> in my experience only 20% of the advanced C++ topics are implemented
> by 80% of the shop, and the other code is given the black-art
> skepticism...

Hmmm... Who assesses the code quality? I guess if it is some very critical
software, its a case. But in other cases, who are there to tell you your
code is inacceptable as long as it works?

> in retrospect, that's a similar glare some folks get when
> mentioning Python. *sigh*)  "Making Python look like C++" might not be
> smart if you have the opportunity to keep the Python "Prototype".
> 
> I was hoping to seperated design from implementation in my
> enumerations, but it doesn't look like that's going to happen at this
> hour.
> 
> - Exception Handling & Asserts: Here I hope it's very similar.  Sorry,
> wasn't a requirement in my world, no specific advice. Project advice
> Include it in your original code & design prototype --adding it in
> later is the dreams turn sour, but it sounds like you've got
> appropriate time for up-front design.

Oh... 
 
> > Maybe this has already been discussed
> 
> I didn't go looking. If you do, please post the links.    I'm waiting
> for a software development book which uses Python for
> design/prototyping, unit testing, building & system testing *non
> Python projects*.
> 
> I do have lofty dreams of Python filling the role of Windows VBScript
> & friends to Unix scripting tools.

Python is used extensively by RedHat Linux, for example.
Even install is governed by Python script (anaconda).

What is probably missing, is Python-based shell. 
(I recall there is Perl-based shell.)
(Hmmm. Can you imagine:

>>> cp file1 file2
>>> 

)
 
> Back to the topic...
> 
> - I got bit by Python shortcuts.  Looping and Sorting an array.  It
> was so simple in python that it turned into an "assumption" in my
> prototype and was dropped when I coded the C++ version.  A smack on
> the forehead later and I started figured out how to sort my vector...

Aren't there any standard thing to sort in C++?
Or are you speaking about sorting some fancy data?
 
> - I'm not sure if this needs saying, but... STL is nifty and
> translates, but debugging is much harder than the comparable idioms in
> Python. Schdule time  for creating debugging support code in those
> objects. (if not for you, your peers)

> - I succesfully tested a sub-system in a Python test harness that
> allowed me to find bugs in it, protocol improvements, and that first
> wave of implementation perspective before my writing any of my C++
> code.
> I didn't get to SWIG/Boost/etc. wrapping my C++ for further tests.
> *sigh*    On that note, you  may want to peruse the boost.org/python
> documentation ( http://boost.org/libs/python/doc/index.html ) and the
> mailing list for a closer look at mapping C++ to Python.
 
> If you DO built unit tests in Python for your Python API, the majority
> should be re-usable in your python wrapped C++ implementation.

Interesting idea.

> - Finally, note data inputs as reading and interacting from different
> sources (existing libraries, goofy database connections, ??) might
> take double time mapping prototype (python) and your implementation
> (c++) to the source, along with who-knows-what "marshalling" issues.
> 
> Worse comes to worse, keep that shell handy as working out a small
> problem in python sure beats Visio and Excel. (actually, unless your
> reallly fast, I still used a spreadsheet for tabular perspective &
> printing on data dumps.)
> 
> As you can see, I've contributed more cue's than facts.  You're quest
> is a noble one, good luck in balancing research time with
> hammer-it-out-ignore-redudancies-just-get-it-done time.

> Maybe I should go back and read your code some more...
> ...maybe I should sleep.
> 
> Night,
> 
> Dean

Sincerely yours, Roman A.Suzi
-- 
 - Petrozavodsk - Karelia - Russia - mailto:rnd at onego.ru -
 





From gerhard at bigfoot.de  Mon Jun  3 12:11:48 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 3 Jun 2002 16:11:48 GMT
Subject: different gui-toolkits pros/cons...
References: <adfsoo$dmik4$1@hades.rz.uni-sb.de> <_rLK8.12$fd2.979@news.ecrc.de>
Message-ID: <slrnafn583.j2i.gerhard@lilith.my-fqdn.de>

Markus Jais wrote in comp.lang.python:
> with pyQT there might be problems with the license
> if you want to make your programm freely available

No, there are only problems when you legally want to use GPL'd Python
modules with it and want to distribute the result of your effort.
There is a "free for noncommercial use" edition of Qt/win32 that
should be fine for most in-house projects and for most other use
cases, too.

Qt/X11 is available under the GPL, which might be an ok license for
many people. Those that don't find the GPL and its implications
acceptable can buy a commercial Qt license. AFAIR, for PyQt use, the
Blackadder IDE includes such a license.

> pyGTK might be faster, because wxWindows is AFAIK an additional
> layer on top of Gtk. but only in Unix/Linux platforms, AFAIK

We had that discussion recently. If your algorithm isn't bad, it's
very unlikely that a GUI toolkit will be "too slow". Unless it is
Swing or maybe Tkinter ;-)

> pyGtk seems to be pretty fast, but I have made no tests

Define "fast". PyGTK loads a lot faster than wxPython or PyQt. AFAIC,
this is because PyQt and wxPython wrap _all_ of a very large toolkit,
that both (wxWindows and Qt) even include a lot of non-GUI stuff. This
might be solved by trying to split the one shared library into many
small ones, that are loaded on demand.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From deblNonooosmawahppmy at theworld.com  Mon Jun 17 01:09:45 2002
From: deblNonooosmawahppmy at theworld.com (David Lees)
Date: Mon, 17 Jun 2002 05:09:45 GMT
Subject: Minor install annoyance on Activestate Python 2.2.1 build 222
References: <3D0BF7B4.21096694@bellatlantic.net> <aehmr5$sc8$1@reader1.tiscali.nl>
Message-ID: <3D0D6F16.24952322@theworld.com>

A day later (Early Monday now) and I just went and started a fresh
download and all is fine with the name.

David


GerritM wrote:
> 
> "David Lees" <debl2Nonospawhammy at bellatlantic.net> schreef in bericht
> news:3D0BF7B4.21096694 at bellatlantic.net...
> > I just downloaded and installed the latest ActiveState build of Python
> > 2.2.1 off their site and could not get it to install on my win98SE box
> > even though I had previously run the windows installer.  The problem
> > turns on to be that the name of the file you download is
> > ActivePython-2.2.1-222.msi.exe
> > Simply remove the '.exe' extension so the name is:
> > ActivePython-2.2.1-222.msi and the installation goes just fine.
> >
> > David Lees
> 
> Quite strange, I installed ActivePython-2.2.1-222 last thursday, freshly
> downloaded and did not have this (or any other) problem.
> 
> regards Gerrit
> --
> www.extra.research.philips.com/natlab/sysarch


From mhammond at skippinet.com.au  Mon Jun  3 18:48:40 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 03 Jun 2002 22:48:40 GMT
Subject: New win32all versions for Python 2.1 and 2.2
Message-ID: <3CFBF2BE.6020508@skippinet.com.au>

Hi all,
   I have released new win32all builds.  They have a number of changes, 
most notably the ASP refresh problem has been fixed (properly, so ASP 
with Python should even be faster).  The Pythonwin interactive window 
also has word-wrap!

These binaries are new (and have been built with WISE 8 - thanks Tim!) 
so are still considered experimental.  They seem to work wekk though :)

All the details, including binaries, can be found via 
http://starship.python.net/crew/mhammond

Mark.



From hesterloli at hotmail.com  Tue Jun 18 16:56:03 2002
From: hesterloli at hotmail.com (George Hester)
Date: Tue, 18 Jun 2002 20:56:03 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com>
Message-ID: <D7NP8.46301$GY.14487785@twister.nyroc.rr.com>

I just cannot believe what I ran into in this newsgroup.  If you people want to
be paid for your suggestions then that is fine.  But this newsgroup does not
have the facility set up for that to happen.  Then since it don't it is
courteous not to give advice if that is what you really want. It almost seems
that I was given half an answer so that the other half could be provided by
someone that would try to make a buck off me.

Thank you Duncan for finally seeing the error of thier ways.

Don't worry I won't return here.  This is not a group to get help but a group of
individuals that think the world owes them something when they participate here.

--
George Hester
_________________________________
"George Hester" <hesterloli at hotmail.com> wrote in message
news:K7PO8.16057$uk2.8406537 at twister.nyroc.rr.com...
> What is a simple script that will tell me the version of python I am currently
> running?  IIS 5.0 Windows 2000.
>
> --
> George Hester
> _________________________________
>
>




From eikeon at eikeon.com  Tue Jun 18 14:47:15 2002
From: eikeon at eikeon.com (Daniel 'eikeon' Krech)
Date: 18 Jun 2002 14:47:15 -0400
Subject: pychecker + XEmacs
In-Reply-To: <20020618175005.GA613@lilith.my-fqdn.de>
References: <3d0de483$0$3881$e4fe514c@dreader4.news.xs4all.nl>
	<aekok8$7p118$1@ID-59885.news.dfncis.de>
	<20020618165548.GB29257@matijek.plusseven.com>
	<20020618175005.GA613@lilith.my-fqdn.de>
Message-ID: <un0ts4e6k.fsf@eikeon.com>

Gerhard H?ring <gerhard at bigfoot.de> writes:

> * Alex Polite <m2 at plusseven.com> [2002-06-18 18:55 +0200]:
> > Interesting.  Where is the CVS repository for python-mode.el?
> 
> It's in the Python CVS repository. Directory dist/src/Misc/

I had trouble finding it when I was looking around too. When does the
one on http://www.python.org/emacs/python-mode/ get updated... 

Could a link to the CVS version be added to the python-mode page? 

--eikeon, http://eikeon.com/





From peter at engcorp.com  Sun Jun 23 22:46:45 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 23 Jun 2002 22:46:45 -0400
Subject: Suggestions for good programming practices?
References: <WEuR8.2074$5l4.75657@ozemail.com.au>
Message-ID: <3D168815.ADA19BCE@engcorp.com>

Dianne van Dulken wrote:
> 
> I'm fairly new to python, coming from a perl background, and was wondering
> if anyone has a list of things that they consider as good programming
> practice in Python (I thought this might be an interesting topic for
> discussion, anyway)
> 
> For example, in perl you obviously are always encouraged to use strict, and
> we always use eval around our main code to pick up any unexpected errors.  I
> presume the equivalent in python is the try statement.  Would you suggest
> that this should be used, or would that be redundant here?  Any other tips
> that you would suggest a new python user should be always doing as a good
> programming practice?

While it's not quite an answer to the question you asked, you should
read this:

  http://www.python.org/doc/essays/styleguide.html

I would say in addition to that, that unlike many other languages you
don't need to worry too much about other good programming practices
if you just write simple, clean Python code.  Make your goal to 
write the most readable and maintainable code you can, and all else
will follow.

As for try/except... if you want you could put a non-specific 
try/except around the entire application, but all that would be good
for is to mask fatal errors and prevent the interpreter from printing
them properly to show you what went wrong.  Not sure why you'd want
to do that in smaller programs at all, though in large ones you might 
catch the exceptions and write them to an error file for later analysis.

-Peter


From fredrik at pythonware.com  Sat Jun 29 13:16:48 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 29 Jun 2002 17:16:48 GMT
Subject: I'm an idiot
References: <mailman.1025314959.23032.python-list@python.org>
Message-ID: <4YlT8.45607$n4.10833068@newsc.telia.net>

Sean 'Shaleh' Perry wrote:
> the python idiom is usually either:
>
> while 1:
>   line = f.readline()
>   if not line: break

no, the idiom is:

    while 1:
        line = f.readline()
        if not line:
            break
        ...


</F>




From opengeometry at yahoo.ca  Thu Jun  6 17:53:49 2002
From: opengeometry at yahoo.ca (William Park)
Date: Thu, 6 Jun 2002 17:53:49 -0400
Subject: Efficient python programming...
In-Reply-To: <adoa9u$vqed$1@ID-146704.news.dfncis.de>
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de>
Message-ID: <20020606215349.GA13474@node1.opengeometry.net>

On Thu, Jun 06, 2002 at 08:40:15PM +0200, Shagshag13 wrote:
> 
>  I'm looking for any state of art, optimizing techniques, tips, that a
> python beginner couldn't know by itself, but should learn and
> would be aware to write "efficient and good looking"python's code...

Actually, I don't recommend this approach.  I recommend you look at your
algorithm and try to translate that to Python syntax.  Optimization should
occur at the modelling and algorithm side of things.

> 
>  (not the kind of thing which made you an obfuscated python nerd,
> but that "map" should be preferred to "for in loop" if possible, that
> dictionary are very well optimized, etc.)

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin




From andymac at bullseye.apana.org.au  Wed Jun  5 04:32:41 2002
From: andymac at bullseye.apana.org.au (Andrew MacIntyre)
Date: Wed, 5 Jun 2002 19:32:41 +1100 (edt)
Subject: libpython as a shared library?
In-Reply-To: <20020604143402.GB22993@hal.mediasupervision.de>
Message-ID: <Pine.OS2.4.32.0206051931120.1678-100000@tenring.andymac.org>

On Tue, 4 Jun 2002, Gregor Hoffleit wrote:

> Debian has had libpython as a shared lib since a while. Recently, we got
> a report that the performance of the shared version is dramatically
> slower than the statically linked version (e.g. about 30%). I made some
> pystone tests on different architectures of our GNU/Linux, and tried
> different flags for building and linking the library (and different
> Python versions), and in fact there seems a significant speed penalty
> involved as soon as the interpreter is linked against the shared
> library. The penalty has some correlation to the architecture, but it is
> significant on all tested constellations.

A result of PIC code in the shared library?  I've not been motivated
enough to try building a shared lib on FreeBSD.

--
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andymac at bullseye.apana.org.au  | Snail: PO Box 370
        andymac at pcug.org.au            |        Belconnen  ACT  2616
Web:    http://www.andymac.org/        |        Australia





From boud at valdyas.org  Mon Jun  3 15:41:01 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Mon, 03 Jun 2002 21:41:01 +0200
Subject: different gui-toolkits pros/cons...
References: <adfsoo$dmik4$1@hades.rz.uni-sb.de> <_rLK8.12$fd2.979@news.ecrc.de> <adg4e0$dq6mb$3@hades.rz.uni-sb.de>
Message-ID: <3cfbc67b$0$52024$e4fe514c@dreader3.news.xs4all.nl>

Uwe Schmitt wrote:

> Markus Jais <mjais at web.de> wrote:
> | with pyQT there might be problems with the license
> | if you want to make your programm freely available
> | (but I am not a lawyer, so I might be wrong)
> 
> That's a very important point: what do I have to pay if I used pyQt
> for development ??? I did not find any explicit prices at the
> trolltec-page...

Trolltech doesn't sell PyQt: you need to go to http://www.thekompany.com
for BlackAdder. 

-- 
Boudewijn Rempt | http://www.valdyas.org


From tim.one at comcast.net  Mon Jun 10 17:10:25 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 10 Jun 2002 17:10:25 -0400
Subject: Where is Berkeley DB on your system? Can you live without DB 1.85?
In-Reply-To: <83wut6q2i3.fsf@panacea.canonical.org>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEOIPLAA.tim.one@comcast.net>

[Kragen Sitaker, to David LeBlanc]

[on the bsddb.pyd installed by the PythonLabs Windows installer]
> That itself is probably the Python module that interfaces to whatever
> version of Berkeley DB you have, not Berkeley DB itself (although it
> could be statically linked in there),

Sam Rushing's Windows port of Berkeley DB 1.85 is statically linked in
there.  Read PCbuild/readme.txt in the source distribution; for the
convenience of compiler-less Windows users, we build and ship several
precompiled 3rd-party components in the Windows distro.

[on what the pitfalls of 1.85 are]
> I don't know what they are because I use sane versions of Berkeley DB.

See "2. Are there known problems with the 1.85 and 1.86 versions?" at

    http://www.sleepycat.com/historic.html

1.85-is-hopeless-ly y'rs  - tim





From peter at engcorp.com  Sat Jun 22 20:36:34 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 22 Jun 2002 20:36:34 -0400
Subject: position of matches
References: <653cf871.0206221452.6cb17ef2@posting.google.com> <3D15175E.74934C43@engcorp.com>
Message-ID: <3D151812.4796A10D@engcorp.com>

Peter Hansen wrote:
> 
> les wrote:
> >
> > i am new to python and want to do the following:
> > given a string str='abcABCDefsgRSTVderDae'
> > i would like to find all the internal lowercase strings between 2 caps
> > i.e. pattern=re.compile(r'[A-Z]([a-z]+)[A-Z]')
> >
> > match_obj=pattern.search(str)
> > begin,end=match_obj.span()
> >
> > however i would like to get all the begining and end positions
> > of the pattern,
> > i.e.
> > efsg  begin=7 end=10
> > der   begin=15 end 17
> 
> Try this:
> >>> for x in m.finditer(s):
> ...   print '%s\tbegin=%s end=%s' % ((x.group(1),) + x.span(1))

Oops, if it wasn't obvious, "m" is "pattern" from above...
sorry.

-Peter


From gnufnork at fuckspam.hetdigitalegat.nl  Mon Jun 10 23:00:52 2002
From: gnufnork at fuckspam.hetdigitalegat.nl (Wladimir)
Date: Tue, 11 Jun 2002 05:00:52 +0200
Subject: newbie question - python blocking on int()?
References: <4LaN8.2488$Vr2.586019@news20.bellglobal.com> <Xns922A149F19723cliechtigmxnet@62.2.16.82>
Message-ID: <pan.2002.06.11.03.00.52.273636.25803@fuckspam.hetdigitalegat.nl>

On Tue, 11 Jun 2002 01:59:33 +0200, Chris Liechti wrote:

> also note that split() might not be safe for every case. e.g when the
> user name contains a space (if possible) and surely when the filename
> has a space in it.

Parsing the output of ftp.dir in any way is never comphrehensively safe
because it it platform dependent according to the FTP RFC. Spaces in
usernames are not possible though AFAIK, and most servers use a unix ls
like semantic for dirs so it's *pretty safe*. (although I've seen ugly DOS
like listings as well)

Wladimir


From mt at open2web.com  Thu Jun 20 23:24:58 2002
From: mt at open2web.com (Don Low)
Date: 21 Jun 2002 03:24:58 GMT
Subject: newbie string question
References: <aeu40s$9jg2s$1@ID-145503.news.dfncis.de> <3D129BB7.C7EA15D2@engcorp.com>
Message-ID: <aeu6aa$a93ut$1@ID-145503.news.dfncis.de>

Thanks for the great explanation.

Mark


From jason-exp-1025465152.e12cd5 at mastaler.com  Tue Jun 25 15:25:52 2002
From: jason-exp-1025465152.e12cd5 at mastaler.com (Jason R. Mastaler)
Date: Wed, 26 Jun 2002 00:25:52 +0500 (AQTST)
Subject: bug in os.getgroups?
Message-ID: <1025033152.17806.TMDA@nightshade.la.mastaler.com>

I've noticed that if you set both the real and effective uid and gid
of a process (as root), os.getgroups() still shows root's supplemental
groups listed, yet you cannot delete a file that one of the
supplemental groups has permission to delete.  Any explanations?

For example:

# id johndoe
uid=667(johndoe) gid=31(guest) groups=31(guest)

# id root
uid=0(root) gid=0(wheel) groups=0(wheel), 2(kmem), 3(sys), 4(tty), 5(operator), 20(staff), 31(guest)

# ls -l /tmp/junk
-rw-rw-r--  1 root  staff  0 Jun 25 13:19 /tmp/junk

# python
Python 2.2.1 (#1, Apr 22 2002, 10:19:01) 
[GCC 2.95.3 20010315 (release) [FreeBSD]] on freebsd4
Type "help", "copyright", "credits" or "license" for more information.
>>> import os
>>> os.setgid(31)
>>> os.setuid(667)
>>> os.getgroups()
[31, 0, 2, 3, 4, 5, 20, 31]
>>> os.unlink('/tmp/junk')
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
OSError: [Errno 1] Operation not permitted: '/tmp/junk'
>>> 




From cliechti at gmx.net  Sun Jun 23 16:59:43 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 23 Jun 2002 22:59:43 +0200
Subject: Python hits the spot
References: <mailman.1024772248.9399.python-list@python.org> <yppfsn3dlpco.fsf@hotlips.cs.chalmers.se>
Message-ID: <Xns9236EA5AAE01Fcliechtigmxnet@62.2.16.82>

kurlberg at math.chalmers.se (P?r Kurlberg) wrote in
news:yppfsn3dlpco.fsf at hotlips.cs.chalmers.se: 
> Funny you should mention this.  On my laptop (IBM Thinkpad with 128
> megs of ram) the following two-liner gets me a lockup (i.e. keyboard
> and mouse completely unresponsive, powertoggle needed):
> 
> for i in range(50*1000*1000):
>    pass
> 
> (I am serious.)  Happened some time ago, but I never got around to
> filing a bug report (this thread was a good reminder.)  Seems like the
> python code exercises the linux VM/swap to the breaking point.  It is
> rather annoying that a user space program can do this...
> 
> Can someone else reproduce this behaviour?  (Make sure to quit all
> important applications before you try!)

i had to lower te number to not get a MemoryError on my machine (96MB)
it runs uses some time, begins to use swap and then gets killed by the OOM 
Killer.
(Kernel 2.4.17, Debian Woody Linux)

the old kernel 2.2 went crazy when no more memory was left. thats fixed 
with the Out-Of-Memory-Killer in newer kenels.

it is _not_ a python problem!

for such loops like the above 'xrange' was invented.

chris

-- 
Chris <cliechti at gmx.net>



From carel.fellinger at iae.nl  Sun Jun 16 19:52:50 2002
From: carel.fellinger at iae.nl (Carel Fellinger)
Date: Mon, 17 Jun 2002 01:52:50 +0200
Subject: Snort alert tail...
In-Reply-To: <AB0P8.107$W91.886@news1.global-one.fi>
References: <AB0P8.107$W91.886@news1.global-one.fi>
Message-ID: <20020616235250.GA5686@schaduw.felnet>

On Sun, Jun 16, 2002 at 03:51:45PM +0300, Jan-Eric wrote:
> HEllo !
> I'm trying to write a dynamic firewall script in python that scans the Snort 
> alert file like 'tail -f' and takes action based on the infomation it gets 
> from that file. But I can't get the 'tail' function to work.It reads the 
> file, but any new information that Snort is writing to the file doesn't 
> show up to the script.
> 
> ex.
> file = open('/var/log/snort', 'r')
> while 1:
>         file = file.read()
>         print file

I expect you're getting exceptions here, or you don't show the real code.
The line `file = file.read()' turns your file opject into a string object
and that string object doesn't have a read method, so the next round in
the while loop should raise an exception.

Others showed you more complete ways of achieving this, but the following
might work for you too:

   f = open('/var/log/snort')
   while 1:
      data = f.read()
      if data:
          print "GOT DATA:", `data`
      else:
         time.sleep(1)


I left out the secand argument in the open call as it defaults to "r".
I prefer to use f instead of file as file is a buildin (aka open) in
newer Pythons. And finally, it's advisable to pause after reading
nothing from the log file to let others fill the logfile again.

-- 
groetjes, carel





From dave5774 at hotmail.com  Tue Jun 18 16:51:42 2002
From: dave5774 at hotmail.com (Dave)
Date: Tue, 18 Jun 2002 16:51:42 -0400
Subject: Win32com Python script/MS-Access Please help!!
Message-ID: <aeo6h1$itf$1@netnews.upenn.edu>

Hey all,
I'm having trouble writing a Python script relating to Microsoft
Access.  If
anyone could help me I would greatly appreciate it.  Here's the
problem :

I built an MS-Access table with 5 columns(fields) : OrderID, OrderDate,
PlantID, ProductID, OrderQty, with OrderID as the primary key.  All
fields
are numeric, except for OrderDate(which is a date field).

Now I want to populate this table with data from a web page
(http://opim.wharton.upenn.edu/~opim101/spring02/Plant1.html).  please
check
the site.

My object is to read the the data and insert it record by record into my
Access table.

Here's what I have so far :

import urllib, string
Plant1 =
urllib.urlopen
('http://opim.wharton.upenn.edu/~opim101/spring02/Plant1.html'
)
lines1 = Plant1.readlines()
import win32com.client
engine = win32com.client.Dispatch("DAO.DBEngine.35")
db=engine.OpenDatabase(r'd:\Plants.mdb')

I have several problems :
Firstly, when I read the lines, I notice that not all of them are
relevant.
I need to ignore <html>, <body>, etc. at the beginning of the file, as
well
as the column headers, so that the action really begins from line 4 of
the
file and ends with </body>
Also, I need to eliminate <p> from every line in the file in order to
get to
the five comma-delimited fields that follow, and then split the rest of
the
string to get to each field.  Here's what I think part of the code
should
look like :

for line in lines1.readlines():
    line=line.rstrip()
    inOrderID,inOrderDate,inRegionID,inProductID,inOrderQty =
line.split(",")

Something like that, I'm not sure...
Lastly, now that I've read all the fields, I need to insert them each
as a
single record into my table of my access database using Python DAO
using the
Insert statement.  Also, the OrderDate field that I read in, looks like
a
text type to Access, so I need to pad it with the # sign on either side
before inserting it into the Access database, so that Access interprets
it
as a Date/Time type.

Finally, I need to get the same information from 2 other plants and tag
them
onto the same table...from the web page
http://opim.wharton.upenn.edu/~opim101/spring02/PlantPortal.html
where plant1's data shows up first, plant2's data shows up below it, and
plant3's data shows up last...so all plants' data in the same table.

I would like
to use more standard commands.  I can use SQL, and I know Python -
Access
commands like OpenRecordSet(), TableDefs.count(), Execute(), while not
rs.EOF, Fields().....
Is there any other way to write the program without SGML?  To give you
more
of an idea of the way I'd like the program, I wrote another program for
doing a similar thing in Excel, not Access, for a file named sales.txt
and
the same 5 columns, but put from a file into 5 columns on excel, and
certain
parameters to enter at the beginning... :

ProductID = float(raw_input("Please enter the ProductID: "))
OrderQty = float(raw_input("Please enter the cutoff order quantity: "))
CSV = open('c:/sales.txt','r')
import win32com.client
x1=win32com.client.Dispatch('Excel.Application')
x1.visible=1
x1.Workbooks.Add()
x1.Range("A1").Select()
x1.Range("A1:B1").Value = ("ProductID:",ProductID)
x1.Range("A2:B2").Value = ("OrderQty:",OrderQty)
x1.ActiveCell.Offset(4).Activate()
x1.ActiveCell.Range("A1:E1").Value=
("OrderID","OrderDate","RegionID","Produc
tID","OrderQty")
x1.ActiveCell.Offset(2).Activate()
for line in CSV.readlines():
    line=line.rstrip()
inOrderID,inOrderDate,inRegionID,inProductID,inOrderQty = line.split
(",")
if float(inProductID) == ProductID:
       if float(inOrderQty) >= OrderQty:
           xl.ActiveCell.Range("A1:E1").Value =
(inOrderID,inOrderDate,inRegionID,inProductID,inOrderQty)
xl.ActiveCell.Offset(2).Activate()

I just don't know how to adjust this type of thing for Access...

If anyone could help me by writing the script I would greatly appreciate
it...
Thank you so much.
Sincerely,
Dave








From peter at engcorp.com  Wed Jun  5 22:42:38 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 05 Jun 2002 22:42:38 -0400
Subject: Interfacing to 1-Wire LAN, especially Dallas temperature sensors?
References: <a7iefuk650noamrqcmpvroqj60mtgbpf9v@4ax.com> <mailman.1023319090.27530.python-list@python.org>
Message-ID: <3CFECC1E.FECDF675@engcorp.com>

"James J. Besemer" wrote:
> 
> It's possible you'll have better results via the parallel port:
> 
>     http://www.google.com/search?hl=en&q=i2c+parallel+port
> 
> "I2C" is similar to the two-wire protocol (data plus clock) used on the
> dallas parts.

True, but just to avoid possibly misleading anyone: there is almost
nothing in common between two-wire protocols like I-squared-C and the
One-Wire protocol from Dallas.  The latter is highly timing-sensitive,
as well as requiring a bidirectional pin, or one input plus one
open-collector output pin.  Tricky stuff.

-Peter


From gerhard.haering at gmx.de  Wed Jun  5 13:39:55 2002
From: gerhard.haering at gmx.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Wed, 5 Jun 2002 19:39:55 +0200
Subject: [Pypgsql-users] Re: [ANN] pyPgSQL 2.1 released
In-Reply-To: <20020605164010.GB32960@gargamel.hqd-internal>
References: <mailman.1023270308.20913.python-list@python.org> <23891c90.0206050721.49b31f50@posting.google.com> <20020605164010.GB32960@gargamel.hqd-internal>
Message-ID: <20020605173955.GA33711@gargamel.hqd-internal>

* Gerhard H?ring <gerhard.haering at gmx.de> [2002-06-05 18:40 +0200]:
> * Paul Boddie <paul at boddie.net> [2002-06-05 08:21 -0700]:
> > [...] Another thing: pyPgSQL seems to be fine with Unicode query strings
> > (unlike psycopg, apparently), but it doesn't like Unicode parameter values.
> > Is there some explicit encoding step I should be performing?
> 
> I've fiddled with adding Unicode support for some time now, but I didn't get
> enough testers yet to make me comfortable enough with it for adding it to a
> pyPgSQL release. On the Sourceforge patches section, I've uploaded a patch
> for full Unicode support, including getting Unicode results back from the
> database, insert Unicode strings, and automatic encoding conversions. I'll
> try to update this patch today to work with the latest pyPgSQL 2.1 release
> and drop you a note when it's ready.

Done now. I like vimdiff :-)

You can get the experimental Unicode patch that applies cleanly to pyPgSQL 2.1
at the following long URL:

http://sf.net/tracker/index.php?func=detail&aid=484468&group_id=16528&atid=316528

There isn't much docs (read: any) yet, but you can look into the Unicode test
suite (test/unicode_tests.py) that the patch will install for usage of the
various Unicode features.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From max at alcyone.com  Tue Jun 25 04:14:26 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 25 Jun 2002 01:14:26 -0700
Subject: 'parent object'?
References: <3d18228c$0$221$4d4ebb8e@news.nl.uu.net>
Message-ID: <3D182662.CBC925DB@alcyone.com>

"Guyon Mor?e" wrote:

> Is there, like 'self, also a 'parent' object of some sort?
> 
> The problem i'm having is that i have a class, which contains a list
> of
> instances of another class. How can i let this 'childclass' call a
> function
> which is on the 'parent class'?

An instance has access to its class (.__class__ attribute), and a class
has access to its parent classes (.__bases__ attribute).

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ See the son in your bad day / Smell the flowers in the valley
\__/ Chante Moore
    Bosskey.net: Aliens vs. Predator 2 / http://www.bosskey.net/avp2/
 A personal guide to Aliens vs. Predator 2.


From merkosh at hadiko.de  Fri Jun  7 05:18:34 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Fri, 7 Jun 2002 11:18:34 +0200
Subject: common approach in socket programming
Message-ID: <MPG.176a98d24f0689e1989680@news.rz.uni-karlsruhe.de>

hi,

i want to use the module SocketServer to create a tcp server. with it i 
want to work on arbitrary packet lengths. f.e. from the contents of each 
packet an md5 hash is created and send back. so it makes a difference 
weather i hash each single packet or collect packages and then create 
the hash.
SocketServer offers two possibilities to deal with incomming 
connections. either use self.request which is a socket object, or use 
the file wrapper.
now my problem is how to know that 

a) a connection is still open

i couldn't find a opened() or isopen() function or a similar attribute 
in the documentation. not for the socket object. for the file object 
there should be the open attribute.

b) how do i know the size of each packet?

using the file wrapper the method calls block until the buffer has been 
filled. i could set the maximum buffer size to 1 byte, but thats somehow 
odd to read data bytewise - and: how do i know i haven't crossed packet 
boundaries to a newly arrived packet while i was busy receiving single 
bytes from the tcp/ip layer?


what is the usual solution to something like that?
i don't really want to use own header information to determine packet 
lengths nor would udp packeges suffice because transmission isn't 
certified.

Any ideas?
Thanks in advance
Uwe


From BPettersen at NAREX.com  Fri Jun 21 19:45:26 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 21 Jun 2002 17:45:26 -0600
Subject: do...while
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192F1A76D@admin56.narex.com>

> From: Oleg Broytmann [mailto:phd at phd.pp.ru] 
> 
> On Fri, Jun 21, 2002 at 12:53:02PM -0700, Chris Barker wrote:
> > while line = file.readline():
> 
>    In Python it is spelled
> 
> for line in file.xreadlines():
>    ...

Actually, now it's spelled:

  for line in file:
     ...

<wink>

-- bjorn




From garrett at bgb.cc  Mon Jun 10 20:57:11 2002
From: garrett at bgb.cc (Don Garrett)
Date: Tue, 11 Jun 2002 00:57:11 GMT
Subject: Why does Python mix OO concepts and non OO concepts for operation
  s  on basic types?
References: <mailman.1022097199.9970.python-list@python.org> <3CEC1506.B0DA1787@earthlink.net>
Message-ID: <3D054A98.3090708@bgb.cc>

Hans Nowak wrote:
> Besides, while Python is object-oriented, that isn't the only
> paradigm in the language. People coming from a functional
> background may find the len() function more natural than a
> method.
> 
> YMMV,
> 

   That's one thing I don't get. Why wasn't len() added as a member to the 
various types during the recent rework that added so many other members to the 
native types?

-- 
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc



From whisper at oz.net  Sun Jun  9 03:31:18 2002
From: whisper at oz.net (David LeBlanc)
Date: Sun, 9 Jun 2002 00:31:18 -0700
Subject: Newby: Dictionaries as file objects?
In-Reply-To: <Xns9228AB206D06CAusYourStandingInIt@130.133.1.4>
Message-ID: <GCEDKONBLEFPPADDJCOECEKDDHAA.whisper@oz.net>

Take a look at Pickle, cPickle and Shelve in the pythondoc - these packages
support saving arbitrary python objects in files (with some restrictions)

HTH,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of netvegetable
> Sent: Sunday, June 09, 2002 0:15
> To: python-list at python.org
> Subject: Newby: Dictionaries as file objects?
>
>
> Hi.
>
> I want to be able to store a list on a file object, and I want each list
> item to be a dictionary, complete with its own set of keys and values.
>
> I know I can store a list in a file object, simply by writing
> each item to
> the file as a string , and making sure there's a "\n" character
> at the end
> of it. When I read the file object again, I merely need to open it as
> "open(object).readlines()" and hey presto, my list is back in order.
>
> So is there a way to conveniently store a dictionary in each item as well?
>
> Ta.
>
> --
> netvegetable at excite.com
> --
> http://mail.python.org/mailman/listinfo/python-list





From none at example.com  Fri Jun 28 13:24:31 2002
From: none at example.com (Reclusive Schmoe)
Date: Fri, 28 Jun 2002 17:24:31 GMT
Subject: Can i write chating system using python
References: <3d1c8661$1_1@news.iprimus.com.au>
Message-ID: <jZ0T8.10194$qf1.550435012@newssvr21.news.prodigy.com>

There are libraries for this on freshmeat and sourceforge.
Also... search Google.

In article <3d1c8661$1_1 at news.iprimus.com.au>, "tritran"
<tritran2001 at iprimus.com.au> wrote:

> Can anyone provide me the website or resource to create the chating
> system or instance messeger using python.
> 
> 


-- 
"Don't pretend you want to talk to me, I know you hate me."


From wangzhao_wz at hotmail.com  Fri Jun 14 03:20:10 2002
From: wangzhao_wz at hotmail.com (zhao)
Date: Fri, 14 Jun 2002 10:20:10 +0300
Subject: How can Python bindings for the Qt toolkit?
Message-ID: <aec5fa$e3u$1@bowmore.utu.fi>

Hi,
I don't understand how the Sip create Python bindings for C++ libraries. How
does Sip Python module provides support functions to the automatically
generated code? I have installed Sip, Python2.2.1 and Qt3.0.4 in Windows.
But if I write python code, how can it be changed to C++ code so that I can
use it under Qt software. Can anybody give me the explanation in details and
how to install it? And when I install PyQt and run pyuic.exe, it always says
that 'the dynamic link library qt-mt230nc.dll could not be found in the
specified path'. Why?

Thanks for your help!

zhao




From James_Althoff at i2.com  Wed Jun  5 16:25:27 2002
From: James_Althoff at i2.com (James_Althoff at i2.com)
Date: Wed, 5 Jun 2002 13:25:27 -0700
Subject: Thought on PEP 204 and 276
Message-ID: <OFE3927B8E.9BE67D0C-ON88256BCF.006F9838@i2.com>

[phil hunt]
>I used to think that
>
>   fir i in x:
>
>where (x) is an intreger, was confusing. But lately, I've
>found myself accidently spontaneously typing it myself
>instead of range(x) so i guess it makes intuitive sense
>after all.

I "accidently spontaneously" type it too.  But of course, I'm not very
objective (oops, dangerous wording <wink>) about this.

Thanks for making my day!

Jim






From cliechti at gmx.net  Thu Jun  6 18:19:43 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 7 Jun 2002 00:19:43 +0200
Subject: writing a filter in python
References: <adm43t$p5j$1@news.asu.edu>
Message-ID: <Xns92263B9D2392cliechtigmxnet@62.2.16.82>

David Bear <david.bear at asu.edu> wrote in news:adm43t$p5j$1 at news.asu.edu:

> I'm writing a filter to be used by lprng.  Since there is no stdout I
> am wondering two things:
> 
> 1) good strategies for showing error message (was thinking of syslog)

i created my own /var/log/myfilter and redirected sys.stdout adn std err to 
that file. that way you don't miss any important exception, which makes 
debuging easier.

> 2) an easy way to catch all exceptions and quickly return back to the 
> caller (lpd)  

try:
    	... your code here
except:
    	traceback.print_exc(file=yourlogfile_or_sysstderr)

why do you need to be quick? i thing lpr only tries to restart your job if 
it doesn't run properly. but when it runs it may take a long time.
(atleast it wrork for my pdf printer, which i controlol through a python 
script and lprng)

chris

> I'm noticing that if my script doesnt run fast enough I may end up
> have lpd try multiple times to run it.  
> 
> Any advice?



-- 
Chris <cliechti at gmx.net>



From grimesj at wharton.upenn.edu  Mon Jun 10 12:52:43 2002
From: grimesj at wharton.upenn.edu (Grimes, Jessica C)
Date: Mon, 10 Jun 2002 12:52:43 -0400
Subject: activating a webpage's click command from python
Message-ID: <8CFDEDD569991D43BC84141B2A93A5C802F77E8D@stock1.wharton.upenn.edu>

How do I activate a click command on Yahoo's website from Python?    

I want to create a  program that will ask the user what stock he/she wants to look up.  The program will then extract financial data from yahoo and import it into a spreadsheet.  

My problem is that I don't know how to make the application take the user's input (the stock's ticker symbol) and place the user's input from python into the space where yahoo accepts input for a ticker symbol and activate yahoo's "Look up" click command. 

I hope I have explained my question clearly.  If my proposed method is impossible, can you suggest another approach to designing the program? 

Thank you, 

Jessica Grimes




From claird at starbase.neosoft.com  Mon Jun  3 19:12:44 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 3 Jun 2002 18:12:44 -0500
Subject: In search of Frank Stajano
Message-ID: <DEB184D73D3D88C8.2C5F43B7D4B1FC1B.1C8CEBF4D8D19EB5@lp.airnews.net>

I'd welcome help contacting Frank Stajano.
-- 

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


From aahz at pythoncraft.com  Mon Jun 17 22:59:39 2002
From: aahz at pythoncraft.com (Aahz)
Date: 17 Jun 2002 22:59:39 -0400
Subject: Import caches names (was Re: Problems with importing; must do it twice?)
References: <fgphgus0pv2c7hp8l27uaenjlerr97tdta@4ax.com> <aeat13$pli$1@panix1.panix.com> <3D0E98BF.B2A9A6C2@engcorp.com>
Message-ID: <aem7mr$47m$1@panix1.panix.com>

In article <3D0E98BF.B2A9A6C2 at engcorp.com>,
Peter Hansen  <peter at engcorp.com> wrote:
>
>Should it not be considered a bug that a failed import leaves 
>a broken module around?  

Lots of people think so, but every time this comes up on python-dev,
someone fumbles around and eventually remembers exactly why it has to
work this way -- but I'm never the someone.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From siegfried.gonzi at kfunigraz.ac.at  Mon Jun 24 05:23:52 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Mon, 24 Jun 2002 11:23:52 +0200
Subject: Python and remote sensing
Message-ID: <3D16E528.5040207@kfunigraz.ac.at>

Hi,


a technician of us asked me whether it is possible to do the following:


Starting a script on Unix at lets say 2 am and calling a remote site 
(http://www.whatever...) and that site should become passively processed 
(the site holds 41 observing stations). Is it possible to start a remote 
script (automatically) connect to a http page and retrieve information 
from that html-page?

I answerd Perl or Python can do the job. But I am not sure whether that 
is correct.

I am sure the above topic has been asked 1000 times. Or is there a tool 
or program availabe which can do the job better?; I mean our technican 
has never heard of Python and I am not interested to make web based (put 
in the right term here) programming or the like.

If Python can do it - okay; if not please say so (if Python is dead - 
okay; if Python lives - okay)!


Thanks,
S. Gonzi



From sholden at holdenweb.com  Mon Jun  3 08:51:01 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 3 Jun 2002 08:51:01 -0400
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <06da01c208f4$d69003c0$6601a8c0@boostconsulting.com> <200206010121.g511LxX19223@pcp742651pcs.reston01.va.comcast.net> <073501c20960$23f367e0$6601a8c0@boostconsulting.com> <200206011334.g51DY1D21669@pcp742651pcs.reston01.va.comcast.net> <07d801c20970$cdebe050$6601a8c0@boostconsulting.com> <mailman.1022940122.3421.python-list@python.org> <Id5K8.171932$Po6.340972@rwcrnsc52.ops.asp.att.net> <oWIK8.179611$%u2.31494@atlpnn01.usenetserver.com>
Message-ID: <cHJK8.181202$%u2.93890@atlpnn01.usenetserver.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:oWIK8.179611$%u2.31494 at atlpnn01.usenetserver.com...
> "Emile van Sebille" <emile at fenx.com> wrote ...
> > <jepler at unpythonic.net> wrote:
> > > IMO the following is an odder behavior:
> > > >>> t
> > > ([1], [2], [3])
> > > >>> t[1] += [3]
> > > Traceback (most recent call last):
> > >   File "<stdin>", line 1, in ?
> > > TypeError: object doesn't support item assignment
> > > >>> t
> > > ([1], [2, 3], [3])
> > >
> >
> > Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
> > >>> class myTuple(tuple):
> > ...     def __setitem__(self, attr, value):
> > ...             print 'in setitem:', attr
> > ...
> > >>> a = myTuple(([1],[2],[3]))
> > >>> a
> > ([1], [2], [3])
> > >>> a[1] += [4]
> > in setitem: 1
> > >>> a
> > ([1], [2, 4], [3])
> > >>>
> >
> >
> > Here's another that's differently wrong:
> > >>> a = ('aaa',)
> > >>> a[0]+='bbb'
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> > TypeError: object doesn't support item assignment
> > >>> a
> > ('aaa',)
> > >>>
>
>
> Why do you consider either of these examples wrong? The whole point of
> tuples is that if d is a dictionary, then d[t1] == d[t2] if, and only if,
t1
> == t2. Both of these examples are trying to make changes that will modify
an
> (immutable) tuple. There's every reason to treat these modifications as
> illegal, since a before the fiddling is not equal to a after the fiddling,
> but tuples aren't mutable so such changes are illegal.
>
> regards
> --
> -----------------------------------------------------------------------
> Steve Holden                                 http://www.holdenweb.com/
> Python Web Programming                http://pydish.holdenweb.com/pwp/
> -----------------------------------------------------------------------
>
The dictionary example is somewhat misleading (because my assertion is
wrong). However, I stand by the statement that you are both trying to modify
tuples, which is fundamentally against the nature of the type.

regards
 Steve






From vvainio at tp.spt.fi  Wed Jun 12 09:44:52 2002
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 12 Jun 2002 06:44:52 -0700
Subject: Why does Python mix OO concepts and non OO concepts for operation  s  on basic types?
References: <mailman.1023826878.17168.python-list@python.org>
Message-ID: <ad496f8.0206120544.c92f181@posting.google.com>

Roman Suzi <rnd at onego.ru> wrote in message news:<mailman.1023826878.17168.python-list at python.org>...

> I only forgot how will I do this:
> 
> map(string.split, s)
> 
> if GvR will deprecate string module...

List comprehension is as pretty IMO.

At least it will be if/when the iteration variable (or whatever) will
no longer be visible after the list comprehension.

Name it ____listcomp_____var, like in class private vars, and del it
after running the comprehension?

-- Ville


From shagshag13 at yahoo.fr  Tue Jun  4 03:57:24 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Tue, 4 Jun 2002 09:57:24 +0200
Subject: Pop a list from beginning ? and memory saving...
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de> <3CFA0676.68F80DE6@engcorp.com> <add289$vud42$1@ID-146704.news.dfncis.de> <Xns9221CE2AF76D0cliechtigmxnet@62.2.16.82>
Message-ID: <adhrsl$118kvn$1@ID-146704.news.dfncis.de>

> - use "del var" instead of "= None".

see the "reset" method below, if is a good way to free memory ???
or is there a better way/tips ?

class Shag:
 def __init__(self, v = []):
  self._v = v

 def __str__(self):
  return str(self._v)

 def reset(self):
  del self._v
  self._v = []

thanks,

s13.




From skip at pobox.com  Mon Jun 10 11:25:32 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 10 Jun 2002 10:25:32 -0500
Subject: Could someone submit this to the sf-bug tracker, please?(Was: Re: __imul__ broken for 'objects' in 2.2.1)
In-Reply-To: <yfs660r8dgl.fsf_-_@black132.ex.ac.uk>
References: <yfsd6v3qa86.fsf@black132.ex.ac.uk>
        <lk660tlnon.fsf@pc150.maths.bris.ac.uk>
        <yfs660r8dgl.fsf_-_@black132.ex.ac.uk>
Message-ID: <15620.50412.292478.254369@12-248-41-177.client.attbi.com>

    >> Hmm.  You should file a bug on sf so this doesn't get forgotten
    >> (unless it's already been reported, but I don't think so).

    Alexander> Well yes, but then I *can't* which is why I've posted the
    Alexander> previous message.

    Alexander> I tried again today, but logging in from the submission form
    Alexander> has no effect, I'm back to the same form that tells me to
    Alexander> "please log-in" *after* I have completed the log-in step (if
    Alexander> I submit the wrong password, however I do get an error
    Alexander> message). I've tried 3 different browsers (mozilla .94,
    Alexander> konqueror and netscape 4.7)

Do you have cookies enabled?

    Alexander> If my problems don't reflect the weirdness of the proxy I'm
    Alexander> forced to use, I'd really suggest that either the decision
    Alexander> not to allow anonymous bug reports should be reviewed again.

This is really not an option.  Far too often an incomplete bug report would
be filed.  When a python developer wanted more input there was nobody
listening.   There was then no choice but to close the bug report
unresolved.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html





From emile at fenx.com  Mon Jun 24 23:52:56 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 25 Jun 2002 03:52:56 GMT
Subject: Numerical Python array math with different sized arrays
References: <3d173f56$1@news.mhogaming.com>
Message-ID: <sORR8.60560$EP.6287@sccrnsc03>

Michael Weiss
> I'm using Numeric Python's arrays. I want to perform the various math
> operators between arrays (+-*/ etc...) .
> While all my arrays are 1 dimensional, they may be different lengths.
The
> result I'd like in a mis-matched size case is the following:
> >>> a = Numeric.array([1,2,3,4])
> >>> b = Numeric.array([1,2])
> >>> print a+b
> [2,4,3,4]
>

Looks like these work.

For two Numeric arrays, if you want a op b to return the full length of
a:

add:
a[:len(b)] = map(Numeric.sum, zip(a,b))

subtract:
a[:len(b)] = map(Numeric.sum, zip(a,b*-1))

multiply:
a[:len(b)] = map(Numeric.product, zip(a,b))

divide:
a[:len(b)]=Numeric.divide(a[:len(b)],b)


HTH,

--

Emile van Sebille
emile at fenx.com

---------



From wurmy at earthlink.net  Wed Jun 26 13:50:42 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Wed, 26 Jun 2002 17:50:42 GMT
Subject: question od default args
References: <o3vjhuguen6gfm8vnq7rbk5volmi80shl5@4ax.com>
Message-ID: <3D19FEA3.6020804@earthlink.net>

Gon?alo Rodrigues wrote:
> Hi,
> 
> When I want default args I usually do
> 
> def (arg = None):
>     ...
> 
> But what if I want to make None a reasonable argument too? That is, I
> want to know if the user has in fact passed an argument, and if not to
> do something special, with None (or whatever object) being a reasonable
> arg to be passed.

Hmm, so use a default argument that is not reasonable to pass... some weird 
string, or a complex number, or even an instance of a (singleton?) class 
written especially for this... although that may be a bit overkill. Of course, 
the question is what is "reasonable" here and what not...

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/



From peter at engcorp.com  Mon Jun 24 21:01:04 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 24 Jun 2002 21:01:04 -0400
Subject: Suggestions for good programming practices?
References: <WEuR8.2074$5l4.75657@ozemail.com.au> <bko7fa.9t.ln@grey.aerojockey.localdomain> <af7sch$dcp$1@panix1.panix.com>
Message-ID: <3D17C0D0.439C5269@engcorp.com>

Aahz wrote:
> 
> In article <bko7fa.9t.ln at grey.aerojockey.localdomain>,
> Carl Banks  <imbosol at vt.edu> wrote:
> >
> >* Learn how references and the object system work as soon as possible.
> 
> ObBrokenRecord: "bindings", not "references".

I haven't had a record player for a while :-) so would you please
expand on that?  I've adopted "references" without giving it any
thought.  Why is that not as good as "bindings"?

Thanks.

-Peter


From cliechti at gmx.net  Tue Jun 25 15:05:21 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 25 Jun 2002 21:05:21 +0200
Subject: Threads on embedded Python
References: <1024952585.111238@seux119> <Xns9237EECAA9CC5cliechtigmxnet@62.2.16.82> <1024957605.771843@seux119> <Xns9238944DA0FEcliechtigmxnet@62.2.16.82> <1025028733.290520@seux119>
Message-ID: <Xns9238D6F8D5BBAcliechtigmxnet@62.2.16.82>

"Ugo Garc?a" <plugin at supercable.es> wrote in
news:1025028733.290520 at seux119: 
> But, one question: you say that I don't have to have a main loop in
> order to have threads running using Py_BEGIN_ALLOW_THREADS, but don't
> known exactly how to use it. I explain: from the main loop in my C
> application I launch threads in python using PyRun_SimpleString. But
> these threads doesn't run unless I call Python to do something. How I
> have to use Py_BEGIN_ALLOW_THREADS in order not to call Python
> continuosly? 

(im realy not sure if it works that way, but im hoping)

normaly Py_BEGIN_ALLOW_THREADS is used when python calls a C function in an 
extension module. while the C function is executed, all python activities 
are stopped. to reactivate python, while still beeing in the C func. you 
can call that macro. this allows long running C funcs, without blocking 
python. (note that you must call Py_END_ALLOW_THREADS before leaving the C 
func).

what you want is somewhat the other way round... the macros mentioned above 
won't work directly, but if you look what they do, they are 
saving/restoring of thread state. it could possibly work if you use those 
functions to release the global interpreter lock. but i haven't done such 
things (i've started a native thread in an extension that called back to 
python on events, but the main loop is in python)

i think its much easier to lauch a loop in python and let it call back to c 
functions...

http://python.org/doc/current/api/threads.html

chris

-- 
Chris <cliechti at gmx.net>



From achim.domma at syynx.de  Fri Jun  7 07:54:16 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Fri, 7 Jun 2002 13:54:16 +0200
Subject: Memory leaks when  embedding python (debug-version)  ?
References: <1eed5b9.0206070157.6d281f65@posting.google.com>
Message-ID: <adq6cp$s9o$00$1@news.t-online.com>

On which OS are you working on? Are you shure that you are linking against
the correct libraries? On windows it's quite easy to mix release/debug
versions and different runtime versions.

greetings
Achim





From nospam at bigfoot.com  Tue Jun 18 07:09:45 2002
From: nospam at bigfoot.com (Gillou)
Date: Tue, 18 Jun 2002 13:09:45 +0200
Subject: Can python have lib to display PDF ?
References: <a6a2957e.0206180246.511fb932@posting.google.com>
Message-ID: <aen3to$2a1$1@norfair.nerim.net>

I'm afraid there's no "pythonic" solution for displaying.
You can run acrobat reader or ghostview from a python script.
If you're running a Win32 box, and are a lucky owner of an Adobe Acrobat
licence, you should inspect the COM objects exposed by this application. If
there are solutions, here they are.
AFAIK, the - excelent - reportlab library does not provide features for
changing an existing PDF.

www.reportlab.com

HTH

"coolmenu" <menucool at yahoo.com.cn> a ?crit dans le message news:
a6a2957e.0206180246.511fb932 at posting.google.com...
> i want a python application to display end edit pdf.
> where are the lib pdf for python?
> thank



From martin at v.loewis.de  Sun Jun  9 15:48:21 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 09 Jun 2002 21:48:21 +0200
Subject: Python 2.2.1 vs. gcc 3.1?
References: <yu99vg8v6orb.fsf@europa.research.att.com> <15616.57963.519636.730290@12-248-41-177.client.attbi.com> <mailman.1023468673.31246.python-list@python.org> <1radq4bmey.fsf@fimm.ifi.uio.no>
Message-ID: <m3k7p8b5dm.fsf@mira.informatik.hu-berlin.de>

Kjetil Torgrim Homme <kjetilho at haey.ifi.uio.no> writes:

> >   PS: I can't find a "ldconfig" command on my system.
> 
> it's called crle (Solaris 8 and newer).  I do not recommend its use
> over rpath except in the most desparate situations.

Indeed. Atleast up to and including Solaris 8, crle is not comparable
in functionality and convenience of GNU ldconfig. crle in Solaris 9 is
slightly better, allowing incremental updates of the loader path.

So for the moment, the best choice is still to pass -R options to
ld(1) (or using LD_RUN_PATH at build time).

Regards,
Martin


From martin at v.loewis.de  Sun Jun  9 17:03:02 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 09 Jun 2002 23:03:02 +0200
Subject: Socket error problem
In-Reply-To: <GCEDKONBLEFPPADDJCOEOELGDHAA.whisper@oz.net>
References: <GCEDKONBLEFPPADDJCOEOELGDHAA.whisper@oz.net>
Message-ID: <m3elfg9ncp.fsf@mira.informatik.hu-berlin.de>

"David LeBlanc" <whisper at oz.net> writes:

> I've put print statements around all this before I first tryed the ml, but
> here they are again:

This is still not the right place.

>         print 'poplib._init__: self.host = %s ' % self.host, 'self.port =
> %s' % self.port

Notice that host and port are not passed directly to socket. Instead,
getaddrinfo is used

>         for res in socket.getaddrinfo(self.host, self.port, 0,
> socket.SOCK_STREAM):
>             af, socktype, proto, canonname, sa = res
>             try:

Please add a line printing res inside of this try statement. Also add
a print statement between the socket call and the connect call, so
that we know which call it is that fails.

Regards,
Martin




From gerhard at bigfoot.de  Wed Jun 19 03:13:45 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 19 Jun 2002 09:13:45 +0200
Subject: Caching query resuls
In-Reply-To: <aentug$m8d$1@news.onet.pl>
References: <aentug$m8d$1@news.onet.pl>
Message-ID: <20020619071345.GB9430@lilith.my-fqdn.de>

* Pawel Lewicki <lewy0lewy at poczta.onet.pl> [2002-06-18 20:25 +0200]:
> Hallo,
> Is there any good solution to cache the results of SQL queries?

Your database has one. It should save at least the query execution plan.

> I found the great cache manager in dmtools (http://csl.anu.edu.au/ml/dm/), but I wonder
> if there is any alternative. I will start with MySQL database, but it would
> be great to be able to plug it into other RDBMS's. I am interested in both -
> file and memory caching. I also know ZSQL Method in Zope.

I'm sceptical. How does this help? How does the cache manager get
notified of changes in the database? How is it different from just
storing the queries I'm interested in away on disk if they deliver large
resultsets?

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 25.4 ?C      Wind: 2.9 m/s




From calves at coelce.com.br  Wed Jun 19 15:24:36 2002
From: calves at coelce.com.br (Alves, Carlos Alberto - Coelce)
Date: Wed, 19 Jun 2002 16:24:36 -0300
Subject: latex-tex-packages
Message-ID: <29A97D00F387D411AC7900902770E148058E5A11@lcoeexc01.coelce.net>

Hi all,
You forgive me, c'os I know this issue it's not to this list; But, anybody
can tell me where can I get tex/latex packages to install. I just browse
some url but can't find'em.

Thanks, in advance.



Carlos Alberto
COELCE/DPRON-Departamento de Projetos e Obras Norte
Fone: 677- 2228
e-mail: calves at coelce.com.br
\|||/
(o o)
--ooo0-(_)-0ooo--

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020619/f13efc34/attachment.html>

From tdelaney at avaya.com  Thu Jun 20 01:27:12 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 20 Jun 2002 15:27:12 +1000
Subject: Variable Interpolation - status of PEP 215
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A3A9@natasha.auslabs.avaya.com>

> From: aahz at pythoncraft.com [mailto:aahz at pythoncraft.com]
> 
> Peter already told you what you really wanted to know, but for the
> benefit of other people, PEP 215 is dead and PEP 292 is currently the
> subject of a LARGE thread on python-dev.

And as Barry predicted, it's starting to get emotional ;)

Tim Delaney




From sameer_ at email.com  Tue Jun 11 09:42:14 2002
From: sameer_ at email.com (sameer)
Date: 11 Jun 2002 06:42:14 -0700
Subject: Medium to Large Scale Python Deployments
References: <mailman.1023460091.14067.python-list@python.org>
Message-ID: <6bd9f01b.0206110542.f5e0670@posting.google.com>

pyljclient: an interesting blend of python and wxpython

http://pyljclient.sourceforge.net

Michael Chermside <mcherm at destiny.com> wrote in message news:<mailman.1023460091.14067.python-list at python.org>...
> Folks, I think this is a pretty useful thing to collect.
> 
> In fact, it might be really nice to collect a list of large (successful) 
> Python projects and post it somewhere so the Python community could use 
> it for advocacy.
> 
> So do me a favor... if you have (or know of) a large Python project 
> which you think our community ought to be bragging about, please either 
> post it to this thread or email it to me. I'll collect the list, and 
> then see if I can get it posted somewhere appropriate. Please run 
> pycount (http://starship.python.net/crew/gherman/playground/pycount/ ) 
> as "pycount.py -R *.py" on the top of your source tree, (or modify that 
> appropriately for your project!).
> 
> The exact definition of "large" is fairly vague... I'd say 30K loc is 
> borderline and 50K loc is definitely big.
> 
> Please note that the project does NOT need to be open-source or freely 
> available to qualify!!!
> 
> If I get at least 8 responses, I'll assume there's sufficient interest, 
> and arrange to get it posted.
> 
> Alternatively, if such a list already exists, let me know where, and 
> I'll volunteer to help update and promote it.
> 
> Here's what I have so far:
> 
> Project      Size(Kloc)  URL
> -------        ----      ---------------------
> Zope            ???      http://www.zope.org/
> twisted         >50      http://www.twistedmatrix.com
> ekit            ~50      http://www.ekit.com/
> Domenic's
>    ERP system   >100      --
> Something
>    by aahz      30-40     ???
> 
> 
> Sources:
>    http://groups.google.com/groups?th=b7ec40fc188a2a52&rnum=1


From whisper at oz.net  Wed Jun 26 03:15:11 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 26 Jun 2002 00:15:11 -0700
Subject: Python and Windows XP
In-Reply-To: <20020625233842.29966.00000386@mb-ma.aol.com>
Message-ID: <GCEDKONBLEFPPADDJCOEAEDODMAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Marc
> Sent: Tuesday, June 25, 2002 20:39
> To: python-list at python.org
> Subject: Python and Windows XP
>
>
> Hello all,
>    I apologise for this newbie question, but I use Windows XP and
> wonder how I
> can open a shell session to use Python?  There used to be the DOS command
> option, but now...
> --
> Marc

If it's like Windows 2000, either the command promt has to be added to the
start menu, or it's been relegated to Start | Accessories or a sub menu of
Accessories (and the parachute is in the baggage compartment for your
convenience). If, at the Start | Run prompt, you can enter "cmd /c" and you
get a dos box, then you can probably go to c:\winnt\system32, find cmd.exe,
create a shortcut (select, ctrl-c, ctrl-v), drag it onto the desktop and
rename it - and from there onto your start menu if you want it there. I
found that I could drag it into the left end of the taskbar and have it show
up as a mini-icon that single clicks to a dos box.

HTH,

Dave LeBlanc
Seattle, WA USA





From gashton at cmedltd.com  Tue Jun 11 15:41:28 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 11 Jun 2002 20:41:28 +0100
Subject: using m2crypto to encrypt credit card numbers
In-Reply-To: <NCBBLFCOHHDIKCAFGCFBOEPLLDAA.marklists@mceahern.com>
References: <NCBBLFCOHHDIKCAFGCFBOEPLLDAA.marklists@mceahern.com>
Message-ID: <1023824489.22503.11.camel@walter>

On Tue, 2002-06-11 at 14:05, Mark McEahern wrote:

> If I don't store the credit card number, how do I send it to the payment
> processor to collect payment?

I've not followed the thread, but I'm assuming that your problem is that
you don't want to store the card number in plain text, and that if you
encrypt and store it in a manner that will allow automatic decryption by
your software, then a cracker who gains access to your servers will be
able to decrypt the card numbers with ease.

The only sensible solution to this that I've ever thought of involves
getting the payment processor (i.e. online transaction processing
company) to store a hash for each of your customers' credit cards. Then
you could send them the hash each time you want to put another
transaction on the card, and they would store the card number instead of
you (and I'm fairly confident that they have to store it anyway).

Consequently you wouldn't need to store the card number at all, just the
encrypted hash. Job's a good'un; card numbers would be nicked from far
fewer online web sites with shoddy security.

Admittedly the eggs all go in one basket, but if the payment processor
stores them all any way then all the eggs are already sat in the same
basket.

But I don't know of any companies that provide such a service. :(

And I may also have completely misread the direction of the thread from
the last post....
-- 
Graham Ashton





From marknews at mceahern.com  Sat Jun 29 11:41:49 2002
From: marknews at mceahern.com (mark mceahern)
Date: 29 Jun 2002 08:41:49 -0700
Subject: generate all combinations of a list (with variable output length)
References: <mailman.1025295140.24921.python-list@python.org> <afip0i$8c3$0@216.39.172.122>
Message-ID: <a93021de.0206290741.6f8aab5a@posting.google.com>

bokr at oz.net (Bengt Richter) wrote in message news:<afip0i$8c3$0 at 216.39.172.122>...
> Or, implementing your base-5 digits idea, and leaving out '':
> 
>  >>> def adD5r(n):
>  ...     codes='AD$5R' # <=> 01234
>  ...     out = []; m = n
>  ...     while 1:
>  ...         n,d = divmod(n,5)
>  ...         out.append(codes[d])
>  ...         if not n:break
>  ...     out.reverse()
>  ...     return '%d:%s' % (m,''.join(out))
>  ...

Bengt, this is smoov.  I took your code and, using nested scopes, made
a function that returns the equivalent of adD5r for an arbitrary set
of codes:

def make_number_generator(codes):
    def number_generator(n):
        out = []
        m = n
        while 1:
            n, d = divmod(n, len(codes))
            out.append(codes[d])
            if not n:
                break
        out.reverse()
        return '%s' % (''.join(out))
    return number_generator

codes='AD$5R' # <=> 01234
number_generator = make_number_generator(codes)
n = 0
print "%d:%s" % (n, number_generator(n))
n = len(codes) ** 10 - 1
print "%d:%s" % (n, number_generator(n))

Cheers,

// mark


From maxm at mxm.dk  Tue Jun 18 14:22:43 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 18 Jun 2002 20:22:43 +0200
Subject: why len(list) instead of list.len()  ?
References: <EhBP8.1$Qn1.1124@news.ecrc.de> <c5HP8.243362$cQ3.10419@sccrnsc01> <mailman.1024420102.1620.python-list@python.org>
Message-ID: <3D0F7A73.1080709@mxm.dk>

holger krekel wrote:

>>Because a two-line wrapper goes a long way?
>>
>>Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32


I thought it was historical, because tupples didn't have methods.

I also find obj.len() more natural/logical. But I have learned to live 
with it...

regards Max M



From hesterloli at hotmail.com  Tue Jun 18 23:08:00 2002
From: hesterloli at hotmail.com (George Hester)
Date: Wed, 19 Jun 2002 03:08:00 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com>
Message-ID: <kASP8.47640$GY.14966135@twister.nyroc.rr.com>

Expected

--
George Hester
_________________________________
"George Hester" <hesterloli at hotmail.com> wrote in message
news:K7PO8.16057$uk2.8406537 at twister.nyroc.rr.com...
> What is a simple script that will tell me the version of python I am currently
> running?  IIS 5.0 Windows 2000.
>
> --
> George Hester
> _________________________________
>
>




From mhammond at skippinet.com.au  Wed Jun  5 08:50:04 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 05 Jun 2002 12:50:04 GMT
Subject: manipulating child processes under Windows
References: <lcptz7zw0m.fsf@gaffa.mit.edu> <uadqb3q9v.fsf@online.no> <lcy9duhksg.fsf@gaffa.mit.edu> <d7226884.0206050341.61fa2ab0@posting.google.com>
Message-ID: <3CFE0975.8010106@skippinet.com.au>

>Syver Enstad <syver-en+usenet at online.no> writes:
>
>
>>I'll have to check the msdn
>>library to find out how to get the owning process handle from a
>>window handle.

FYI, GetWindowThreadProcessId was added in the win32all-147/8 
win32process module.

Mark.



From jwbaxter at spamcop.net  Mon Jun 10 20:36:37 2002
From: jwbaxter at spamcop.net (John Baxter)
Date: Mon, 10 Jun 2002 17:36:37 -0700
Subject: Iterators vs. Generators
References: <ae3a6f$9d9$1@panix1.panix.com>
Message-ID: <jwbaxter-5ABAFE.17363610062002@corp.supernews.com>

In article <ae3a6f$9d9$1 at panix1.panix.com>, aahz at pythoncraft.com (Aahz) 
wrote:

> So when would one actually write an iterator instead of a generator?
> I've been trying to think of an example and failing.

One answer is...

One would write an interator if one comes from a language which uses 
iterators.

One would write a generator if one comes from a language which uses 
generators.

And one dithers if one is comfortable with both forms.

  ;-)

--John


From jbublitzNO at SPAMnwinternet.com  Sat Jun 29 21:24:57 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Sun, 30 Jun 2002 01:24:57 GMT
Subject: Newbie that don't understand
References: <AIlT8.13914$p56.4478653@newsb.telia.net> <3d1e262f.3755054@news.clara.net> <3D1E4F80.D8EFBB3E@engcorp.com> <3D1E56EC.7080203@SPAMnwinternet.com> <3D1E5BE9.86189DA9@engcorp.com>
Message-ID: <3D1E5DB3.7080909@SPAMnwinternet.com>

Peter Hansen wrote:
> Jim wrote:

>>Peter Hansen wrote:

>>>Just a dude wrote:

>>>>My son is an electronics
>>>>design engineer, and I still can't get my head around negative
>>>>voltages, how the hell do you get a -12 V supply... to him it's basic
>>>>but to me..............

> [snip my own reply which rather missed the mark this time,
> because it's too embarrassing to leave in :-)]

>>While this is way off topic and way pedantic, electrons
>>always leave and enter the same terminals (think electro-
>>chemistry, like a battery). 

> D'oh!  What was I thinking?  And they call me an engineer. :(

> Jim is quite right of course.  Grumble grumble.

Embarassingly I had to think about it for a little
bit (won't say how long). Having just finished 9th
grade General Science recently (my daughter's)
really helped.

Now I suppose I'll have bad dreams again about them
coming to take away my EE degree.


Jim



From phr-n2002b at NOSPAMnightsong.com  Thu Jun 27 23:10:19 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 27 Jun 2002 20:10:19 -0700
Subject: Is Python growing?
References: <3D1B92FB.59B590A0@pop.ntlworld.com> <slrnahnif7.1c6.matt@happy-hour.mondoinfo.com>
Message-ID: <7x66042j50.fsf@ruckus.brouhaha.com>

Matthew Dixon Cowles <matt at mondoinfo.com> writes:
> On the other hand, some features make Python much easier to use and
> read. A while ago, I replaced:
> 
> def lengthOfLongestStr(list):
>   l=len(list[0])
>   for count in range(1,len(list)):
>     if len(list[count])>l:
>       l=len(list[count])
>   return l
> 
> with
> 
> def lengthOfLongestStr(myList):
>   return max([len(s) for s in myList])
> 
> To my eye, the second one is much easier to read. 

Well, you could have said

    def lengthOfLongestStr(myList):
      return max (map (len, myList))

without needing any new features added.


From m2 at plusseven.com  Wed Jun 12 16:28:29 2002
From: m2 at plusseven.com (Alex Polite)
Date: Wed, 12 Jun 2002 22:28:29 +0200
Subject: Something faster then sgmllib for sucking out URLs
Message-ID: <20020612202828.GB16934@matijek.plusseven.com>

I'm working on a webspider to fit my sick needs. The profiler
tells me that about 95% of the time is spent in sgmllib. I use sgmllib
solely for extracting URLs. I'm looking for a faster way of doing
this. Regular expressions, string searches? What's the way to go? I'm
not a python purist. Calling some fast C program with the html as
argument and getting back a list of URLs would be fine by me.

-- 
Alex Polite
http://plusseven.com/gpg/




From edream at tds.net  Sat Jun 22 21:53:44 2002
From: edream at tds.net (Edward K. Ream)
Date: Sun, 23 Jun 2002 01:53:44 GMT
Subject: ConfigParser & converting strings to lists
Message-ID: <3D152A28.2177B054@tds.net>

Hi all,

I have a question concerning lists in the ConfigParser module.  What I
would like to do is something like:

config = ConfigParser.ConfigParser()
config.set("recent files", "recentFiles, files)

where files is a _list_ of file names.  Actually, this works just fine,
and the config file contains:

[recent files]
recentFiles = [file1, file2, ...]

as expected.  However,

files = config.get("recent files", "recentFiles")

sets files to the _string_ "[file1, file2, ...]".  I suppose this must
be how get should work; after all, settings could be arbitrary
strings...

Anyway, I wonder whether there is a clever way of converting the string
representation of a list containing strings into a true list of
strings.  In any event, wouldn't a getlist convenience method be nice?

Thanks!

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


From BPettersen at NAREX.com  Tue Jun 11 00:06:41 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 10 Jun 2002 22:06:41 -0600
Subject: Changing the path on Win2k/XP
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192F1A3F6@admin56.narex.com>

Is there a way to change the path on Win2k/XP through Python? (I'm
trying to write a .dll deployment script for our production department,
and I'd like to give them a "one button" install...)

-- bjorn




From shalehperry at attbi.com  Wed Jun 19 13:50:12 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 19 Jun 2002 10:50:12 -0700 (PDT)
Subject: python __main__ is similar to java main()
In-Reply-To: <3015ca5e.0206190901.3cb0b726@posting.google.com>
Message-ID: <XFMail.20020619105012.shalehperry@attbi.com>

On 19-Jun-2002 David Nunes wrote:
> Well I've read the docs and searched in the newsgroup archives but
> I'm still confused about __main__
> 
>         I  expected it to startup the program, like main in java but
> when I feed it to the interpreter he just "eats" the file loads wathever
> is in the file but doesn't run main.
> 
>         So am I suposed to explicitly call it like in the file below or is
> there a better (pythonic) way to do it??
> 

the idiom is this:

class MyClass:
  pass

if __name__ == '__main__':
  # do stuff with MyClass

what is happening here is __name__ is defined to be the name of the module when
the file is called via 'import' or 'from' or it is called '__main__' if being
used as a script and called directly by the interpreter.  Note __name__ is at
the module scope, not part of any class.

This allows for each module to have test code in it that can be executed easily.

If you are simply writing a #!python style script there is no need for main,
the interpreter starts at the first line and executes everything it sees.




From backflip at gmx.de  Wed Jun  5 02:21:42 2002
From: backflip at gmx.de (Backflip)
Date: 4 Jun 2002 23:21:42 -0700
Subject: writing a text in hyphenless justification
Message-ID: <32d0f89.0206042221.1910798c@posting.google.com>

Hi everybody,
i?ve got a little problem again......
i want to take a raw_input and makes it to a hyphenless justification output.....
how is this possible?????
The most beautiful solution would be the one, that i can define the width....
has got smb a idee??? it would be very nice, if smb can help me with my problem.....
THankfull 
Backflip


From graham_guttocks at yahoo.co.nz  Fri Jun 14 20:17:50 2002
From: graham_guttocks at yahoo.co.nz (Graham Guttocks)
Date: Fri, 14 Jun 2002 17:17:50 -0700 (PDT)
Subject: getgrouplist() ?
Message-ID: <20020615001750.94759.qmail@web10304.mail.yahoo.com>

Greetings,

I need a way to get a back a list of supplemental group ids for a
specified uid.  Something like getgrouplist() in my C standard lib.  I
don't see a convenient method for this, have I missed it?

The os module contains getgroups(), but that doesn't allow me to
specify a uid; it uses only the uid of the current process.

Regards,
Graham


__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com




From michael at damaru.com  Sun Jun  9 12:13:53 2002
From: michael at damaru.com (Michael Davis)
Date: Sun, 09 Jun 2002 12:13:53 -0400
Subject: libftp.size function gives wrong results!
References: <RNKM8.560$Yw3.96921@news20.bellglobal.com>
Message-ID: <_6LM8.603$Yw3.104253@news20.bellglobal.com>

Hi,

I just realized that this isn't a python question, it's an ftp question. I 
tried doing:

size index.html

from the ftp prompt, and it too gave me the wrong size!

So I apologize for being off-topic. The trouble is, I can't seem to find a 
newsgroup specific to ftp problems. The existing ftp newsgroups seem to be 
just full of announcements about servers with adult content...

Cheers,
Michael

Michael Davis wrote:

> Hi,
> 
> I'm writing a python program which will upload or download files via ftp.
> It's intended to be used to deploy a web site from a local machine to a
> server. It tries to determine which files are modified on the local
> machine by comparing file sizes and modification times.
> 
> I try to compare files using os.stat for the local file and FTP.size for
> the remote. The problem is, FTP.size is giving me incorrect results. If I
> use the ftp command line client, and look at one of my files on the server
> (the output looks like 'ls -l') then I can see that the file size, is,
> say, 1200. But the FTP.size function returns, say, 1242. This happens for
> all the files, they're all reported to be about 5% larger than they really
> are.
> 
> Is there a better way of doing this, aside from capturing the output of
> FTPLIB.dir (which goes to stdout) and parsing it?
> 
> Thanks in advance!

-- 
Michael Davis
Damaru
Custom Programming - Web Development - Database Design
http://www.damaru.com
416-540-1284


From geoff at gerrietts.net  Fri Jun  7 20:26:19 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Fri, 7 Jun 2002 17:26:19 -0700
Subject: Python MTA
In-Reply-To: <GCEDKONBLEFPPADDJCOECEHHDHAA.whisper@oz.net>
References: <20020607232709.GA12947@lilith.my-fqdn.de> <GCEDKONBLEFPPADDJCOECEHHDHAA.whisper@oz.net>
Message-ID: <20020608002619.GB18346@isis.gerrietts.net>

Quoting David LeBlanc (whisper at oz.net):
> <ahem> and the documentation (something more then help(smtpd) please) for
> smtpd is where? :P)
> 
> It's not in pythondocs for 2.2.1, unless it's well hidden. smtplib is
> documented, but that's a client interface.
> 
> TMDA: BTDT - Like ASK, requires an external Linux/Unix based MTA. Thanks for
> the pointer though :)

The comments at the top of the library file --
/usr/local/lib/python2.2/smtpd.py on my box -- should give you much of
what you're looking for in the way of documentation. The source may
prove equally enlightening if you just skip over the stuff you don't
understand. :)

I can't be of further help on the topic, though; I don't know anything
about THIS module, just have done this drill myself a few times. Lots
of undocumented goodies in the standard libraries, if you crack them
open and have a look inside.

-- 
Geoff Gerrietts        <geoff at gerrietts dot net>
"Ordinarily he was insane, but he had lucid moments 
when he was merely stupid."        --Heinrich Heine




From nospam at bigfoot.com  Tue Jun 18 07:01:39 2002
From: nospam at bigfoot.com (Gillou)
Date: Tue, 18 Jun 2002 13:01:39 +0200
Subject: Unicode to default encoding...
Message-ID: <aen3ej$1ko$1@norfair.nerim.net>

Hi,

I'm looking for a way to encode an Unicode string to the default encoding
for the output (console, file...).

someunicodestring.encode('iso8859-1')

works for a file (windows & unix) but gives ugly characters on a windows
console.

Thanks in advance

--Gilles
glenfant AT bigfoot DOT com




From bokr at oz.net  Sat Jun 15 21:22:27 2002
From: bokr at oz.net (Bengt Richter)
Date: 16 Jun 2002 01:22:27 GMT
Subject: Pronouncing '__init__'
References: <20020615132519.A14452@hishome.net> <20020615134410.A15079@prim.han.de> <mailman.1024149620.26897.python-list@python.org>
Message-ID: <aegp8j$f2s$0@216.39.172.122>

On Sat, 15 Jun 2002 16:59:15 +0300, Oren Tirosh <oren-py-l at hishome.net> wrote:

>On Sat, Jun 15, 2002 at 01:44:10PM +0200, holger krekel wrote:
>> for oral language:
>> 
>>     init constructor
>> 
>> and if the context allows
>> 
>>     init
>> 
>> for written languange __init__
>> 
>> Isn't it obvious? :-)
>> 
>>     holger
>
>I used __init__ just as an example.  I meant how do you pronounce names 
>with leading and trailing double underscore in general.  The magic name 
>__init__ may be by far the one most commonly encountered in everyday 
>programming but __there__ __are__ __many__ __others__.
>
Since they are formally defined and documented (somewhere, if not always obviously :),
how about calling a.__xxx__ "the formal xxx method of a?" And maybe "formal xxx" for short?

Thus you could say, "you need a formal getitem," or
"call formal new instead of init".

Of course, len('formal xxx') > len('__xxx__'), so it's more for talking
or reading silently aloud to yourself. Written, it should remain '__xxx__'.

Regards,
Bengt Richter


From logiplexsoftware at earthlink.net  Thu Jun 20 15:15:01 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Thu, 20 Jun 2002 12:15:01 -0700
Subject: How to represent the infinite ?
In-Reply-To: <3d11ee9b$1@news.swissonline.ch>
References: <3d11ee9b$1@news.swissonline.ch>
Message-ID: <20020620121501.5e8498c1.logiplexsoftware@earthlink.net>

On Thu, 20 Jun 2002 17:02:33 +0200
erreur wrote:

> Will somebody have an idea, to represent the infinite one?
> 
> I have variables to initialize with is +inf (or - inf).  To be sure that
> later, all will be smaller (or larger) than my variables.
> I tried to redefine the operators on an object.  It goes for Inf>10 but I do
> not arrive for 10>Inf (because that takes > of Int).
> 
> --------------------
> 
> def __gt__(self, val):
> 
>     return 1
> 
> --------------------

Hm, for some reason, my last two posts on this thread didn't seem to go
through...  one more try (with a couple of little enhancements):

from __future__ import generators

class Infinity(int):
    def __init__(self):
        int.__init__(self)
        self.sign = 1

    def __str__(self):
        return "%sInfinity" % {-1: "-", 1: ""}[self.sign]

    def __repr__(self):
        return "<%s>" % self

    def __cmp__(self, n):
        if isinstance(n, Infinity):
            return cmp(self.sign, n.sign)
        return self.sign
    
    def __neg__(self):
        retval = Infinity()
        retval.sign = self.sign * -1
        return retval

    def __mul__(self, n):
        retval = Infinity()
        retval.sign = self.sign * cmp(n, 0)
        return retval

    def __div__(self, n):
        if isinstance(n, Infinity):
            # or should this raise an exception?
            return self.sign / n.sign
        return self * n

    def __add__(self, n):
        if isinstance(n, Infinity):
            if n.sign != self.sign:
                return 0
        return self

    def __sub__(self, n):
        return self + -n

    def __iter__(self):
        return self.next()

    def next(self):
        n = 0
        while 1:
            yield n
            n += self.sign            

    
if __name__ == '__main__':
    import sys
    
    m = -Infinity()
    n = Infinity()

    print m
    print n
    print m * n
    print n * n
    print m * m

    assert(m < n)
    assert(n > m)
    assert(m == m)
    assert(n == n)
    assert(m != n)
    assert(m < 0)
    assert(n > 0)
    assert(m != 0)
    assert(m < -sys.maxint)
    assert(n > sys.maxint)
    assert(m < -m)
    assert(n == -m)

    assert(n * 1 == n == -m)
    assert(n * 1000 == n)
    assert(n * -1 == m)
    assert(n * -1000 == m)
    assert(n * 0 == 0)
    assert(m * 1 == m)
    assert(m * 1000 == m)
    assert(m * -1 == n)
    assert(m * -1000 == -m == n)
    assert(m * 0 == 0)
    
    assert(n / 1 == n)
    assert(n / -1 == m)
    assert(n / -1000 == m)
    assert(m / 1 == m)
    assert(m / 1000 == m)
    assert(m / -1 == n)
    assert(m / -1000 == n)

    assert(m + 1 == m)
    assert(m - 1 == m)
    assert(n + 1 == n)
    assert(n - 1 == n)
    
    assert(n + m == 0)
    assert(n - m == n)
    assert(n - n == 0)
    assert(m + m == m)
    assert(m - m == 0)
    assert(m - n == m)

    print m / m
    print n / n
    print m / n
    print n / m
    
    try:
        print 1 / m
    except ZeroDivisionError:
        print "Can't divide by %s" % m

    try:
        print 1 / n
    except ZeroDivisionError:
        print "Can't divide by %s" % n

    for i in Infinity():
        if i > 10:
            break
        print i,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From anthony at interlink.com.au  Sun Jun 16 01:44:46 2002
From: anthony at interlink.com.au (Anthony Baxter)
Date: Sun, 16 Jun 2002 15:44:46 +1000
Subject: Pronouncing '__init__' 
In-Reply-To: Message from Oren Tirosh <oren-py-l@hishome.net> 
   of "Sat, 15 Jun 2002 13:25:19 +0300." <20020615132519.A14452@hishome.net> 
Message-ID: <200206160544.g5G5ilv18922@localhost.localdomain>

At the python conf I recall people using 
under under init 
or
double under init

I find the second is what I usually use.

-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From spammers.do.not.bother at void.bogus  Sun Jun 23 16:41:54 2002
From: spammers.do.not.bother at void.bogus (Magnus)
Date: Sun, 23 Jun 2002 20:41:54 GMT
Subject: sys.exc_info() into string
References: <KZoR8.44805$n4.10504231@newsc.telia.net> <Xns9236D90D4E2FAcliechtigmxnet@62.2.16.82>
Message-ID: <moqR8.44823$n4.10512288@newsc.telia.net>

Chris Liechti wrote:

> Magnus <spammers.do.not.bother at void.bogus> wrote in
> news:KZoR8.44805$n4.10504231 at newsc.telia.net:
>> in my program I am catching exceptions that I might have missed by
>> using something like:
>> 
>> try:
>>         doMyStuff()
>> except:
>>         type = sys.exc_info()[0]
>>         syslog.syslog(syslog.LOG_EMERG, "My program: unexpected error"
>>         + type)
>> 
>> The thing is that I can not figure out how to make "type" into a
>> string above so I can use it for e.g. syslog(). Is there an easy way
>> of doing this?
> 
> yes, the traceback module has functions to format an exception as string
> or to print.
> 
> format_exc is a candidate for you. it return a list of lines from which
> you can select to mosr informative or just join it the one big string.
> 
> chris
> 

I think following is working...

except:
        syslog.syslog(syslog.LOG_EMERG, "My program: unexpected er
ror" + str(sys.exc_info()[0]))

...thanks for the suggestion.

/Magnus



From eric.brunel at pragmadev.com  Wed Jun 12 10:15:28 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Wed, 12 Jun 2002 14:15:28 +0000
Subject: AttributeError? Why?
References: <3D07337A.870E82E8@gol.ge>
Message-ID: <ae7dlv$ouk$1@wanadoo.fr>

Giorgi Lekishvili wrote:
> Hi all!
> 
> I have been making a module, which, as I believed, was organized
> correctly. It has __init__.py file, which is empty, and several modules.
> 
> Now, what happens: import goes fine. But when I try to create a class,
>>>> s=SimAn.SimAn([],[],None,0,0)
> Traceback (innermost last):
>   File "<pyshell#18>", line 1, in ?
>     s=SimAn.SimAn([],[],None,0,0)
>   File "D:\Python20\SED\SimAn.py", line 18, in __init__
>     self.Scenario=ARScenario.ARScenario(self.xdata)
> AttributeError: ARScenario
> =====
> Of course, ARScenario is there:
>>>> dir(SimAn)
> ['ARScenario', 'SimAn', '__builtins__', '__doc__', '__file__',
> '__name__', 'copy', 'math']
> =====
> and, of course, the class ARScenario based on the class Scenario, has
> __init__(self,..) method.
> 
> 
> The question: where's my error?

The answer is in the question: why do you try to create an instance of 
ARScenario with:

self.Scenario=ARScenario.ARScenario(self.xdata)

since your ARScenario class is in the SimAn module, as shown by:

>>> dir(SimAn)
['ARScenario', 'SimAn', '__builtins__', '__doc__', '__file__', '__name__', 
'copy', 'math']

If your class ARScenario is in the same module than your SimAn class 
(namely SimAn), just create the ARScenario instance with:

self.Scenario=ARScenario(self.xdata)

It should be enough.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From "aglyportat\" at n-o.spa__mmm.yahoo dott com>  Sat Jun 15 22:44:44 2002
From: "aglyportat\" at n-o.spa__mmm.yahoo dott com> (Anton Graph)
Date: Sat, 15 Jun 2002 19:44:44 -0700
Subject: PyArg_Parse() & review for memory leaks
References: <ugk9amdu63ec36@corp.supernews.com> <aeelab$1gpv$1@agate.berkeley.edu>
Message-ID: <ugntojkjtj7v33@corp.supernews.com>

Daniel Yoo wrote:
> Anton Graph <""aglyportat\"@n-o.spa__mmm.yahoo dott com> wrote:
> 
> [some code cut]
> 
> :        char *pstr;
> :        if(!PyArg_Parse(result, "s", &pstr)) { // the plug-in just
> :                // builds a string containing tag
> :                return 0;
> :        }
> :        int st=strlen(pstr);
> :        if(st>=len) {
> :                delete [] *res;
> :                len=st+1;
> :                *res = new char[len];
> :        }
> :        astrncpy(*res, pstr, len);
> :        Py_DECREF(result);
> :        return st;
> : }
> 
> : do I leak pstr here? Sample code I've googled up seems similar, but I 
> : don't understand how does PyArg_Parse allocate enough space without 
> : going to the heap?
> 
> I don't believe so: PyArg_Parse()ing a string shouldn't be doing any
> allocation because pstr is simply directed to the string pointer that
> Python is using internally to store that String object --- no string
> copying is involved.
> 
Yep, you're right. False alarm. Thanks for the URL.

> Section 5.5 of the Python/C API document explains more about
> PyArg_Parse():
> 
>     http://python.org/doc/api/arg-parsing.html
> 
> Good luck to you!




From peter at engcorp.com  Tue Jun  4 23:36:39 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 04 Jun 2002 23:36:39 -0400
Subject: Pop a list from beginning ? and memory saving...
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de>	<3CFA0676.68F80DE6@engcorp.com>	<add289$vud42$1@ID-146704.news.dfncis.de>	<Xns9221CE2AF76D0cliechtigmxnet@62.2.16.82>	<adhrsl$118kvn$1@ID-146704.news.dfncis.de> <mailman.1023194347.14669.python-list@python.org> <3CFD5B60.90831F21@engcorp.com> <aIeL8.120588$Kp.12235125@bin7.nnrp.aus1.giganews.com> <3CFD86D4.C4F81404@engcorp.com>
Message-ID: <3CFD8747.4AA8D29F@engcorp.com>

Peter Hansen wrote:
> 
> Terry Reedy wrote:
> > >>> l
> > [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
> > >>> l[:] = []
> > >>> l
> > []
> 
> I must have missed the _why_ part...  ah!  enlightenment:
> you are removing the contents of the list, which might
> have other references to it, rather than rebinding the name
> self._v to a new empty list but leaving those other
> references alone.

Oh, and in my opinion, that is obscure and deserves a
comment explaining how it works, and why it is being done.
I might not feel that way if I write that idiom a few times,
though. :)

-Peter


From K.Rdt at TU-Berlin.DE  Thu Jun 27 05:07:07 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Thu, 27 Jun 2002 11:07:07 +0200
Subject: background-processes with win-python?
In-Reply-To: <afegu8$dlvqk$1@ID-71831.news.dfncis.de>
Message-ID: <YIGIEMJKFUO1XA0538IGGB1WHCMJE.3d1ad5bb@FRITZweb>

Am 27.06.02 10:04:23, schrieb "James Kew" <james.kew at btinternet.com>:

>
>system("start plink ... tu-berlin.de") ?
>
---------------------------------------------------------------------
Hi
Thanks; a good tip.
		K at Rdt
---------------------------------------------------------------------







From eric.brunel at pragmadev.com  Tue Jun 11 11:35:30 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 11 Jun 2002 15:35:30 +0000
Subject: Defining a method final
References: <6glN8.35376$86.884025@twister1.libero.it> <mailman.1023796095.1427.python-list@python.org> <phmN8.36164$K4.893500@twister2.libero.it> <KwmN8.153269$352.7754@sccrnsc02>
Message-ID: <ae4tvv$rvd$1@wanadoo.fr>

Emile van Sebille wrote:
> Accepted practice is to use a single underscore for objects to be
> considered private.  This does not prevent their use, but strongly
> signals that re-use ought not to be done.
> 
>>>> class Test:
> ...     def _private(self): pass
> ...
> 

Python now have an "official" means to make attributes private by prefixing 
them with a double-underscore. And it *does* prevent their use or 
overloading:

class A:
  def __private(self): pass
class B(A):
  pass
o = B()
o.__private()

results in:

Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: B instance has no attribute '__private'

Regarding the original question (final methods), it has however obvious 
drawbacks:
- the method can be *redefined*: that doesn't mean it can be overloaded, 
but class B can have its own __private method
- the method is actually *private*, which means that neither sub-classes, 
nor "outside" classes can call it.

I don't think that actual final methods can be defined in Python...
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From mwh at python.net  Thu Jun  6 09:33:30 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 6 Jun 2002 13:33:30 GMT
Subject: introspect c-functions?
References: <mailman.1023366009.13221.python-list@python.org>
Message-ID: <lku1oglefi.fsf@pc150.maths.bris.ac.uk>

holger krekel <pyth at devel.trillke.net> writes:

> is there any way to introspect the *number of arguments*
> of a function implemented at c-level?

In short, no.  To the runtime almost all C functions look roughly like
python functions defined like:

def func(*args):
   pass

& so there is no way of telling what the function will do with the
arguments.  There are others -- METH_O and METH_NOARGS style C
functions -- but I don't think you can tell them apart from Python,
and they don't cover all cases.

It's a common, but hardly universal, convention to have the first line
of the docstring be something like

  func(arg1[, arg2]) -> None

so you /might/ be able to dig some info out that way -- but it's
hardly reliable.

Cheers,
M.

-- 
  ... but I'd rather not reinvent the wheel if I don't have to.  On
  the other hand, if the currently instantiated version of the wheel
  consists of a square rock covered with moss, I might as well just
  start fresh.                          -- Roy Smith, comp.lang.python


From achim.domma at syynx.de  Thu Jun 13 03:09:25 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Thu, 13 Jun 2002 09:09:25 +0200
Subject: How to call a function using apply with keyword args?
Message-ID: <ae9fud$u23$00$1@news.t-online.com>

Hi,

I have a function like this:

def MyFkt(A,B,C): pass

and some parameters from a config file:

params = {'A':5,'B':8,'C':9}

how can I call MyFkt with this parameters? I tried apply(MyFkt,[],params)
but get this error:

TypeError: MyFkt() takes exactly 3 non-keyword arguments (0 given)

any idea?

greetings
Achim




From martin at v.loewis.de  Fri Jun 14 14:54:46 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Jun 2002 20:54:46 +0200
Subject: a possibly foolish question about slices
References: <yu99sn3prbqm.fsf@europa.research.att.com>
Message-ID: <m3vg8lu1vt.fsf@mira.informatik.hu-berlin.de>

Andrew Koenig <ark at research.att.com> writes:

> Suppose x is a string or list, and I want to refer to
> a subsequence of x with length n starting at position p.
> under the assumption that this subsequence is entirely
  ... within the boundaries?

> To do so, I can use x[p:p+n] in all circumstances but one,
> namely when p < 0 and p == -n.  Is there a similarly concise
> expression that works even in this special case?

Perhaps not concise, but compact:

  x[p:p+n or sys.maxint]

Regards,
Martin


From kragen at pobox.com  Mon Jun 10 18:12:35 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 18:12:35 -0400
Subject: Porting python to MVS Host
References: <adhvh4$v95q8$1@ID-28256.news.dfncis.de>
Message-ID: <83ofeioka4.fsf@panacea.canonical.org>

"Steven Varco" <ngcontact at varco.ch> writes:
> In our company we developed a python script thats helps us in a bit complex
> program structure.
> Our problem is now, that we have to run the script on our Host system, that
> runs with MVS OS.
> 
> So we have to port it... Does anyone here have experiences with porting
> python to MVS?

I haven't run Python on MVS, but I have run it on Linux on the
zSeries, and it worked OK (although I never did get the test suites to
pass, unfortunately).  I'm very interested to hear about your
experiences.



From bokr at oz.net  Thu Jun 27 16:45:16 2002
From: bokr at oz.net (Bengt Richter)
Date: 27 Jun 2002 20:45:16 GMT
Subject: Python crashed on me
References: <mailman.1025196737.29740.python-list@python.org>
Message-ID: <afftgs$5p7$0@216.39.172.122>

On Thu, 27 Jun 2002 20:50:42 +0400, Oleg Broytmann <phd at phd.pp.ru> wrote:

>Hello!
>
>   Recently Python (one program that I am dbugging) started to crash.
>FreeBSD kills it with "Bus error", Linux with "Segmentation fault".
>
>   I think the program crashed in the cPickle.dump(file, 1), because after
>the crash the file is corrupted and cannot be load back (upon loading
>cPickle complain EOFile). The error is reproduceable.
>
>   How can I debug the situation? May be my data grew too big? The pickle
>file is about 800K.
>
>   Details. The program is my URL robot
>(http://phd.pp.ru/Software/Python/#bookmarks_db). The robot runs
>urllib.urlretrive() on my bookmarks file and report the results. Between
>runs data may be saved in different formats, including pickle.
>   While trying to understand what is going on and make full run faster I
>replaced the call to urlretrive with fake function that returns nothing.
>The robot crashed with the same segfault.
>
>   Python is
>
>Python 2.2.1 (#2, Apr 10 2002, 15:15:56)
>[GCC 2.95.2 20000220 (Debian GNU/Linux)] on linux2
>
To eliminate pickle, you could substitute a dump pickle.
If it's happening in pickle, you could substite a file object of your own,
to get a clue what it died on, e.g., you might import the following dummy
output class POUT, and first get a file object with opt=None to see if pickle
dies doing dummy output. Then use the 'n' option and some start count to get
nearer the end and see what the last number is. It should just keep printing
on the same line, so you can watch them whiz by without scrolling overhead, at least.

Then back off a few dozen, and run it with opt='all' and start=<whatever, near the end>
to get the last few objects and presumably the one it died on.

====< pout.py >========================================
class POUT:
    def __init__(self, opt=None, start=0):
        self.opt = opt
        self.start = start
        self.n = 0
    def write(self, s):
        if not self.opt: return
        self.n +=1
        if self.n < self.start: return
        if self.opt == 'n':
            print ('\r%d' % self.n),
        else:
            print '%8d: %s' % (self.n, `s`)

def test(opt='all',start=0):
    import pickle
    f = POUT(opt,start)
    print "---< pickle dump binary of 'example object' and 0x12345678 >---"
    pickle.dump('example object', f, 1)
    pickle.dump(0x12345678, f, 1)
    print "---------------------------------------------------------------"

if __name__=='__main__':
    test()
====</pout.py >========================================

HTH

Regards,
Bengt Richter


From a at b.com  Wed Jun 19 21:35:15 2002
From: a at b.com (Thinkit)
Date: Thu, 20 Jun 2002 01:35:15 GMT
Subject: Unicode hexadecimal characters
References: <3eeda89d.0206191615.71b0629d@posting.google.com> <3D112C60.9C3D2F9F@engcorp.com>
Message-ID: <njaQ8.101599$6m5.87417@rwcrnsc51.ops.asp.att.net>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3D112C60.9C3D2F9F at engcorp.com...
> Thinkit wrote:
> >
> > It's a bit OT but does Unicode support the digits ten through fifteen
> > (for hexadecimal)?
>
> Do you mean "does Unicode have unique glyphs representing the
> hexadecimal digits ten through fifteen other than the standard
> A, B, C... ?"

Yes, that's what I mean.  I looked at the spec and saw a lot of different
numerals, including 1-20 for circled numbers.  Will have to examine it
again.

> Unicode of course supports various glyphs for A through F, but
> it has nothing to do with other representations of bases other
> than ten and under, as far as I know.
>
> Not sure if that's exactly what you meant...
>
> -Peter




From jwbaxter at spamcop.net  Wed Jun 19 23:46:39 2002
From: jwbaxter at spamcop.net (John Baxter)
Date: Wed, 19 Jun 2002 20:46:39 -0700
Subject: Unicode hexadecimal characters
References: <3eeda89d.0206191615.71b0629d@posting.google.com> <3D112C60.9C3D2F9F@engcorp.com> <njaQ8.101599$6m5.87417@rwcrnsc51.ops.asp.att.net> <3D1137E2.7CB8F814@engcorp.com>
Message-ID: <jwbaxter-CB588C.20463919062002@corp.supernews.com>

In article <3D1137E2.7CB8F814 at engcorp.com>,
 Peter Hansen <peter at engcorp.com> wrote:

> It's unlikely there is any 
> aspect of Unicode which covers "pretty" representations of hexadecimal
> digits...

No representation I've seen is less pretty than NCR's for the C-315 (ca 
1963).  The 6 digits for the hex values were simply the six characters 
which happened to fall after the digits in the four-byte world.  I think 
10 was @...12 was space, and others were , . and two more.

Memory dumps in which spaces are either significant data or whitespace 
depending on column "can be a nuisance."  ;-)

   --John


From whisper at oz.net  Wed Jun 19 20:22:24 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 19 Jun 2002 17:22:24 -0700
Subject: Threading question
Message-ID: <GCEDKONBLEFPPADDJCOECEFBDKAA.whisper@oz.net>

with a thread sitting in a loop processing intermittant transactions from a
buffer, how does one wait if reading an empty buffer does not block - how do
you relinquish control to other threads/processes?

TIA,

David LeBlanc
Seattle, WA USA





From bokr at oz.net  Tue Jun 25 15:57:56 2002
From: bokr at oz.net (Bengt Richter)
Date: 25 Jun 2002 19:57:56 GMT
Subject: SOLVED (Re: Python hits the spot)
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at> <af4vrl$ffe$0@216.39.172.122> <3D16DD5C.1050708@kfunigraz.ac.at> <uhdume2n1mhbba@corp.supernews.com> <3D171291.2060905@kfunigraz.ac.at> <af8ib7$k51$0@216.39.172.122> <3D181BA2.3090009@kfunigraz.ac.at>
Message-ID: <afai04$97n$0@216.39.172.122>

On Tue, 25 Jun 2002 09:28:34 +0200, Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote:

>Bengt Richter wrote:
>
>> Well, lucky guess on the heat. That _probably_ accounts for laptop
>> going to 87% and then shutting down, but how about the other huge
>> slowdowns in your big loop of 8? Did you get that diagnosed too?
>> Just curious ;-)
>
>On my stationary machine definitely not. The 8 steps remain constant. I 
That can be read two ways ;-) 
1) Definitely no diagnosis, and 8 steps remain different as they were
2) Definitely no slowdowns, and 8 steps now execute with ~constant timing.

Still curious ;-)

>am contemplating to buy a special cooler for may laptop (I saw some in a 
>computer magazine).
>
I think the laptop manufacturer should buy it for you (unless you were
running it outside specs for ambient temperature and ventilation).
Good luck.

Regards,
Bengt Richter


From sketerpot at chase3000.com  Tue Jun 11 22:34:14 2002
From: sketerpot at chase3000.com (Peter Scott)
Date: 11 Jun 2002 19:34:14 -0700
Subject: Cross platform spawning
References: <mailman.1023811574.29386.python-list@python.org>
Message-ID: <659f82ff.0206111834.533adf16@posting.google.com>

"Harvey Thomas" <hst at empolis.co.uk> wrote:
> Not sure if it's exactly what you want but under Win NT you could use
> 
> os.system('start python myscript.py myargument')
> 
> (see the NT help for details of the start command)

Yes, that's exactly what I want. I ended up using code like this:

if sys.platform.startswith('win'):
    os.system('start python myscript.py myargument')  # Windows
else:
    os.system('python myscript.py myargument &')  # Unix

It works perfectly, and that solved the final bug in my program. Thanks! :-)


From bokr at oz.net  Sun Jun 30 18:37:16 2002
From: bokr at oz.net (Bengt Richter)
Date: 30 Jun 2002 22:37:16 GMT
Subject: yield equivalent in C/JavaScript?
References: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk> <mailman.1025460170.29711.python-list@python.org>
Message-ID: <afo16s$imk$0@216.39.172.122>

On Sun, 30 Jun 2002 14:01:39 -0400, Oren Tirosh <oren-py-l at hishome.net> wrote:

>On Sat, Jun 29, 2002 at 07:12:46PM +0100, Robin Becker wrote:
>> Is there any way to do yielding easily in C? I have a bunch of programs
>> which are normally connected together using unix pipes. I would like to
>> merge them and the problem is then how to get them to communicate nicely
>> using a single process.
>
>It's possible to implement this in C using Duff's Device:
>
>#define BEGIN_GENERATOR static int _GenState=0; \
>  switch(_GenState) { \
>    case 0:
>
>#define END_GENERATOR } \
>  _GenState = 0;
>
>#define YIELD(x) _GenState=__line__; return (x); case __line__:
>
>
>int generator(arguments) {
>
>  static int foo, bar, etc;	/* Make all variables static */
>
>  BEGIN_GENERATOR 
>  for(...) {
>      if(...) {
>        ...
>        YIELD(x)
>        ...
>      } else {
>        ...
>      }
>      ...
>      YIELD(y)
>      ...
>   }
>   END_GENERATOR
>}
>
>If you want this to be reentrant use a struct passed as an argument
>to the function instead of static variables.
>
That's cool, but I'm wondering how generally BEGIN_GENERATOR's switch/case hopping
into the inside of the for loop like the above will work. Mind-bogglingly, it seems the
switch/case can slice just about orthogonally through loop code. Normally jumping into loops
is frowned upon, I thought, but it seems like case is practically like a label for a goto with
the minor restriction of being inside a switch(){...}, which your BEGIN_/END_GENERATOR guarantees.
I never explored this particular arena of possible abuses much ;-)

Anyway, you can't generate case numbers from line numbers without some risk of collision:

E.g., looking at the preprocessor output (which doesn't mind your ellipses ;-)
and putting a contrived switch around the second YIELD like
--
      switch(y){
      case 28:
          YIELD(y);
          printf("YIELD will make this case 29 here");
          break;
      case 29: printf("So this won't compile"); break;
      }
--

in place of the plain YIELD(y) line, we get:

--
int generator(arguments) {

  static int foo, bar, etc;     /* Make all variables static */

  static int _GenState=0; switch(_GenState) { case 0:
  for(...) {
      if(...) {
        ...
        _GenState=21; return (x); case 21:
        ...
      } else {
        ...
      }
      ...
      switch(y){
      case 28:
          _GenState=29; return (y); case 29:;
          printf("YIELD will make this case 29 here");
          break;
      case 29: printf("So this won't compile"); break;
      }
      ...
   }
   } _GenState = 0;
}
--

BTW, MSVCC6.0 wants
    #define __line__ __LINE__
at the top to do the above. And for compiling, it doesn't like case __LINE__:
because __LINE__ appears not to be a constant expression to it. At the moment
I can't find the directive to get around that. It's happy to compile the
pre-processor output though ;-)

Regards,
Bengt Richter


From a.schmolck at gmx.net  Tue Jun 18 14:41:06 2002
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 18 Jun 2002 19:41:06 +0100
Subject: "Fatal Python error: GC object already in linked list"
References: <yfsfzzleg6s.fsf@black132.ex.ac.uk> <lkhek0lt79.fsf@pc150.maths.bris.ac.uk>
Message-ID: <yfs3cvke8fx.fsf@black132.ex.ac.uk>

Michael Hudson <mwh at python.net> writes:

> Alexander Schmolck <a.schmolck at gmx.net> writes:
> 
> > Just got the following in an interactive session:
> > 
> >     >>> reload(awmstools)
> >     Exception exceptions.RuntimeError: 'maximum recursion depth exceeded' in  ignored
> >     Fatal Python error: GC object already in linked list
> >     
> >     Process Python aborted
> >     
> > Any ideas as to what might be the cause?
> 
> Is this the lastest version of Python?  Is it reproducible?

Yes (2.2.1). I guess I might be succesful in reproducing it under controlled
circumstances, but I haven't tried so far. It has occured in interactive
sessions while I've been working on pickling (in the form of __get/setstate__
and cpickle) for a class in a module that uses a C++ extension (pymat -- a
matlab wrapper) and doesn't seem to be linked to reload as such (later on
other commands also produced the same error message).

It seemed to occur after something with the pickling went wrong, but after I
completed the routines and they worked correctly (or appeared to) I haven't
had these problems anymore. 

The error message seemed rather bizzarre ("GC object already in linked list"),
so I was wondering if anyone could point me into the right direction.

If the whole thing is just a freak interaction between any of: abortive
(c)pickling, the C++ extension, `inspect`, `weakref` (the modules which I
suspect to be most likely to cause troubles) and only occurs when something is
wrong then I'm not sure I fancy spending much time on tracking down what
exactly is going on, unless doing so would actually be beneficial to someone
(me, in case problems continue even after the pickling routines work or
someone else, in case this might be a python bug).

alex



From ken at hotmail.com  Sat Jun  1 23:56:18 2002
From: ken at hotmail.com (Ken)
Date: Sun, 2 Jun 2002 13:56:18 +1000
Subject: How to call functions inside a cgi script from the html interface?
Message-ID: <adc4sv$vjrdc$2@ID-49758.news.dfncis.de>

How to call functions inside a cgi script from the html interface?

Eg. If there is a link on the html display call "Search", how do I get it to
call the function call "def search( )" inside the cgi script?

Thanks




From ak at silmarill.org  Sat Jun  1 23:16:40 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Sun, 02 Jun 2002 03:16:40 GMT
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org> <3CF8E382.3A44F0DA@engcorp.com> <83ptzavk14.fsf@panacea.canonical.org> <3CF9408E.6015D2A@engcorp.com> <3cf94ca7$0$31235$e4fe514c@dreader1.news.xs4all.nl>
Message-ID: <slrnafj3eo.86f.ak@ak.silmarill.org>

In article <3cf94ca7$0$31235$e4fe514c at dreader1.news.xs4all.nl>,
    Boudewijn Rempt wrote:
> Peter Hansen wrote:
>> 
>> Other options might include making sure modules are not loaded
>> until they are needed, or loading them on a separate thread just
>> after the main program starts.
>> 
> 
> Convincing your users that they need to preload the libs might
> work too -- for instance by creating a 'server' component that
> should be started when the machine is booted that loads the lib.
> 
>> None of which seems to help the OP.  I'm curious why the Linux
>> implementation is so much slower than the Win32.  My own machine
>> is _not_ fast!  (It's only a 233 MHz Pentium MMX.)
>> 
> 
> wxPython has to load gtk, probably, and the bindings. The underlying
> gui library is already present in memory on a windows machine, leaving
> only the bindings to be loaded.
>
gtk-gnutella starts in 0.5 seconds on this system. I timed the
program I posted about, it takes 5.6s on first invocation, 4.7
and 4.6 on 2 subsequent ones.

> 
> I've also noticed that Python loads its own bytecode files rather
> more slowly than other files. So any pixmaps shouldn't be put in the
> code, but loaded separately. That can make quite a difference.
> 


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From johnroth at ameritech.net  Fri Jun 14 17:58:24 2002
From: johnroth at ameritech.net (John Roth)
Date: Fri, 14 Jun 2002 17:58:24 -0400
Subject: Zip makes sense! was: Re: why not "'in' in 'in'"?
References: <mailman.1024086417.5966.python-list@python.org>
Message-ID: <ugkpmkj4f0sid4@news.supernews.com>

"David LeBlanc" <whisper at oz.net> wrote in message
news:mailman.1024086417.5966.python-list at python.org...
> Aaaah, a logger. Obviously the first on his block to be a logger. The
joys
> of bravely loggin where no one has logged before! ;->
>
> Perhaps if he was an english major, or english was his first language,
he
> could tell that the common meaning of yield isn't produce. I have no
idea
> what zip means beyond being a name for a compression utility or a verb
> meaning to exercise a zipper.

But that's what it is: exercising a zipper...
It puts two (or more) sequences together into a single
sequence, just like a zipper puts the two sides
together into one.

John Roth





From irmen at NOSPAMREMOVETHISxs4all.nl  Thu Jun 13 18:07:14 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Fri, 14 Jun 2002 00:07:14 +0200
Subject: Python 2.2.1
References: <II7O8.622$e27.275217555@twister1.starband.net>
Message-ID: <3D091792.3060500@NOSPAMREMOVETHISxs4all.nl>

Dave wrote:
>     Hi, I am 12 years old and would like to begin learning Python over 
> the summer. 

Wow! Talking about good idea's to spend your free time in the summer ;-)
Good luck and I think you will enjoy learning Python.

>     How do I set the path and what should the name of the path be? All 
> help and information would be appreciated. Thanks.:)

You could install ActivePython (as the other reply pointed out). That's
probably the easiest solution. It also comes with the PythonWin
GUI development environment, which saves you from typing a lot at
the command prompt.


If you really want to know... look around a bit for "PATH environment 
variable", "c:\autoexec.bat" and "windows 98".
Maybe it's even in the Win98 help (just press F1)?

Irmen



From cliechti at gmx.net  Mon Jun 24 15:03:30 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 24 Jun 2002 21:03:30 +0200
Subject: Generic Python
References: <3D16E465.F0D2509E@ifib.uni-karlsruhe.de> <uhe1069gs1f4fc@news.supernews.com>
Message-ID: <Xns9237D6A7EEE66cliechtigmxnet@62.2.16.82>

"John Roth" <johnroth at ameritech.net> wrote in 
news:uhe1069gs1f4fc at news.supernews.com:

> 
> "Uwe Mayer" <Uwe.Mayer at ifib.uni-karlsruhe.de> wrote in message
> news:3D16E465.F0D2509E at ifib.uni-karlsruhe.de...
>> Hi,
>>
>> is it possible to to write a class which, f.e. takes an argument in
>> __init__() and that doesn't return an instance object but a new class
>> object?
> 
> Sure, but you have to do it closer to the source level. One
> quibble here: you can't do it with the __init__() method, that's
> defined to always return an instance. You have to do it with some
> kind of generator or factory function.

just to clarify: __init__ is _not_ return anything.
1. an instance is created, you can already call methods on it.
2. __init__ is called if defined. (it's an initializer, notthing more or 
less)

that why it isn't called __new__ [hint, hint]

>>> class A(type):
... 	def __new__(klass, arg):
... 		return int(arg)
... 
>>> A(5)
5
>>> type(A(5))
<type 'int'>


chris


-- 
Chris <cliechti at gmx.net>



From oliver.eichler at dspsolutions.de  Fri Jun 28 03:42:43 2002
From: oliver.eichler at dspsolutions.de (Oliver Eichler)
Date: Fri, 28 Jun 2002 09:42:43 +0200
Subject: Execute python script via C API line-by-line
Message-ID: <afh413$h5f$01$1@news.t-online.com>

Hi

Is it possible to execute a Python script via C API interactively? Like: 

- read the file
- execute 1st line 
- tell next/current line (especially usefull for loops and function calls)
- execute next line
- ...

As far as I can see from the documentation I can use the normal commandline 
interface or execute a file as one single call. There seems to be some 
parsing functionality described, however, the official API documentaion is 
too brief to make much sense to start with. Is there a better C API docu to 
start with (books/link)? Something explaining the concepts in detail?  


Just as background:
I would like to control some hardware via python script for testing 
purposes. That's why I would like to step thru the script line-by-line. The 
aim is to have a text edit widget with the script code and the current line 
to process highlighted.

Thanks for hints, details and stuff

Oliver


From jdhunter at nitace.bsd.uchicago.edu  Thu Jun 20 09:34:49 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Thu, 20 Jun 2002 08:34:49 -0500
Subject: MySQLdb question
References: <aesi66$cgb$1@sunsite.icm.edu.pl>
Message-ID: <m2sn3inkee.fsf@mother.paradise.lost>

>>>>> "Micha?" == Micha? Kurowski <mkur at poczta.gazeta.pl> writes:

    Micha?> I cannot find a way to pass a sql statement a table as a
    Micha?> parameter.

Use the %s thingie:

c.execute('SELECT spam, eggs, sausage FROM %s
           WHERE price < %s', (table_name, max_price))

JDH


From fperez528 at yahoo.com  Thu Jun 20 12:53:25 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 20 Jun 2002 10:53:25 -0600
Subject: Pedagogic advice needed
References: <3D104D35.E0932297@info.unicaen.fr> <aeqfuh$88n$1@peabody.colorado.edu> <ec4802fb.0206192008.318f0e45@posting.google.com>
Message-ID: <aet1a6$gh7$1@peabody.colorado.edu>

Travis N. Vaught wrote:

> Actually, nightly builds (with a test scoreboard) are creating
> downloadable binaries (from the cvs) that make installation much
> nicer.??You?still?need?Numeric?and?wxPython,?but?windows?installs
> (;)), at least, are quite simple now.

Sorry Travis. Goes to show that I haven't built recently since the new 
nightlies system was put in place. My apologies for the misinformation.

cheers,

f.


From benji_york at cknainc.com  Thu Jun  6 10:55:16 2002
From: benji_york at cknainc.com (Benji York)
Date: 6 Jun 2002 07:55:16 -0700
Subject: Medium to Large Scale Python Deployments
References: <KHkL8.21383$eD2.2399923271@newssvr10.news.prodigy.com>
Message-ID: <b10fcc9a.0206060655.443d06c2@posting.google.com>

"Domenic R. Merenda" <dom at edgereport.put_a_c_o_m_here> wrote in message news:<KHkL8.21383$eD2.2399923271 at newssvr10.news.prodigy.com>...
>    I am running a home-grown Enterprise Resource Planning system, written
> entirely in Python...

Do you have any plans to open source your system. I, and others, would
be most interested in using, and contributing to such a project.

--
Benji York
benji_york at cknainc.com


From ftobin at neverending.org  Mon Jun  3 16:34:54 2002
From: ftobin at neverending.org (Frank Tobin)
Date: Mon, 03 Jun 2002 20:34:54 GMT
Subject: What is a classmethod for?
In-Reply-To: <addgv30f4d@enews4.newsguy.com>
References: <addgv30f4d@enews4.newsguy.com>
Message-ID: <Pine.LNX.4.44.0206031559001.8474-100000@palanthas.neverending.org>

On Sun, 2 Jun 2002, Dennis Peterson wrote:

> That's nice, but I can't think of a use for it, given that we have
> inheritance and isinstance(). Can anyone enlighten me?

Personally, I like using classmethods as secondary constructors, generally
gones that are able to construct 'from' something.

e.g.:

class HostPortAddress(tuple):
	def from_str(self, s):
		return self(s.split(':', 1))
	from_str = classmethod(from_str)

-- 
Frank Tobin			http://www.neverending.org/~ftobin/



From twister at raplounge.de  Mon Jun 10 13:59:09 2002
From: twister at raplounge.de (Ben Thursfield)
Date: Mon, 10 Jun 2002 19:59:09 +0200
Subject: Newbie question communicating between 2 progs
References: <3D046BF0.DC36542C@raplounge.de> <3D04A938.627EE705@engcorp.com>
Message-ID: <3D04E8ED.C0708C2B@raplounge.de>


Peter Hansen schrieb:

> The question is ambiguous.  When you say "frames", the only thing
> that comes to mind is that you are running Python on client-side
> Internet Explorer under Windows, as scripts in two different parts
> of a window (the frames) simultaneously.  I think you must have
> meant something different however.
>
> Could you please describe more about your environment?  The
> term "frame" can mean a number of things.

That's right. I am running 2 cgi-scripts in two different frames of the
browser, the one should only print an array into the frame and the other one
communicates with a server. I decided to split the programm up because i want
to refresh the printing frame every second. The main frame changes the array
the second frame prints it out. My host has Python installed, so the
#usr/bin/python method works fine.
So the question is how can the programm responsible for printing out the array
access the array defined in the other programm.

I hope i described the problem better now.

Ben



From peter at engcorp.com  Thu Jun 27 09:59:19 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 27 Jun 2002 09:59:19 -0400
Subject: How to find out operating system
References: <mailman.1025162014.16302.python-list@python.org>
Message-ID: <3D1B1A37.CF516C4C@engcorp.com>

David LeBlanc wrote:
> 
> You might try os.environ:
> 
> os.environ['WINOS'] -> WIN2000
> 
> I don't know if WINOS is defined on '98 or ME, but it should be good for NT,
> 2K and XP

It is not.


From narnett at mccmedia.com  Sun Jun 16 18:49:01 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Sun, 16 Jun 2002 15:49:01 -0700
Subject: Visual Python
In-Reply-To: <337619fa.0206161337.546a958f@posting.google.com>
Message-ID: <NDBBLAEMOMEHKALAMKKJMEOHKPAA.narnett@mccmedia.com>


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Raphael Ribeiro
> Sent: Sunday, June 16, 2002 2:38 PM
> To: python-list at python.org
> Subject: Visual Python
>
>
> When i go to tools/debug and the ms-dos prompt start appears ,  an
> error message appears , anyone knows why?

Which item in the Debug menu are you choosing?  What's the error?

I believe ActiveState says to start a debug run with Step Into or Step
Through, not the other options (just checked this, it's only true if Visual
Studio .NET was not an upgrade to Visual Studio version 6).  However, I
haven't had a problem using Start (and my installation was not an upgrade).
But then again, I'm running it with Python 2.2, which isn't tested, so I'm
breaking all sorts of rules, it seems.

You don't have to run in a shell, you know.  You change that in the
project's properties.  And that's where you also tell it which module to run
at start, which had me baffled for a while, I have to admit.  In the project
properties dialog, un-check "Run in shell window" to get rid of the prompt
where your error is occurring.  Not that that'll fix the error, of course...

Nick





From uncleenzo at hushmail.com  Thu Jun 27 12:48:06 2002
From: uncleenzo at hushmail.com (Uncle Enzo)
Date: 27 Jun 2002 09:48:06 -0700
Subject: COM server threads/unloading
References: <10d944cb.0206261529.741da082@posting.google.com> <3D1A7A9A.4050208@skippinet.com.au>
Message-ID: <10d944cb.0206270848.ed1d70b@posting.google.com>

Thanks for your reply, Mark. I have a couple of follow-up questions...

Mark Hammond <mhammond at skippinet.com.au> wrote in message 
> * Ensure CoUninitialize() is called for each thread (you would have 
> needed to call CoInitialize[Ex] manually for these threads.
> * Call CoUninitialize() on the main thread.

I noticed in the documentation that CoUninitialize() is not called for
the main Python thread automatically. But, how will my server know
when to call it?Should I ask my clients make a Release-like call?

Also, my testcomserver is simple, but it imports the multithreaded
logger. Does proper use of pythoncom require me to add pythoncom
Co(Un)Initialize() calls to the logger?

Thanks in advance,

Enzo


From peter at engcorp.com  Mon Jun  3 07:57:56 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 03 Jun 2002 07:57:56 -0400
Subject: Is there a Text to HTML conversion module?
References: <B9208083.5EF6%sarmstrong13@mac.com>
Message-ID: <3CFB59C4.90BBB459@engcorp.com>

SA wrote:
> 
> Does anyone know of a module that converts a text file to html?
> 
> I started writing my own, but decided I better make sure I'm not
> re-inventing the wheel here.

You are probably looking for what has been labelled "Structured Text".
Try a search on google with that and "python"...

-Peter


From jepler at unpythonic.net  Mon Jun  3 14:24:41 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Mon, 3 Jun 2002 13:24:41 -0500
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
In-Reply-To: <ctKK8.185481$Po6.352104@rwcrnsc52.ops.asp.att.net>
References: <06da01c208f4$d69003c0$6601a8c0@boostconsulting.com> <200206010121.g511LxX19223@pcp742651pcs.reston01.va.comcast.net> <073501c20960$23f367e0$6601a8c0@boostconsulting.com> <200206011334.g51DY1D21669@pcp742651pcs.reston01.va.comcast.net> <07d801c20970$cdebe050$6601a8c0@boostconsulting.com> <mailman.1022940122.3421.python-list@python.org> <Id5K8.171932$Po6.340972@rwcrnsc52.ops.asp.att.net> <oWIK8.179611$%u2.31494@atlpnn01.usenetserver.com> <cHJK8.181202$%u2.93890@atlpnn01.usenetserver.com> <ctKK8.185481$Po6.352104@rwcrnsc52.ops.asp.att.net>
Message-ID: <20020603132440.B19967@unpythonic.net>

Maybe the bytecode for inplace operations like
    f[i] += v
should implement the
pseudocode
    _temp = f[i] 
    _new = inplace_add(_temp, v)
    if _new is not _temp:
	f[i] = _new
    del _temp, _new

Then, the following would work:
    >>> x = ([1], [2])
    >>> x[0] += [3]
    >>> x
    ([1, 3], [2])
and the following would continue to not work
    >>> x = (1, 2)
    >>> x[0] += 3
    TypeError: object doesn't support item assignment

However, this would probably double the bytecode size of the sequence,
and hurt performance similarly....

And personally, I still never use augmented-assignment syntax anyway..

Jeff




From opengeometry at NOSPAM.yahoo.ca  Wed Jun 12 20:19:40 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 12 Jun 2002 20:19:40 -0400
Subject: Wanted:  Potential Python Authors
References: <ae7gke$mhb$1@ccis12.baylor.edu> <db3fgu40sbnt3pno5v1i2h9lnmpcoti2bk@4ax.com> <3d07d4a8@news.eol.ca> <ae8khr$nnn$1@panix1.panix.com>
Message-ID: <3d07e51c@news.eol.ca>

Aahz <aahz at pythoncraft.com> wrote:
> In article <3d07d4a8 at news.eol.ca>,
> William Park  <opengeometry at NOSPAM.yahoo.ca> wrote:
>>
>>There is no such thing as "trained as writer", whether it's for computer
>>books or fiction/nonfiction novels.  You either can write or you can't.
> 
> That's not true.  Writing is a skill -- just like programming -- that
> can be learned and/or taught.  Of course, just like programming, innate
> talent counts for something, but there are few people on this newsgroup
> who are incapable of writing a book.

Writing a computer book is not just grammar and syntax, although poor
grammer/syntax makes it worse.  You have to deliver the info fast and with
some structure (usually with examples).

My biggest peeves about current books is that they don't give me infos fast
enough; and, when they do, they are so disorganized that I have to break my
flow in reading to parse out they are babbling about.  HTML/XML books are
notorious in this regard.

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin


From pyth at devel.trillke.net  Sun Jun  2 05:58:33 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 2 Jun 2002 11:58:33 +0200
Subject: Pop a list from beginning ? and memory saving...
In-Reply-To: <adcpdg$vui53$1@ID-146704.news.dfncis.de>; from shagshag13@yahoo.fr on Sun, Jun 02, 2002 at 11:44:32AM +0200
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de>
Message-ID: <20020602115833.B30380@prim.han.de>

Shagshag13 wrote:
> I had a huge list containing consuming data, as i only need a one pass in a
> for statement,
> i'm wondering if by doing this with a pop i could save memory ? (as i think
> that python
> free it after the pop...)
> and is there a way to pop from the beginning  of the list ? (i think i could
> use reverse but
> i don't know if this memory consuming)
> 
> >>> a = [1,2,3,4,5,6,7]
> >>> while 1:
>  try:
>   f = a.pop()
>  except IndexError:
>   break
>  print f

try a.pop(0) :-)

    holger




From whisper at oz.net  Mon Jun 10 16:43:50 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 10 Jun 2002 13:43:50 -0700
Subject: wxPlotCanvas documentation?
In-Reply-To: <20020610182231.26180.qmail@web21108.mail.yahoo.com>
Message-ID: <GCEDKONBLEFPPADDJCOEOEONDHAA.whisper@oz.net>

In the lib, take a look at wxplotcanvas.py, especially the __main__ routine.
I empirically discovered that this blows up with a memory access error on
line 36 if numeric isn't installed first. Otherwise, it seems to work.

Generally speaking, Python tends to be long on code and shorter on
docuemntation. I think this is true of most open source projects. Compared
to most, Python itself is well documented (and has holes in it's
documentation too!), but it's "groupies" tend to be less so.

The solution is to:
* check out any samples/demos
* check out the wxpython wiki: http://wiki.wxpython.org/, complete with
cookbook!
* grep the source - in particular in Pythonland, look for test/demo scripts
lurking at the bottom of lib files hiding behind __main__
* as a last resort, ask on newsgroups :-)

In the case of WxWindows/WxPython itself, the WxWindows doc describes which
classes are/are not implemented in WxPython and the WxWindows doc itself is
a guide to those that are.

HINT: about your specific questions. EVT_* are events and don't throw
events, they get thrown. PlotCanvases are Frames are Windows and Window has
a GetClientSize().

Is that a WxButtLight? ;)

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of David Brady
> Sent: Monday, June 10, 2002 11:23
> To: Python
> Subject: wxPlotCanvas documentation?
>
>
> I have searched the wxPython and wxWindows
> documentation, peered through the wxdemo files, and
> searched the web in the hopes of finding the answers
> to my questions, but to no avail.
>
> I would like to write a simple graphing calculator
> program with wxPython.
>
> 1. How do I create a wxPlotCanvas in a wxFrame window?
>
> I can see in wxPython/demo/wxPlotCanvas.py  how to
> create a PlotCanvas object if I wanted, say, to run my
> graphing calculator inside the wxDemo framework, but I
> don't.  How do I instantiate this guy and stick him in
> a wxFrame class of my own deriving?
>
> 2. How do I manipulate wxPlotCanvas?
>
> From the demo code I see there are PolyMarker and
> PolyLine objects, as well as a PlotGraphics function.
> However, since I can find exactly zero documentation
> on PlotCanvas, this is all I know about this object.
>
> 3. How do I find anything in the documentation?
>
> My inability to find my wxButt in the wxDark with both
> wxHands (with or without a wxFlashlight) is not
> limited to PlotCanvas, so I'm hoping that the problem
> is on my end.  The wxPython website, as far as I can
> tell, only has one documentation link, and it's a link
> to wxWindows, not the wxPython objects that wrapper
> it.  There's a sample code section, but every last one
> of the examples is marked "To Be Written When I Have
> More Time."
>
> The wxPython homepage proudly shows a long list of
> people who claim that wxPython is a Java and/or
> TkInter killer.  All I want to know is, how the hell
> did these people figure out how to use it?
>
> I'm not complaining about wxPython, mind you.  What
> I've gotten to work makes me believe that it really is
> wonderful.  But I'm at wit's end here trying to find
> out how to do anything that isn't already in a
> tutorial.  Like get the client rect of a wxFrame
> object.  Like find out what events EVT_PAINT might
> throw.  Like use a wxBitmap to double-buffer a wxDC.
> Like do anything beyond the first four tutorial pages.
>
> HELLLLLP!
>
> (thanks)
>
> -dB
>
> =====
> David Brady
> daves_spam_dodging_account at yahoo.com
> I'm feeling very surreal today... or *AM* I?
>
> __________________________________________________
> Do You Yahoo!?
> Yahoo! - Official partner of 2002 FIFA World Cup
> http://fifaworldcup.yahoo.com
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From larooy at xtar.co.nz  Mon Jun  3 02:13:32 2002
From: larooy at xtar.co.nz (John La Rooy)
Date: Mon, 3 Jun 2002 18:13:32 +1200
Subject: Why no open(f, "w").write()?
References: <mailman.1023067625.14591.python-list@python.org>
Message-ID: <20020603181332.2975c585.larooy@xtar.co.nz>

> 
> It is explicitly not a documented part of the interface - the documented
> part is that it is not safe to rely on this behaviour.
> 
> When it is explicitly documented that this behaviour is guaranteed now and
> forever, for all implementations of Python after a specified version, I will
> *consider* using this behaviour. Until then it is not worthy of
> consideration.
> 
> Tim Delaney
> 
> 

It's unlikely to be explicitly documented. from the docs...

The file() constructor is new in Python 2.2. The previous spelling, open(),
is retained for compatibility, and is an alias for file().

So you shouldn't be using open() in new programs anyway

John


From imbosol at vt.edu  Thu Jun 20 23:04:27 2002
From: imbosol at vt.edu (Carl Banks)
Date: Thu, 20 Jun 2002 23:04:27 -0400
Subject: Generating SWFs using Ming via CGI
References: <2wvQ8.19$eH2.59629@ruti.visi.com>
Message-ID: <r35uea.rc.ln@grey.aerojockey.localdomain>

Tim Wilson wrote:
> 
> wilson at beethoven:~/public_html/cgi-bin$ python shape.py
> Content-type: applicaton/x-shockwave.flash
> 
> FWSQp
>     ??
>       C???p
>            0@???P?5,?G????&Q h?@wilson at beethoven:~/public_html/cgi-bin$
> 
> Can anyone provide a hint about how I can make this work?

First, check your spelling.  The MIME type of SWF files is
application/x-shockwave-flash (no period in it).


-- 
CARL BANKS


From owens at c21bowman.com  Fri Jun 28 19:43:03 2002
From: owens at c21bowman.com (Michael Owens)
Date: Fri, 28 Jun 2002 18:43:03 -0500
Subject: exec, namespace, and reference count question
Message-ID: <200206281843.03032.owens@c21bowman.com>

Since exec() allows for passing in a dictionary as the local and global 
namespaces, shouldn't it be possible to clear out all references to (global) 
variables declared in the exec'ed code simply by deleting the namespace dict? 
This doesn't work:

namespace = {}
exec(code, namespace)
del(namespace)

There are still references to the global variables instantiated within 'code' 
here. But this does:

namespace = {}
exec(code, namespace)
for key in namespace.keys():
   del(namespace[key])

Is the this then the correct/safe way to ensure that global variables created 
within the code are cleared so that there is no refcount buildup as a result 
of running the code block through exec more than once (i.e. the same global 
variables are reinstantiated and/or have their reference counts build up)? I 
am having some class object references increase after each exec, so I am 
wondering if this is the correct way to address it.





From pyth at devel.trillke.net  Thu Jun 13 06:09:28 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 13 Jun 2002 12:09:28 +0200
Subject: Why does Python mix OO concepts and non OO concepts for operation s  on basic types?
In-Reply-To: <20020613195038.281d513d.larooy@xtar.co.nz>; from larooy@xtar.co.nz.trillke.net on Thu, Jun 13, 2002 at 07:50:38PM +1200
References: <3D054A98.3090708@bgb.cc> <Pine.LNX.4.33.0206120008180.9668-100000@rnd.onego.ru> <mailman.1023827360.23248.python-list@python.org> <20020613195038.281d513d.larooy@xtar.co.nz>
Message-ID: <20020613120928.B6609@prim.han.de>

John La Rooy wrote:
> On Tue, 11 Jun 2002 22:28:44 +0200
> holger krekel <pyth at devel.trillke.net> wrote:
> 
> > Roman Suzi wrote:
> > > I only forgot how will I do this:
> > > 
> > > map(string.split, s)
> > > 
> > > if GvR will deprecate string module...
> > 
> > what about
> > 
> >     map(str.split, s) 
> > 
> > :-) holger
> > 
> > 
> what about 
> 
> map(string.capwords, s)
> 
> I'm guessing it's an oversight that there is no str.capwords??

probably. With python 1.6 many methods of the string *module* were
included included directly with the *type* string ('str'). Most 
python developers think that the string module should be deprecated 
some time in the future. But there are some issues which haven't been 
resolved so far. 'capwords' is one of them IMO and not very difficult. 

the 'string' module might even be never deprecated because so many 
programs (including more than 20 stdlib-modules) use it. 

    holger




From anoop79 at myrealbox.com  Mon Jun 10 11:55:25 2002
From: anoop79 at myrealbox.com (Anoop P B)
Date: Mon, 10 Jun 2002 21:25:25 +0530
Subject: Python 2 Monolithic Executable
Message-ID: <3D04CBED.8070803@myrealbox.com>

Hi,

	I need to write a program which will take user input and mail it to a 
predesignated email address. The constraint is that it needs to be a 
single, small executable (without requiring installation and no 
dependencies) (Windows). Is there a way to create such an executable 
from a python program?

Thanks in advance,
anoop





From occeanlinux at linuxmail.org  Thu Jun 27 14:34:27 2002
From: occeanlinux at linuxmail.org (SiverFish)
Date: Thu, 27 Jun 2002 18:34:27 +0000
Subject: How to get rid the new line
Message-ID: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org>

Anyone tell me how to skip the empty line when you read through the file
and copy the information in to the list(no empty line )


From jb at cascade-sys.com  Fri Jun 28 13:26:52 2002
From: jb at cascade-sys.com (James J. Besemer)
Date: Fri, 28 Jun 2002 10:26:52 -0700
Subject: speech synthesis
References: <idCS8.378$yY2.9108@ozemail.com.au>
Message-ID: <3D1C9C5C.8A196336@cascade-sys.com>

Alistair Campbell wrote:

> I am a newbie with Python so any response to this may be way over my head.
>
> Basically, I am wanting to know whether there are any python modules that
> allow you to write programs which include voice synthesis. [...]
> So, I want to be able to convert
> typed text and stored blocks of text into audible speech using standard
> sound card interfaces.

> I am using Python 2.2 on a Windows OS. I have a Linux OS as well but I am

What I have done is I downloaded the ViaVoice software from IBM.  This is a
large package that includes speech recognition, although presently I use it ONLY
for text to speech generation.  I'll go out on a limb here and suggest that
Speech Recognition and Analysis is something better done by an auxilliary
package outside of Python, rather than in Python itself.

ViaVoice is presently FREE on Linux, downloadable from IBM's web pages.   THe
package also runs on Windows, though IBM markets it as a retail product, so it's
not free.  They have a 'personal' version for $30 which I *believe* includes all
the necessary parts.  Once you have the Retail version, you can download the
Text to speech API for free and access the speech generator programmatically, if
you wanna go that route (may have to on Windows).

I have text to speech running on Linux.  I have not yet installed or tried it on
Windows.

ViaVoice installs pretty easily on Linux and works with its standard Audio
devices.  The only glitch I have encountered is that Linux dinks with device
permissions when people log in and out and this sometimes screws up long running
apps.  I need to change that permission thingy.

For the simple things I am doing I did not attempt to interface Python to the
API.  ViaVoice comes with some command line utilities that perform all the basic
tasks.  On Linux, I simply use os.popen2() to write the text to the command line
program.  This works fine for me.  There is a tiny latency in starting to talk
but I am not sure that would be speeded up via the API, as I suspect there is
some considerable computation to generate the speech.  I did a similar thing to
play WAV files via the Linux "play" command.

The generated speech is quite inteligible, though clearly mechanical.  The
default voice sounds a lot like Stephen Hawkings, though it's easy to configure
it for other voices and the install comes with a variety.  Even easier is to
speed up or slow down speech, change pitch and simple things like that.  It
comes with a customizable pronunciation dictionary but I haven't taken a look at
it as it does pretty great by itself.  If you have important proper nouns or
technical terms, you may need to work with it.  It does a pretty good job of
Understanding the text before pronuncing, thus it does the 'right thing' with
things like

    Dr. Smith lives on Elm Dr.

As I said, I merely put some simple python wrappers around the out of the box
IBM command line.  You could dig deeper along these lines if you wish or, as I
did, instead focus on the rest of your Python app.

First step would be to get the IBM software install it and get it to work.

You can look at my low-level Python code via

    http://www.cascade-sys.com/~jb/Pythonetics/vv/index.htm

As I said, the IBM sofware for Linux is available for download.  I forget the
URL.  A Google search for "IBM ViaVoice Linux" (maybe add "download"). should
get you close.  IIRC you sort of have to dink around their web pages to find the
exact right download page.  You also have to "register" before they'll let you
have the free code.

I got a copy of the Windows Personal version at FRY's for $30.  It's still in
the shrink wrap, as my associated project is on Linux.




--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com






From abidnego25 at yahoo.com  Sat Jun 22 18:52:11 2002
From: abidnego25 at yahoo.com (les)
Date: 22 Jun 2002 15:52:11 -0700
Subject: position of matches
Message-ID: <653cf871.0206221452.6cb17ef2@posting.google.com>

Hi,
i am new to python and want to do the following:
given a string str='abcABCDefsgRSTVderDae'
i would like to find all the internal lowercase strings between 2 caps
i.e. pattern=re.compile(r'[A-Z]([a-z]+)[A-Z]')

match_obj=pattern.search(str)
begin,end=match_obj.span()

however i would like to get all the begining and end positions
of the pattern, 
i.e.
efsg  begin=7 end=10
der   begin=15 end 17
thanks


From theller at python.net  Wed Jun 12 15:38:59 2002
From: theller at python.net (Thomas Heller)
Date: Wed, 12 Jun 2002 21:38:59 +0200
Subject: Can Distutils include .pyd files directly?
References: <mailman.1023906796.16318.python-list@python.org>
Message-ID: <ae8843$4t7d5$1@ID-59885.news.dfncis.de>

> "Bjorn Pettersen" <BPettersen at NAREX.com> wrote in message news:mailman.1023906796.16318.python-list at python.org...
> > From: Thomas Heller [mailto:theller at python.net]
> >
> > "Bjorn Pettersen" <BPettersen at NAREX.com> wrote in message
> > news:mailman.1023837255.23228.python-list at python.org...
> > > I've built my extension modules using the MS Devstudio IDE
> > and now I
> > > would like to use Distutils to create a Windows installer
> > for me that
> > > puts both the debug and release versions of the extension in the
> > > python22\DLLs directory. Is this possible? If not, is there
> > a way to
> > > have Distutils build my extension in both debug and release
> > mode and
> > > install both?
> > >
> > > -- bjorn
> >
> > Sure. The 'build' command builds a 'pseudo installation tree'
> > in the directory build\lib.win32-2.2, all of this will later
> > be installed by the 'install' command. So you can
> >   python setup.py build
> > to build the release version,
> >   python setup.py build -g
> > to build the debug version, and
>
> Is there any way I can skip having Distutils building the extension? If
> there isn't, is there a way to provide a separate 'library_dirs' option
> for debug libraries? (We have our release libraries in \active\release
> and the debug libraries in \active\debug, and they have the same
> names...)

It would be possible using custom distutils command classes -
but IMO it would be a shame to do so.
In your case I would recommend to build an installer using
Inno setup or something comparable.

Thomas




From elabuschagne at gims.com  Tue Jun 25 11:03:58 2002
From: elabuschagne at gims.com (GimsMail)
Date: Tue, 25 Jun 2002 17:03:58 +0200
Subject: Sending string var with apply sends all chars as params!???
Message-ID: <002501c21c5a$16f514f0$6f0d10ac@integer>

Hi all,

I am having problems with the apply function in Python.  When applying a string variable to a method, the string var gets chopped up into all it chars and each char gets sent as a parameter!  The result is getting a TypeError because I am sending too much parameters to the function (more detail below if you still don't understand)

Regards
Etienne

Detailed explanation:
When running my program I get the following:

My print line debugging string output:
Formatting value '20020625140609' of type <type 'string' with formatter <function TimeFormatter at 00D613F4>

as you can see, I am sending the string '20020625140609' to the function TimeFormatter.

The code that does this is:
print "UCP: Formatting value '%s' of type '%s' with formatter '%s'" %(value, type(value), formatter)
value = apply(formatter,(str(value)))

in this case, formatter is the function TimeFormatter (as defined in a mapping table)

When this code executes I get a TypeError:

TypeError: TimeFormatter() takes exactly 1 argument (14 given).

Why is the string (which is 14 chars long) broken up into 14 values when passed using apply!?

The weird thing is this code has been working for MONTHS and have only recently stopped working after some refactoring in a completely different part of code!

The problem goes away when I don't use apply:
value = formatter(value)

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020625/3443d575/attachment.html>

From peter at engcorp.com  Fri Jun 28 08:53:00 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 28 Jun 2002 08:53:00 -0400
Subject: feeding a prompt with python
References: <3D1C5341.9070001@huno.net>
Message-ID: <3D1C5C2C.2BE3C660@engcorp.com>

thomas wrote:
> 
> is there a way with python to feed a prompt with data? i call sth. with
> system($some_cmd) and the cmd gives me a prompt which of course waits
> for user input. however i want to automate this, is there a way?
> i browsed through the essential reference 2nd edt. but didn't find anything.

Maybe you could post an actually code sample, since "system($some_cmd)"
is not Python code.  We need more information to answer well.

-Peter


From roelsch at comtrak.com  Mon Jun  3 11:39:53 2002
From: roelsch at comtrak.com (Robert L. Oelschlaeger)
Date: 3 Jun 2002 08:39:53 -0700
Subject: Version strings in executable created by Py2exe
Message-ID: <4323edf0.0206030739.329a3629@posting.google.com>

Running with Python 2.2. I've created myAppName.exe using Py2exe, with
the "version_companyname" and "version_legalcopyright" specified in
the setup .cfg file as follows:

##########################
[py2exe]
.
.
.
# version information
version_companyname = My Company Name BLAH BLAH
version_filedescription = My File Description BLAH BLAH
version_fileversion = 0.0.2
version_legalcopyright = BLAH BLAH
version_legaltrademarks = None
version_productname = py2exe myAppName
version_productversion = 0.0.2
##########################

Using Windows Explorer, when I check "Properties" on the resulting
.exe file, I get:

------------
File Version: 0.0.2.0
Description: My File Description BLAH BLAH
Copyright:

Company Name:
Language: Language Neutral
Legal Trademarks:
Original Filename: myAppName
Product Name: myAppName
Product Version: 0.0.2
------------

Why are the version_companyname and version_legalcopyright strings not
installed for the "Company Name" and "Copyright"?


Bob Oelschlaeger


From sarmstrong13 at mac.com  Sat Jun  8 13:09:13 2002
From: sarmstrong13 at mac.com (SA)
Date: Sat, 08 Jun 2002 12:09:13 -0500
Subject: Tkinter help on MacOSX please.
References: <B9266129.6930%sarmstrong13@mac.com> <1fdga3f.1kt9ridf2v94yN%yvon.thoraval@free.fr>
Message-ID: <B927A469.6B30%sarmstrong13@mac.com>

On 6/8/02 1:27 AM, in article 1fdga3f.1kt9ridf2v94yN%yvon.thoraval at free.fr,
"Yvon Thoraval" <yvon.thoraval at free.fr> wrote:

> SA <sarmstrong13 at mac.com> wrote:
> 
>> I've installed python, tcl, tk on my MacOSX 10.1.5 system using the
>> --enable-frameworks configuration. I have a framework for tcl,tk,and python
>> under my /Library/Frameworks directory.
> 
> i've read somewhere that tkinter isn't - yet - compatible with Carbon ,
> don't know if that apllies to os X but to os 9 for sure !
That is not wholly correct. I know Tkinter works in OSX and Apple is
supporting it. You may be correct about OS9, but my problem is getting the
Apple version of Tcl/TK to work with Python's Tkinter module in OSX. I can
run the "Wish Shell" app in OSX and run a tcl program that calls the correct
widgets in Tk for OSX (ie Tk.framework), but I can't get Python's Tkinter
module to recognize the Tk libsand headers in Tk.framework. Part of the
problem is that I do not fully grasp the Apple frameworks idea.

Anyways, the main issue is that I can't get Python, from the command, line
to interface with TK properly. The error pops up in _tkinter in the Tkinter
module. Any help would be most appreciated.

Thanks.
SA



From dalke at dalkescientific.com  Mon Jun  3 17:19:24 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Mon, 3 Jun 2002 15:19:24 -0600
Subject: Is there a Text to HTML conversion module?
References: <mailman.1023132485.29235.python-list@python.org>
Message-ID: <adgn01$d18$1@slb5.atl.mindspring.net>

Michael Chermside:
>def textToHtml(text):
>     return '<html><body><pre>%s</pre></body></head>' % text

import cgi
def textToHtml(text):
     return '<html><body><pre>%s</pre></body></html>' % cgi.escape(text)

><wink>

<wink/>

>(sorry... couldn't resist)

Neither could I.  :)

                    Andrew
                    dalke at dalkescientific.com





From thor__00 at yahoo.com  Tue Jun 11 09:25:35 2002
From: thor__00 at yahoo.com (Thor)
Date: Tue, 11 Jun 2002 15:25:35 +0200
Subject: Filling a space...
Message-ID: <ae4v9h$3qg34$9@ID-108351.news.dfncis.de>

Having this tupe: (2,3,4), which one would be the msot efficient way to get 
this (parentheses ommited):

0,0,0  0,0,1  0,0,2  0,0,3
0,1,0  0,1,1  0,1,2  0,1,3
0,2,0  0,2,1  0,2,2  0,2,3
1,0,0  1,0,1  1,0,2  1,0,3....

I mean, to get all the possible integer points in a defined space.

Thanks in advance.
-- 
Thor -- Stockholm -- Sverige


From vze3myxh at verizon.net  Mon Jun 17 18:54:03 2002
From: vze3myxh at verizon.net (vze)
Date: Mon, 17 Jun 2002 22:54:03 GMT
Subject: ImageMagick for Python incl. Source
References: <aelmhc$35m$03$1@news.t-online.com> <aelo0t$sog$1@peabody.colorado.edu>
Message-ID: <fMtP8.57199$uX3.27041@nwrddc01.gnilink.net>

> 
> Sorry to display my ignorance here, but what on earth is a "Jamfile"?

Looks like jam is a make -like utility from Perforce
( http://public.perforce.com/jam/index.html )


From BPettersen at NAREX.com  Wed Jun 19 15:19:02 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Wed, 19 Jun 2002 13:19:02 -0600
Subject: Sending email from a Python script
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158EE4@admin56.narex.com>

> From: Jeffrey Keil [mailto:keilj_33 at yahoo.com] 
> 
> Is it possible to send an email from a Python script?

Read The Fine Manual. The first example in the email module should show
you the basics, but you probably also need to read the section on
smtplib.

-- bjorn




From peter at engcorp.com  Sat Jun 22 09:55:42 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 22 Jun 2002 09:55:42 -0400
Subject: The best GUI library in python
References: <8d3f4438.0206220457.b820418@posting.google.com>
Message-ID: <3D1481DE.C0331204@engcorp.com>

d2002xx wrote:
> 
> I want to learn GUI in python, who would tell me which GUI library is
> the best? (don't tell me Tkinter, it's too slow......)

How fast do you need your GUI to be?  Are you talking about 
an unacceptably slow loading time, or response time within 
the application?

-Peter


From hobiewon at hotmail.com  Tue Jun  4 00:14:02 2002
From: hobiewon at hotmail.com (Topy)
Date: Tue, 4 Jun 2002 00:14:02 -0400
Subject: Please help with script...
Message-ID: <adheqc$nhn$1@netnews.upenn.edu>

Hey all,
I'm having trouble with a problem I have with this python program.  It's
kind of long, so if anyone has the patience, I would gladly appreciate your
time(the programming is fairly simple I think.)  Anyway, the problem I need
to do is this :

I am given a floppy disk containing CSV (comma separated values).  Each
record (row) of the data contains 5 data-elements:
o OrderID (a five digit number) indicating the order number,
o OrderDate - a date-field indicating the date the order was placed
o RegionID - a single digit indicating the region where the order was
manufactured
o ProductID - a two-digit code indicating the product number
o OrderQty - an integer indicating the order size.

    I am interested in analyzing this sales data based on the product type
and the order quantity and would like to have this information extracted
from the CSV file into a spreadsheet. For example, only orders for ProductID
9 having an OrderQty greater than 1600 are to be included in the worksheet.

Write a Python program that will first ask the user:
1) Which ProductID is of interest
2) What is the minimum OrderQty of interest.

The program will then read the file (it will always be "a:\sales.txt") and
create a new workbook in excel named "ProductSales.xls" (in the directory
where the CSV file is) with a sheet named after the ProductID of interest,
e.g., "9".  The first row of the sheet should list the minimum amount of the
sales, e.g., "Minimum OrderQty: 1600" in Cell(A1). Each row thereafter
should contain the OrderID (in Column1), OrderDate (in Column 2) and the
OrderQty (in Column 3) for each record in the sales file that meets the
requirements.

The program should work on a file based on any number of sales transactions;
test it on the file included.  Below, are the first seven lines of the CSV
file:

19034,4/1/02,1,7,1732
19035,4/1/02,1,9,1888
19036,4/1/02,1,4,1048
19037,4/1/02,1,5,1708
19038,4/1/02,1,6,876
19048,4/2/02,1,10,472
19049,4/2/02,1,9,1596
..

If the Sales Manager is interested in orders for ProductID 9 that are
greater than 1600, then based on the data in the first seven lines, your
Python program should select line 2 and enter the required elements from
this line in the spreadsheet.

Here is an approximate outline of what my program is supposed to look like:
1. Use the function raw_input() to get the ProductID and the cut-off order
quantity from the user.
2. Note that the raw_input() function will return the order quantity as a
string while you want to use order quantity as a float when you make
comparisons. So you will need an additional line of code that uses the
float() function to convert the "string" OrderQty into a "float" OrderQty
3. Open the CSV file using the open() function
4. Import the Excel client, open an instance of the Excel application and
add a workbook to it. (Refer to class notes or IDT book for this).
5. You could use the readlines() function along with a for loop OR the
readline() function with a while loop to read through the lines in the CSV
file that you opened in Step4.
6. For each line read from the CSV file:
a. Use the split() function (with ',' comma-delimiter) to break the line
into its components (OrderID, OrderDate, etc).
b. Note that the last item returned by the split() function, which is
OrderQty, will have a newline character appended to it. You will have to
strip off the newline character and convert the OrderQty into a float (like
we did in step 2), before you will be able to use it.
c. If the ProductID and OrderQty for the current record satisfy the criteria
specified by the user, then insert this information into the Excel
spreadsheet


Anyway, so far this is what I have :

ProductID = float(raw_input("Please enter the ProductID: "))
OrderQty = float(raw_input("Please enter the cutoff order quantity: "))
CSV = open('a:/sales.txt', 'r')
import win32com.client
x1 = win32com.client.Dispatch('Excel.Application')
x1.Visible=1
x1.Workbooks.Add()


It's not much, but I'm not sure where to go from here...can anyone help?

Thanks so much!
Topy






From skip at pobox.com  Mon Jun 10 13:00:30 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 10 Jun 2002 12:00:30 -0500
Subject: module name differences depending on how generated
In-Reply-To: <5.1.0.14.2.20020610171311.037b8c98@pop.ftel.co.uk>
References: <5.1.0.14.2.20020610171311.037b8c98@pop.ftel.co.uk>
Message-ID: <15620.56110.602850.26627@12-248-41-177.client.attbi.com>

    Richard> The problem I then had was that in patching nismodule.c to fix
    Richard> a documented problem in the nismdoule.c code running under
    Richard> Linux running make install to generate a revised nis support
    Richard> module produced and installed a new nismodule.so. But this had
    Richard> no effect as the originally generated nis.so was being picked
    Richard> up in preference. Of course nis.so was not being regenerated
    Richard> and installed (presumably because setup.py detected that
    Richard> nismodule.so was being produced) so that at first blush the
    Richard> patch appeared to be ineffective.

At the moment, setup.py doesn't do a good job of deducing dependencies.  You
will have to manually delete the nismodule.o and nis.so files to get them to
rebuild.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html





From grisha at ispol.com  Mon Jun  3 16:24:28 2002
From: grisha at ispol.com (Gregory (Grisha) Trubetskoy)
Date: Mon, 03 Jun 2002 20:24:28 GMT
Subject: libpython as a shared library?
Message-ID: <Pine.BSF.4.32.0206031620070.82931-100000@localhost>

Has there been any progress in making an option of building libpython as a
shared lib (.so)? I saw some feature requests, but couldn't quite figure
out the ultimate status of this.

Thanks,

Grisha



From aahz at pythoncraft.com  Tue Jun 25 12:31:08 2002
From: aahz at pythoncraft.com (Aahz)
Date: 25 Jun 2002 12:31:08 -0400
Subject: Suggestions for good programming practices?
References: <mailman.1024971992.26429.python-list@python.org> <roy-2CF124.07480625062002@reader2.panix.com> <mailman.1025012263.2719.python-list@python.org> <3D18868A.4050002@mxm.dk>
Message-ID: <afa5sc$ptl$1@panix1.panix.com>

In article <3D18868A.4050002 at mxm.dk>, Max M  <maxm at mxm.dk> wrote:
>
>You can also make it even more general:
>
>class MyData:
>
>     def __init__(self, **args):
>         self.__dict__.update(args)
>
>d = MyData(firstName='Max M', lastName='Rasmussen')

My preference is for

class MyData:
    pass

d = MyData()
d.firstName = None
d.lastName = 'Aahz'

At the very least, I think you ought to rewrite your example as

d = MyData(
    firstName = 'Max M',
    lastName = 'Rasmussen
    )
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From narnett at mccmedia.com  Fri Jun 14 14:02:48 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Fri, 14 Jun 2002 11:02:48 -0700
Subject: executemany & UPDATE (MySQL & Python)
In-Reply-To: <slrnagk7tu.114.gerhard@lilith.my-fqdn.de>
Message-ID: <NDBBLAEMOMEHKALAMKKJKEJJKPAA.narnett@mccmedia.com>


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Gerhard H?ring
> Sent: Friday, June 14, 2002 9:55 AM
> To: python-list at python.org
> Subject: Re: executemany & UPDATE (MySQL & Python)
>
>
> In article <mailman.1024071438.5095.python-list at python.org>, Nick
> Arnett wrote:
> > Can anyone confirm that it is *not* possible to pass a list to
> the Python
> > MySQLdb module in UPDATE operations?
>
> executemany accepts a sequence of sequences, like:
>
> self.dbh.execute("UPDATE Foo SET bla=%s, blorg=%s WHERE snork=%s",
>     [(1,2,3), (4,5,6)])

I know that (obviously, since I'm doing it with INSERTs).  It doesn't seem
to work with UPDATEs, however.  That was my question.  Have you made
something like your example work?

> > Doing the latter with a loop in Python is much slower than I'd
> expect that
> > doing it with a list would be.
>
> For database modules that don't support prepared statements, the db module
> implements executemany with multiple calls to execute, anyway.

Well, since MySQLdb is much faster at inserting that way, I'm assuming that
if it'll do it on an UPDATE, that will also be much faster.

Anyone know if it's supposed to?

Nick





From martin at came.sbg.ac.at  Tue Jun 11 11:59:37 2002
From: martin at came.sbg.ac.at (Martin Krallinger)
Date: 11 Jun 2002 08:59:37 -0700
Subject: python-gui-tk
Message-ID: <e0b1d6f6.0206110759.157c37b8@posting.google.com>

Dear all,

I am interested in developing a gui to a programm written in c and
python.
i would like to construct a gui for it using tk.

what i need is basically a widget which has an option to load a file
or to past a certain text (protein sequence) in a proper field and
take it as an input run on it the c programm ( which needs just the
name of the input file and the name of the outputfile). there should
be an option to get a gnuplot out of the results,..(but this is at
this stage resudial).

if anyone has a  sample code similar to my problem (python-tkgui)it
would be very nice to have a look at it to get some hints, as i am not
very familiar to tk under python.

thanks a lot,

Martin


From prema at prema.co.nz  Sat Jun  8 09:09:16 2002
From: prema at prema.co.nz (Prema)
Date: 8 Jun 2002 06:09:16 -0700
Subject: RE Module How to escape $
Message-ID: <ae58a20e.0206080509.44da4bcb@posting.google.com>

Hi People !
I'm trying to use the re.sub function to replace several instances of a
PHP Variable eg "$Amount". Could someone please help with an example
Thanks very much
Mike


From rjones at ekit-inc.com  Sun Jun 23 22:58:04 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Mon, 24 Jun 2002 12:58:04 +1000
Subject: PIL, Python and CGI
In-Reply-To: <Pine.GSO.4.33.0206232047450.27090-100000@infinity.infineon.net>
References: <Pine.GSO.4.33.0206232047450.27090-100000@infinity.infineon.net>
Message-ID: <200206241258.04096.rjones@ekit-inc.com>

On Mon, 24 Jun 2002 12:51, Sascha Ferley wrote:
> I actually got some nice little traceback that might be of use ..
>
> Traceback (most recent call last):
>   File "/export/server/web/CPSC461/cgi-bin/test4.py", line 10, in ?
>     import _imaging
> ImportError: ld.so.1: /usr/local/bin/python: fatal: libjpeg.so.6: open
> failed: No such file or directory
>
> though:
>
> infinity:/usr/local/lib->ls -ld libjpeg*
> lrwxrwxrwx   1 root     other         12 Jun 16 18:58 libjpeg.so ->
> libjpeg.so.6*
> lrwxrwxrwx   1 root     other         12 Jun 16 18:58 libjpeg.so.1 ->
> libjpeg.so.6*
> -rwxr-xr-x   1 root     bin       137924 May 21  2001 libjpeg.so.6
> infinity:/usr/local/lib->
>
> Which is interesting ... :)

Your web server environment most likely doesn't have /usr/local/lib in its 
LD_LIBRARY_PATH.


   Richard





From gerhard.haering at gmx.de  Sat Jun 29 01:10:09 2002
From: gerhard.haering at gmx.de (=?ISO-8859-1?Q?Gerhard=20H=E4ring?=)
Date: Sat, 29 Jun 2002 07:10:09 +0200 (CEST)
Subject: Is Python growing?
In-Reply-To: <3D1D3D34.6080001@hotmail.com>
References: <mailman.1025267736.21565.python-list@python.org> <slrnahoucn.140.gerhard@lilith.my-fqdn.de>
 <3D1D3D34.6080001@hotmail.com>
Message-ID: <Mahogany-0.64-3668-20020629-071009.00@lilith.my-fqdn.de>

Am Sat, 29 Jun 2002 04:53:23 GMT schrieben Sie:
> Gerhard H?ring wrote:
> > Mark McEahern wrote:
> > 
> >>[...] What's so
> >>beautiful about Python is that all I had to do was fire up the
> interactive
> >>interpreter and do this to remind myself of what map does:
> >>
> >> >>> print map.__doc__
> > 
> > 
> > help(map) in 2.2.
> 
> Or 2.1.1.

No. Of course, for anything more recent than 2.2.0, yes.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))





From donn at u.washington.edu  Mon Jun 10 14:59:07 2002
From: donn at u.washington.edu (Donn Cave)
Date: 10 Jun 2002 18:59:07 GMT
Subject: M2Crypto: select() behaves weird on SSL sockets
References: <3D04ED17.5070606@NOSPAMREMOVETHISxs4all.nl>
Message-ID: <ae2str$b0q$1@nntp6.u.washington.edu>

Quoth Irmen de Jong <irmen at NOSPAMREMOVETHISxs4all.nl>:

| I'm using m2crypto 0.07 snapshot 3 with Python 2.2.1
| and I'm experiencing weird behavior when using select().
|
| Basically, it seems that a SSL socket is no longer in ready
| state when entering select() again, after reading *a part*
| of the available data on the socket.
|
| Using plain sockets, the select() will keep reporting that
| the socket is ready for reading, until all data has been read.
| With SSL sockets, the second select() blocks eternally...
|
| Any clue? I really need the regular select() behavior,
| also on SSL sockets...

I'm no expert on this, but my theory is that you must call
SSL_pending(self->ssl) to see if any data remains in its buffer.
Then only when no data remains, you would call select.

I don't know if SSL guarantees it will work.  If the data in
the buffer is incomplete for SSL cryptography, then you might
not be able to exhaust it without causing a recv().  I hope
SSL_pending() returns 0 when no more decrypted data can be read,
but I don't have any documentation at all.

I don't think the socket module supports SSL_pending(), so
it would be up to you to put that in.

	Donn Cave, University Computing Services, University of Washington
	donn at u.washington.edu


From gustafl at algonet.se  Mon Jun 10 19:16:46 2002
From: gustafl at algonet.se (Gustaf Liljegren)
Date: Mon, 10 Jun 2002 23:16:46 +0000 (UTC)
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
References: <Xns9229EE722467Fgustafl@195.100.94.182> <mailman.1023746539.13373.python-list@python.org>
Message-ID: <Xns922AD3C74280gustafl@195.100.94.182>

Cliff Wells <logiplexsoftware at earthlink.net> wrote: 

Thanks both of you. I tried both solutions. Since some intervals are very 
high (1000000+), it takes too long to add them to a list. Cliff's solution 
was really fast, but it took a while to understand. :)

Gustaf


From radix at twistedmatrix.com  Tue Jun 11 10:32:46 2002
From: radix at twistedmatrix.com (Christopher Armstrong)
Date: 11 Jun 2002 10:32:46 -0400
Subject: Source formatting for long expressions.
In-Reply-To: <6bd9f01b.0206110543.168b0f83@posting.google.com>
References: <mailman.1023570972.27430.python-list@python.org>
	<m3it4srhq5.fsf@mira.informatik.hu-berlin.de>
	<874rgcxy99.fsf@twistedmatrix.com>
	<mailman.1023668414.25769.python-list@python.org>
	<6bd9f01b.0206110543.168b0f83@posting.google.com>
Message-ID: <878z5lx4vl.fsf@twistedmatrix.com>


>>>>> "sameer" == sameer  <sameer_ at email.com> writes:

    sameer> why not just do it in xml?  

Twisted already has an XML persistence mechanism, and that is exactly what
inspired me (with disgust) to write AOT (this source-persistence mechanism). 
AOT is better in every possible respect than the XML persistence IMNSHO, save 
for the fact that it requires (minimal) understanding of Python syntax,
and the format for References is _slightly_ uglier (Ref(num, obj) -- since not
every object is represented with a "tag" (or class call), I can't insert
reference numbers "into" arbitrary objects).

XML is a terrible format for object persistence, due to the lack of robust
mappings. Twisted's XML uses attributes in some cases, but that is only
possible when you have string:string mappings. In *any* other case, you
have to resort to children nodes, with such ugly syntax as::

<instance class="twisted.some.Class">
<dictionary>
  <string role="key" value="someAttr" />
  <list>
    <string value="foo" />
  </list>
</dictionary>

Compare that with the equivalent AOT format::

Instance("twisted.some.Class", 
  someAttr=["foo"],
)

*That* is why I don't just use XML.

-- 
                                Chris Armstrong
                         << radix at twistedmatrix.com >>
                http://twistedmatrix.com/users/carmstro.twistd/





From peter at engcorp.com  Sat Jun  1 17:55:19 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 01 Jun 2002 17:55:19 -0400
Subject: symbolic python
References: <3CEFAE4C.38A8022B@doc.ic.ac.uk> <3CEFD881.D7992095@doc.ic.ac.uk> <3cf8ce86$1@giga.realtime.net> <83sn46vkgi.fsf@panacea.canonical.org>
Message-ID: <3CF942C7.2E4AA4CF@engcorp.com>

Kragen Sitaker wrote:
> 
> "Ira Baxter" <idbaxter at semdesigns.com> writes:
> > If you want program transformation capability,
> > you can get it in the form the of DMS Software Reengineering
> > Toolkit.
> > See http://www.semdesigns.com/Products/DMS/DMSToolkit.html.
> 
> It sounds very interesting.  How can I get a copy?

No doubt by sending them a pile of money.  :-)  (Given the customer
list mentioned, it would probably be a large pile.)

Also, no support for Python is (yet) mentioned in the list of 
supported languages.

-Peter


From r_proietti at yahoo.com  Thu Jun 20 07:38:42 2002
From: r_proietti at yahoo.com (NOPollution)
Date: Thu, 20 Jun 2002 11:38:42 GMT
Subject: module string.py
Message-ID: <3d11bd6a.11177520@news.polito.it>

Hi, I'm a Python newbye so I'm sorry if my questions will be a little
stupid...

I have two trouble:

1)  looking at the module "string.py" I have seen that are described
some functions like "lower", "rjust", etc. Ther problem is that it
seems not to describe exactly how this functions are made, but there
is only a comment about their properties. So in which file are these
functions described?

2) I tried to cancel the file "string.py". After that, the Python
Interpreter doesn't work anymore (it can't even start working). Why?

Thank you very much, Nop.


From jwbaxter at spamcop.net  Sat Jun 29 21:36:26 2002
From: jwbaxter at spamcop.net (John Baxter)
Date: Sat, 29 Jun 2002 18:36:26 -0700
Subject: Python needs better error reporting
References: <mailman.1025350057.14486.python-list@python.org>
Message-ID: <jwbaxter-F0DA69.18362629062002@corp.supernews.com>

In article <mailman.1025350057.14486.python-list at python.org>,
 Tim Peters <tim.one at comcast.net> wrote:

> I once used a
> Pascal compiler whose favorite trick was to point a caret at a trailing
> semicolon, and then complain "semicolon expected"; the message was always
> correct when it appeared, while appearing dead wrong, and was useless in
> determining the *real* problem regardless.  It did appear friendlier than
> "syntax error" the first 719 times I saw it, though <wink>.

So did I, which makes me think it was probably USCD Pascal or a relative.

I used a C compiler which said things like
"A typedef was a complete surprise to me at this point in your program"

Sometimes, it even meant it.

  --John


From * at spam.ruud.org  Fri Jun 14 13:30:58 2002
From: * at spam.ruud.org (ruud de rooij)
Date: 14 Jun 2002 13:30:58 -0400
Subject: I feel stoopid ... this code is to verbose
References: <3D0A1791.3060307@mxm.dk>
Message-ID: <871yb922el.fsf@bunty.ruud.org>

Max M <maxm at mxm.dk> writes:

> 42.000.000,00 which is 42 millions
> 
> So I need to insert dot's at every three character from end of the
> string for the integer value of the number.

import locale
locale.setlocale(locale.LC_NUMERIC, 'da_DK')
locale.format('%.2f', 42000000.00, 1)

batteries-included-ly y'rs,

        - ruud
-- 
ruud de rooij | *@spam.ruud.org | http://ruud.org


From peter at engcorp.com  Wed Jun 26 08:51:17 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 26 Jun 2002 08:51:17 -0400
Subject: system-output to wxpython-window
References: <mailman.1025068954.14462.python-list@python.org>
Message-ID: <3D19B8C5.1544B995@engcorp.com>

Klaus Reinhardt wrote:
> 
> I have some batch-files, which I want to
> invoke from a nice wxpython-frame. So
> I did   system("0m.bat");   but the output
> is coming to the DOS-shell-window. Is
> there a simple way to do the job. Best case
> would be one menue-window and another
> for output.

Look into os.popen or the popen2 module.

-Peter


From peter at engcorp.com  Wed Jun 19 19:35:38 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 19 Jun 2002 19:35:38 -0400
Subject: python __main__ is similar to java main()
References: <3015ca5e.0206190901.3cb0b726@posting.google.com>
Message-ID: <3D11154A.40F49CB4@engcorp.com>

David Nunes wrote:
> 
> class Lancamento:
>     ReqID = 0
> 
>     #PDNConsulta
> 
>     def ___init__(self):
>         #inicializacao do PDNConsulta
>         pass
> 

Not answering the immediate question, but pointing out a bug:

  def __init__(self):

is the correct spelling.  If it was just a typo, you're okay,
but if you actually cut-and-pasted the text you've got a latent
problem in your code with the triple leading underscore.

-Peter


From ngiven at lanl.gov  Tue Jun 25 12:53:36 2002
From: ngiven at lanl.gov (Nathan Given)
Date: Tue, 25 Jun 2002 10:53:36 -0600
Subject: Integrals with Python
References: <3D188551.6BA00DED@lanl.gov> <3D189307.1417ECD5@info.unicaen.fr>
Message-ID: <3D18A010.CAF3FA9E@lanl.gov>

Matlab has quad, quadl, and quad8...


does Numeric have anything like those?



Here is my current (lame) project...

I am comparing the Monte Carlo method for single integrals against the
answers given by quad, quadl, and quad8 in Matlab.  I coded everything
in matlab, and was able to do it just fine... the problem is that it was
SLOW... so, someone suggested that I run python, so I converted all my
code to python (except the quad, quadl, and quad8 parts), and was able
to run my code and get the monte carlo answer much faster....

now, i am going to be doing some double integrals, and I was just
wondering if python had some built in functions that will calculate some
double integrals (and single integrals) for me... if there are different
methods, great... I will be able to compare the monte carlo answer to
different methods of calculating the integral...

that's what I'm looking for...

Thanks!
--
Nathan


From whisper at oz.net  Fri Jun  7 15:47:49 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 7 Jun 2002 12:47:49 -0700
Subject: Working with different word processing formats?
In-Reply-To: <3D01065B.7070505@verio.net>
Message-ID: <GCEDKONBLEFPPADDJCOEMEGFDHAA.whisper@oz.net>

http://www.wvware.com/

http://word2x.alcom.co.uk/

There are a variety of format converters (pdf, latex, xml, rft etc.) on
Sourceforge.

AbiWord (also on sourceforge) might also offer some pointers.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of VanL
> Sent: Friday, June 07, 2002 12:16
> To: python-list at python.org
> Subject: Working with different word processing formats?
>
>
> Hello,
>
> Does anyone know of a solution for exporting editing (and
> ideally, importing) common word processing formats?
>
> I am trying to find any existing work in this area, pure
> python or a wrappable library, that I could use or extend.
> I don't even need to go straight to the target format -- I
> could output XML or LaTeX that could then be transformed.
>
> I would like this to be as cross-platform as possible (which
> is why I'm just not using Pythonwin + com).  I'd like to hit
> Linux, Windows, and MacOS X.
>
> If anyone knows of anything, please speak up!  I just cannot
> find *anything*...  The closest thing that google turned up
> was an apparently defunct Python + LaTeX scientific text
> editor, and that was only a mention, not any actual code.
>
> Thanks,
>
>
> VanL
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From printers at sendme.cz  Thu Jun 27 13:03:34 2002
From: printers at sendme.cz (A)
Date: Thu, 27 Jun 2002 19:03:34 +0200
Subject: How to find out DNS ?
Message-ID: <3D1B6186.3142.295959@localhost>

Hi,
Is there a way how to find out, from Python , what primary or 
secondary DNS I use when connecting to internet?
Thanks for help
Ladislav





From sholden at holdenweb.com  Tue Jun  4 10:49:19 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 4 Jun 2002 10:49:19 -0400
Subject: Thought on PEP 204 and 276
References: <ckp4fuc4fskjcpilnn2mrt3665fqvld92g@4ax.com> <mailman.1022528744.19859.python-list@python.org> <gm17fu40rh6r675g3a6d1ovl49tb9pd822@4ax.com> <mailman.1022594566.27265.python-list@python.org> <slrnaf7mo4.dfe.philh@comuno.freeserve.co.uk>
Message-ID: <6w4L8.229234$%u2.120502@atlpnn01.usenetserver.com>

"phil hunt" <philh at comuno.freeserve.co.uk> wrote in message
news:slrnaf7mo4.dfe.philh at comuno.freeserve.co.uk...
> On Tue, 28 May 2002 16:00:12 +0200, holger krekel <pyth at devel.trillke.net>
wrote:
> >
> >that was precisely my argument on the python-dev list but
> >it got rejected by GvR personally ... "Trust me. Don't go there."
> >
> >I realize that mapping types and sequence types are quite different
> >despite appearances.  It may not be a good idea to blur the lines
> >between these two too much.
>
> I think they are essentially the same. A sequence tpye is just a
> mapping type where the keys are restricted to the integers from
> 0 to however many elements there are minus 1.
>
No matter what you think, they are essentially different. Insertion into a
mapping doesn't alter the key to which other elements are mapped, but
insertion into a sequence (potentially) does that.

[ ... ]
regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From mrspnk at yahoo.com  Wed Jun 19 10:38:27 2002
From: mrspnk at yahoo.com (Mr Spannik)
Date: Wed, 19 Jun 2002 14:38:27 GMT
Subject: web based e-mail
References: <83e21f2.0206190635.14617b0a@posting.google.com>
Message-ID: <DH0Q8.4080$yj7.65598@news4.ulv.nextra.no>

http://bobomail.sourceforge.net/

Thomas

"Rod G" <reason5000 at yahoo.com> wrote in message
news:83e21f2.0206190635.14617b0a at posting.google.com...
> Hello. Does anyone know of a tutorial or how-to on creating
> a web based e-mail system using Python? I'm wanting to
> develop my own system as a learning project and need
> a starting point.
>
> Thanks,
> Rod




From achim.domma at syynx.de  Thu Jun 13 07:22:14 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Thu, 13 Jun 2002 13:22:14 +0200
Subject: How to call a function using apply with keyword args?
References: <ae9fud$u23$00$1@news.t-online.com> <mailman.1023954797.13466.python-list@python.org>
Message-ID: <ae9uoe$obj$01$1@news.t-online.com>

"Cliff Wells" <logiplexsoftware at earthlink.net> wrote in message
news:mailman.1023954797.13466.python-list at python.org...

> Python 2.1.1 (#1, Aug 13 2001, 19:37:40)
> [GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-96)] on linux2
> Type "copyright", "credits" or "license" for more information.
> >>> def MyFkt(A,B,C):
> ...     print A, B, C
> ...
> >>> params = {'A':5,'B':8,'C':9}
> >>> apply(MyFkt,(),params)
> 5 8 9
> >>>

that's exactly what I wanted to do, but I missed one parameter. It works
now, thanks !




From peter at engcorp.com  Thu Jun 27 10:02:21 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 27 Jun 2002 10:02:21 -0400
Subject: Strings across lines
References: <3D1AF1CD.8010503@thomas-guettler.de> <3D1AF36B.7000202@wedgetail.com> <3D1AFF78.8060704@thomas-guettler.de>
Message-ID: <3D1B1AED.DE04C612@engcorp.com>

Thomas Guettler wrote:
> 
> Derek Thomson wrote:
> 
> > Thomas Guettler wrote:
> 
> [cut]
> 
> > print ( 'foo ................. %s ' +
> >         'bar ................. %s '   ) % (foo, bar)
> 
> That's better than "foo" + \ "bar". Thank you!

And better yet, now that you have two adjacent strings you don't even
need to use the + ...  (but it only works because of the brackets, 
again).

-Peter


From jjl at pobox.com  Tue Jun  4 16:51:09 2002
From: jjl at pobox.com (John J. Lee)
Date: Tue, 4 Jun 2002 21:51:09 +0100
Subject: urllib, urllib2, httplib -- Begging for consolidation?
In-Reply-To: <477762c2.0206041204.6993e7e2@posting.google.com>
References: <mailman.1020962597.19539.python-list@python.org> <b0f083db.0205091346.397e35db@posting.google.com> <477762c2.0206041204.6993e7e2@posting.google.com>
Message-ID: <Pine.SOL.4.30.0206042145300.10073-100000@mimosa.csv.warwick.ac.uk>

On 4 Jun 2002, Dave Brueck wrote:
[...]
> together. For example, what is the correct way to do a HTTP HEAD
> request that follows redirects? It's not hard, but it's silly to have
> to code it yourself if somebody already did the work for you. Well,
> httplib doesn't know how to follow an HTTP redirect, but lo and
> behold, urllib does. Unfortunately, there's no way to use it because
> in this case because urllib is for opening URLs (the GET is hardcoded
> and not easily overridable).

I suppose the standard answer applies here: it isn't there because
nobody has written it yet.  I don't think there is any disagreement about
what would be nice to have -- just about who is going to do it.

> I don't disagree that a layering or division of functionality can be
> helpful, I just disagree that the current layering/division makes
> sense.

There is no layering issue AFAICS: httplib does the basic HTTP stuff.
urllib adds some higher-level functions.  urllib2 is a essentially a
rewrite of urllib (though some of the urllib2 is implemented using
urllib), with some missing bits -- like HEAD, as you point out.

> I think it's great that opening a URL via urllib will automatically
> follow redirects, it's just a shame that you can't take advantage of
> that functionality if what you're doing falls outside the narrow
> design domain of urllib transactions. :(

We're agreed, I think.


John



From boud at valdyas.org  Sat Jun 15 06:43:52 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Sat, 15 Jun 2002 12:43:52 +0200
Subject: IDE recommendations??
References: <b9veea.l85.ln@majestix.gallien.de> <3d0b06bd$0$52043$e4fe514c@dreader3.news.xs4all.nl> <aef2lb$6h5bg$1@ID-71985.news.dfncis.de>
Message-ID: <3d0b1e54$0$3883$e4fe514c@dreader4.news.xs4all.nl>

John R. Marshall wrote:

> On Saturday 15 June 2002 04:18 am, Boudewijn Rempt wrote:
> 
>> At one
>> point there was a version of KDevelop that supported Python development,
>> but that has disappeared, it seems.
> 
> It's (Gideon) still there :-). Instructions for getting it from CVS can be
> found here
> http://www.kdevelop.org/index.html?filename=branches_compiling.html
> 
> 

Right, I'm compiling it just now. Took me an upgrade of automake (and then a 
downgrade, because I hadn't read the instructions well enough), but it's 
working now. Last time I looked at Gideon (hasn't there been some alpha 
release at one point?), it crashed my X server when I tried to create a 
Java project. Hopefully it's better now.

-- 
Boudewijn Rempt | http://www.valdyas.org


From cliechti at gmx.net  Wed Jun 26 15:39:06 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 26 Jun 2002 21:39:06 +0200
Subject: geting python files for use on web
References: <k6oS8.191997$6m5.159601@rwcrnsc51.ops.asp.att.net>
Message-ID: <Xns9239DCB2C57A0cliechtigmxnet@62.2.16.82>

"Ali K" <akakakhel at attbi.com> wrote in 
news:k6oS8.191997$6m5.159601 at rwcrnsc51.ops.asp.att.net:
> I would like to know where I can download the nessecary files for usin
> python in ASP and/or CGI.

the standard python distro from python.org should do for CGI.

> Please reply.

what do you mean? by email?

chris

-- 
Chris <cliechti at gmx.net>



From Chris.Barker at noaa.gov  Thu Jun 20 15:30:47 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Thu, 20 Jun 2002 12:30:47 -0700
Subject: do...while
References: <slrnah40l1.12v.msoulier@pmerd071.ca.nortel.com>
Message-ID: <3D122D35.F91D242A@noaa.gov>


"Michael P. Soulier" wrote:

>     How do you people handle the lack of a do...while in Python?

>     line = filehandle.readline()
>     while len(line) > 5:
>         line = filehandle.readline()

The most common idiom is:

while 1:
    line = filehandle.readline()
	if not len(line > 5): 
		break
	do stuff....

I try to avoid "break" but is it is at the top of the loop, the meaning
is pretty clear.

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From johnroth at ameritech.net  Fri Jun 14 11:33:05 2002
From: johnroth at ameritech.net (John Roth)
Date: Fri, 14 Jun 2002 11:33:05 -0400
Subject: [development doc updates]
References: <mailman.1024057266.26182.python-list@python.org>
Message-ID: <ugk341aivqpl27@news.supernews.com>

"Fred L. Drake" <fdrake at acm.org> wrote in message
news:mailman.1024057266.26182.python-list at python.org...
> The development version of the documentation has been updated:
>
>     http://www.python.org/dev/doc/devel/
>
> Updated to reflect recent changes.

I'm not certain I understand item 5. Extended Slices.
Does this mean that the standard sequence types now support
the stride operand, or that the stride operand is being
depreciated and will eventually vanish?

John Roth
>
>




From bdesth at nospam.free.fr  Fri Jun 21 16:36:01 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Fri, 21 Jun 2002 16:36:01 -0400
Subject: [newbie] import fails first time, then works :-[
References: <3D1379D1.5020705@nospam.free.fr> <3D132C65.34B38F8C@engcorp.com>
Message-ID: <3D138E31.4070409@nospam.free.fr>

Peter Hansen wrote:
> laotseu wrote:
> 
>>Hi Everybody
>>
>>I've got this strange problem :
>>when importing modules that wrap external modules (like socket and
>>_socket etc), the first time, I get an ImportError
>>(no module named _[module]).
>>If i try a second time, it works ok ?-o
>>
>>Exemple :
>> >>> import socket
>>Traceback (most recent call last):
>>   File "<stdin>", line 1, in ?
>>   File "/usr/local/lib/python2.2/socket.py", line 41, in ?
>>     from _socket import *
>>ImportError: No module named _socket
>> >>> import socket
>> >>> socket
>><module 'socket' from '/usr/local/lib/python2.2/socket.pyc'>
> 

Hi Peter
Wow ! faster than lightnin !

> See recent postings in the archives for explanations about why this
> is actually _not_ working the second time, but only appears to.

Testing a little bit further, I discovered that fact...


> In this case, despite that you say the external modules are in the
> right place, there is probably something wrong with the _socket
> external module 

I guess not, I had the same problem with all external modules.

>or it cannot be found and that's the root cause
> of your trouble.

Well... I added '/usr/local/lib/lib-dynload' to my PYTHONPATH and 
'import' does not complain anymore. Everything seems just fine now. But 
that is puzzling me ! [being a python newbie,] I thought the 
'/lib-dynload/' directorie, that is in the '/lib/python<version>/' 
directories, was automagically searched for external modules. Was I wrong ?

Thanks anyway for your fast answser, and may the force be with you (and 
your favorite snake)

laotseu

> -Peter




From thomas at cbs.dtu.dk  Fri Jun 14 04:59:21 2002
From: thomas at cbs.dtu.dk (Thomas Sicheritz-Ponten)
Date: 14 Jun 2002 10:59:21 +0200
Subject: Translating Tcl/Tk code into Python/Tkinter help please...
References: <B92E4241.6F6B%sarmstrong13@mac.com>
Message-ID: <m3sn3qs0ba.fsf@fkogtsp1.bmc.uu.se>

SA <sarmstrong13 at mac.com> writes:

> Hi Everyone-
> 
>     I am trying to translate a tcl program that use tk into python. I seem
> to be going along fine until I get to the process that wants to "catch" the
> results onto one text screen. Here is what I am looking at:
> 
> proc python {} {
>     catch {exec /sw/bin/python -c [.t1 get 0.0 end]} output
>     .t2 delete 0.0 end
>     .t2 insert end $output
> }
> 
> Basically, this process catches the output from running the code in the text
> widget .t1 through python -c and places it into the variable ouput.
> The text widget .t2 is then cleared and the contents of the variable $output
> are displayed in the text widget .t2.
> 
> I can translate all of the widgets into python/tkinter code, but I'm not
> sure about how to translate this process. Is there a function similar to
> "catch" in Python?

"try" and "exec" is used instead of "catch"
but in this case you could use commands.getoutput

Untested code:

import commands
from Tkinter import *


def execute_python_code1():
    global t1, t2

    txt = t1.get(0.0, END)
    com = "/sw/bin/python -c %s" txt
    
    try: 
        output = commands.getoutput(com)
    except:
        output = 'Error running command: %s' % com

    t2.delete(0.0, END)
    t2.insert(END,output)


    
# getoutput returns errors as catch does, so there is no need to use try/except
# IMHO, you should strip newlines from the command returned from the text widget

def execute_python_code2():
    t2.delete(0.0, END)
    output = commands.getoutput(com = "/sw/bin/python -c %s" % t1.get(0.0, END).strip())
    t2.insert(END,output)
        
        

-- 
Sicheritz-Ponten Thomas, Ph.D, thomas at biopython.org      (
Center for Biological Sequence Analysis                   \
BioCentrum-DTU, Technical University of Denmark            )
CBS: +45 45 252485      Building 208, DK-2800 Lyngby  ##----->
Fax: +45 45 931585      http://www.cbs.dtu.dk/thomas       )
                                                          /
     ... damn arrow eating trees ...                     (


From BPettersen at NAREX.com  Thu Jun  6 15:52:17 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 6 Jun 2002 13:52:17 -0600
Subject: 64-bit Python?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158EB4@admin56.narex.com>

> From: Kragen Sitaker [mailto:kragen at pobox.com] 
> 
> loewis at informatik.hu-berlin.de (Martin v. L?wis) writes:
> > I believe there is another limit: the maximum length of a string, 
> > list, or tuple is 2**31-1 on 64-bit platforms where a C int is 4 
> > bytes. That, in particular, includes mmapped objects as well.
> 
> Yes, this is correct.  Even the buffer interface, which my 
> arrayfrombuffer package[0] uses to provide access to mmapped 
> files as Numeric arrays inexplicably uses int instead of size_t.
> 
> The problem is that all of the 64-bit platforms I have access 
> to (various Linuxes on Alpha and IA-64, Tru64 on Alpha) have 
> 4-byte ints and 8-byte longs, at least by default.  This is 
> called "LP64": longs and pointers are 64, but ints are 32, so 
> all the code that assumes ints are 32 will continue to work.

Are you using Win64 on IA-64? Last time I checked (using the MS compiler) a long was still 32 bits, and you had to use __int64 to get a 64 bit integer...

-- bjorn




From sabine at textraeume.de  Mon Jun  3 14:07:29 2002
From: sabine at textraeume.de (Sabine Richter)
Date: Mon, 03 Jun 2002 20:07:29 +0200
Subject: word wrap in regex
Message-ID: <3CFBB061.66443266@textraeume.de>

Hello,

I have a regex which is a bit long (200 characters) and I want to add
word wraps and comments to make it more clear. I'm sure I read about
that in the references before but I don't find the passage with the word
wrap no more. May anybody give me a hint?

Sabine




From tew at wiencko.com  Mon Jun  3 21:54:15 2002
From: tew at wiencko.com (Tom Wiencko)
Date: Tue, 04 Jun 2002 01:54:15 +0000
Subject: win32com under Cygwin
References: <3CF9846B.579606ED@wiencko.com>
	 <Xns9221CB7B32FDcliechtigmxnet@62.2.16.82>
	 <%5JK8.180011$%u2.88943@atlpnn01.usenetserver.com> <mailman.1023130086.17065.python-list@python.org>
Message-ID: <3CFC1DC7.4F56E302@wiencko.com>

Heavy sigh.  That is what I was afraid of.

I would love to help out in such a project, but I took a solemn oath
10 years ago to never work with Windows libraries.  I had so hoped that
someone else had already ventured into that valley of death so I would
not have to.

I think I'll figure out another way.  Thanks for the response.

-Tom

Jason Tishler wrote:
> 
> On Mon, Jun 03, 2002 at 08:11:25AM -0400, Steve Holden wrote:
> > > Tom Wiencko <tew at wiencko.com> wrote:
> > > > I noticed that the Python distribution with Cygwin does not include
> > > > the win32 extensions.  The only distribution I can find for them is
> > > > a Windows executible, which does not seem to know about Cygwin.
> > > >
> > > > Can these libraries be loaded under Cygwin?
> 
> No (*).
> 
> > > > Do they even work under Cygwin?
> 
> No (*).  They were built against a completely different Python --
> Win32 Python.
> 
> > > > Can somebody point me toward how to load them if it is possible?
> 
> One would have to build them first.  See below...
> 
> > But don't expect a cygwin port to be a simple recompile...
> 
> Agreed...
> 
> Besides converting MSVC workspaces/projects into Makefiles and dealing
> with normal platform porting issues there are also ABI issues that
> will make the porting of win32com even more challenging.  Unless, one
> compiles using the gcc "-fnative-struct" option, it will require source
> level changes to get COM to work.  IIRC, *everything* (e.g., the Cygwin
> DLL, Python, win32com, etc.) must be compiled with "-fnative-struct"
> or nothing.  So, until Cygwin gcc defaults to "-fnative-struct", I think
> this effort may be doomed. :,(
> 
> Nevertheless, I would *love* to be able to use win32com under Cygwin
> Python.  It has been on my list to try since I first got shared extensions
> to work under Cygwin Python.  If anyone would like to collaborate, please
> let me know.  If successful, I would definitely include win32com in the
> standard Cygwin Python package.
> 
> Jason
> 
> (*) Or, at least I doubt it.

-- 
------------------------------------------------------------------------
Tom Wiencko                                              tew at wiencko.com
President - Wiencko & Associates, Inc.                    (404) 255-2330
Telecom Consulting & Project Development -- Wireline, Wireless, Internet


From tchur at optushome.com.au  Mon Jun 10 06:27:11 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Mon, 10 Jun 2002 20:27:11 +1000
Subject: Confidence intervals, stats modules?
References: <ug2r3aac4etnef@corp.supernews.com> <adt0hs$3bl$1@news7.svr.pol.co.uk> <mailman.1023574153.24062.python-list@python.org> <3D047944.7070305@cardiff.ac.uk>
Message-ID: <3D047EFF.D71700A8@optushome.com.au>

Alan James Salmoni wrote:
> 
> Problem is that I don't think stats.py has confidence intervals (it does
> have mean, stddev etc). I am working on it myself for my own stats
> package, but it might be a few weeks as I am concentrating upon non
> parametric stuff.

Assuming your are talking about a normal variate, the stats.py module
has a function 
for the sample std deviation but not, as far as I recall, for the
Student's t distribution,
which is needed to calculate confieence intervals where the popuation
std deviation is
unknown (which is most of the time).

Thus, I think that R, via RPy, is a better bet. R also has all the
machinery you need to 
calculate confidence limits for non-normal variates too. Even simple
statistics only remain
"elementary" if you don't think too hard about them.

Tim C

> 
> Alan.
> 
> Tim Churches wrote:
> > Duncan Smith wrote:
> >
> >>"Stephen Boulet" <stephen at theboulets.net> wrote in message
> >>news:ug2r3aac4etnef at corp.supernews.com...
> >>
> >>>Are there any stats modules to calculate things like confidence intervals,
> >>>mean, standard deviation? Thanks.
> >>>
> >>>-- Stephen
> >>
> >>http://www.nmr.mgh.harvard.edu/Neural_Systems_Group/gary/python.html
> >>
> >
> >
> > Gary's stats module is excellent and is probably sufficient for the uses
> > you
> > describe. There are also some elementary statistical functions in Konrad
> > Hinsen's
> > equally excellent Scientific Python package - see
> > http://starship.python.net/crew/hinsen/scientific.html
> >
> > In fact, you could just use NumPy (Numerical Python) - have a look at
> > the MLab
> > module in NumPy (which is at http://www.pfdubois.com/numpy/ ).
> >
> > However, if you are interested in more advanced statistical analysis
> > and/or in
> > drawing statistical graphs, then have a look at RPy, by Walter Moreira,
> > at http://rpy.sourceforge.net
> >
> > RPy embeds the R stats environment within Python. R is good for most
> > things
> > statistical, from the elementary to the advanced and experimental, and
> > it does
> > truly beautiful graphics. It is a mature project backed by a number of
> > eminent
> > statisticians in five continents.
> >
> > Note that R is multi-platform (see http://www.r-project.org ) but RPy
> > only
> > works under Linux and Unix (and Mac Os X, perhaps) at this stage. I
> > believe
> > a Windows version is in the works.
> >
> > Tim C
> >
> >
> 
> --
> http://mail.python.org/mailman/listinfo/python-list




From sholden at holdenweb.com  Mon Jun 17 02:08:16 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 17 Jun 2002 02:08:16 -0400
Subject: SyntaxWarning: name 'mailcounter' is used prior to global declaration
References: <ae74k3$49l33$1@ID-79019.news.dfncis.de> <ae74tc$49l33$2@ID-79019.news.dfncis.de>
Message-ID: <2srP8.127335$Fp1.66963@atlpnn01.usenetserver.com>

"Felix Seeger" <seeger at sitewaerts.de> wrote in message
news:ae74tc$49l33$2 at ID-79019.news.dfncis.de...
> Ahh, I have it.
> global was used after a mailcounter was used.
>
> Felix Seeger wrote:
>
> > Hi
> >
> > I get this warning.
[ ... ]

See how effective comp.lang.python is, Felix? You just have to post to the
group and you instantly understand what's wring with your program!

support-just-doesn't-get-more-effective-that-that-ly y'rs  - steve
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From nico at tekNico.net  Fri Jun  7 11:52:31 2002
From: nico at tekNico.net (Nicola Larosa)
Date: Fri, 07 Jun 2002 17:52:31 +0200
Subject: Creating Dynamic Web Pages (ZOPE)
References: <46ca81a0.0206062016.789d3af6@posting.google.com> <mailman.1023437832.11588.python-list@python.org> <46ca81a0.0206070654.787ac0f7@posting.google.com>
Message-ID: <3D00D6BF.5070308@tekNico.net>

> Thanks for the reply and all the good information and links.

You're welcome, nothing like a good jumpstart to save time.


> You have not only helped in making my decision easier (not that it
> seems I need to decide in the end)

Now, I would slow down a bit. Surely Zope is a good Web tool, and the most 
known in the Python world, but it's not the only one.

Based on your requirements it is likely the best choice for you, but it's 
*my* opinion, so you could do worse than to give a look to a few alternative 
options:

Webware (the most J2EE-like, and quite featureful)
http://webware.sourceforge.net/

SkunkWeb (smaller, but well endowed)
http://skunkweb.sourceforge.net/

Quixote (the smallest, a little jewel, PTLs are a charm)
http://www.mems-exchange.org/software/quixote/

All of them within the reach of the Snake, of course. :^) And there's also 
Paul Boddie's description of some of them, to save another little bit of 
your time:

http://www.boddie.org.uk/python/web_frameworks.html


> Management here is not exactly that computer literate and don't really
> understand Open Source software (like a lot of people) and as we run
> primarily a Microsoft shop it's good to know in advance that ZOPE will
> work nicely with what's already in place (I know that the Tech guys
> will raise concerns for sure).

Tell them to sleep soundly, they are accustomed to weaker stuff, after all. 
[No smile here]


> P.S. Thanks also for the weekend reading -- I will grab the ZOPE book
> and look through it this weekend!!

Have fun with all the examples, you'll need a working install to make the 
most of it.


-- 
"Too much cleverness in the parser can turn against you."
   Guido Van Rossum

Nicola Larosa - nico at tekNico.net





From mwh at python.net  Thu Jun 20 04:28:40 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 20 Jun 2002 08:28:40 GMT
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com>
Message-ID: <lkznxqjqx2.fsf@pc150.maths.bris.ac.uk>

"John Roth" <johnroth at ameritech.net> writes:

> There's no equivalent of a goto command in Python, for good reason,
> which I won't repeat here. Prof. Djikstra did it very well in 1974,
> in his letter "Goto Considered Harmful."

There are also technical difficulties implementing it (the
blockstack), or I'd have probably done goto functionality in
bytecodehacks :)

Cheers,
M.

-- 
  The ability to quote is a serviceable substitute for wit.
                                                -- W. Somerset Maugham


From kragen at pobox.com  Thu Jun  6 13:37:16 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 06 Jun 2002 13:37:16 -0400
Subject: Medium to Large Scale Python Deployments
References: <KHkL8.21383$eD2.2399923271@newssvr10.news.prodigy.com>
Message-ID: <833cw0gvg3.fsf@panacea.canonical.org>

"Domenic R. Merenda" <dom at edgereport.put_a_c_o_m_here> writes:
>    I am running a home-grown Enterprise Resource Planning system, written
> entirely in Python, to manage the operations and manufacturing aspects of a
> medium ($100,000,000 annual sales) manufacturing organization.

That's fascinating!

>    We have found Python to be a robust solution for the problems we've
> encountered to date, and pass approximately half a gigabyte of daily data
> through our custom systems.  We have merged the AS/400 (Python 2.1, soon to
> go to 2.2.1) with midrange (PIII700) PC's to create an interesting
> heterogenous environment.

Are you running Python natively on the AS/400?  How hard was that?
Does it support dynamic loading of extensions?



From noah at noah.org  Thu Jun  6 14:25:17 2002
From: noah at noah.org (Noah Spurrier)
Date: Thu, 6 Jun 2002 11:25:17 -0700
Subject: programatically changing windows desktop background
References: <c4e6b17d.0206060716.6afae229@posting.google.com>
Message-ID: <010601c20d87$8225b310$0400a8c0@saysyou>

The Win32 API call you need is "SystemParametersInfo", but that is
not exposed in the Python win32 extensions (as far as I can tell). 
This does not mean that you can't do it, but it isn't a no-brainer.

Sorry to be unhelpful.

Yours,
Noah

----- Original Message ----- 
From: "Bob Roberts" <bobnotbob at byu.edu>
Newsgroups: comp.lang.python
To: <python-list at python.org>
Sent: Thursday, June 06, 2002 8:16 AM
Subject: programatically changing windows desktop background


> I want to write a python program that periodically takes an image
> file, writes some useful infomation on it (like the Calendar, etc.),
> saves the file, and then sets it as the current windows background
> image.
> 
> Most of it is pretty straightforward, but does anybody know how I can
> change the windows background imagage from a python program?
> 




From com-nospam at ccraig.org  Fri Jun 21 13:46:29 2002
From: com-nospam at ccraig.org (Christopher A. Craig)
Date: 21 Jun 2002 13:46:29 -0400
Subject: I'd give up Perl tomorrow if only...
In-Reply-To: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au>
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au>
Message-ID: <t1wlm98y16y.fsf@kermit.wreck.org>

While there isn't any single Python repository as nice as CPAN, you
can find a module to do just about anything.  Given your list I can
find a module for everything but Tie on the first try using
Google. (and I'd guess the fact that I have no idea what Tie is
contributed to my not finding it)

Given the hardware and bandwidth, it shouldn't be all that hard to
create a CPAN for Python.  I suspect one of the main reasons it hasn't
been done is that Google works so well.

-- 
Christopher A. Craig <com-nospam at ccraig.org>
"In the beginning the Universe was created.  This has made a lot of people
very angry and been widely regarded as a bad move." Douglas Adams




From gerson.kurz at t-online.de  Sat Jun  8 01:08:36 2002
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Sat, 08 Jun 2002 05:08:36 GMT
Subject: Newby: How do I strip HTML tags?
References: <Xns92278CC566FEAusYourStandingInIt@130.133.1.4> <3d00e1b1.47853937@news.t-online.de> <mailman.1023469751.21489.python-list@python.org>
Message-ID: <3d0190e7.5774640@news.t-online.de>

On Fri, 7 Jun 2002 10:07:19 -0700, Andy McKay wrote:

>This of course assumes that there are no "<" or ">" anywhere else in
>attributes etc...

Standalone "<" and ">" indicate invalid HTML code, one should use &lt;
and &gt; instead. You are of course right, in the end a use of
predefined classes is almost always better than reinventing the wheel
yourself.

>>>> html2text("<b>This is a <i>test</i></b>")
>'This is a test'

Works with both of my funcs, too ;)

Cheers,
Gerson 



From phd at phd.pp.ru  Wed Jun 19 15:24:48 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 19 Jun 2002 23:24:48 +0400
Subject: Sending email from a Python script
In-Reply-To: <67f6bbe5.0206191026.72e14e8e@posting.google.com>; from keilj_33@yahoo.com on Wed, Jun 19, 2002 at 11:26:46AM -0700
References: <67f6bbe5.0206191026.72e14e8e@posting.google.com>
Message-ID: <20020619232448.A16160@phd.pp.ru>

On Wed, Jun 19, 2002 at 11:26:46AM -0700, Jeffrey Keil wrote:
> Is it possible to send an email from a Python script?

   Of course. Use "email" module.

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From dreed at capital.edu  Sat Jun  8 10:48:51 2002
From: dreed at capital.edu (Dave Reed)
Date: Sat, 8 Jun 2002 10:48:51 -0400
Subject: Removing ^M
In-Reply-To: <Pine.LNX.4.21.0206091005170.7346-100000@localhost.localdomain>
	(message from Michael Hall on Sun, 9 Jun 2002 10:08:44 +0930 (CST))
References: <Pine.LNX.4.21.0206091005170.7346-100000@localhost.localdomain>
Message-ID: <200206081448.g58Empj11746@localhost.localdomain>

> X-Authentication-Warning: localhost.localdomain: olc owned process doing -bs
> From: Michael Hall <olc at ninti.com>
> X-Sender: olc at localhost.localdomain
> Date: Sun, 9 Jun 2002 10:08:44 +0930 (CST)
> 
> 
> I am trying remove ^M characters (some kind of newline character) from an
> HTML file. I've tried all sorts of string.replace and sed possibilities
> but the
> things just won't go away. Does anyone have a way of removing such
> characters?
> 
> TIA 
> 
> Mick

There's a program called dos2unix that will do this. From your mail
headers it looks to me like you have a Unix/Linux system in which case
it may already be installed. If not do a google search on dos2unix.

Dave





From mfranklin1 at gatwick.westerngeco.slb.com  Sun Jun  9 16:44:42 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Sun, 9 Jun 2002 20:44:42 +0000
Subject: Screen resolution again
In-Reply-To: <3D039794.9549C54@bellatlantic.net>
References: <3D039794.9549C54@bellatlantic.net>
Message-ID: <200206091947.g59JltB14412@helios.gatwick.geco-prakla.slb.com>

On Sunday 09 Jun 2002 5:59 pm, you wrote:
> I am still trying to figure out how to automatically create a canvas
> that is sized according to the screen resolution using tkinter on a
> win95 box and python 2.2.1.  The hack below sort of does that, but I
> would appreciate advice, suggestions or pointers to a url on how to do
> it correctly.  Note that I create an extra canvas get the root height
> and windth.
>
> david lees
> -------------------
>
> from Tkinter import *
> root = Tk()
>
> root.title('Drawing')
> canvas = Canvas(root,width=20,height=20,bg='white')
> canvas.pack()
>
> h = canvas.winfo_vrootheight()
> w = canvas.winfo_vrootwidth()
>
>
> print w,h
> canvas2 = Canvas(root,width=w-30,height=h-30,bg='white')
> canvas2.pack()
>
> root.mainloop()

Why not call winfo_vrootheight() on root, that at least gets rid of the first 
canvas.  I used this to test and it seemed to work but did not fill the 
entire screen I guess there is a bit of slack for the window decorations etc.
however i'm running RedHat linux 7.2 with KDE What happens when you run this 
exactly?

import Tkinter
root=Tkinter.Tk()
root_height=root.winfo_vrootheight()
root_width=root.winfo_vrootwidth()

print root_height, root_width # on my screen this prints 768 1024

canvas=Tkinter.Canvas(root, width=root_width, height=root_height)
canvas.pack()
root.mainloop()











From kkto at csis.hku.hk  Sun Jun  2 00:30:57 2002
From: kkto at csis.hku.hk (Isaac To)
Date: 02 Jun 2002 12:30:57 +0800
Subject: Why no open(f, "w").write()?
References: <mailman.1022702940.7544.python-list@python.org> <slrnafagek.64g.huaiyu@gauss.almadan.ibm.com> <slrnafcsor.12v.msoulier@pmerd071.ca.nortel.com> <20020531225504.248c92ce.larooy@xtar.co.nz> <slrnafg8jn.12v.msoulier@pmerd071.ca.nortel.com> <20020601192511.7ec4a585.larooy@xtar.co.nz>
Message-ID: <7iptzati7y.fsf@enark.csis.hku.hk>

>>>>> "John" == John La Rooy <larooy at xtar.co.nz> writes:

    John> I've often used the 'open(f, "w").write()' right at the end of a
    John> program where I know the file is about to get closed anyway. From
    John> what I've seen here about Jython, it seems that I can't even count
    John> on that behaviour. eeek!

    John> Perhaps the specification will some day change to make the objects
    John> returned by open() to be *smarter* so we can count on them knowing
    John> when they should flush and when they should close the toilet seat.

    John> A few people seem to be getting stuck on the explicit instead of
    John> implicit idea but I think that is mostly because we are used to be
    John> told to close files and we've all had a few late nights because we
    John> forgot to. Just as lots of us learned we should always free() our
    John> mallocs() and now we think nothing of letting python do that
    John> housekeeping for us.

That's possible if you just want to make sure that all files are closed at
the end of execution: just keep a list of opened files objects somewhere
(perhaps using weak references?), and, when the program exits, close all of
them.  This is standard OS technique (all process has a list of opened file,
which are closed on exit).  It is even easy to implement at library level.
But that won't solve all the problems.  In particular, if the output of your
files are read by somebody else, you have to know at some time that the
content is valid for reading.

Regards,
Isaac.


From sholden at holdenweb.com  Fri Jun  7 09:39:31 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 7 Jun 2002 09:39:31 -0400
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <20020606215349.GA13474@node1.opengeometry.net> <mailman.1023401650.18434.python-list@python.org> <adpktk$17j3k$1@ID-146704.news.dfncis.de>
Message-ID: <oM2M8.348515$%u2.332829@atlpnn01.usenetserver.com>

"Shagshag13" <shagshag13 at yahoo.fr> wrote in message
news:adpktk$17j3k$1 at ID-146704.news.dfncis.de...
>
> "gabor" <gabor at realtime.sk> a ?crit dans le message de news:
mailman.1023401650.18434.python-list at python.org...
>
> > but if you want to know about specific python issues, then you have to
> > learn things like "dictionary are very well optimized etc.."
>
> That's what i'm asking : learning and understanding... but where should i
go ???
>
You may not want to hear this, but experience and thought are always going
to be your best guides. I am a much better programmer now than I was thirty
years ago. Why? Because I've done things wrong, and corrected my errors.
I've repeated the same tasks many times, and found out by practical
experience what works best (for me) and what doesn't. The meat between your
ears will provide you with important knowledge if you take the time to
analyse your experience. If you don't then you should remember that "those
who cannot remember history are doomed to repeat it".

> I'm wondering if there is a place to go to find "why this solution is
better in this
> appropriate case... ". For example, i was looking for the best pythoner's
way to
> remove duplicate in a list, and after reading many posts, i find at the
end of a
> cookbook recipe that it was :
>
> set = {}
> map(set.__setitem__, alist, [])
> alist = set.keys()
>
> But that's not trivial !!! And quite difficult to find...
>
... and not always appropriate. One of the things experience will teach you
is that when your list only contains ten elements the more obvious solutions
are to be preferred, since they will execute in maybe 100 microseconds
instead of 63.

> > those things doesn't make your code obfuscated..
>
> I agree :)
>
> > [file.printsize() for file in directory]
> >
> > they're the same, and everyone would recommend you the last one..
>
> Yes but i would have asked why this is the best... And i think that
> your answer would have helped me to become a better pythoner...
>
Of course, comp.lang.python itself is a good place to ask questions like
that. The more trivial the question, the more people will have their own
opinions, of course, so you will still need to use your intelligence to make
sense of the profusion of replies.

> I mean there should be a place to go where you could read generalities
> like "as dictionary are very well optimized, you can use it without
> trouble and for example it's best in : removing duplicates, ..." or
> "if you need to do an operation on a list use map because it's
> c-optimized", there are "object overhead if you...", and so on
> that's not in the books, that's not in 'dive in python',
> it's spreaded in Guido's essays, and posts over there... But i worry,
> you could miss it if you don't spend many time to search, and say
> "there should be a place for that"...
>
Perhaps you should think about writing a book? I did, and I know there are
lots of things I would have liked to include but didn't have space and/or
time.

> > so to summarize this: what do you want to know/learn?
>
> What a beginner should learn to do it the pythoner way...
>
> s13.
>
> ps : and sorry for my poor english and typos...
>
Don't worry about that. Your concern for knowing these things suggests that
you will be a good Python programmer. But the most important part of
becoming a good Python programmer is becoming a good programmer. Knowing the
right question sto ask is as important as getting the answers, so you are
already a good way down the path. Many of the required skills aren't
specific to Python...

regards
 Steve
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From jadestar at idiom.com  Tue Jun 18 05:37:37 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 18 Jun 2002 09:37:37 GMT
Subject: "Fatal Python error: GC object already in linked list"
References: <yfsfzzleg6s.fsf@black132.ex.ac.uk>
Message-ID: <aemv11$2i38$1@news.idiom.com>

Alexander Schmolck <a.schmolck at gmx.net> wrote:

> Just got the following in an interactive session:

>  >>> reload(awmstools)
> Exception exceptions.RuntimeError: 'maximum recursion depth exceeded' in  ignored
> Fatal Python error: GC object already in linked list
     
>     Process Python aborted
> Any ideas as to what might be the cause?

> alex

 I suspect that you invoked a runaway recursive function --- it called
 itself enough times to exceed Python's max recursion depth; just like
 the message says.

 I don't recommend reload(), just exit and restart (and import).
 I think reload is a hack and I'm sure I've seen caveats regarding
 this in the docs.





From derek at wedgetail.com  Fri Jun 21 04:22:02 2002
From: derek at wedgetail.com (Derek Thomson)
Date: Fri, 21 Jun 2002 08:22:02 GMT
Subject: converting an array of chars to a string
References: <3d12dda2_6@news.newsgroups.com>
Message-ID: <3D12E228.6030501@wedgetail.com>

JB wrote:
> How can I convert an array of chars to string?
> 
> The problem is this: I should like to perform 8 bit 
> calculations in a buffer. After this has been done, the 
> buffer has to ba written to a binary file (and file.write 
> takes strings instead of lists).
> 

Easy, just use the "join" function to concatenate a sequence of strings 
into a single string. For example:


import string

list = ['a', 'b', 'c', 'd']

s = string.join(list, '')

print s


Regards,
Derek.



From ponderor at lycos.com  Mon Jun  3 03:15:07 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 3 Jun 2002 00:15:07 -0700
Subject: prototyping good OOdesign in Python?
References: <mailman.1022961484.24782.python-list@python.org>
Message-ID: <e81be8b2.0206022315.3a0bf9a9@posting.google.com>

Ken Seehof <kseehof at neuralintegrator.com> wrote in message \
> > Maybe this was already discussed, but I'd liked to raise this topic again.
> >
> > Python has superb OOProgramming support. Lets not dispute this for a
> > moment. Now let's consider that we need to make well-designed OOP solution
> > first prototyped in Python.
> >
> > What troubles are there? Python is too dynamic and all that. And all
> > those nifty features need to be translated properly into, say, C++.
> >

Wow! A topic that I have a keen interest in and slight experience in!

re troubles...

- Implementing the design in Python using idioms that translate into
stuff only possible through extraneous C++ code and/or STL.
(range->foreach'isms, function pointers, ... mind is blanking ...
python's data types and support sure are handy, but it does take a few
extra brain cycles to write them in a C/C++ shop fashion (note the C,
in my experience only 20% of the advanced C++ topics are implemented
by 80% of the shop, and the other code is given the black-art
skepticism...in retrospect, that's a similar glare some folks get when
mentioning Python. *sigh*)  "Making Python look like C++" might not be
smart if you have the opportunity to keep the Python "Prototype".

I was hoping to seperated design from implementation in my
enumerations, but it doesn't look like that's going to happen at this
hour.

- Exception Handling & Asserts: Here I hope it's very similar.  Sorry,
wasn't a requirement in my world, no specific advice. Project advice
Include it in your original code & design prototype --adding it in
later is the dreams turn sour, but it sounds like you've got
appropriate time for up-front design.

> Maybe this has already been discussed

I didn't go looking. If you do, please post the links.    I'm waiting
for a software development book which uses Python for
design/prototyping, unit testing, building & system testing *non
Python projects*.

I do have lofty dreams of Python filling the role of Windows VBScript
& friends to Unix scripting tools.

Back to the topic...

- I got bit by Python shortcuts.  Looping and Sorting an array.  It
was so simple in python that it turned into an "assumption" in my
prototype and was dropped when I coded the C++ version.  A smack on
the forehead later and I started figured out how to sort my vector...

- I'm not sure if this needs saying, but... STL is nifty and
translates, but debugging is much harder than the comparable idioms in
Python. Schdule time  for creating debugging support code in those
objects. (if not for you, your peers)

- I succesfully tested a sub-system in a Python test harness that
allowed me to find bugs in it, protocol improvements, and that first
wave of implementation perspective before my writing any of my C++
code.
I didn't get to SWIG/Boost/etc. wrapping my C++ for further tests.
*sigh*    On that note, you  may want to peruse the boost.org/python
documentation ( http://boost.org/libs/python/doc/index.html ) and the
mailing list for a closer look at mapping C++ to Python.

If you DO built unit tests in Python for your Python API, the majority
should be re-usable in your python wrapped C++ implementation.

- Finally, note data inputs as reading and interacting from different
sources (existing libraries, goofy database connections, ??) might
take double time mapping prototype (python) and your implementation
(c++) to the source, along with who-knows-what "marshalling" issues.

Worse comes to worse, keep that shell handy as working out a small
problem in python sure beats Visio and Excel. (actually, unless your
reallly fast, I still used a spreadsheet for tabular perspective &
printing on data dumps.)

As you can see, I've contributed more cue's than facts.  You're quest
is a noble one, good luck in balancing research time with
hammer-it-out-ignore-redudancies-just-get-it-done time.

Maybe I should go back and read your code some more...
...maybe I should sleep.

Night,

Dean


From duncan at NOSPAMrcp.co.uk  Thu Jun 13 05:31:22 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Thu, 13 Jun 2002 09:31:22 +0000 (UTC)
Subject: How to check version of Python under CGI
References: <rbmggus7ln1uiutktf5libf4bbngu29ji7@4ax.com> <873cvrin1q.fsf@tux.ntw23.fr> <cgnggu40ethjhpqcmltnfb06d5qvuf078k@4ax.com>
Message-ID: <Xns922C692677D80duncanrcpcouk@127.0.0.1>

Johann <observer at NOSPAM.space.pl> wrote in 
news:cgnggu40ethjhpqcmltnfb06d5qvuf078k at 4ax.com:

> 
> Thanx. Do you know how to check what modules are installed? E.g. I
> would like to check if PYANA or other xml modules are installed.
> 

Try importing them and handle the exceptions that are thrown if they don't 
exist. Either wrap the import directly in a try:except: block, or define a 
function using __import__:

>>> def modAvailable(name):
	try:
		__import__(name)
	except ImportError:
		return 0
	return 1

>>> print modAvailable('sys')
1
>>> print modAvailable('junk')
0

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From xah at xahlee.org  Sat Jun 15 02:51:36 2002
From: xah at xahlee.org (Xah Lee)
Date: 14 Jun 2002 23:51:36 -0700
Subject: python install through fink, mac os x
Message-ID: <7fe97cc4.0206142251.164d9ec5@posting.google.com>

Python installation through fink failed with the following message.

can anyone help?

 Xah
 xah at xahlee.org
 http://xahlee.org/PageTwo_dir/more.html


[xah at localhost ~][Fri Jun 14,23:48:52]
fink install python
sudo /sw/bin/fink install python
Password:
Reading package info...
Information about 435 packages read in 5 seconds.
The following 9 additional packages will be installed:
 db3 db3-shlibs expat gdbm gdbm-shlibs gmp gmp-shlibs manconf tcltk
Do you want to continue? [Y/n] y
curl -f -L -s -S -O http://www.sleepycat.com/update/3.3.11/db-3.3.11.tar.gz
curl: (22) The requested file was not found
### curl failed, exit code 22
Downloading the file "db-3.3.11.tar.gz" failed.

(1)  Give up
(2)  Retry

How do you want to proceed? [2]


From gerhard at bigfoot.de  Sat Jun  1 09:28:43 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 1 Jun 2002 13:28:43 GMT
Subject: Migrating from PHP to Python...
References: <xWyk7.4824$v%4.492814@news20.bellglobal.com> <l4ihfu0nhp1a9p53vgch3n3kcc25s8c00p@4ax.com>
Message-ID: <slrnafhj0d.dp4.gerhard@lilith.my-fqdn.de>

Walnut wrote in comp.lang.python:
> On Sun, 2 Sep 2001 19:12:00 -0400, "Rick Shin" <rick at slaveware.com>
> wrote:
> 
>> Hi,
>> 
>> I'm a PHP man looking to migrate to Python for a change of scenery. I've
>> skimmed over the specs, but missed a few PHP constructs that I find
>> essential.
> 
> Forget them both and go with Perl - the most flexible and powerful
> language anywhere in the world today.

This statement in its absoluteness looks quite ridiculous.

> You already know the syntax.

It may look familiar, but the devil's in the details. So the seemingly
familiar syntax may even hurt more than help.

X-Post and F'up to appropriate group. Please continue trolling over
there.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From lsmith at aracnet.com  Thu Jun  6 15:51:09 2002
From: lsmith at aracnet.com (Lindsey Smith)
Date: Thu, 6 Jun 2002 12:51:09 -0700
Subject: win32com problem: No Such Interface Supported
Message-ID: <adoed4012g2@enews3.newsguy.com>

I'm having trouble using a COM object from Python 2.2.1 using the
win32all-143 extensions. I want to access ISubscriptionMgr, which is
implemented by webcheck.dll, a part of Internet Explorer 5+.
ISubscriptionMgr is defined by subsmgr.idl and subsmgr.h in the MS Visual C
vc98\include directory.

What I did was:
] midl subsmgr.idl

which yields subsmgr.tlb, then
] python makepy.py -i path\subsmgr.tlb

which yields:
{C54FD88A-FFA1-11D0-BC5E-00C04FD929DB}, lcid=0, major=1, minor=0
>>> # Use these commands in Python code to auto generate .py support
>>> from win32com.client import gencache
>>> gencache.EnsureModule('{C54FD88A-FFA1-11D0-BC5E-00C04FD929DB}', 0, 1, 0)


then making test.py:

from win32com.client import gencache
mod = gencache.EnsureModule('{C54FD88A-FFA1-11D0-BC5E-00C04FD929DB}', 0, 1,
0)

o = mod.SubscriptionMgr()
print o


but test.py bombs out in this way:

Traceback (most recent call last):
  File "test.py", line 5, in ?
    o = mod.SubscriptionMgr()
  File
"D:\DEV\PYTHON22\lib\site-packages\win32com\gen_py\C54FD88A-FFA1-11D0-BC5E-0
0C04FD929DBx0x1x0.py", line 37, in __init__
    if oobj is None: oobj = pythoncom.new(self.CLSID)
pywintypes.com_error: (-2147467262, 'No such interface supported', None,
None)

Any ideas on what I'm doing wrong?

thanks,
Lindsey





From rdacker at pacbell.net  Thu Jun 20 13:30:59 2002
From: rdacker at pacbell.net (rdack)
Date: 20 Jun 2002 10:30:59 -0700
Subject: urllib POST question
Message-ID: <644f6688.0206200930.48110837@posting.google.com>

i am doing a POST to a server using urllib and it works.
is there some way i can see exactly what is being posted?
a way to get the post object out of urllib - the exact lines it is sending?


From bokr at oz.net  Tue Jun 11 06:01:04 2002
From: bokr at oz.net (Bengt Richter)
Date: 11 Jun 2002 10:01:04 GMT
Subject: Does Python need a '>>>' operator?
References: <mailman.1023707533.19855.python-list@python.org>
Message-ID: <ae4hp0$3v6$0@216.39.172.122>

On Mon, 10 Jun 2002 14:11:31 +0300 (IDT), Beni Cherniavksy <cben at techunix.technion.ac.il> wrote:

>On 2002-06-10, Ken Seehof wrote:
>
>> Beni Cherniavsky <cben at tx.technion.ac.il> wrote:
>> > I just got another idea: use 0x1234 for 0-filled numbers and 1xABCD for
>> > 1-filled ones.  That way you impose no restrictions on what follows the
If there are no restrictions, 1x7 must have a meaning.
What do you want it to mean?

>> > prefix and keep backward compatibility.  0xFFFFFFFF stays a 2^n-1
>> > _positive_ number, as it should be.  The look of 1x is weird at first but
>> > it is very logical...
[...]
Well, consider the logic of a different format: The prefix is always 0h
and the hex data part always includes just enough duplicated-as-necessary
sign bits to make the first hex digit 0 for positive numbers and f for negative.

IOW, you could pretend to store the integer value in a 4*k bit wide register,
where k is just large enough so the top 4 bits are either all 0 or all 1.
Then just print with k hex digits for the data.

A previous version had a bug for sure, and I never got around to fixing it
and posting it to answer Martin's last objection. Still no guarantees, but
the version below illustrates the format...

[Martin]
>Unfortunately, you cannot represent such a number in hex, atleast not
>under any of the usual conventions: the convention is that you can
>strip infinitely many leading zeroes, but stripping off infinitely many
>leading ones is confusing.
>
With Greg Ewing's suggestion (to make sure leading hex digits are 0 or F rather
than just having 0 or 1 in the msb of the first hex digit to indicate sign, as
I had first suggested), I think it can be done very clearly:

Pretend that you have printed with extra width (e.g., infinite) and just strip
leading duplicate 0's or F's from the left of the data part until you have either
0h0.... or 0hF....

This illustrates the format, anyway:

 >>> def newhex(n, width=0):
 ...     """prints hex bits with leading 0 for positive and leading f for negative"""
 ...     if n<0: nh = ((~long(n))<<4)&long(n)
 ...     else:   nh = (( long(n))<<4)&(~long(n))
 ...     nh = len('%x' % nh)
 ...     ret =  '0h%s%x' % ((n>0 and '0' or ''),long(n)&((1L<<(nh*4))-1))
 ...     if width>len(ret):
 ...         return '0h'+ret[2]*(width-len(ret))+ret[2:]
 ...     else:
 ...         return ret
 ...
 >>> for i in range(12): print '%16s%16s' % (newhex( 1L<<i   ), newhex(-1L<<i   ))
 ...
             0h01             0hf
             0h02            0hfe
             0h04            0hfc
             0h08            0hf8
            0h010            0hf0
            0h020           0hfe0
            0h040           0hfc0
            0h080           0hf80
           0h0100           0hf00
           0h0200          0hfe00
           0h0400          0hfc00
           0h0800          0hf800
 >>>
 >>> for i in range(12): print '%16s%16s' % (newhex( 7L<<i   ), newhex(-7L<<i   ))
 ...
             0h07            0hf9
             0h0e            0hf2
            0h01c           0hfe4
            0h038           0hfc8
            0h070           0hf90
            0h0e0           0hf20
           0h01c0          0hfe40
           0h0380          0hfc80
           0h0700          0hf900
           0h0e00          0hf200
          0h01c00         0hfe400
          0h03800         0hfc800
 >>> for i in range(12): print '%16s%16s' % (newhex( 1L<<i,5), newhex(-1L<<i,5))
 ...
            0h001           0hfff
            0h002           0hffe
            0h004           0hffc
            0h008           0hff8
            0h010           0hff0
            0h020           0hfe0
            0h040           0hfc0
            0h080           0hf80
           0h0100           0hf00
           0h0200          0hfe00
           0h0400          0hfc00
           0h0800          0hf800
 >>> for i in range(12): print '%16s%16s' % (newhex( 1L<<i,10), newhex(-1L<<i,10))
 ...
       0h00000001      0hffffffff
       0h00000002      0hfffffffe
       0h00000004      0hfffffffc
       0h00000008      0hfffffff8
       0h00000010      0hfffffff0
       0h00000020      0hffffffe0
       0h00000040      0hffffffc0
       0h00000080      0hffffff80
       0h00000100      0hffffff00
       0h00000200      0hfffffe00
       0h00000400      0hfffffc00
       0h00000800      0hfffff800
 >>> for n in xrange(-32,33,4): print '%3d %16X %s' % (n,n,newhex(n))
 ...
 -32         FFFFFFE0 0hfe0
 -28         FFFFFFE4 0hfe4
 -24         FFFFFFE8 0hfe8
 -20         FFFFFFEC 0hfec
 -16         FFFFFFF0 0hf0
 -12         FFFFFFF4 0hf4
  -8         FFFFFFF8 0hf8
  -4         FFFFFFFC 0hfc
   0                0 0h0
   4                4 0h04
   8                8 0h08
  12                C 0h0c
  16               10 0h010
  20               14 0h014
  24               18 0h018
  28               1C 0h01c
  32               20 0h020

After 0h0 or 0hf there are no restrictions to the hex that follows (or doesn't),
and the abstract integer value is unambiguously determinable.

Regards,
Bengt Richter


From amuys at shortech.com.au  Thu Jun 27 20:29:01 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 27 Jun 2002 17:29:01 -0700
Subject: Help with an algorithm wanted
References: <rowen-CE3C0F.13020226062002@nntp1.u.washington.edu>
Message-ID: <7934d084.0206271628.665b03b3@posting.google.com>

"Russell E. Owen" <rowen at cesmail.net> wrote in message news:<rowen-CE3C0F.13020226062002 at nntp1.u.washington.edu>...
> I'd like some help solving a problem in Python.
> 
> The basic idea is I have an item of data that can be represented in any 
> of four ways: A,B,C and D. I also have conversion functions between 
> certain pairs of representations, such that there's always a path 
> between any two data representations. In this case suppose I know A<->C, 
> B<->C and C<->D, or graphically:
> 
> A <--+
>      +--> C <---> D
> B <--+
> 
> I'd like to create objects that have one of these items (A,B,C,D) as a 
> given, then be able to query any such object for any data representation.
> 
> For instance, an object with A as the given would return D by solving 
> A->C->D.
> 
> What I'm trying to figure out is a simple and reasonably elegant way to 
> code such objects.
> 

So effectively you have a conversion graph, where each vertex
represents a known representation, and each edge represents a
conversion (presumably with a conversion function as an annotation to
the edge).

What you are asking for appears to be an algorthm to find the shortest
path between to representations in the graph.  Fortunately for you,
Dijkstra provided a solution to this problem a number of years ago,
and implementing it is a standard algorthms assignment (I do hope
you're not asking for help with a homework assignment?).

You'll find an implementation in the Python Cookbook @
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/119466

The recipie was submitted by David Eppstein.  It dosn't look
optimised, but that's ok because it is legible (much more important)
and the algorithm is efficient.

Of signifigant interest to you is that it represents the graph as a
dictionary of edges, and as you will want to annotate the edges with
conversion functions, this is a good thing.

Also note that Dijkstra's algorthm uses weighted edges, so you get to
consider differing conversion costs for free.

Andrae Muys


From syver-en+usenet at online.no  Sat Jun  1 20:24:51 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Sun, 02 Jun 2002 00:24:51 GMT
Subject: ANN: PyUnitTestBrowser 003
References: <63604d2.0205310937.15d47972@posting.google.com> <uit54nem7.fsf@online.no> <ZSSJ8.10245$Om3.446676066@newssvr21.news.prodigy.com> <ad9nvr$hav@dispatch.concentric.net>
Message-ID: <uit52sf2z.fsf@online.no>

Phlip <phlip_cpp at yahoo.com> writes:

> >> Have you any hints on how to hack browser.py so that it run
> >> unittest.TestCase derived classes in a file without a suite method
> >> that generates it?
> > 
> > Someone sent me a patch that does this, so it's going in the next
> > "release".
> 
> This tweak and a few other cosmetic issues are in the bag.

Great, I'll try it out at once.
 

-- 

Vennlig hilsen 

Syver Enstad


From tim.one at comcast.net  Thu Jun 13 21:57:12 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 13 Jun 2002 21:57:12 -0400
Subject: why not "'in' in 'in'"?
In-Reply-To: <ae8dg70217g@drn.newsguy.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEPBPMAA.tim.one@comcast.net>

[Grant Griffin]
> One of the things I like about Python is that it works as expected.
> However, I sometimes accidently expect it to allow one to use "in"
> to determine if one string is "in" another, e.g.:
>
> >>> 'in' in 'in'
>
> which yields a traceback:
>
>    Traceback (most recent call last):
>      File "<interactive input>", line 1, in ?
>    TypeError: 'in <string>' requires character as left operand
>
> ...
>
> So why is the "in" operator limited to single characters when used as a
> membership operator?

In The Beginning, "x in y" was implemented as if via:

def like_in(x, y):
    for element in y:
        if cmp(x, element) == 0:
            return 1
    return 0

That's still close(*) to what it does if the class of y doesn't implement
the newer-fangled __contains__ method.  Ponder what like_in(x, y) does when
x and y are strings:  it can't possibly return 1 unless x is a 1-character
string, because each value element takes on is a 1-character string.
Therefore "str1 in str2" raised an exception when len(str1) != 1, as you
almost certainly didn't intend to use this as an expensive and obscure way
to spell 0.  If that's what you really wanted, fine, but then Guido wanted
to help you achieve that goal by forcing you to use some way even more
expensive and obscure <wink>.

But that was then.  Python does allow defining __contains__ methods now,
precisely so that a type can define the meaning of "x in y" in whatever way
makes most sense for type(y).  The meaning of "in" is no longer *defined* by
like_in, and consequently Guido agrees that the string behavior is much
harder to justify now.

So if somebody wants this change enough to submit a patch (code, doc, + test
suite changes), it's likely to be accepted for 2.3.  As code changes go,
this one should be particularly easy.  We have an internal bet going as to
whether this invitation is enough to kill the idea <wink>.

some-parts-of-this-were-actually-true-ly y'rs  - tim


(*) The default implementation today uses "x == element" instead of cmp, so
that a type implementing only __eq__ works fine.  Other than that, same
thing.





From in-call at gmx.net  Fri Jun 21 15:11:05 2002
From: in-call at gmx.net (Pablo Ripolles)
Date: 21 Jun 2002 12:11:05 -0700
Subject: Passing inner functions as arguments
References: <d72779c7.0206201628.33d6a657@posting.google.com> <aetsf4$sam$1@peabody.colorado.edu>
Message-ID: <d72779c7.0206211111.62255c6c@posting.google.com>

> Nested scopes are not on by default in 2.1 You need AT THE TOP of your file to 
> put 'from __future__ import nested_scopes'.

Thanks, It worked great!

Just one more thing: jythonc doesn't build the *.jar  file when
executing 'jythonc --core --deep --jar kmoment.jar *.py'

Any idea?

Thanx again,
Pablo.


From bokr at oz.net  Sun Jun 30 18:53:12 2002
From: bokr at oz.net (Bengt Richter)
Date: 30 Jun 2002 22:53:12 GMT
Subject: arguments to functions
References: <MPG.178955a4fc8bfc1c989680@news.rz.uni-karlsruhe.de> <afnj4g$4l5$0@216.39.172.122>
Message-ID: <afo24o$l3t$0@216.39.172.122>

On 30 Jun 2002 18:37:04 GMT, bokr at oz.net (Bengt Richter) wrote:
[...]
>You could get what I guess is the effect you want (keyword value takes priority) something like:
>
> >>> def test(_expected=None, *misc, **kwargs):
> ...     expected = kwargs.get('expected',None) or _expected
> ...     print 'expected:%s, misc:%s, kwargs:%s' % (`expected`,`misc`,`kwargs`)
> ...
Oops, not quite:
 >>> test(expected=0)
 expected:None, misc:(), kwargs:{'expected': 0}

Shortcut evaluation is dangerous ;-/

 >>> def test(_expected=None, *misc, **kwargs):
 ...     if kwargs.has_key('expected'):
 ...         expected = kwargs['expected']
 ...     else:
 ...         expected = _expected
 ...     print 'expected:%s, misc:%s, kwargs:%s' % (`expected`,`misc`,`kwargs`)
 ...
 >>> test(expected=0)
 expected:0, misc:(), kwargs:{'expected': 0}

Regards,
Bengt Richter


From cliechti at gmx.net  Sun Jun 30 17:45:43 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 30 Jun 2002 23:45:43 +0200
Subject: I'd give up Perl tomorrow if only...
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au> <mailman.1024681648.27687.python-list@python.org> <Pine.LNX.4.44.0206301550120.7807-100000@palanthas.neverending.org>
Message-ID: <Xns923DF22CCC54Ccliechtigmxnet@62.2.16.82>

Frank Tobin <ftobin at neverending.org> wrote in
news:Pine.LNX.4.44.0206301550120.7807-100000 at palanthas.neverending.org: 
> Christopher A. Craig, on 2002-06-21, wrote:
> 
>> Given the hardware and bandwidth, it shouldn't be all that hard to
>> create a CPAN for Python.  I suspect one of the main reasons it
>> hasn't been done is that Google works so well.
> 
> I don't think you have a good grasp of what CPAN really is.  It's much
> more than simply a big FTP site.
> 
> The key thing about CPAN is not only the breadth of modules it has,
> but the strong testing system they employ, namely cpan-testers.  I am
> the author of a module on CPAN, and it's great to be able to upload
> module, and have it tested on a diverse set of architecures.
> 
> http://testers.cpan.org/
> 
> Note that in order to have something be as successful as cpan-testers,
> modules needs to have:
> 
> 1) a standard extraction mechanism (Perl and Python do have)
> 2) a standard testing mechanism (Perl has ('make test'), Python doesn't)

well it has. the module unittest provides all thats needed. the module just 
needs a class derrived from unittest.TestSuite and the unittest module can 
find the tests.
its also a convention to have a test() function, but the former is more 
powerful. yes, it should be used more widely.

> 3) a standard installation mechanism (Perl and Python do have)
> 4) a standard README system (Perl has, Python doesn't)

i don't know the perl way but with python every object has a __doc__ 
attribute, every module, class, function and method etc. and with pydoc 
there is a standard extraction tool right in the standard distro. ever 
tried it?

> What is most missing for Python is a standard testing suite system. 
> And until it gets this, it's pretty much not possible to get anything
> close to the quality of CPAN with cpan-testers.

as metioned above, python has all it needs, now if you talk of an 
organization that runs test on non-standard extension modules, no we don't 
have that right now. the python distro has good tests for the standard lib, 
but as there is no organization that distributes extensions, everyone is 
responsible for testing its modules. (there are collections of modules, 
like The Vaults of Parnassus: http://www.vex.net/parnassus/ or ActiveStates 
Programmer's Package Manager (PPM) but thats not as strong or established 
as CPAN)

> Another great benefit of CPAN is that the modules on it are somewhat
> 'sanctioned'.  Sure, it doesn't take much to get onto CPAN, but the
> difference between the 'something' it does take, and 'nothing', is
> enormous in speaking for the quality of the modules you'll find in
> contrast to searching Google and coming across Joe Schmoe's Python
> module. A similar scenario is comparing random RPM's you might find on
> Google to FreeBSD/Gentoo's ports trees; the centralization somewhat
> sanctions the software, and this is critical, IMO.

agreed, but nothing aginst non registed modules, found on google etc... not 
everyone likes the work to get a module registerd but it might be of great 
quality and right the thing one needs.

> Furthermore, I find the de-facto multi-tier naming system that Perl
> employs for modules to be a good thing; it encourages users dividing
> up their modules in the global namespace.  This is critical for
> finding what you need.  For example, I can go into the DBD branch on
> CPAN and find every database driver it has.  Python doesn't have an
> equivalent (packages just aren't the same thing; they are owned by
> 'one' thing, most of the time).

thats one solution, java uses the reversed URL of the company to avoid 
naming conflicts.
python should select one system if it comes to a CPAN equivalent. but the 
current situation has worked very well.

chris

-- 
Chris <cliechti at gmx.net>



From jepler at unpythonic.net  Thu Jun 27 15:41:20 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Thu, 27 Jun 2002 14:41:20 -0500
Subject: Pythoniac: Thoughts on a hardware Python processor
In-Reply-To: <GCEDKONBLEFPPADDJCOEGEIHDMAA.whisper@oz.net>
References: <F7DB8D13DB61D511B6FF00B0D0F06233015DE661@mail.hebmex.com> <GCEDKONBLEFPPADDJCOEGEIHDMAA.whisper@oz.net>
Message-ID: <20020627194119.GB5616@unpythonic.net>

On Thu, Jun 27, 2002 at 12:36:00PM -0700, David LeBlanc wrote:
> Ahh! Haven't you heard? Forth came in Fifth and the Forth chip is dead. In
> fact I don't think anyone is still offering a stack based computer (only HP
> did AFAIK, the HP 3000).

Intel and AMD still offer a chip with a register stack in the
floating-point unit.  Does that count?

Jeff




From mgilfix at eecs.tufts.edu  Wed Jun 12 19:31:25 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Wed, 12 Jun 2002 19:31:25 -0400
Subject: Wanted:  Potential Python Authors
In-Reply-To: <200206130924.20528.rjones@ekit-inc.com>; from rjones@ekit-inc.com on Thu, Jun 13, 2002 at 09:24:20AM +1000
References: <ae7gke$mhb$1@ccis12.baylor.edu> <db3fgu40sbnt3pno5v1i2h9lnmpcoti2bk@4ax.com> <3d07d4a8@news.eol.ca> <200206130924.20528.rjones@ekit-inc.com>
Message-ID: <20020612193125.D10542@eecs.tufts.edu>

On Thu, Jun 13 @ 09:24, Richard Jones wrote:
> On Thu, 13 Jun 2002 09:09, William Park wrote:
> > Lothar Scholz <llothar at web.de> wrote:
> > >>- Author of articles in the various trade magazines
> > >
> > > You really mean "trade" magazines ?
> > > Normally they are written by marketing peoples who don't
> > > have the competence to write more then a 4 page overview.
> > > Hands off !!!!
> >
> > Bull.  I agree about trade magazines above, though there isn't any magazine
> > covering Python.
> 
> Ehem.   http://www.pyzine.com/

  ... and the guy who runs it has a good sense of humour too!

                 -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From cliechti at gmx.net  Mon Jun 24 18:52:05 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 25 Jun 2002 00:52:05 +0200
Subject: Threads on embedded Python
References: <1024952585.111238@seux119> <Xns9237EECAA9CC5cliechtigmxnet@62.2.16.82> <1024957605.771843@seux119>
Message-ID: <Xns9238944DA0FEcliechtigmxnet@62.2.16.82>

"Ugo Garc?a" <plugin at supercable.es> wrote in
news:1024957605.771843 at seux119: 

> Let see if I get something clear... I don't have any main loop in
> Python. I only have some functions that 'control' certains object in
> my operation. To be a little more clear, the task of the functions are
> for example, move an object in the world generated by a game with a
> little intelligence. For example: (a simple example)
> 
> FUNCTION 1:
>     - Wait until main character is in correct position
>     - Play a sound
>     - Move an enemy from pos X1,Y1 to X2,Y2
>     - Play an animation
>     - ....
>     - ....
> 
> So, I don't have a main loop, only function like that (and other kind
> of functions) that a run as a thread to have all them running
> simultaneous. Is there other way to do that without having threads?

sure, it's called polling ;-)

and i think this one could be an iteresting read for you:
http://www-106.ibm.com/developerworks/library/l-pythrd.html
it's a good way to distribute CPU power to different 'microthreads' while 
having  them well separated. altough you will need polling as microthreads 
don't support blocking IO.


why not hanging it up the other way round? i think if you have the main 
loop in python the thing will get easier. you can then call back to your C 
module to do the screen updates etc.

> Do
> I have to have a main loop in order to have the threads running?

when you're in a C module the Global Interpreter Lock stopps all python 
activity util you enable it explicitly (Py_BEGIN_ALLOW_THREADS) or you 
return to python.
the GIL makes writing extensions to python peanuts...

> Perhaps running Python in a separate thread could be a good idea.

yes someone has to spend the CPU cycles...

> Thanks,
> --ugo
> "Chris Liechti" <cliechti at gmx.net> escribi? en el mensaje
> news:Xns9237EECAA9CC5cliechtigmxnet at 62.2.16.82...
>> "Ugo Garc?a" <plugin at supercable.es> wrote in
>> news:1024952585.111238 at seux119:
>>
>> > Hi all. I have an applicattion who uses Python to control some
>> > processes. I've made a little script library to work with threads
>> > and each proccess is a Python thread. The main application make
>> > calls to Python in order to create these proccesses. The problem is
>> > that the proccess doesn't run unless I call Python to do something.
>> > Now in the main loop I have a line like this:
>> >     PyRun_SimpleString("dummy=1");
>> >
>> > This make Python the threads to run. The problem is that this is a
>> > little slow 'cause if I call Python from tha main application too
>> > many times the proccesses run well but not the main application,
>> > and if a make a big pause between calls the proccesses doesn't work
>> > at a good speed. Is posible to have some threads running in an
>> > embedded Python without making callings to it continuosly? It would
>> > be a better idea to have another thread who's only task is to call
>> > Python (a line that the one written before) in order to have the
>> > Python proccesses running?
>>
>> why not run python in a separate thread (created by your app) and
>> doing 
> the
>> loop in python, like PyRun_SimpleString("while 1: mainloop()") or so?
>> you do have a main loop in python, don't you? otherwise you wouldn't
>> need python threads, right? :-)
>>
>> you could possibly also release the GIL, but don't forget to grab it
>> again when accessing the python interpreter. i'm not sure about this
>> one, so i would try the above.
>>
>> chris
>>
>> --
>> Chris <cliechti at gmx.net>
>>
> 
> 
> 



-- 
Chris <cliechti at gmx.net>



From raven_lists at hotmail.com  Wed Jun 19 14:39:44 2002
From: raven_lists at hotmail.com (Raven)
Date: Wed, 19 Jun 2002 14:39:44 -0400
Subject: isometric engine
Message-ID: <uh1nfitkhs2s36@corp.supernews.com>

Hi, I was wondering if there were any open source iso engines for python. I
found the old tkclient for worldforge, but its page is gone now and I havent
found the source anywhere and I really didn't want to have to write one up
from scratch if their was already one available. Any help wouold be greatly
appreciated. :)

Eli




From pedro_rodriguez at club-internet.fr  Sun Jun 30 11:55:08 2002
From: pedro_rodriguez at club-internet.fr (Pedro RODRIGUEZ)
Date: Sun, 30 Jun 2002 17:55:08 +0200
Subject: Problem with re group in python 2.1 and 2.2
Message-ID: <pan.2002.06.30.15.55.07.677916.2568@club-internet.fr>

Greetings,

While writing some unit tests for a rfc1738 parser, I found the following
strange behaviour when using parenthesis in a named group regular
expression.

The following expression (?P<name>.*) and (?P<name>(.*)) don't behave in 
the same way. When the matching fails, the first group will be None, but
the last one will contain an empty string.

The problem occurs with python 2.1.1 and 2.2. Python 1.5.2 OTH works fine.

I checked the documentation but didn't find any hint on this. So, before
submitting a bug report, I wanted to know if I missed something.


# Code exhibiting the problem
import sys
print sys.version

import re

# [ user [ : password ] @ ] host

x = re.compile("((?P<user>(.*))(:(?P<password>.*))?@)?(?P<host>.*)") 
m = x.match("host")
print m.groupdict()

x = re.compile("((?P<user>.*)(:(?P<password>.*))?@)?(?P<host>.*)") 
m = x.match("host")
print m.groupdict()


# Outputs 1.5.2
1.5.2 (#1, Jul  5 2001, 03:02:19)  [GCC 2.96 20000731 (Red Hat Linux 7.1 2
{'password': None, 'host': 'host', 'user': None}
{'password': None, 'host': 'host', 'user': None}

# Outputs 2.1.1
2.1.1 (#1, Sep 27 2001, 13:48:33)
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-81)]
{'user': '', 'password':None, 'host': 'host'} 
{'user': None, 'password': None, 'host': 'host'}

# Outputs 2.2
2.2 (#2, Jan 15 2002, 20:55:15)
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-96)] 
{'host': 'host', 'password': None, 'user': ''} 
{'host': 'host', 'password': None, 'user': None}

Regards,
Pedro


From kurugant at denshi.ece.utk.edu  Wed Jun 26 20:46:15 2002
From: kurugant at denshi.ece.utk.edu (Teja Sastry)
Date: Wed, 26 Jun 2002 20:46:15 -0400
Subject: Accessing C++ variables from Python
Message-ID: <3D1A6057.509@denshi.ece.utk.edu>

I have a c++ program storing a binary stream into a buffer, i want to 
access that buffer from python. The buffer is actually a cPickled binary 
stream from another program loaded in to c++ 'cuz we use a routing layer 
called diffusion, and its API is in c++ so i have to put the variable 
into C++ to transport it over the network, at the receiving end i have 
another C++ diffusion API program holding the variable, i have to 
transfer it as object back to python, i'm not quite sure about how to do 
this could you please clarify or is there any mailing list for python users

Thanks,
Teja Sastry

P.S: I know how to use SWIG and creating shared modules for python





From nookieNO_SP_AM at online.no  Tue Jun  4 18:57:52 2002
From: nookieNO_SP_AM at online.no (Erlend J. Leiknes)
Date: Tue, 04 Jun 2002 22:57:52 GMT
Subject: directory 
References: <3cfcfd4a_1@news.iprimus.com.au>
Message-ID: <QBbL8.9844$fG3.335957@news2.ulv.nextra.no>

you forgot the ending /

"Gold Fish" <occeanlinux at linuxmail.org> wrote in message
news:3cfcfd4a_1 at news.iprimus.com.au...
> How to read the subdirectories in the directory
> Support i have the directory /home/ which have some subdirectories
> /home/sub1
> /home/sub2
> /home/sub3
> I am using os.listdir('/home/') to read in the list so it will be
> ['sub1','sub2','sub3'] now i want to read inside the sub directory sub2
> once again i using os.listdir(os.listdir('/home/')[1]) but it came out
with
> an errors: not a file or directory. Can anyone tell me what did i wrong.
> Anh how to fix it.




From ken at hotmail.com  Sat Jun  1 06:12:16 2002
From: ken at hotmail.com (Ken)
Date: Sat, 1 Jun 2002 20:12:16 +1000
Subject: Python and HTML frames
Message-ID: <ada6hf$vico3$1@ID-49758.news.dfncis.de>

How do I output HTML onto frames? This code below doesn't work. Can someone
help me?

Thanks

==
def mainpagelayout():
        print """
        <head><title>Languages Dictionary</title></head>

        <frameset rows="70,40,*" scroll=no frameborder="NO" border="0">
        <frame src="title.html" name=Title scrolling=no noresize>
        <frame src="topmenu2.html" name=TopMenu scrolling=no noresize>

        <frameset cols="*, 200" frameborder="NO" border="0">
        <frame src="main2.html" name=Main scrolling=auto>
        <frame src="sidemenu2.html" name=SideMenu scrolling=auto>
        </frameset> </frameset>

        <body>This page doesn't work without a frames-enabled
browser.</body>
        </noframes></body></html>
        """




From Eivind.Bergh.Nilssen at dnv.com  Tue Jun 25 02:41:40 2002
From: Eivind.Bergh.Nilssen at dnv.com (Eivind B. Nilssen)
Date: 24 Jun 2002 23:41:40 -0700
Subject: ntohl/htonl socket transfer between c and python
Message-ID: <ff4066da.0206242241.f614118@posting.google.com>

Hi,

I have a bit of a problem I cant seem to solve. I am new at
programming python,
and my task is to make a client in python for fetching data from a
remote server written in c.

The buffer sendt to me from the server is generated as follows:

*(int*)&bfr[bix] = htonl(*(int*)&f); #where f is a float value

The buffer received in my python program is a lot of strange asciis
I am not able to cast this buffer to an int or a long and this again
makes
me unable to apply the ntohl command.

Does anybody have an idea how to extract the data from this buffer?
Do I have to write this part in C?

Thanks a lot in advance.


From teyc at cognoware.com  Fri Jun 28 23:58:44 2002
From: teyc at cognoware.com (Chui Tey)
Date: Sat, 29 Jun 2002 13:58:44 +1000
Subject: Execute python script via C API line-by-line
References: <afh413$h5f$01$1@news.t-online.com>
Message-ID: <3d1d3139$0$31824$afc38c87@news.optusnet.com.au>

You could consider subclassing the python debugger. Have a look at
lib/bdb.py and lib/pdb.py. You get pretty fine grained control over the
interpreter this way.




From cbbrowne at acm.org  Mon Jun 10 14:38:31 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: 10 Jun 2002 18:38:31 GMT
Subject: activating a webpage's click command from python
References: <mailman.1023727993.12908.python-list@python.org> <3D04EA96.B3518A37@engcorp.com> <mailman.1023733034.32199.python-list@python.org>
Message-ID: <ae2rn7$37mpf$2@ID-125932.news.dfncis.de>

Quoth Andy McKay <amckay at merlintechnologies.com>:
>> does the job for Nortel on the TSE (disclaimer: anyone even 
>> thinking of buying Nortel should probably just keep thinking :-).
>
> Oh come on its a bargain, a fraction of what it was, good buying
> opportunity ;)
>
> not-putting-my-money-there'ly yours

The normal "Efficient Market" hypothesis is that the present price at
which a stock trades on a public market is a fair price.

So if it's trading at $2, then there's no fairer price to be found.
It is a pretty good theory, all in all.
-- 
(reverse (concatenate 'string "gro.mca@" "enworbbc"))
http://www.ntlug.org/~cbbrowne/internet.html
Rules of the  Evil Overlord #96. "My door  mechanisms will be designed
so that blasting  the control panel on the outside  seals the door and
blasting  the control panel  on the  inside opens  the door,  not vice
versa." <http://www.eviloverlord.com/>


From gerhard.haering at gmx.de  Thu Jun 20 08:17:38 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Thu, 20 Jun 2002 14:17:38 +0200
Subject: Java to Python converter?
In-Reply-To: <YrjQ8.34$g32.995@nreader1.kpnqwest.net>
References: <YrjQ8.34$g32.995@nreader1.kpnqwest.net>
Message-ID: <20020620121738.GB1043@lilith.my-fqdn.de>

* Martin Oberhuber <martin.oberhuber at windriver.com> [2002-06-20 13:58 +0200]:
> Hi all,
> 
> I've got some Java source code that I would like to convert into Python.

Why?

> Does anybody know of a converter that can do this?

Brain 1.0.

Btw. the Jython incarnation of Python can use Java libraries.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From dswegen at software.plasmon.com  Tue Jun 25 05:52:49 2002
From: dswegen at software.plasmon.com (Dave Swegen)
Date: Tue, 25 Jun 2002 10:52:49 +0100
Subject: Python daemon
In-Reply-To: <3d182dd6$1@news.mt.net.mk>
References: <3d182dd6$1@news.mt.net.mk>
Message-ID: <20020625095249.GD25507@software.plasmon>

On Tue, Jun 25, 2002 at 10:46:24AM +0200, ?????? ?. wrote:
> Hello, 
> 
> I'm writing a Python daemon on Unix and I need the folowing features:
>  1. going to background, diassociate from controling tty (done)

At this point I'd save the PID of the daemon into a file in
/var/run/<daemonname>.pid, say.

>  2. watchdog, the program should never be stuck

Have a shell script called by cron that checks that the PID still exists
in the output of /bin/ps, and have it called at appropriate intervals.
Or instead of using cron call both the watcher and the daemon from a
starter script. Personally I'd go with the cron option.

The above solution would work if the daemon went away, but not if the
it hung. If that might happen have the daemon update the timestamp on
the pid file at a regular interval (say each iteration of a main loop).
Then use the timestamp to determine if the daemon seems to have hung or
not.

This is a simple but quite robust way, that I would think should be
robust enough for most apps. If that isn't finegrained enough you could
look at using named pipes, signals or a gazillion other methods of
communicating with the daemon.

Perhaps there is a more pythonic way of doing it, but this is how I'd do
it.

>  3. restarting of the program if it fails, or the watchdog doesn't trigger

Use a wrapper script that starts, stops (using the pidfile) and restarts the
daemon. The wrapper can then be called at bootime or by the watcher
script.

>  4. Logging

Don't know what the best python way of doing this is.

> 
> 
> The issue 1. I solved, it's on ActiveState's site. Issue 4. is not that
> hard to do I guess. Any suggestions about 2 and 3 ?
> 
> 
> 
> -- 
> ??????
> 
>    "One world, one web, one program"  -- Microsoft promotional ad
>          "Ein Volk, ein Reich, ein Fuehrer"  -- Adolf Hitler
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From denpeterson at yahoo.com  Sun Jun  2 12:26:07 2002
From: denpeterson at yahoo.com (Dennis Peterson)
Date: Sun, 2 Jun 2002 12:26:07 -0400
Subject: What is a classmethod for?
Message-ID: <addgv30f4d@enews4.newsguy.com>

class C:
    def foo(cls):
        print cls
    foo = classmethod(foo)

class D(C):
    pass

c = C()
d = D()
c.foo() prints __main__.C
d.foo() prints __main__.D

That's nice, but I can't think of a use for it, given that we have
inheritance and isinstance(). Can anyone enlighten me?

Thanks,
Dennis




From CousinStanley at HotMail.Com  Thu Jun 13 10:05:48 2002
From: CousinStanley at HotMail.Com (Cousin Stanley)
Date: 13 Jun 2002 14:05:48 GMT
Subject: PIL _imaging
References: <Xns922AA829A8C5FHotMailcs@130.133.1.4> <3D06E694.4040103@mxm.dk> <Xns922B3150A6B32HotMailcs@130.133.1.4> <3D074529.E848AA6@ipm.fhg.de> <Xns922B561EDB212HotMailcs@130.133.1.4> <3D07BCC1.2000002@mxm.dk> <Xns922BB340B788BHotMailcs@130.133.1.4> <9C_N8.11930$p56.3660854@newsb.telia.net>
Message-ID: <Xns922C48A439D35HotMailcs@130.133.1.4>

|| and you are 100% sure that you're using a version
|| of PIL built for Python 2.2 ?

Fredrik ...

Thanks for the reply ...

The tests I tried were using PIL v1.1.2 
which I see  NOW  after reading your question
and re-checking at the web site is indeed 
NOT compatible with Python 2.2 ... 

However, I'm confused as to whether or not
part of  PIL v1.1.3  must be re-compiled/re-built
before using it ... 

Cousin Stanley


From insyte-clp at emt-p.org  Thu Jun 20 16:06:09 2002
From: insyte-clp at emt-p.org (Ben Beuchler)
Date: 20 Jun 2002 20:06:09 GMT
Subject: urllib POST question
References: <644f6688.0206200930.48110837@posting.google.com>
Message-ID: <slrnah4dbq.r77.insyte-clp@petra.bitstream.net>

In article <644f6688.0206200930.48110837 at posting.google.com>, rdack wrote:
> i am doing a POST to a server using urllib and it works.
> is there some way i can see exactly what is being posted?
> a way to get the post object out of urllib - the exact lines it is sending?

I'm a big fan of Ethereal...
"Sniffing the glue that holds the internet together!"

http://www.ethereal.com/

-- 
Ben Beuchler                                           There is no spoon.
insyte at emt-p.org                                            -- The Matrix


From pearu at cens.ioc.ee  Thu Jun 20 05:57:00 2002
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Thu, 20 Jun 2002 12:57:00 +0300 (EEST)
Subject: Looking for Python tools for extracting information from PDF files.
Message-ID: <Pine.LNX.4.21.0206201249470.4516-100000@cens.kybi>

Hi,

Can anyone suggest a (preferably Python) tool that can be used to extract
various information from PDF files? In particularly, I am interested in
extracting graphics from the PDF files.

I tried to search the web but all I found were tools to
create/modify/transform PDF files but I could not find any tools that can
extract graphics. Any hints are appreciated.

Thanks,
	Pearu






From whisper at oz.net  Tue Jun  4 02:41:46 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 3 Jun 2002 23:41:46 -0700
Subject: Please help with script...
In-Reply-To: <adheqc$nhn$1@netnews.upenn.edu>
Message-ID: <GCEDKONBLEFPPADDJCOEMEHGDGAA.whisper@oz.net>

Go here: http://www.vex.net/parnassus/

Type "CSV" (without quotes) into the find box at the bottom, then hit Find.
This will get you pointers to Python packages that do CSV processing.

Dave LeBlanc
Seattle, WA USA





From yvon.thoraval at free.fr  Sat Jun  1 17:46:40 2002
From: yvon.thoraval at free.fr (Yvon Thoraval)
Date: Sat, 1 Jun 2002 23:46:40 +0200
Subject: wxPython
Message-ID: <1fd4hyj.bno0s91x0bz9cN%yvon.thoraval@free.fr>

does a macos 9 wxMacPython exists ?

if yes, where ?


From rnd at onego.ru  Tue Jun 18 05:36:13 2002
From: rnd at onego.ru (Roman Suzi)
Date: Tue, 18 Jun 2002 13:36:13 +0400 (MSD)
Subject: unicode strings and strings mix
In-Reply-To: <j4ofe9t67z.fsf@informatik.hu-berlin.de>
Message-ID: <Pine.LNX.4.33.0206181327340.30160-100000@rnd.onego.ru>

On 18 Jun 2002, Martin v. [iso-8859-1] L?wis wrote:

>Gerhard H?ring <gerhard at bigfoot.de> writes:
>
>> 'x' and 'A' are in the ASCII range, so this shouldn't produce an
>> exception. I also cannot reproduce it with sys.getdefaultencoding() ==
>> "ascii".

sorry. message was in koi8-r but I just wanted to have 
chars with high bit set.

>These where not 'x' and 'A', but '\xd7\xc1\xd7\xc1\xd7'. Since the
>article was posted in KOI8-R, Roman probably meant those bytes to
>denote CYRILLIC SMALL LETTER VE and CYRILLIC SMALL LETTER A,
>respectively.
>
>Of course, when Python add strings, it can't possibly know that this
>is how the byte string was meant to be interpreted, so you need to
>write
>
>unichr(0x3345) + unicode('\xd7\xc1\xd7\xc1\xd7', 'koi8-r')
>
>The result string cannot be represented in KOI8-R, though, since it
>contains SQUARE MAHHA.

That was randomly choosen. 
What I wanted to ask is that it should be considered UNSAFE
to mix strings and Unicode-strings without explicit {en,de}code
methods, because the program must not depend on locale setting.

For example, a developer is using "ASCII" 128-255 for his national
alphabet (Oleg pointed siteconfig.py) and writes a program which runs just
fine for him. However, the same program will fail for almost everyone
else.

I think, it is a source for very subtle errors in Python programs as it
easily makes them not 8-bit clean! Any astray 

u"unicode-string" + "simple \234string"

can raise UnicodeException.

What do you think?

Sincerely yours, Roman Suzi
-- 
\_ Russia \_ Karelia \_ Petrozavodsk \_ rnd at onego.ru \_
\_ Tuesday, June 18, 2002 \_ Powered by Linux RedHat 7.2 \_
\_ "Laughing stock: cattle with a sense of humour." \_





From ykingma at accessforall.nl  Sun Jun 23 15:43:59 2002
From: ykingma at accessforall.nl (Ype Kingma)
Date: Sun, 23 Jun 2002 20:43:59 +0100
Subject: TimeoutQueue.py
References: <mailman.1024698810.22387.python-list@python.org>
Message-ID: <3D1624E8.AF7A9BF1@accessforall.nl>

Chris, Tim,

Tim Peters wrote:
> 
> [Chris Liechti]
> > Anyone ever needed a Queue with timeout?

Yes, and I found a solution for jython, see below.
 
> Yes.  I encourage you to submit a patch to SourceForge, adding optional
> timeout arguments to the std Queue module (include the requisite doc changes
> and test suite changes too, and I'll secretly ram it in <wink>).  Queue was
> written long before threading.py, and, as you noticed, the blocking
> operations in threading.py generally do support optional timeouts.  They're
> a Good Idea for all blocking operations.

For jython users: a queue with timeout on get() and put() operations
was done by Doug Lea in his util.concurrent java package. See:
http://aspn.activestate.com/ASPN/Mail/Message/Jython-users/828840
for a python interface to this package.

Btw. I would expect Chris' patch to work in jython, too.

Chris, did you add a timeout to the put() method as well?

Regards,
Ype


-- 
email at xs4all.nl


From rnd at onego.ru  Sat Jun 22 17:29:17 2002
From: rnd at onego.ru (Roman Suzi)
Date: Sun, 23 Jun 2002 01:29:17 +0400 (MSD)
Subject: Python hits the spot
In-Reply-To: <3D14D0FE.9070801@kfunigraz.ac.at>
Message-ID: <Pine.LNX.4.44.0206230106560.31863-100000@rnd.onego.ru>

On Sat, 22 Jun 2002, Siegfried Gonzi wrote:

>Roman Suzi wrote:
>
>> It's no wonder that 1000 lines of uncommented code behave like that, taking
>> into account that you were trying to integrate several (Tcl/Tk, Fortran, C,
>> Python) systems in a hurry, and, I presume, not knowing most of them.
>
>I am up to the point where you made my day (really). Are you sure do you 
>know what you are talking about?

If you are unhappy with Python - be happy with whatever tools you like more,
but please do not make strange illogical statements here.
I am sorry, but after reading your posts I concluded that you are
completely clueless about your computing environment(s)
(or have fun playing fool).

Have fun!

>Regards,
                  >S. Gonzi

Sincerely yours, Roman Suzi
-- 
rnd at onego.ru =\= My AI powered by Linux RedHat 7.2





From pierpaolo.glave at ciaolab.com  Tue Jun 25 09:04:59 2002
From: pierpaolo.glave at ciaolab.com (Pier Paolo Glave)
Date: 25 Jun 2002 06:04:59 -0700
Subject: Embedding in Tkinter (reprise)
References: <1907a5e4.0206240803.6ca0a4c@posting.google.com> <mailman.1024936355.12699.python-list@python.org>
Message-ID: <1907a5e4.0206250504.4b417e9a@posting.google.com>

I tried to follow the hint by Jeff, but it seems that the TkSteal
package hasn't been updated for a long time: the last available
version is for Tk 4.x.

Anyway, I made some further investigation on Tcl/Tk, and I found out
that there exists an extension package to Tk 8.x, named BLT, which has
a "container" object that can fit for my needs.

I tried this Tcl code sample, and I succeeded in embedding an "xclock"
into a Tk window:


#!/usr/bin/bltwish
package require BLT
if { $tcl_version >= 8.0 } {
    namespace import blt::*
    namespace import -force blt::tile::*
}
frame .f 
bgexec myVar xclock &
container .f.c
pack .f -fill both -expand yes
pack .f.c -fill both -expand yes -padx .1i -pady .1i
.f.c configure -relief raised -bd 2 -name "xclock"


Now, my problem is that I'm not an expert in Python/Tk integration,
and I don't know how to wrap this into a Python class.

Does anyone have any suggestions?

Thank you and regards,
--
Pier Paolo Glave

Jeff Epler <jepler at unpythonic.net> wrote in message news:<mailman.1024936355.12699.python-list at python.org>...
> Search for the "tksteal" package.  This adds a Tk widget that can
> "swallow" or "steal" an arbitrary window when its window ID is given.
> Once you've compiled this and it is available to your Tcl/Tk
> installation, you need merely define a Python class that wraps this new
> widget type, like the existing ones in Tkinter.py.
> 
> I found a manpage here:
> http://www-komo.ise.eng.osaka-u.ac.jp/misc/tcl/tk_steal_man/tksteal.n.html
> One copy of the source code for this widget seems to be at
> ftp://ftp.neosoft.com/languages/tcl/sorted/packages-7.6/devel/TkSteal4.0c.tar.gz
> but there are a lot of dead links related to this piece of software.
> 
> Jeff


From whisper at oz.net  Fri Jun  7 20:35:33 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 7 Jun 2002 17:35:33 -0700
Subject: Python MTA
In-Reply-To: <20020607232709.GA12947@lilith.my-fqdn.de>
Message-ID: <GCEDKONBLEFPPADDJCOECEHHDHAA.whisper@oz.net>

<ahem> and the documentation (something more then help(smtpd) please) for
smtpd is where? :P)

It's not in pythondocs for 2.2.1, unless it's well hidden. smtplib is
documented, but that's a client interface.

TMDA: BTDT - Like ASK, requires an external Linux/Unix based MTA. Thanks for
the pointer though :)

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Gerhard H?ring
> Sent: Friday, June 07, 2002 16:27
> To: Python-List at Python. Org
> Subject: Re: Python MTA
>
>
> * David LeBlanc <whisper at oz.net> [2002-06-07 15:44 -0700]:
> > Has anyone developed a simple/stupid Mail Transfer Agent using
> Python that
> > does't rely on any external mail agents (and thus would be platform
> > independent)?
>
> Python 2.2.1 (#1, Apr 29 2002, 11:12:54)
> [GCC 2.95.3 20010315 (release) [FreeBSD]] on freebsd4
> Type "help", "copyright", "credits" or "license" for more information.
> >>> import smtpd
> >>>
>
> Looks like so ;-)
>
> Also check out TMDA.
>
> Gerhard
> --
> mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
> web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
> public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
> reduce(lambda x,y:x+y,map(lambda
> x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From haakon at fast.no  Mon Jun 24 07:44:34 2002
From: haakon at fast.no (Haakon Brugaard)
Date: Mon, 24 Jun 2002 13:44:34 +0200
Subject: Python and remote sensing
References: <3D16E528.5040207@kfunigraz.ac.at>
Message-ID: <af70d7$o0p$1@news.broadnet.no>

On Mon, 24 Jun 2002 11:23:52 +0200, Siegfried Gonzi wrote:
> If Python can do it - okay; if not please say so (if Python is dead -
> okay; if Python lives - okay)!
> 

A cron job and a simple Python script could easily do the trick.

> 
> Thanks,
> S. Gonzi
 
H?kon


From klausman-un040602 at tuts.net  Wed Jun  5 09:12:27 2002
From: klausman-un040602 at tuts.net (Tobias Klausmann)
Date: 5 Jun 2002 13:12:27 GMT
Subject: python cookbook at oreilly's web site
References: <dCnL8.5$566.1250@news.ecrc.de>
Message-ID: <adl2nq$3hrm$1@ID-41653.news.dfncis.de>

Markus Jais <mjais at web.de> wrote:
> just found
> 
> http://www.ora.com/catalog/pythoncook/

Does this have any relation to ASPNs Python cookbook?
http://aspn.activestate.com/ASPN/Cookbook/Python

Wondering, 
Tobias



From aahz at pythoncraft.com  Thu Jun 20 15:43:29 2002
From: aahz at pythoncraft.com (Aahz)
Date: 20 Jun 2002 15:43:29 -0400
Subject: Variable Interpolation - status of PEP 215
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <aermlt$in5$1@panix1.panix.com> <7xu1ny4cvz.fsf@ruckus.brouhaha.com>
Message-ID: <aetb91$a7r$1@panix1.panix.com>

In article <7xu1ny4cvz.fsf at ruckus.brouhaha.com>,
Paul Rubin  <phr-n2002b at NOSPAMnightsong.com> wrote:
>
>What's the deal with the curly braces in 292?  That's not discussed
>at all in the PEP.  I don't see what's wrong with
>
>   print "My name is $name".sub()
>
>(which is similar to what 215 allowed) instead of
>
>   print "My name is ${name}".sub()

Consider

    "$numbers of items on sale"

versus

    "${number}s of items on sale"
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From mcfletch at rogers.com  Wed Jun 12 12:48:36 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Wed, 12 Jun 2002 12:48:36 -0400
Subject: Python Virtual File System: A spike test
References: <mailman.1023872902.2540.python-list@python.org> <Xns922B70C12302duncanrcpcouk@127.0.0.1>
Message-ID: <3D077B64.3030605@rogers.com>

discussion intermixed below...

Duncan Booth wrote:
> "Mike C. Fletcher" <mcfletch at rogers.com> wrote in
> news:mailman.1023872902.2540.python-list at python.org: 
...
> My initial thoughts:
> What is your use-case for equality testing? Especially given that it is 
> virtually impossible to implement this reliably.
> For example the normcase implementation in the standard library lowercases 
> all pathnames under windows. While this is useful most of the time, not all 
> filesystems are case insensitive, even on windows.

Equality-testing is useful for:

	Match previously known file against new file (unify resources on load to 
avoid textures getting loaded multiple times)
	Compare URLs which include: . .. ~ $ to determine if they point to the 
same file/directory/resource (unify URLs to avoid extra downloads)
	Filter duplicates from file lists (unify names in manifests to avoid 
including duplicate resources)
	Stop-processing-checks for exhaustive search (e.g. one that follows 
symbolic links)

it definitely is tricky, but it's one of the major things I want in the 
package (and, after all, if there's no tricky code in the system, why is 
someone going to use it instead of rolling their own).  I wind up with 
code to
do this stuff all over my packages otherwise, I'd like to localise it 
somewhere.

> Your class should support iterators: probably three flavours, one for 
> contained files, one for contained directories and one for everything. The 
> list, walk, and remove methods would probably then benefit from reusing the 
> iterators.

I started working with an iterator class, but the end result was that 
the iterator object was just in the way compared with the 
callback/entry-exit approach (for the things I was doing).  Room for 
both approaches (see Holger's posting, for example).

Implementing the functionality of the three-callback approach using 
iterators is ugly, because your iterators need to generate each 
directory twice, with flags to determine whether the directory is being 
entered or exited.

> The FileSystemPath seems a bit windows oriented with drive() and unc() 
> methods.

This was taken directly from the os.path module on my development 
machine.  Not sure what's available on unix-ish machines.  For the test, 
I was more interested in making sure the basic functionality was 
possible than I was with making a portable module.

> 
> The root method returns a drive such as 'c:'. This means that:
...
The drive method had a bug, it was supposed to return the root drive 
path, not just the drive name.  Fixed now.  Test added.

> There seems to be a problem with the implementation of canonical() when 
> working with some UNC pathnames (watch w2):
>  >>> from filepath import path
>  >>> w = path(r'\\osprey\work')
>  >>> w1 = path(r'd:\work')
>  >>> w2 = path('\\.\d:\work')
>  >>> w==w1
>  0
>  >>> w1==w2
>  0
>  >>> w.root()
>  FileSystemPath('\\\\osprey\\work')
>  >>> w1.root()
>  FileSystemPath('d:')
>  >>> w2.root()
>  FileSystemPath('d:')
>  >>> w2.canonical()
>  'd:\\d:\\work'
>  >>> w.canonical()
>  '\\\\osprey\\work'
> 
> I suspect however, this is a problem with the underlying os.path functions.
> 
Will need to add special-casing for "local machine" unc, though
I don't see how to do it with cases other than \\.\ given that the local 
machine name isn't immediately available as far as I know.

Thanks for the feedback,
Mike
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From ffjhenon at club-internet.fr  Mon Jun 24 13:36:52 2002
From: ffjhenon at club-internet.fr (Fabien =?iso-8859-1?Q?H=E9non?=)
Date: Mon, 24 Jun 2002 19:36:52 +0200
Subject: Embed terminal into a Tkinter application
Message-ID: <3D1758B4.8FA63EBE@club-internet.fr>

The title says it all.
Has someone already managed to include a Linux terminal into a Tkinter
GUI ?

Redirecting the output of a console into the application does not do
what I want.

Thanks for any help.

Fabien




From mgerrans at mindspring.com  Fri Jun  7 00:35:47 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Thu, 6 Jun 2002 21:35:47 -0700
Subject: How do you read unicode files?
Message-ID: <adpd7h$6sv$1@slb3.atl.mindspring.net>

How do you read in a unicode file and convert it to a standard string?

It seems that when you open a file and read it, what you get is a string of
single-byte characters.   I've tried all kinds of permutations of calls to
unicode(), decode(), encode(), etc. with different flavors of encoding
('utf-8',  'utf-16' and so on).

I could parse the data myself (skipping the initial two bytes and then every
other one -- I'm only working with ASCII in double byte format, so the high
order byte is always 0), but I imagine there must be a way to get the
existing tools to work.

What I want to be able to do is write a search and replace tool that will
work equally well on ANSI and Unicode (or double-byte) text files (without
changing the file type, of course)...




From vvainio at tp.spt.fi  Mon Jun 10 10:24:55 2002
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 10 Jun 2002 07:24:55 -0700
Subject: Configuration recommendations
References: <adra81$qf9$1@nntp2-cm.news.eni.net>
Message-ID: <ad496f8.0206100624.506e16a5@posting.google.com>

"Stuart D. Gathman" <stuart at bmsi.com> wrote in message news:<adra81$qf9$1 at nntp2-cm.news.eni.net>...
> I have an application - of the background daemon variety - written in
> python.  It has some configuration data like the following:

...

> What is standard practice for incorporating such data?  I would like it


I do it like this:

import new
config = new.module("config")
sys.modules['config'] = config

execfile('./mykillerapp.cfg',config.__dict__)


And then use the vars via the config module.

-- Ville


From mwh at python.net  Thu Jun  6 08:49:02 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 6 Jun 2002 12:49:02 GMT
Subject: array, list, performance...
References: <mailman.1023364388.4129.python-list@python.org>
Message-ID: <lky9dslghm.fsf@pc150.maths.bris.ac.uk>

gabor <gabor at realtime.sk> writes:

> hi,
> 
> how fast is list?
> 
> i mean accessing an item at an arbitrary position is O[n] or O[1]?

They are more like what tend to be called vectors in other languages,
in particular accessing an arbitrary element is O(1) (ignoring cache
effects and other little pesky details).

Cheers,
M.

-- 
  Gullible editorial staff continues to post links to any and all
  articles that vaguely criticize Linux in any way.
         -- Reason #4 for quitting slashdot today, from
            http://www.cs.washington.edu/homes/klee/misc/slashdot.html


From rnd at onego.ru  Wed Jun 12 01:59:13 2002
From: rnd at onego.ru (Roman Suzi)
Date: Wed, 12 Jun 2002 09:59:13 +0400 (MSD)
Subject: JIT Complier for python?
In-Reply-To: <8d3f4438.0206111905.7934664b@posting.google.com>
Message-ID: <Pine.LNX.4.33.0206120952520.12337-100000@rnd.onego.ru>

On 11 Jun 2002, d2002xx wrote:

>Is there any JIT Compiler for Python?

Hmmm... Python always compiles it's modules 'just-in-time'. If you do not
import some module, it will not be compiled. Also if compiled version
exists, it will be used (if it is not older than code version).

On the other hand, there are no Python compilers (to native machine code)
I know of. So presumable there are no JIT compilers too.

(I am trying to imagine what JIT compilation could do in Python.)

Sincerely yours, Roman Suzi
-- 
\_ Russia \_ Karelia \_ Petrozavodsk \_ rnd at onego.ru \_
\_ Tuesday, June 11, 2002 \_ Powered by Linux RedHat 7.2 \_
\_ "I can't use Windows. The cat ate my mouse." \_





From mwh at python.net  Fri Jun  7 12:44:09 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 7 Jun 2002 16:44:09 GMT
Subject: Python 2.2.1 vs. gcc 3.1?
References: <yu99vg8v6orb.fsf@europa.research.att.com>
Message-ID: <lkadq7kpid.fsf@pc150.maths.bris.ac.uk>

Andrew Koenig <ark at research.att.com> writes:

> I just installed gcc 3.1 on my Solaris 8 machine
> and am trying to rebuild Python 2.2.1

Ah.

> Installation is running into some trouble:

Yes...

> I do not believe that these problems happened with gcc 3.0
> 
> Any suggestions as to how to go about troubleshooting it?

I think this is a problem relating to how configure detected gnu
binutils.  You might want to grap configure from 2.2.1 cvs.

Cheers,
M.

-- 
  I'm not sure that the ability to create routing diagrams 
  similar to pretzels with mad cow disease is actually a 
  marketable skill.                                     -- Steve Levin
               -- http://home.xnet.com/~raven/Sysadmin/ASR.Quotes.html


From alf at logilab.fr  Mon Jun 10 05:53:09 2002
From: alf at logilab.fr (Alexandre Fayolle)
Date: Mon, 10 Jun 2002 09:53:09 +0000 (UTC)
Subject: [ANN] constraint-0.2
References: <mailman.1023468735.788.python-list@python.org> <slrnag50sd.gac.mlh@vier.idi.ntnu.no>
Message-ID: <slrnag8to5.dr.alf@virgo.logilab.fr>

Magnus Lie Hetland a ?crit :
> In article <mailman.1023468735.788.python-list at python.org>, Alexandre wrote:
>>How well does it perform?
>> * 8-queens problems solved in 15 seconds on my 1GHz PC
> 
> Interesting... The package looks really cool, but the performance in
> this example is truly abysmal. The simple queens.py found in the
> standard Python distribution solves the 8-queens problem (finds all
> solutions) in 0.11 seconds on an 800MHz PC. I guess maybe that't the
> price you have to pay for generality... On the other hand, a speedup
> factor of well over 100 seems to indicate a potential for
> streamlining... Maybe?

I do agree to all of this.

As you said, your comparison is not fair at all ;o), because 
Demo/scripts/queens.py  is targeted specifically at the n-queens
problem.

<tong-in-cheek mode="on">
Please note that if your main interest is solving the N-queens
problems, searchers have found an O(N) algorithm that will certainly
perform much better than queens.py in the Python distribution.
</tong-in-cheek>

Now, seriously.

This script uses a simple backtracking approach and shows that
it can be very efficient with some problems. The constraint package uses
a different approach (constraint propagation), which is supposed to be 
*at least* as efficient as backtracking, in terms of algorithmic
complexity.

So how comes the constraint package doesn't toast the backtracking
approach?
 * the example solution in the constraint package uses the constraint
 class MathematicConstraint which deals with arbitrary N-ary mathematic
 constraint. I can write an subclass optimized for binary constraints,
 which are the only thing we need for n-queens
 * the constraint package adds some overhead in processing constraint
 queues, and it is possible that some of the inner structures are not
 efficient. Blame this on the young age of the package.
 * maybe there is some horribly inneficient code in my package that I
 haven't seen yet. Be assured that as soon as I spot it, it will get
 refactored.

It's nice to have a comparison anyway. It helps to get an idea of what
our target is. Just as a sidenote, the very first (unreleased) version
of the package took 35 minutes to solve the 8 queens problem, so I
already got a 140x speed increase. Achieving another 140x speed increase
will certainly be harder, but it's nothing out of reach. 

And then, once constraint-1.0 is considered stable, maybe will go for a
rewrite in C.

> - Magnus (who hasn't looked at the internals, and who looks forward to
>   playing around with the package :)

Please do so, and join us on the python-logic mailing list.
http://lists.logilab.org/mailman/listinfo/python-logic

Alexandre Fayolle
-- 
LOGILAB, Paris (France).
http://www.logilab.com   http://www.logilab.fr  http://www.logilab.org
Narval, the first software agent available as free software (GPL).


From tim.one at comcast.net  Tue Jun  4 12:28:38 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 04 Jun 2002 12:28:38 -0400
Subject: self
In-Reply-To: <BI5L8.78$Vv1.4155449@newnews.cc.stevens-tech.edu>
Message-ID: <BIEJKCLHCIOIHAGOKOLHEELBDDAA.tim.one@comcast.net>

[Vojin Jovanovic]
> I have been working on a project trying to implement some advanced
> functionality.  Now, one thing that has been a constraint for me is the
> concept of self.  Why is it that within a class one has to use self. (or
> "any word". ) to refer to instance variables? Would it not be more
> logical to just use the name of the variable like it is for example in
> C++?

...
class C:
    ...
    def vojin(self):
        i = j + k

Which of i, j and k are instance variables?  Which are locals?  Which are
globals?  For all its virtues, Python isn't telepathic <wink>.





From matt at mondoinfo.com  Fri Jun  7 17:13:41 2002
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Fri, 07 Jun 2002 21:13:41 GMT
Subject: email.Message question.  including email attachments
References: <6bd9f01b.0206070744.64cb30c0@posting.google.com>
Message-ID: <slrnag28gg.1sf.matt@happy-hour.t-island.com>

On 7 Jun 2002 08:44:25 -0700, sameer <sameer_ at email.com> wrote:

Dear Sameer,

> I am trying to use the new (since Python 2.2) email module to send
> email with a pdf attachment.  I have tried various ways to send the
> pdf file, but always end up corrupting the file.

I think that the examples in the documentation at:

http://www.python.org/doc/current/lib/node389.html

are pretty good (but then I would). If you haven't looked at them, you
may find them helpful. If you have looked at them and something
remains obscure, let us know what it is.

Regards,
Matt


From K.Rdt at TU-Berlin.DE  Thu Jun 27 10:19:51 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Thu, 27 Jun 2002 16:19:51 +0200
Subject: crypt for windows?
Message-ID: <441A7JEE0J54TR6ZJF93XUGC764Z.3d1b1f07@FRITZweb>

---------------------------------------------------------------------
Hi
Is there a crypt for windows-python, so
I can manage some password-files?
		K at Rdt
---------------------------------------------------------------------







From tdelaney at avaya.com  Wed Jun 12 21:15:33 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 13 Jun 2002 11:15:33 +1000
Subject: Code checksums at compile time?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A383@natasha.auslabs.avaya.com>

> From: gat at jpl.nasa.gov [mailto:gat at jpl.nasa.gov]
> 
> My question is: is there a way to arrange for the module hash to be
> computed at compile time rather than run time?  This would 
> seem to be a
> straightforward adaptation of my current code, except that I 
> don't know
> how to get a handle on "the module that is currently being 
> compiled" (and
> its associated file).

You will need to write an import hook. Look up __import__ in the python
library reference (specifically, 2.3 Built-in Functions).

Tim Delaney




From emile at fenx.com  Thu Jun 27 12:05:01 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 27 Jun 2002 16:05:01 GMT
Subject: "multi" zip
References: <lrGS8.13892$bt.307@nwrddc04.gnilink.net>
Message-ID: <NIGS8.326499$352.35968@sccrnsc02>

Mark
> 1) instead of producing tuples, I'd like to produce lists.
>
> 2) I'd like to extend this to an arbitrary number of lists.  If it
were
> possible, I'd do zip(line[0], line[1], ..., line[n]) but this doesn't
seem
> to compile :).
>
> Is there a way to make zip() do this, or should I stop trying to
hammer with
> a wrench?
>


>>> line = [range(ii*10,11*10+5) for ii in range(5) ]
>>> map (list,zip(*line))


HTH,

--

Emile van Sebille
emile at fenx.com

---------



From edream at tds.net  Thu Jun 20 11:51:02 2002
From: edream at tds.net (Edward K. Ream)
Date: Thu, 20 Jun 2002 15:51:02 GMT
Subject: Inhibiting duplicate apps in Windows?
Message-ID: <3D11F9E8.DFFFC8E4@tds.net>

Hi All,

I have a pure Python app called Leo.  In Windows, I can open Leo by
double-clicking a .leo file.  I would like to do more: if Leo is open
and I double-click a .leo file, I would like the _already open_ copy of
Leo to open the second .leo file, rather than (permanently) opening a
second copy of Leo.

What I would like is for the second copy of Leo to detect that another
copy of leo.py is already running, send a message to the first copy and
then exit. I wonder if anyone knows how this might be done?

The C++ version of Leo works as follows:

1. It inhibits multiple copies of leo.py from running using a mutex.  I
think this part is easy because the Python library supports mutexes. 

2. It sends a message from the second copy of leo.py (the copy that will
shortly close), instructing the first copy of leo.py to open a file. 
This is the part that I'm not sure can be done in pure Python.  I don't
see anything applicable in section 22 (MS Windows Specific Services) of
the Python Library.  The C++ version of Leo does this by doing:

SendMessage(hWnd, WM_COPYDATA, 0, (LPARAM)&cds);

where hWnd is a handle to the previously open copy of Leo and cds is a
COPYDATASTRUCT containing the command line used to open the second copy
of Leo.

Perhaps this problem can be solved in Python using processes instead of
messages, and I don't see how.  Are there any demos that might be of
use?  Thanks!

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


From jcw at equi4.com  Wed Jun 19 10:47:53 2002
From: jcw at equi4.com (Jean-Claude Wippler)
Date: Wed, 19 Jun 2002 16:47:53 +0200
Subject: Tcl adept wanted for Stackless problem
References: <mailman.1024494035.10441.python-list@python.org>
Message-ID: <jcw-C9517F.16475319062002@news1.xs4all.nl>

Christian Tismer <tismer at tismer.com> wrote:

> My big question is:
> When does Tcl use C stack entries as globals, which
> are passed as function arguments to interpreter calls?
> This is of special interest when Python callbacks
> are invoked.
> The problem is, that I use stack slicing, which moves
> parts of the C stack away at some times, and I need
> to know when I am forbidden to do that. It is solved
> to some extent, but not all.

This may be caused by a CallFrame struct allocated on the stack in 
tclObjInterpProc (file generic/tclProc.c).  Am following up by email, 
have also CC'd Miguel Sofer on that email discussion.  The solution to 
this particular issue is within reach, IMO.

The point is that Tcl can access dynamic scope through its upvar and 
uplevel commands, but AFAICT only that CallFrame struct chain has any 
problems with the (terrific) stack-switching tricks you're playing.

> That's unfortunately not all.
> A friend is running the Tcl/Tk mainloop in one real thread,
> and stackless tasklets are running in another one (of course
> calling into Tcl). When tasklet switches are performed
> (again, this is moving stack contents around), there appear
> to be crashes, too.
> This kind of slicing should be allowed IMHO, since these
> are different contexts, which shouldn't interact at all.
> Are there any structures which are shared between threads
> that use Tcl/Tk? Something that may not disappear during
> some operations?

Evidently, Tcl better be compiled with thread support.  If it is, then 
that ought to do it.  Tcl threads run in parallel, with all shared data 
protected by mutexes.  Ought to be quite stable, given that AOLserver 
can use an unmodified threaded Tcl nowadays.

Not sure, perhaps it's the same issue...

One thought: Tcl 8.4b1 is due out any day now, might be a good way to 
make sure you're not hitting some recently resolved snag.

-jcw


From mhammond at skippinet.com.au  Thu Jun 20 18:44:41 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 20 Jun 2002 22:44:41 GMT
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <aer2fn$frc$1@peabody.colorado.edu> <3D1214B6.E9BB77D2@noaa.gov>
Message-ID: <3D125B79.6090903@skippinet.com.au>

Chris Barker wrote:

> On that note: does anyone know if there is a way to turn a Python script
> into something that acts like an application on Windows...without
> resorting to Py2exe and the like. What I want is something just like:

Add the .py extension to your PATHEXT environment variable, and you can 
execute "foo.py" at the command prompt.  Like much of Windows, this is 
based on the extension rather than file attributes.

Mark.



From hesterloli at hotmail.com  Tue Jun 18 04:12:53 2002
From: hesterloli at hotmail.com (George Hester)
Date: Tue, 18 Jun 2002 08:12:53 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com> <gYgP8.42356$GY.12361719@twister.nyroc.rr.com> <aekcug$7js8n$1@ID-114079.news.dfncis.de>
Message-ID: <9YBP8.44418$GY.13865843@twister.nyroc.rr.com>

I think you guys don't how to do it.

--
George Hester
_________________________________
"Jeremy Yallop" <jeremy at jdyallop.freeserve.co.uk> wrote in message
news:aekcug$7js8n$1 at ID-114079.news.dfncis.de...
> * George Hester
> | Well yours I haven't the slightest idea what you are getting at.
>
>   http://www.python.org/doc/current/tut/tut.html




From Chris.Barker at noaa.gov  Tue Jun 25 12:27:00 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 25 Jun 2002 09:27:00 -0700
Subject: SOLVED (Re: Python hits the spot)
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at> <af4vrl$ffe$0@216.39.172.122> <3D16DD5C.1050708@kfunigraz.ac.at> <uhdume2n1mhbba@corp.supernews.com> <3D171291.2060905@kfunigraz.ac.at> <mailman.1024950033.16332.python-list@python.org> <Xns9237E4F0DB10Dcliechtigmxnet@62.2.16.82> <3D17BFE1.B55C689F@engcorp.com>
Message-ID: <3D189972.D1477AA7@noaa.gov>

Peter Hansen wrote:

> > you have to admit that it was a nasty failure... is bad hardware your first
> > thought when your program fails? 

I have to say, I got pretty suspicious when he said that he had crashed
Linux. I have been using Linux for years and have never crashed it. When
I have had Python accidentally eat up more memory than the physical
memory of the system, it has ground to an almost halt, but if I was
patient, I could always find a way to kill the process and recover the
system. I suspected that he had in fact not crashed Linux, but didn't
have the patience or the knowledge to know how to recover the system.

By the way, if you are having a memory leak problem, you can use
"ulimit" to limit the amount of memory a process can use, so it won't
grind your machine to a halt.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From marklists at mceahern.com  Mon Jun 10 18:28:14 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 10 Jun 2002 17:28:14 -0500
Subject: using m2crypto to encrypt credit card numbers
Message-ID: <NCBBLFCOHHDIKCAFGCFBIEOGLDAA.marklists@mceahern.com>

I'm considering using M2Crypto (http://www.post1.com/home/ngps/m2/) to
encrypt credit card numbers.  My part of the problem starts when the credit
card arrives at the web server.  I plan to encrypt it with a public key and
send it to a database that the web server only has write access to.

A separate server will periodically use the private key to read the credit
card data and transmit it to the payment processor.

I've attached a demo below showing the basic outlines of how I plan to use
M2Crypto.  Key management is the part I'm most leery of.  Obviously, I'll
need to:

1.  Generate the key pair.
2.  Store the public key on the web server(s).
3.  Store the private key on the payment server.  Figure out how to unlock
it with the passphrase securely.

I'm interested to hear any and all feedback on the problem and/or solution
space:

1.  What libraries, if any, are you using to encrypt credit card numbers?
2.  How do you manage keys?

Specific technical questions that reveal the abundance of my ignorance:

1.  Do I need to salt the plaintext before I encrypt it or is that what
padding (M2Crypto.RSA.pkcs1_padding) is all about?
2.  Do I need to seed M2Crypto with random data before generating keys -- or
does it take care of that via /dev/[u]random?

Thanks,

// mark

#! /usr/bin/env python
# --------------------------------------------------------------------------
---
# $Id: demo.py,v 1.2 2002/06/10 21:31:20 mmceahern Exp $
# --------------------------------------------------------------------------
---

__doc__ = \
"""
This is just a very quick hack to see if I can understand M2Crypto.

When this runs, you'll be prompted to enter and verify a passphrase.
Use the single letter p (or change passphrase_callback below).  I
haven't figured out how to override the genkey_callback to skip
passphrase prompting--that doesn't sound like a good idea anyway.

The passphrase_callback is for decrypting, which we can override.

I think the use of pkcs1_padding suggests that I may not need to worry
about salt--that may be something that the RSA algorithm(s) in openssl
take care of for us.

I think this shows that I can store the binascii.b2a_hex version of
the binary ciphertext in the database.

Outstanding questions:

1.  Is this the right way to generate a key?
2.  Do I have to explicitly seed the system with random data?
3.  How will we store the passphrase and feed it to the decryption routine?

"""

# --------------------------------------------------------------------------
---
# Third party imports
# --------------------------------------------------------------------------
---
import M2Crypto

# --------------------------------------------------------------------------
---
# Standard library imports
# --------------------------------------------------------------------------
---
import binascii

# --------------------------------------------------------------------------
---
# Helper methods
# --------------------------------------------------------------------------
---
def passphrase_callback(v, prompt1='Enter passphrase:',
                        prompt2='Verify passphrase:'):
    """passphrase_callback(verify, prompt1, prompt2) --> passphrase

    Return the passphrase (we're overriding util.passphrase_callback).
    """
    return "p"

def debug_print(description, text):
    """debug_print(description, text)

    Print the description and text.
    """
    print "<%s>" % description
    print text
    print "</%s>" % description
    print

def generate_key(privkey_file, pubkey_file, key_length):
    e = 65537
    new_key = M2Crypto.RSA.gen_key(key_length, e)
    new_key.save_key(privkey_file)
    new_key.save_pub_key(pubkey_file)

def encrypt(pubkey_file, plaintext):
    """encrypt(pubkey_file, plaintext) --> hexstring

    Encrypt plaintext and return ascii version of ciphertext.
    """
    pub_key = M2Crypto.RSA.load_pub_key(pubkey_file)
    encrypted = pub_key.public_encrypt(plaintext,
M2Crypto.RSA.pkcs1_padding)
    return binascii.b2a_hex(encrypted)

def decrypt(privkey_file, ciphertext):
    """decrypt(privkey_file, ciphertext) --> string

    Decrypt ciphertext (assume hex string) and return the decrypted string.
    """
    encrypted = binascii.a2b_hex(ciphertext)
    priv_key = M2Crypto.RSA.load_key(privkey_file, passphrase_callback)
    decrypted = priv_key.private_decrypt(encrypted,
M2Crypto.RSA.pkcs1_padding)
    return decrypted

def main():
    privkey_file = "priv.pem"
    pubkey_file = "pub.pem"
    key_length = 1024
    generate_key(privkey_file, pubkey_file, key_length)

    plaintext = "5105105105105100"
    encrypted = encrypt(pubkey_file, plaintext)
    decrypted = decrypt(privkey_file, encrypted)

    assert plaintext == decrypted

    debug_print("plain text", plaintext)
    debug_print("cipher text", binascii.b2a_hex(encrypted))
    debug_print("decrypted text", decrypted)

# --------------------------------------------------------------------------
---
# Main
# --------------------------------------------------------------------------
---
if __name__ == "__main__":
    main()

-





From K.Rdt at TU-Berlin.DE  Wed Jun 26 01:11:49 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Wed, 26 Jun 2002 07:11:49 +0200
Subject: mail-problem over t-online
In-Reply-To: <slrnahdh4s.123.usenet@fohnhab.minus1.de>
Message-ID: <GBA7VWSGA72XVHEXWWTL93UR7632UP.3d194d15@FRITZweb>

Am 24.06.02 09:05:00, schrieb usenet at minus1.de (Konrad Anton):

>telnet mailszrz.zrz.tu-berlin.de 25
>HELO meinhostname.meinedomain.de
>MAIL FROM: soundso at tu-berlin.de
>RCPT TO: blabla at tu-berlin.de
>DATA
>Dies ist ein Test

---------------------------------------------------------------------
Hi
Danke. With putty-tunneling it's working.
		K at Rdt
---------------------------------------------------------------------







From jadestar at idiom.com  Sun Jun 16 20:21:17 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 17 Jun 2002 00:21:17 GMT
Subject: Writing a simple linux daemon type thing...
References: <K9eO8.393$LN6.261@newsread1.prod.itd.earthlink.net>
Message-ID: <aeja1t$2ebi$1@news.idiom.com>

Jones <jones at jones.jones> wrote:

> Hi everyone,
> I hope this is a simple question for you guys.  I'm new to Python and Linux.
> I'm wondering if Python (and this group) can help me.
> I want a program to always be running in the background that, connects to
> PostgreSQL every few seconds, queries a table, and for each record it finds,
> shells out to the system to execute program.

> My research indicates that "nohup" might do the trick?  So would it be as
> simple as writing the program in Python and then puting something like
> "nohup python myprog" in the Linux startup sequence (wherever that may be)?

 Technically a daemon is different than a nohup'd background process.
 The differences have to do with signal handling and session/process groups.

> That's my primary goal.  Now... assuming I get past that part, I may want to
> take it a few steps further later on.  I'd be happy if you guys could take a
> stab at telling me if these are possible and any tips too...
> a) Make it multi-threaded...   As I said before, the number of times it
> executes the external program equals the number of records it finds in the
> table.  It would be nice if I could execute the program multiple times
> concurrently in a *controlled* manner (depending on my system resources)
> instead of having to wait for each one to be done before starting another
> one.

 It is unlikely that you need to be "multi-threaded" for at least
 two reasons.  First the time you spend establishing new connections
 will probably dominate the program, it might make sense for your
 deamon (or background process) to keep an open connection to the 
 db rather than closing and reopening it every few seconds.  If the
 interval was several minutes or was to be event driven then I might
 offer a different suggestion.   It if was responding to signals or
 especially to socket connection then I might see a case for a 
 multi-threaded set of connections (in a pool).

> b) Make it fail-safe.  I would like to figure out how I could re-launch the
> program if it was to crash.

 The most straightforward way to do this on a Linux or SysV UNIX system
 is to have your program launched under a respawn directive from 
 the /etc/inittab (but this requires sysadmin (root) intervention to
 configure).  Another way is to use a program like DJ Bernstein's 
 "supervise" (which is a daemon written in C that will launch and 
 "supervise" another process, respawning it as necessary.

> P.S.  You're probably wondering why I don't just use a trigger.  One reason
> is that I don't think I can execute an external program from PostgreSQL and
> I don't know C to write a wrapper function.  Second reason is that because I
> am running an external program (out of my "control"), I'm afraid it might
> crash, etc and that record will never get processed.

> Thank You!!!

 Searching google on "python daemon os.setsid" if find the following:

http://mlf.linux.rulez.org/Archivum/l-code-l-199911/msg00060.html

 Which contains some text in a language that I don't recognize
 (Swedish?, Slavic?, who knows) and some code in Python:

from posix import geteuid 
import os import sys 

def daemon(nochdir = 0,noclose = 0): 
    """Runs server as INIT daemon."""  
    if geteuid() != 0:  
      print "You should be root to run an INIT daemon"  #
      sys.exit() 
    if os.fork(): os._exit(0)
    os.setsid()
    if nochdir == 0: os.chdir("/")
    if noclose == 0: 
        fp = open("/dev/null", "rw")
        sys.stdin = sys.__stdin__ = fp
        sys.stdout = sys.__stdout__ = fp
        sys.stderr = sys.__stderr__ = fp
        del fp
    if os.fork(): os._exit(0)

 This looks reasonable.  In your case you don't want to run this
 as root so you can reduce that a bit to:


import os import sys 

def daemon(nochdir = 0,noclose = 0): 
    """Runs server as INIT daemon."""  
    if os.fork(): os._exit(0)
    os.setsid()
    if nochdir == 0: os.chdir("/")
    if noclose == 0: 
        fp = open("/dev/null", "rw")
        sys.stdin = sys.__stdin__ = fp
        sys.stdout = sys.__stdout__ = fp
        sys.stderr = sys.__stderr__ = fp
        del fp
    if os.fork(): os._exit(0)

 First this code spawns a child and has the parent
 commit suicide; so the child is detached from the 
 immediate parent.  Then it makes itself the session
 leader of a new process group (divorcing it from some
 TTY signal handling that it inherited).  BTW: the setsid
 call may *require* us to do the if os.fork(): os._exit(0)
 --- I got an error when I tried to skip that part in 
 one text and the (Linux) man page for setsid(2) suggests
 that it is required. 
 
 Then it puts itself in the root directory (so the process 
 doesn't interfere with umounting of filesystems by keeping a 
 directory "busy") and replaces its file descriptors with 
 some on /dev/null.  I'd quibble about this (they're all 
 opened wih "rw" which is technically wrong, stdin should be
 in 'r' mode and stdout/stderr should be 'w').  I might
 also I think the extra 
 do something more like: 


#!/usr/bin/env python2.2
def daemonize():
	"""Become a Linux/UNIX daemon"""
	import os,sys
	os.chdir('/')
	if os.fork(): os._exit(0)
	os.setsid()
	sys.stdin  = sys.__stdin__  = open('/dev/null','r')
	sys.stdout = sys.__stdout__ = open('/dev/null','w')
	sys.stdout = sys.__stderr__ = os.dup(sys.stdout.fileno())

def daemonize_log_demo():
	"""Demonstrate deamonize() with trivial syslog output""" 
	import time,syslog,os
	log = syslog.openlog("JimDaemon[%d]" % os.getpid())
	t=0
	while 1:
		t+=1
		syslog.syslog(syslog.LOG_INFO,"Interval %d  reached" % (t) )
		time.sleep(5)

if __name__=='__main__':
	daemonize()
	daemonize_log_demo()

  Obviously you'd replace daemonize_log_demo with a function that
  opened your db connection, periodically performed your queries
  spawned off your child process and (preferably) read back the 
  child process exit results.

  You might want to avoid os.system() and manage your own 
  child processes directly with os.fork() and os.execv() or
  os.execl().  You can then perform your own process management,
  spawning multiple concurrent processes when necessary, keeping 
  a list of outstanding/running processes, and reaping their results
  with os.waitpid().  You might also want to use a signal handler
  so that your child processes will notify you when they've exited
  (SIGCHLD) and your process reaping can then be event driven.

  Under those conditions you could also implement a timeout for 
  your children.  If a process doesn't exit in a reasonable time
  you can kill it and log the error (via syslog or some other 
  mechanism).  Naturally you could report errors via e-mail or
  some instant messaging mechanism if you wanted.  Actually at
  that point you can do pretty much anything that you could 
  concievably code up.

  Perhaps I'll write the write of this (a PostgreSQL query/poll
  and some process management and error handling functionality)
  later.  For now I have a pool party to get to.




From xah at xahlee.org  Sat Jun 15 03:25:37 2002
From: xah at xahlee.org (Xah Lee)
Date: 15 Jun 2002 00:25:37 -0700
Subject: you guys needs a logo
Message-ID: <7fe97cc4.0206142325.4df3a174@posting.google.com>

you guys need a logo.

the light-bulb is quite lame as far as logo goes.

I suggest that the O'Reilly's python book cover a good choice. It
should not be something funny and cute.

The MacPython's some 16 Ton weight block is also a good logo.

Perhaps you guys can run a logo competition.

 Xah
 xah at xahlee.org
 http://xahlee.org/UnixResource_dir/complang.html#python


From never at mind.info  Wed Jun  5 20:04:26 2002
From: never at mind.info (G. Willoughby)
Date: Thu, 6 Jun 2002 01:04:26 +0100
Subject: win32: popen works, popen3 causes hang
References: <930dfcf0.0205301058.6917c599@posting.google.com>
Message-ID: <adm8u2$8cu$1@news7.svr.pol.co.uk>

i think os.popen3 returns a tuple of file objects, so to use this command
correctly it should be coded like this:

childstdin, childstdout, childstderr = os.popen3("spectrum.exe -file
script.tcl")

to read the error stream use childstderr.read() etc..

--G. Willoughby

"Alex Eggenberger" <eggy at island.net> wrote in message
news:930dfcf0.0205301058.6917c599 at posting.google.com...
> I am trying to launch a program called leonardo spectrum from python
> and retrieve the stdout and stderr that results.  spectrum.exe is
> strictly command line under win32.  Here is what I have encountered:
>
> launching spectrum.exe using the following call:
> output = os.popen('spectrum.exe -file script.tcl')
>
> The result is as expected, the program runs and I use output.read() to
> get stdout.  I need some info from stderr.  So I would like to use
> popen3 like this:
> output = os.popen3('spectrum.exe -file script.tcl')
>
> This causes spectrum.exe to hang.  I can tell it is hung on
> spectrum.exe because I can see the process persists indefinitely in
> Task Manager.
>
> Is there a difference in the popen and popen3 operate that could be
> causing this problem.  Everything else except the popen call is
> identical in the two cases.
>
> Any help or suggestions would be appreciated.
>
> Alex




From gerhard.haering at gmx.de  Wed Jun 26 04:05:08 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 26 Jun 2002 08:05:08 GMT
Subject: Static py2exe?
References: <mailman.1025070513.3741.python-list@python.org>
Message-ID: <slrnahitdj.f8.gerhard.haering@lilith.my-fqdn.de>

Klaus Reinhardt wrote in comp.lang.python:
> Is there a possibility to make a 'static binary',
> so that the *.exe hasn't to rely on python2.dll?

Without recompiling Python yourself: no.  With recompiling Python
yourself: yes, but you gotta do it all yourself.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From pyth at devel.trillke.net  Sun Jun  9 05:17:15 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 9 Jun 2002 11:17:15 +0200
Subject: Correct code/unit testing (was Re: Efficient python programming...)
In-Reply-To: <3D02FB3B.CF75E925@engcorp.com>; from peter@engcorp.com on Sun, Jun 09, 2002 at 02:52:43AM -0400
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com> <834rgd5isv.fsf@panacea.canonical.org> <3D02FB3B.CF75E925@engcorp.com>
Message-ID: <20020609111715.K6609@prim.han.de>

Peter Hansen wrote:
> Kragen Sitaker wrote:
> > I think (and extensive experimentation has
> > shown) that careful reasoning and code reviewing is necessary and
> > sufficient to write correct code.
> 
> Whoa!  "extensive experimentation has shown"??  I would suggest that
> what extensive experience (I change the word deliberately) has shown
> many of us is that careful reasoning and code reviewing, no matter
> how extensive, will *always* leave bugs in the code.  Have you never
> spent a large amount of time carefully designing and coding a wonder
> little program, and checked it over very carefully, even with a peer,
> and later found a bug in it?  Maybe you're just a better programmer
> than I, and many others, but I would say this inevitable outcome is 
> the source of the term "egoless programming" -- if you have a big
> ego, you'll be devastated by the repeated proof that you will never,
> ever get it right the first time, or even after lots of reviewing.

I mostly agree and there is even another point that makes tests
important: making sure that later changes do not break
the basic semantics of your program. You would have to do another
code review of the *whole* part of a program to gain confidence, again.
This method doesn't scale well with the number of changes to a program.
Only if you ever touch code once in a year would you affort to gain 
confidence by 'closely examining the code'. Most often, people just
assume that their clever modification will not do any harm...

I must admit that writing tests for C++ or Java code really 
is a verbose pain.  But with python it's almost fun.

just my 2 test cent,

    holger




From gminick at hacker.pl  Sat Jun  1 02:17:48 2002
From: gminick at hacker.pl (Wojtek Walczak)
Date: Sat, 1 Jun 2002 06:17:48 +0000 (UTC)
Subject: Generating unique numbers?
References: <3CF7EBC8.4080802@verio.net>
Message-ID: <slrnafgpij.b1f.gminick@hannibal.localdomain>

Dnia Fri, 31 May 2002 15:31:52 -0600, VanL napisa?(a):
> Is there a good algorithm for generating unique numbers 
> (ints, specifically) that can be used as object identifiers?
Try: http://underground.org.pl/gminick/pliki/veryrand.py

-- 
[ Wojtek gminick Walczak ][ http://hacker.pl/gminick/ ]
[ gminick (at) hacker.pl ][ gminick (at) klub.chip.pl ]



From theller at python.net  Thu Jun 13 10:39:51 2002
From: theller at python.net (Thomas Heller)
Date: Thu, 13 Jun 2002 16:39:51 +0200
Subject: filter2
References: <mailman.1023975618.6474.python-list@python.org>
Message-ID: <aeaava$56l8p$1@ID-59885.news.dfncis.de>

"Oleg Broytmann" <phd at phd.pp.ru> wrote in message news:mailman.1023975618.6474.python-list at python.org...
> Hello!
>
>    I want to have new python builtin - filter2. It is exactly like filter,
> but it returns also a list of items that didn't pass the test.
>

IIRC, in Smalltalk they were named 'select' (like filter), and 'reject' (like your filter2).
IMO these are much better names then filter and filter2.

Thomas




From dkirby at orchestream.com  Fri Jun 28 12:45:37 2002
From: dkirby at orchestream.com (Dave Kirby)
Date: Fri, 28 Jun 2002 17:45:37 +0100
Subject: Python COM problem - accessing properties with optional parameters
Message-ID: <1025282888.585264@lave>

I am using Python to access the Microsoft Accessibility interface.  I have
got as far as getting an IAccessible object, but am having problems calling
some of its methods.

The problem is that some of the methods are declared as properties with
[propget] or [propput] in the IDL, but also have optional parmeters.  For
example:

  [hidden, propget, id(DISPID_ACC_NAME)] HRESULT accName([in, optional]
VARIANT varChild,[out, retval] BSTR* pszName);

if I have a an object called obj I can access the property accName like
this:

>>> obj.accName
u'Some Object Name'
>>>

which is all fine and good.  However there are times when I need to call the
accName method with an int parameter, e.g.

>>> obj.accName(0)
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
TypeError: object of type 'unicode' is not callable
>>>

this obviously doesnt work, since it treating it as a property and returning
a string, then trying to call that string as a function.

Is there any way to access the function without it being treated as a
property?  I tried calling the IDispatch functions directly, but that didn't
work:

>>> obj.GetIDsOfNames('accName')
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
  File "<COMObject AccessibleObjectFromWindow>", line 2, in GetIDsOfNames
com_error: (-2147352573, 'Member not found.', None, None)

I am stumped on this, so any pointers would be appreciated.

I am using ActivePython build 2.1.212, by the way.

Also, are there any mailing lists or discussion groups specifically for
COM/Windows Python programming anywhere?

Regards,


                Dave Kirby





From urnerk at qwest.net  Mon Jun 17 15:59:16 2002
From: urnerk at qwest.net (Kirby Urner)
Date: Mon, 17 Jun 2002 15:59:16 -0400
Subject: If that interests you...
References: <3D0E3226.48E6D338@club-internet.fr>
Message-ID: <pan.2002.06.17.19.59.14.8910.3048@qwest.net>

On Mon, 17 Jun 2002 15:01:58 -0400, Fabien H?non wrote:

> I have written an editor for POV-RAY ( the raytracer) using Tkinter that
> 
> features syntax highlighting.
> It fires/stops up raytraces, save,....
> 
> It can be found at :
> http://pyvon.sourceforge.net
> 
> Fabien

Hey Fabien, this is really cool!  I just downloaded 
your work, which installed perfectly and runs smoothly
(I only had to repoint it to /usr/bin/povray instead
of /usr/local/bin/povray -- but your interface makes
that easy).

As someone used to the slick Povray GUI in Windows, 
I am very happy to have something similar in Linux,
and the fact that it's a Python application is very
appropriate for me, since a lot of my Povray text
files are outputs from Python programs, as per
http://www.inetarena.com/~pdx4d/ocn/numeracy0.html [1]

Kirby

(Mandrake distro 8.2, KDE3, 
Persistence of Vision(tm) Ray Tracer Version 3.1g.Linux.gcc

[1] unrelated note:  I notice this web page renders 
very oddly in Konqueror 3.0.1 but looks the way I 
intended in Mozilla 1.0 -- I wonder why.



From jonathan at onegoodidea.com  Wed Jun 26 12:57:09 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Wed, 26 Jun 2002 17:57:09 +0100
Subject: Python 2.2 __slots__ Bug?
In-Reply-To: <mailman.1025105074.31122.python-list@python.org>
References: <mailman.1025089474.17780.python-list@python.org> <B93F7990.CECB%jonathan@onegoodidea.com> <mailman.1025105074.31122.python-list@python.org>
Message-ID: <B93FB0F5.D00D%jonathan@onegoodidea.com>

On 26/6/2002 16:23, in article
mailman.1025105074.31122.python-list at python.org, "Glyph Lefkowitz"
<glyph at twistedmatrix.com> wrote:

> Hmm.  Sprinkling print statements about shows it dying on the 'del' (this is
> how I encountered the bug originally, I was trying to time the [de]allocation
> of large numbers of objects under 2.1 vs. 2.2), but the stacktrace is still
> the
> same; is there some difference between the way allocation works on OS X
> vs. Linux? (I assume if you can run OS X we're using the same hardware.)
> 
>> This seems to fail for me at 5430 objects. I couldn't get the allocation to
>> crash at all (or at least up to a million objects, after which I got bored).
> 
> Yep, same here.  Allocation is no problem.

I'm not sure about allocation differences, but if I up the stacksize limit
then I can run the example up to much higher numbers of objects. I think
it's basically a flaw in the deallocation mechanism. I don't understand
enough about the __slots__ stuff to know why it should only affect that.

No-one else seems to have noted it as a known bug, so I'd guess it's fair to
log it to Sourceforge as a new one (besides which there's no harm in
duplication). If you can't get it logged, let me know and I'll have a go
myself.

Jonathan



From cancelmi at cli.di.unipi.it  Thu Jun  6 18:51:21 2002
From: cancelmi at cli.di.unipi.it (Gianluca Cancelmi)
Date: Thu, 6 Jun 2002 23:51:21 +0100
Subject: Python as a Game Dev tool on GBA
Message-ID: <3cffe769$1_1@nnrp1.news.uk.psi.net>

Hi,

I've been using python for a lot of little tools for Gameboy Advance.

At the moment I'm using it to convert multiple bitmaps into a large shared
tile set and multiple maps.

I'm using PIL to load the bitmaps and generate the tiles.

At the moment I'm having a little bit of trouble with memory usage.

I'm cutting the 8x8 pixels tiles using Image.crop().
I'm creating 3 other versions of the current tile for Horizontal, Vertical
and H+V flipping.
I convert the tiles in strings and use this string as a key for the tile
dictionary. The value for that key is an object (TileImage).

That's the source code:

class TileImage:
    def __init__(self, image, tile_num, ref_counter = 0):
        # Init the members
        self.image = image
        self.tile_num = tile_num
        self.ref_counter = ref_counter

    def inc_reference(self):
        # Increase the reference number
        self.ref_counter += 1

    def dec_reference(self):
        # Decrease the reference number

        # Check if it's 0
        if (self.ref_counter > 0):
            self.ref_counter -= 1

    .................................
    .................................
    # Main part of the loop

    # We have a unique tile set for all layers but a separate map (a map for
each layer)
    tile_set = {}
    corr_num_tile = {}
    maps = []
    total_tile_counter = 0
    tile_unique_counter = 0

    # Process each image (img_list is a list of images loaded with PIL)
    for img in img_list:
        # Map is the current map
        map = []
        tile_counter = 0

        s_x = img.size[0]/8
        s_y = img.size[1]/8

        # Generate tile set and map
        for y in range(0, s_y):
            for x in range(0, s_x):
                # Generate the tile
                # Normal version of the tile
                tile = img.crop((x*8, y*8, x*8+8, y*8+8))

                # Convert to string for fast compare
                tile_string = tile.tostring()
                tile_v_string =
tile.transpose(Image.FLIP_TOP_BOTTOM).tostring()
                tile_h_string =
tile.transpose(Image.FLIP_LEFT_RIGHT).tostring()
                tile_vh_string =
tile.transpose(Image.FLIP_TOP_BOTTOM).transpose(Image.FLIP_LEFT_RIGHT).tostr
ing()

                # Convert to 16x16 colors mode for palette reduction
                tile_16_string = module_string_fast(tile_string)
                tile_16_v_string = module_string_fast(tile_v_string)
                tile_16_h_string = module_string_fast(tile_h_string)
                tile_16_vh_string = module_string_fast(tile_vh_string)

                # Add the tile in the tile set (check for H,V,H+V)
                if (tile_set.has_key(tile_16_string)):
                    # Set current value for map
                    current_key = [tile_set[tile_16_string].tile_num,
get_pal_idx(tile_string, 16), 0, 0]
                elif (tile_set.has_key(tile_16_v_string)):
                    # Set current value for map
                    current_key = [tile_set[tile_16_v_string].tile_num,
get_pal_idx(tile_v_string, 16), 1, 0]
                elif (tile_set.has_key(tile_16_h_string)):
                    # Set current value for map
                    current_key = [tile_set[tile_16_h_string].tile_num,
get_pal_idx(tile_h_string, 16), 0, 1]
                elif (tile_set.has_key(tile_16_vh_string)):
                    # Set current value for map
                    current_key = [tile_set[tile_16_vh_string].tile_num,
get_pal_idx(tile_vh_string, 16), 1, 1]
                else:
                    # Add a new tile into the dictionary
                    tile_set[tile_16_string] = TileImage(tile,
tile_unique_counter)
                    corr_num_tile[tile_unique_counter] = tile_16_string

                    # Set current value
                    current_key = [tile_unique_counter,
get_pal_idx(tile_string, 16), 0, 0]

                    # Increment unique tile counter because we found a
unique tile
                    tile_unique_counter += 1

                # Add the tile into the map
                map.append(current_key)

                # Increment tile counter
                tile_counter += 1

        # Save the map in the map array and add the tiles to the
total_tile_counter
        total_tile_counter += tile_counter

        # Save the map for this layer
        maps.append(map)

    .................................
    .................................

The problem is that is taking too much memory and I get something like
memory error in python.
The test images are 9600x512 pixels each (4 images).

My PC has 512M of RAM and I'm running the script in win2000.

Is there any way to debug memory usage?

Anyone has an idea of what I'm doing wrong?

Any help would be really appreciated.

Cheers,

Gianluca




From radix at twistedmatrix.com  Sun Jun  9 11:33:38 2002
From: radix at twistedmatrix.com (Christopher Armstrong)
Date: 09 Jun 2002 11:33:38 -0400
Subject: Source formatting for long expressions.
In-Reply-To: <m3it4srhq5.fsf@mira.informatik.hu-berlin.de>
References: <mailman.1023570972.27430.python-list@python.org>
	<m3it4srhq5.fsf@mira.informatik.hu-berlin.de>
Message-ID: <874rgcxy99.fsf@twistedmatrix.com>

>>>>> "mvl" == Martin v Loewis <martin at v.loewis.de> writes:

    mvl> As you can see, it tries to align further arguments to a function
    mvl> together with the opening parenthesis. I would assume that any other
    mvl> auto-formatters use the same style, so you won't get any "sane"
    mvl> indentation until you drop the nesting level (perhaps in favour of
    mvl> method invocations).


See http://twistedmatrix.com/users/carmstro/pypersist.py.txt for a hand-written
sample of what I'm looking for. It's basically what you quoted with a lot more
linebreaks, and instead of aligning to left-parenthesis, I just indent two
spaces for each depth-level.

I'm probably going to end up doing this myself, and I thought of something
while reading your message. I could have two steps for formatting: template-
generation and indentation. the first step would generate something like a
repr() of all of my objects, but with \ns and \ts embedded at strategic
locations. Then I could replace all \ts with current_indent_level, calculating
current_indent_level by counting (s, {s, and [s. I'm going to try this out now.
:-)

-- 
                                Chris Armstrong
                         << radix at twistedmatrix.com >>
                http://twistedmatrix.com/users/carmstro.twistd/





From jepler at unpythonic.net  Tue Jun 25 09:42:47 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 25 Jun 2002 08:42:47 -0500
Subject: Embedding in Tkinter (reprise)
In-Reply-To: <1907a5e4.0206250504.4b417e9a@posting.google.com>
References: <1907a5e4.0206240803.6ca0a4c@posting.google.com> <mailman.1024936355.12699.python-list@python.org> <1907a5e4.0206250504.4b417e9a@posting.google.com>
Message-ID: <20020625134246.GB4435@unpythonic.net>

You first need to make sure your Python's tkinter is linked with blt.
This can be a bit of a mess ...

Then, guessing, you could start by trying to define a Container class
in Python, something like
    class Container(Widget):
	    def __init__(self, master=None, cnf={}, **kw):
		    Widget.__init__(self, master, 'blt::container', cnf, kw)

Now, the program would look something like
    t = Tk()
    t.tk.call("package", "require", "BLT")
    f = frame(t)
    os.system("xclock &")
    c = Container(f)
    f.pack(fill="both", expand="yes")
    c.pack(fill="both", expand="yes", padx=".1i", pady=".1i")
    c.configure(relief="raised", bd=2, name="xclock")
    t.mainloop()

Jeff




From emile at fenx.com  Sun Jun  2 10:40:10 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 02 Jun 2002 14:40:10 GMT
Subject: How to delete file?
References: <add9ai$107clv$1@ID-49758.news.dfncis.de> <FXpK8.90000$352.4832@sccrnsc02>
Message-ID: <e7qK8.178374$Po6.345396@rwcrnsc52.ops.asp.att.net>

(forgot my sample overwrite)

>>> import string
>>> open(r'c:\test.txt', 'w').write(string.letters)
>>> fd = open(r'c:\test.txt', 'r+')
>>> fd.seek(12)
>>> fd.write('hiHI!')
>>> fd.flush()
>>> fd.close()
>>> fc = open(r'c:\test.txt', 'rw').read()
>>> fc
'abcdefghijklhiHI!rstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
>>>

--

Emile van Sebille
emile at fenx.com

---------



From tismer at tismer.com  Tue Jun 18 10:38:26 2002
From: tismer at tismer.com (Christian Tismer)
Date: Tue, 18 Jun 2002 16:38:26 +0200
Subject: streaming audio
References: <mailman.1020652989.23833.python-list@python.org> <cb4ba455.0205070702.53ee32c7@posting.google.com>
Message-ID: <3D0F45E2.9010700@tismer.com>

Bill Tate wrote:
> Rob Brown-Bayliss <rob at zoism.org> wrote in message news:<mailman.1020652989.23833.python-list at python.org>...
> 
>>Hi, 
>>
>>I am looking for some information on the basics of audio streaming, or
>>any sort of data streaming really.

...

> A typical model used in network programming is to use one thread per
> socket.   For you situation, you may wish to consider an alternative
> to threads that is lighter-weight, can be instantiated quickly and
> which can be stopped and restarted with ease.  If you will be
> processing many client requests for different audio files, you might
> check out the story behind the reincarnation of stackless (and
> continuations in particular).  Continuations offer some advantages
> well suited to media streaming.  They are extremely lightweight, can
> be quickly instantiated, there is no "real" limit on their numbers,
> they can be stopped and restarted with ease, and work extremely well
> with sockets.  I can't speak to how the reinvented stackless works
> with python 2.2 but others may have suggestions on other alternatives.
>  At the very lest, its would be worth familarizing yourself with the
> background on stackless and why it was created.

Saw this message a little late...

Stackless is going to support async I/O for a wide range of
operating systems. This is the next major goal to achieve.
Currently, we are trying to design a most protable, but
also most effective solution. Using kqueue where it is
available, falling back to select where necessary...

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From derek at wedgetail.com  Fri Jun 21 06:43:39 2002
From: derek at wedgetail.com (Derek Thomson)
Date: Fri, 21 Jun 2002 10:43:39 GMT
Subject: I'd give up Perl tomorrow if only...
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au> <3D12F6FC.3000608@wedgetail.com> <7xd6ukex9m.fsf@ruckus.brouhaha.com>
Message-ID: <3D130359.1050906@wedgetail.com>

Paul Rubin wrote:
> Derek Thomson <derek at wedgetail.com> writes:
> 
>>Among other things, it's one of the reasons I'm tinkering with a Perl ORB.
> 
> 
> Wow, that sounds cool.
> 
> 
>>Yes, it's stop-gap, but when you already have a non-trivial Perl
>>module, either on CPAN or in-house, it can save a lot of
>>rewriting. What we *really* need to do is dynamically create CORBA
>>interfaces for Perl modules, but that might be impossible given that
>>Perl is like Python in that module and class interfaces aren't
>>pre-declared :(
> 
> 
> Why not some more dynamic remote object protocol instead of messing
> with Corba?

Like what, exactly? There's nothing that I've heard of.

Anyway, you've just described CORBA :)

It supports "dynamic [interfaces and invocations]", "remote [calls]", 
and "objects". Yes you have to define IDL, but given the nature of Perl 
you can't avoid laying down the interface spec *somehow*, so it might as 
well be with something well-defined and standardized.

Regards,
Derek.



From tchur at optushome.com.au  Thu Jun 20 15:39:14 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Fri, 21 Jun 2002 05:39:14 +1000
Subject: Inter Process Communication with Numeric arrays
References: <wx1yb2uvt3.fsf@velocita.nasoftware>
Message-ID: <3D122F62.BF165EC2@optushome.com.au>

Ian McConnell wrote:
> 
> I want to send the results of some calculation on a Numeric array to a
> second process. These arrays will be fairly large (about 1Mb) so can anyone
> recommend a package for simple IPC?
> 
> I'm writing both processes (though one might be in C) so ORB/CORBA seems to
> be overkill and I can't get through to the memory mapped files at
>         http://pobox.com/~kragen/sw/arrayfrombuffer/
> (Internal server error)
> 
> Suggestions?

Have you thought about using MPI? MPI is usually used for cluster
computing but
it also works well for interprocess communication on a single machine -
it is 
commonly used for such purposes on multi-processor SMP machines, but you
can tell
MPI how many processes you want even on a single CPU machine. If you use
LAM/MPI,
which runs in user mode, it is a snap to install and you don't have to
recompile your kernel
- just run it when you need it. Because MPI was written with high
perfromance computing in mind,
it is very efficient. And if you ever want to split your processes
across different machines,
MPI will handle sending the data across the network transparently. My
understanding is that
MPI uses in-memory IPC where it can (on a shared-memory machine),
otherwise it uses network
transport layers - but it looks after all the messy details.

But best of all, Ole Nielsen at the Australian National University has
written a very
nice Python wrapper for MPI called PyPar which makes sending Python data
structures between
processes a snap, and it will even send NumPy arrays directly, without
serialisation. And you
don't need to modify your Python interpreter to use it - just import the
module.
See http://datamining.anu.edu.au/~ole/pypar/ for details.

Tim C




From tritran2001 at iprimus.com.au  Fri Jun 28 11:51:32 2002
From: tritran2001 at iprimus.com.au (tritran)
Date: Sat, 29 Jun 2002 01:51:32 +1000
Subject: Can i write chating system using python
Message-ID: <3d1c8661$1_1@news.iprimus.com.au>

Can anyone provide me the website or resource to create the chating system
or instance messeger using python.




From jepler at unpythonic.net  Sun Jun 23 21:49:23 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Sun, 23 Jun 2002 20:49:23 -0500
Subject: File IO read returning a string -> but I want a byte -> Question: Is there a faster more efficient way?
In-Reply-To: <3f8faafc.0206231627.570b8602@posting.google.com>
References: <3f8faafc.0206231627.570b8602@posting.google.com>
Message-ID: <20020623204923.B1172@unpythonic.net>

On Sun, Jun 23, 2002 at 05:27:52PM -0700, Jay wrote:
> #Question: Is there a faster more efficient way?
> 
> from time import clock
> 
> len_data = 20*1024*1024
> data = fp.read(len_data)

bindata = array.array("b", data)

OR
bindata = array.array("b")
bindata.fromfile(fp, len_data)

Jeff




From fperez528 at yahoo.com  Wed Jun  5 10:28:41 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 05 Jun 2002 14:28:41 +0000
Subject: self
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu> <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com> <mailman.1023294728.2322.python-list@python.org> <sWsL8.82$v_7.4260962@newnews.cc.stevens-tech.edu>
Message-ID: <adls9d$7ji$1@peabody.colorado.edu>

Vojin Jovanovic wrote:

>> For every name in such a value you need to evaluate its value again.
>> IOW it's recursive. The obvious solution is to intercept the
> namevalue-lookups
>> during evaluation of an equation. But this doesn't work for exec/eval:
>> They ignore the interception points.
> 
> Correct.
> 
> Every evaluation has to go through __getattr__ method in order to access
> the instance variables of the object, but that, unfortunately, can only be
> done with self.
> This is needed since the left hand sides of the equations must be the
> variables of the instance
> and not just some key values in a dictionary.

You seem to be trying to set up a system whose evaluation rules are very 
similar to those in Mathematica. I've long toyed with this idea myself, as an 
addition I'd love to make to IPython.  What I've had in mind is allowing a 
new definition symbol in the IPython prompt ':=' which like in Mathematica 
represents 'delayed evaluation'. This part would be trivial, as the hooks for 
special syntax processing are already in place (an example for physical 
quantities with units is provided and one for pasting code from python 
prompts). You'd then have something like (imagine this is at an interactive 
prompt):

y=5  # normal variable
x:=y+1   # x is a delayed evaluation variable, stored in a special class
x # prints 6 on screen
y=9
x  #prints 10 on screen

So ':=' variables would first try to evaluate their expressions in their 
normal namespace (that of the class, the problem you're dealing with now) and 
would then look for 'normal' variable names to resolve. Only if a name 
wasnt't found in either would a NameError be raised.

Unfortunately I don't have the time right now to work on this (and won't for a 
while). But if you want to add the syntax processor and the special class 
using IPython (which gives you finer control over the namespaces wherer 
things are evaluated than you get with the normal interpreter) I'll be glad 
to include it in future releases.

Cheers,

f.

ps. IPython is at http://www-hep.colorado.edu/~fperez/ipython/


From greg at cosc.canterbury.ac.nz  Mon Jun 17 00:16:12 2002
From: greg at cosc.canterbury.ac.nz (greg)
Date: Mon, 17 Jun 2002 16:16:12 +1200
Subject: ANN: Pyrex 0.3.1
Message-ID: <mailman.1024325000.4172.clpa-moderators@python.org>

Pyrex 0.3.1 is now available:

   http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/

Some fairly major problems with the new extension
types facility have been fixed.

There are also some new features:

  -  Default argument values, * and ** arguments are
     now supported in Pyrex-defined Python functions.

  -  External C functions with variable argument
     lists can be declared and called.

  -  Full C type declaration syntax now supported,
     including pointer-to-function types.

  -  More special methods for extension types:
     __delitem__, __delslice__, __getattr__,
     __setattr__, __delattr__

  -  ctypedef statement.

  -  String literal handling improved (triple
     quoted strings, escape sequences in strings
     should now work).

  -  Hex, octal and imaginary literals.

  -  Docstrings.


What is Pyrex?
--------------

Pyrex is a new language for writing Python extension modules.
It lets you freely mix operations on Python and C data, with
all Python reference counting and error checking handled
automatically.




From emile at fenx.com  Sat Jun  1 11:32:11 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 01 Jun 2002 15:32:11 GMT
Subject: Jython: Getopt does not throw exception - 1 attachments
References: <Xns9220AFD3CD394jbodemanngmxde@62.72.90.23>
Message-ID: <%N5K8.80912$352.4183@sccrnsc02>

Joern Bodemann asks:
> does anybody know why the following code does not throw an excetion if
the
> script is started without parameter?
>

The relevant code from getopt (on my version) is:

    while args and args[0].startswith('-') and args[0] != '-':
        [snip]
    return opts, args

which simply drops through to the return because args is empty.

>From your code:

opts, args = getopt.getopt(sys.argv[1:], "s:e:", ["start=", "ende="])

if you intend that  ["start=", "ende="] are required (as the equals
signs indicate), you could do:

opts, args = getopt.getopt(sys.argv[1:3], "s:e:", ["start=", "ende="])

instead, as this will cause an error when called without options.  You
could also check the length or argv.

HTH

--

Emile van Sebille
emile at fenx.com

---------



From ponderor at lycos.com  Mon Jun 24 20:20:42 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 24 Jun 2002 17:20:42 -0700
Subject: Windows versions of Python---pros and cons?
References: <mailman.1024613669.14676.python-list@python.org> <3D138293.85892240@noaa.gov> <mailman.1024941211.13683.python-list@python.org>
Message-ID: <e81be8b2.0206241620.5f962e3d@posting.google.com>

> 
> > Now I have to figure out how to change my PATHEXT environment variable,
> > which is pretty OT for this group...
> 
> Win9x: autoexec.bat
> WinME: dunno
> WinNT/2000/XP: Start->Settings->Control Panel->System
>                     ->Advanced->Environment Variables
> 

WinME: MSConfig


From claird at starbase.neosoft.com  Thu Jun 13 20:27:17 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 13 Jun 2002 19:27:17 -0500
Subject: What language - platform use to data acquisiton/numerical work?
References: <1c849ea9.0206130130.6cdaa36f@posting.google.com> <Xns922CCF37357A2cliechtigmxnet@62.2.16.82>
Message-ID: <AE10C8EA3C636BF9.54C1CA86E3055345.98BCF21D6F432D04@lp.airnews.net>

In article <Xns922CCF37357A2cliechtigmxnet at 62.2.16.82>,
Chris Liechti  <cliechti at gmx.net> wrote:
>mdtorre at freemail.it (Matteo) wrote in 
>news:1c849ea9.0206130130.6cdaa36f at posting.google.com:
			.
			.
			.
>> The data analisys numerical procedures have to be done in matlab due
>> to their complexity and specialization, and I fear even the plotting
>> will have to be done in Matlab. I have seen there are some bridges
>> from Phyton to Matlab, how they are?
>
>there are some modules but there are better solutions. e.g. Numpy is very 
>powerful for work with vectors and matrices (it has some sililaities with 
>matlab). SciPy and ScientificPython give additional toolboxes. (you should 
>find them easily using google.com)
As wonderful as Matlab is, the experience of many
people is that, when they learn Python, they largely
abandon Matlab in its favor.  Even though Matlab
seems targeted at just your situation, Python is so
strong in its cleanliness and expressive power that
it manages to best Matlab in several of Matlab's
strengths (let alone the latter's weaknesses).
>
>> My problem here is where to start developing in Pyyhon and where to
>> stop in ML. ML has a very simple and good sistem of building GUIs, how
>> is Pithon?
>
>you have many choices. Tk is supplied, wxWindows, pyQT and pyGTK are often 
>used ones. (or try anygui if you don't want to decide :-)
>
>> Finallly I hope to develop the wrapping GUI layer in Phyton. Again,
>> has Python some simple and good GUI support? I've seen that the most
>> widely used GUI module is something in tcl/tk (which I don't know at
>> all). Considering that I don't know python nor tcl/tk, how is the
>> learnig and developing curve?
>
>getting started with tkinter should be easy. i often use wxWindows which is  
>nastier to debug but good looking with nice license.
>you might be able to find a visualisation component (SciPy and co deliver 
>some tools i think) if you find such a component you might want to choose 
>the toolkit the component uses :-)
			.
			.
			.
You can reasonably expect to install and begin
using Python in a couple of hours.  If you want
to begin with GUIs, you can have small example
Tkinter programs running almost immediately.

Tkinter is the Python version of Tcl/Tk.  You
can reasonably expect Tkinter's capabilities to
closely match those of Tcl/Tk--in fact, it's
possible, in general, to ask Tkinter to interpret
Tcl scripts.
-- 

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


From mwh at python.net  Thu Jun 20 04:33:13 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 20 Jun 2002 08:33:13 GMT
Subject: Flame fests (was Re: python version?)
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <3D1010C3.C321F8B0@engcorp.com> <raXP8.47926$GY.15260255@twister.nyroc.rr.com> <3D107948.B3FEE4BB@engcorp.com> <aer53s$7is$1@panix1.panix.com>
Message-ID: <lkr8j2jqpg.fsf@pc150.maths.bris.ac.uk>

aahz at pythoncraft.com (Aahz) writes:

> Everybody have fun playing "kick the newbie"?

Huh?  That thread vanished under gnus' adpative scoring long ago...

Cheers,
M.

-- 
  Every now and then, Google doesn't throw up what I need so I start
  checking Altavista, Yahoo, etc.  In almost every single case, I am
  brutally reminded why I use Google in the first place.
                                                  -- John Riddoch, asr


From spam at melkor.dnp.fmph.uniba.sk  Wed Jun 19 05:49:50 2002
From: spam at melkor.dnp.fmph.uniba.sk (Radovan Garabik)
Date: Wed, 19 Jun 2002 11:49:50 +0200
Subject: Writing a simple linux daemon type thing...
References: <K9eO8.393$LN6.261@newsread1.prod.itd.earthlink.net> <aeja1t$2ebi$1@news.idiom.com> <_CTP8.4345$Fv1.455767@newsread2.prod.itd.earthlink.net>
Message-ID: <u3kpea.992.ln@127.0.0.1>

Jones <jones at jones.jones> wrote:
 : I would have to guess Hungarian.... but I could guess by looking at the
 : words.. just the email addresses :-)

definitely hungarian 


-- 
 -----------------------------------------------------------
| Radovan Garabik  http://melkor.dnp.fmph.uniba.sk/~garabik |
| __..--^^^--..__         garabik @ fmph . uniba . sk       |
 -----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!


From skip at pobox.com  Mon Jun  3 23:42:33 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 3 Jun 2002 22:42:33 -0500
Subject: ANN: Pure Python strptime()
In-Reply-To: <f85b5ab1.0206031354.132f3148@posting.google.com>
References: <f85b5ab1.0206031354.132f3148@posting.google.com>
Message-ID: <15612.14121.529921.388554@12-248-41-177.client.attbi.com>


    Brett> Do I ... change the time module into a Python module with the
    Brett> current module renamed _time and then just have a stub module to
    Brett> globally import the _time module?

Yes, that's the usual technique used to get a mixed Python/C module into the
standard library.

    Brett> Do I just need to convince Guido to include it?  

I've heard that often helps. ;-)

    Brett> I have posted this as a patch at SF (patch #474274), but I am
    Brett> wondering if I have to do anything that I just previously
    Brett> mentioned on top of this.  Any guidance on this would be helpful.

You just did it.  I assigned it to myself and will try to take a look at it.

(I don't suppose you have a unit test for this beast laying about do you?)

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html





From spam at eatshit.net  Mon Jun 17 02:10:31 2002
From: spam at eatshit.net (paul colby)
Date: Mon, 17 Jun 2002 06:10:31 GMT
Subject: Newbie Help Please.(OO and Tkinter)
References: <B93209DC.7281%sarmstrong13@mac.com>
Message-ID: <r3fP8.401$uH2.132@newsread1.prod.itd.earthlink.net>

SA wrote:

> Hi Everyone-
> 
>     I've had some problems finishing this code. I think it stems from my
> lack of OOP understanding. Below is the code:
> 
> from Tkinter import *
> import os
> import commands
> import tkSimpleDialog
> 
> class PyShell(tkSimpleDialog.Dialog):
>     def clearin(self):
>         self.t1.delete(0.0,END)
>     def clearout(self):
>         self.t2.delete(0.0,END)
>     def expyth(self):
>         self.t2.delete(0.0, END)
>         self.output = commands.getoutput(self.t1.get(0.0, END))
>         self.t2.insert(END, self.output)
>     def __init__(self, top):
>         self.t1 = Text(top, height="12", width="84", font="Courier 12")
>         self.t1.pack(side=TOP, pady=2)
>         f = Frame(top)
>         f.pack()
>         self.b1 = Button(f, text="Execute", command=self.expyth)
>         self.b1.pack(side=LEFT)
>         self.b2 = Button(f, text="Clear Input", command=self.clearin)
>         self.b2.pack(side=LEFT)
>         self.b3 = Button(f, text="Clear Output", command=self.clearout)
>         self.b3.pack(side=LEFT)
>         self.b4 = Button(f, text="Save Input", command=Saving.savin)
>         self.b4.pack(side=LEFT)
>         self.t2 = Text(top, height="12", width="84", bg="lightblue",
> font="Courier 12")
>         self.t2.pack(side=TOP, pady=2)
>         
> class Saving(PyShell):
>     
>     def savin(self, question):
>         Label(question, text="Directory:").grid(row=0)
>         Label(question, text="Filename:").grid(row=1)
>         self.e1 = Entry(question)
>         self.e2 = Entry(question)
>         self.e1.grid(row=0, column=1)
>         self.e2.grid(row=1, column=1)
>         def apply(self):
>             self.dir = self.e1.get()
>             self.fn = self.e2.get()
>             self.sav = self.dir + self.fn
>             self.savfile = open(self.sav, "w")
>             for line in self.t1.readlines():
>                 self.savefile.write(line)
>             self.savefile.close()
>             
> root = Tk()
> app = PyShell(root)
> root.mainloop()
> 
> 
> 
> Everything runs until I try to save the text in t1 with the "Save Input"
> button. I end up with the following error:
> 
> Exception in Tkinter callback
> Traceback (most recent call last):
>   File "/sw/src/root-python-2.2.1-6/sw/lib/python2.2/lib-tk/Tkinter.py",
> line 1292, in __call__
>     return apply(self.func, args)
> TypeError: unbound method savin() must be called with Saving instance as
> first argument (got nothing instead)
> 
> What am I doing wrong? Saving should inherit from PyShell correct? And in
> order to work correctly saving needs tkSimpleDialog.Dialog, so PyShell
> inherits this correct?
> 
> Or am I totally wrong here?
> 
> Thanks.
> SA

When you set command=Saving.savin you are passing a member 
function which has no define instance pointer "self". Thus 
the "self" thingy in the definition of savin isn't there to 
be used when the button is pressed. Note that your other 
buttons likely work since you use constructs like 
command=self.clearin. Now I haven't tried it but I 
would expect if you included something like

        self.SAVE = Saving(self)

in the __init__ function definition for PyShell then you 
could use command=self.SAVE.savin for the Save Input button. 
Now when the button is pressed self = self.SAVE so there
is an instance.


Hope this helps

Regards
Paul Colby


From bokr at oz.net  Thu Jun  6 10:50:01 2002
From: bokr at oz.net (Bengt Richter)
Date: 6 Jun 2002 14:50:01 GMT
Subject: simple pattern matching question
References: <mailman.1023333548.15465.python-list@python.org>
Message-ID: <adnsqp$amb$0@216.39.172.122>

On Wed, 05 Jun 2002 20:18:04 -0700 (PDT), "Sean 'Shaleh' Perry" <shalehperry at attbi.com> wrote:

>
>On 06-Jun-2002 Corey G. wrote:
>> Gurus,
>> 
>> How would I match with an item from a list?  I would imagine something
>> this simple would not require a regular expression.
>> 
>> web_servers=('www01','www02','www03')
>> 
>>         for server_match in web_servers:
>>            if (trap == "Service:"server_match "State:down"):
>>             print "found match"
>> 
>> The comparsion should be: if trap == Service:www01 State:down
>> 
>> server_match prints "as is, but not from the list".  I tried many 
>> combinations but decided to leave the most obviously wrong one.
>> 
>
>if trap == ("Service:%s State:down" % server_match):
>  print "found match"
>
That should work, if you know that the trap string will never
have an extra blank, or tab instead of space, or something additional
in it, or change case or order.

An alternative to building a match to the entire trap string
is to extract and normalize the parts of interest, according
to the syntax of trap. The syntax will determine if your best
solution is a regular expression, or maybe just .lower().split()
etc. Put a comment in your code referencing where the trap
syntax is defined/determined/guaranteed ;-)

Notice what kinds of changes you'll need if someone later
adds " TimeStamp:..." or such to the trap string, at the front
or back.

Regards,
Bengt Richter


From thenault at nerim.net  Fri Jun 14 11:05:53 2002
From: thenault at nerim.net (Sylvain Thenault)
Date: 14 Jun 2002 17:05:53 +0200
Subject: [ANN] PyReverse 0.3
Message-ID: <mailman.1024086967.9145.clpa-moderators@python.org>

What's new in 0.3 ?
-------------------
        * full architecture reorganisation. Not yet perfect but much more 
	  clean code :-)
	* begin unit tests generator using the unittest framework
	* apply patch from cognite at zianet.com fixing typos and KeyError in 
	  pystats
	* add warning for functions parameters with default to a list or a 
	  dictionnary
	* add a command line tool to generate xmi without argo files
	* add the --ignore option which permits to filter files analyzed
        * use 4Suite cDomlette for XMI-UML generation if it's available

What's PyReverse ?
------------------

Pyreverse is a set of utilities to reverse enginering Python code. So
far, it features dependency analysis tools, unittest generation,
and XMI generation for importation in a UML modeling tool. A special
module can be used to generate files readable by Argo UML. 
It uses a representation of a Python project in a class hierarchy
which can be used to extract any information (such as generating UML
diagrams and make a few statistics from the Python code, as "pyargo"
and "pystats") 


Home page
---------
http://www.logilab.org/pyreverse/

Download
--------
ftp://ftp.logilab.org/pub/pyreverse/

Mailing list
------------
mailto://xml-logilab at logilab.org


-- 
Sylvain Th?nault,  LOGILAB

http://www.logilab.org




From BPettersen at NAREX.com  Tue Jun 25 11:31:04 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 25 Jun 2002 09:31:04 -0600
Subject: Sending string var with apply sends all chars as params!???
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192F1A7F4@admin56.narex.com>

> From: GimsMail [mailto:elabuschagne at gims.com] 
> 
> [snip]
> 
> My print line debugging string output:
> Formatting value '20020625140609' of type <type 'string' with
> formatter <function TimeFormatter at 00D613F4>
> 
> as you can see, I am sending the string '20020625140609' to 
> the function TimeFormatter.
> 
> The code that does this is:
> print "UCP: Formatting value '%s' of type '%s' with formatter 
> '%s'" %(value, type(value), formatter)
> value = apply(formatter,(str(value)))

apply takes a tuple as its second argument. Try:

  value = apply(formatter, (str(value),))

or even better:

  value = formatter(str(value))

-- bjorn




From peter at engcorp.com  Tue Jun 25 23:00:26 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 25 Jun 2002 23:00:26 -0400
Subject: SOLVED (Re: Python hits the spot)
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at> <af4vrl$ffe$0@216.39.172.122> <3D16DD5C.1050708@kfunigraz.ac.at> <uhdume2n1mhbba@corp.supernews.com> <3D171291.2060905@kfunigraz.ac.at> <mailman.1024950033.16332.python-list@python.org> <Xns9237E4F0DB10Dcliechtigmxnet@62.2.16.82> <3D17BFE1.B55C689F@engcorp.com> <3D189972.D1477AA7@noaa.gov>
Message-ID: <3D192E4A.EA3F1DCC@engcorp.com>

Chris Barker wrote:
> 
> Peter Hansen wrote:
> 
> > > you have to admit that it was a nasty failure... is bad hardware your first
> > > thought when your program fails?

Actually, Chris Liechti wrote that; I was just quoting it.

> I have to say, I got pretty suspicious when he said that he had crashed
> Linux. I have been using Linux for years and have never crashed it. When

I suspect many had the same thought, but ignored it in the likelihood
that he was using the term incorrectly.  I can't count the number of
times someone has told me the machine "crashed" or "froze" when what
they meant was more like "my program exited unexpectedly" or "I got
a stack-trace"...

-Peter


From tebeka at cs.bgu.ac.il  Mon Jun 17 07:18:05 2002
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 17 Jun 2002 04:18:05 -0700
Subject: Help: win32gui & threads
Message-ID: <33803989.0206170318.9d3a225@posting.google.com>

Hello All,

I've written a little logger daemon with UI in Windows taskbar.
(See code below)
It seems to work fine but when I choose 'Quit' the windown closes but
the pythonw is still alive and I need to kill it from the TaskManager.

Any idea how to make it close?

Thanks.
Miki.

-----------------------------------------
#!/usr/bin/env python
'''Logger daemon

Opens socket server and write messages to log file
'''

__author__ = 'Miki Tebeka <miki_tebeka at amat.com>'
__version__ = '$Revision: 1.3 $'

# $Source: /home/miki/.cvsroot/logging/logd.py,v $

import SocketServer
import socket
import Queue
import threading
import time
import re

from win32api import *
from win32gui import *
import win32con
import sys
import os, os.path

MESSAGE_QUEUE = Queue.Queue(0) # Thread safe message queue with no
limit

# These two are not protected by mutex since only the main thread
write there and
# the worker thread reads them
WRITE = 1 # Write flag
NEW_LOG = 0 # Create new file

def error(msg):
    '''Error function'''
    MessageBox(win32con.MB_OK, msg)

class SocketListener(SocketServer.BaseRequestHandler):
    '''Socket listener. Reads line from socket and place it in message
queue'''
    
    def __init__(self, request, client_address, server):
        '''Constructor'''
        SocketServer.BaseRequestHandler.__init__(self, request,
client_address, server)

    def handle(self):
        '''Handler function. Reads line from socket and writes to
queue'''
        msg = ''
        while 1:
            try:
                c = self.request.recv(1)
                if not c:
                    return
                if (c == '\n'):
                    MESSAGE_QUEUE.put(msg)
                    msg = ''
                else:
                    msg += c
            except socket.error, e:
                error('Error: %s' % e)
                return

class LoggerDaemon(SocketServer.ThreadingTCPServer, threading.Thread):
    '''Logger Daemon. Listens on given port and threads listeners upon
requests'''
    def __init__(self, port):
        self.port = port
        self.addr = ('', self.port)
        threading.Thread.__init__(self)
        SocketServer.ThreadingTCPServer.__init__(self, self.addr,
SocketListener)

    def run(self):
        self.serve_forever()

class LogWriter(threading.Thread):
    '''Log writer. Get messages from queue and write to file'''
    def __init__(self, log_dir):
        self.log_dir = log_dir
        if not os.path.isdir(self.log_dir):
            try:
                os.makedirs(self.log_dir)
            except OSError, ose:
                error("Can't create log directory %s (error was: %s)"
% (self.log_dir, ose))
                sys.exit(1)
        self.new_log()
        threading.Thread.__init__(self)

    def run(self):
        '''Thread function'''
        global NEW_LOG, WRITE
        while 1:
            msg = MESSAGE_QUEUE.get()
            if NEW_LOG:
                self.new_log()
                NEW_LOG = 0 # This is a sync NO-NO but we can live
with that
            if WRITE:
                print >> self.fp, msg.strip()
                self.fp.flush()

    def new_log(self):
        '''New log file'''
        logname = '%s/%s' % (self.log_dir, re.sub('(\s+)|:', '_',
time.ctime()))
        self.log_file = logname
        self.fp = open(self.log_file, 'a')

class MainWindow:
    '''Main window for handling taskbar notifications'''
    def __init__(self):
        '''Constructor'''
        self.ID_TBAR = win32con.WM_USER + 20
        self.icon_file = './logd.ico'
        self.title = 'BeeLogger Daemon'
        self.IDM_TOGGLE, self.IDM_NEWLOG, self.IDM_QUIT = 1023, 1024,
1025
        self.resume_pause_msg = ('Resume', 'Pause')
        message_map = {
            win32con.WM_DESTROY: self.OnDestroy,
            win32con.WM_COMMAND: self.OnCommand,
            self.ID_TBAR : self.OnTaskbarNotify,
        }
        
        # Register the Window class.
        wc = WNDCLASS()
        hinst = wc.hInstance = GetModuleHandle(None)
        wc.lpszClassName = "BeeLoggerDaemon"
        wc.style = win32con.CS_VREDRAW | win32con.CS_HREDRAW;
        wc.hCursor = LoadCursor( 0, win32con.IDC_ARROW )
        wc.hbrBackground = win32con.COLOR_WINDOW
        wc.lpfnWndProc = message_map # could also specify a wndproc.
        classAtom = RegisterClass(wc)
        # Create the Window.
        style = win32con.WS_OVERLAPPED | win32con.WS_SYSMENU
        self.hwnd = CreateWindow( classAtom, self.title, style, \
                    0, 0, win32con.CW_USEDEFAULT,
win32con.CW_USEDEFAULT, \
                    0, 0, hinst, None)
        UpdateWindow(self.hwnd)

        # Try and find a custom icon
        if not os.path.isfile(self.icon_file):
            icon_flags = win32con.LR_LOADFROMFILE |
win32con.LR_DEFAULTSIZE
            self.hicon = LoadIcon(0, win32con.IDI_APPLICATION)
        else:
            self.hicon = LoadImage(hinst, self.icon_file,
win32con.IMAGE_ICON, 0, 0, icon_flags)

        flags = NIF_ICON | NIF_MESSAGE | NIF_TIP
        nid = (self.hwnd, 0, flags, self.ID_TBAR, self.hicon,
self.title)
        Shell_NotifyIcon(NIM_ADD, nid)

    def OnDestroy(self, hwnd, msg, wparam, lparam):
        nid = (self.hwnd, 0)
        Shell_NotifyIcon(NIM_DELETE, nid)
        PostQuitMessage(0) # Terminate the app.

    def OnTaskbarNotify(self, hwnd, msg, wparam, lparam):
        global WRITE
        if lparam==win32con.WM_LBUTTONUP or
lparam==win32con.WM_LBUTTONDBLCLK or lparam==win32con.WM_RBUTTONUP:
            menu = CreatePopupMenu()
            AppendMenu(menu, win32con.MF_STRING, self.IDM_TOGGLE,
self.resume_pause_msg[WRITE])
            AppendMenu(menu, win32con.MF_STRING, self.IDM_NEWLOG, 'New
Log')
            AppendMenu(menu, win32con.MF_STRING, self.IDM_QUIT,
'Quit')
            pos = GetCursorPos()
            TrackPopupMenu(menu, win32con.TPM_LEFTALIGN, pos[0],
pos[1], 0, self.hwnd, None)
        return 1

    def OnCommand(self, hwnd, msg, wparam, lparam):
        global WRITE, NEW_LOG
        id = LOWORD(wparam)
        if id == self.IDM_TOGGLE:
            WRITE = 1 - WRITE
        elif id == self.IDM_NEWLOG:
            NEW_LOG = 1
        elif id == self.IDM_QUIT:
            DestroyWindow(self.hwnd)
        else:
            raise ValueError('Unknown command: %s' % id)


# MAIN
if __name__ == '__main__':
    if len(sys.argv) != 3:
        print >> sys.stderr, 'usage: logd LOG_DIR PORT'
        sys.exit(1)
    
    # Start writer
    writer = LogWriter(sys.argv[1])
    writer.start()

    # Start TCP server
    logd = LoggerDaemon(int(sys.argv[2]))
    logd.start()
        
    # Start UI
    w=MainWindow()
    PumpMessages()

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


From logiplexsoftware at earthlink.net  Sun Jun 23 01:10:41 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: 22 Jun 2002 22:10:41 -0700
Subject: Python hits the spot
In-Reply-To: <af2ia8$a92$1@panix1.panix.com>
References: <3D1482AE.8FA299D8@kfunigraz.ac.at>
	<3d149f19$0$53133$edfadb0f@dspool01.news.tele.dk>
	<af28vp$j57$1@panix1.panix.com> <3D14AD7F.3D5D243F@engcorp.com> 
	<af2ia8$a92$1@panix1.panix.com>
Message-ID: <1024809043.14694.2.camel@localhost.localdomain>

On Sat, 2002-06-22 at 12:14, Aahz wrote:
> In article <3D14AD7F.3D5D243F at engcorp.com>,
> Peter Hansen  <peter at engcorp.com> wrote:
> >Aahz wrote:
> >> Ya know, you could have saved yourself a lot of typing by simply saying
> >> PBCAK.  ;-)
> >
> >To save us from those who can't Google:  Problem Between Chair And Keyboard.
> >(It was new to me, too.)
> 
> <grin>  Ya coulda ast me, too.

I'd always heard of it called "chair-keyboard interface problem".

Cliff





From pyth at devel.trillke.net  Wed Jun  5 19:59:18 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 6 Jun 2002 01:59:18 +0200
Subject: self
In-Reply-To: <uy9dt79rx.fsf@ctwd0143.fitlinxx.com>; from db3l@fitlinxx.com on Wed, Jun 05, 2002 at 04:25:38PM -0400
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu> <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com> <mailman.1023294728.2322.python-list@python.org> <uy9dt79rx.fsf@ctwd0143.fitlinxx.com>
Message-ID: <20020606015918.E30696@prim.han.de>

[me]
  IOW you cannot successfully wrap or inherit a dictionary type and pass it 
  to exec/eval :-(
 
[David Bolen]
  Yes, I did also find a posting by Guido (didn't save the reference,
  but it was via a google search) where he seemed to indicate that
  subclassing list/dict other than add additional functionality (rather
  than override core operations like lookups) could be problematic -
  presumably because of cases like this where the special Python-exposed
  functions are used.  The message was during one of the later betas of
  2.2, so I don't know if the viewpoint changed before final release.

everything is a google-search away if you know the right codewords :-)
I don't in this case, though ...

anyway, if you call dictionary methods from python you can intercept.
But either generally c-level code ignores python-overriden methods 
or only eval/exec "optimizes" them out. I presume that this is for
speed reasons but there may lurk conceptual reasons as well. Who knows?

  holger




From wurmy at earthlink.net  Tue Jun 25 01:33:17 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Tue, 25 Jun 2002 05:33:17 GMT
Subject: __getattr__ feature
Message-ID: <3D180042.6000902@earthlink.net>

I didn't know this was possible:

class Foo:
     def __init__(self, obj):
         self.obj = obj
     def __getattr__(self, name):
         return getattr(self.obj, name)

foo = Foo([1,2,3])
print foo[2]    # prints 3

I would expect the code to raise an exception, since there is no __getitem__ 
method. A quick test learns that this is apparently a feature new to 2.2;
it doesn't work if I try it in 2.1.

Interestingly, using getattr() directly on a list does *not* work (although 
this is what the method does):

 >>> x = [1, 2, 3]
 >>> getattr(x, 1)
Traceback (most recent call last):
   File "<pyshell#1>", line 1, in ?
     getattr(x, 1)
TypeError: attribute name must be string
 >>> getattr(x, "1")
Traceback (most recent call last):
   File "<pyshell#2>", line 1, in ?
     getattr(x, "1")
AttributeError: 'list' object has no attribute '1'

I didn't find anything about this in the reference manual. Does anybody know 
how/why this works and why it was added?

Cheers,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/



From sholden at holdenweb.com  Thu Jun  6 15:23:11 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 6 Jun 2002 15:23:11 -0400
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet> <3CFDF148.AA2035FF@info.unicaen.fr> <3CFE49CA.46A5FEE1@tds.net> <838z5sgx7w.fsf@panacea.canonical.org> <mailman.1023388273.14192.python-list@python.org>
Message-ID: <cJOL8.326812$%u2.257815@atlpnn01.usenetserver.com>

"Jeff Epler" <jepler at unpythonic.net> wrote ...
> On Thu, Jun 06, 2002 at 12:58:59PM -0400, Kragen Sitaker wrote:
> > "Edward K. Ream" <edream at tds.net> writes:
> > > BTW, a compiled version of a Python program would be valuable
> > > commercially, as it would allow companies to protect their source
code,
> > > as well as (possibly) improve performance.
> >
> > I'm worried about this myself; perhaps the compiler could include the
> > source code in the resulting executable to prevent this.
>
> Your sarcasm is lost on me.
>

Lost, perhaps, but (since you recognised it) not entirely wasted ;-)

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From joesalt at ireland.com  Mon Jun 10 14:41:03 2002
From: joesalt at ireland.com (JS)
Date: 10 Jun 2002 11:41:03 -0700
Subject: Newbie question communicating between 2 progs
References: <3D046BF0.DC36542C@raplounge.de> <3D04A938.627EE705@engcorp.com>
Message-ID: <d9deee30.0206101041.4e8de6a2@posting.google.com>

Don't fully understand the question, but could you write the list to a
temporary file. It could then be read as needed??


Peter Hansen <peter at engcorp.com> wrote in message news:<3D04A938.627EE705 at engcorp.com>...
> Ben Thursfield wrote:
> > 
> > I am running two python scripts in 2 different frames simultanously, and
> > one of the programm needs to access a list defined in the other
> > programm. How do i manage to access that list?
> 
> The question is ambiguous.  When you say "frames", the only thing
> that comes to mind is that you are running Python on client-side
> Internet Explorer under Windows, as scripts in two different parts
> of a window (the frames) simultaneously.  I think you must have
> meant something different however.
> 
> Could you please describe more about your environment?  The
> term "frame" can mean a number of things.
> 
> -Peter


From whisper at oz.net  Tue Jun 11 10:12:06 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 11 Jun 2002 07:12:06 -0700
Subject: [OT] Antispamming the email address
In-Reply-To: <3D05FF31.4060701@araknos.it>
Message-ID: <GCEDKONBLEFPPADDJCOEOECHDIAA.whisper@oz.net>

I think both of them are too easy to programatically detect and unwind. Even
"fred_nospam_ at some.org is only going to have a limited deterrent lifetime.

Something based on a public key system might have legs...

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Nicola Larosa
> Sent: Tuesday, June 11, 2002 6:46
> To: Christos TZOTZIOY Georgiou
> Cc: python-list at python.org
> Subject: [OT] Antispamming the email address
>
>
>  > Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')
>
> That's nifty. :^)
>
> You could encode it in rot13 format instead, so it'll keep the
> shape of an email address and *will* be harvested, while still being
> invalid. Let's throw more wrenches in the wheels of spammers!
>
> Like this:
>
>  > Real email address: 'gmbg at fvy-grp.te'.decode('rot13')
>
>
> --
> "Too much cleverness in the parser can turn against you."
>      Guido Van Rossum
>
> Nicola Larosa - nico at tekNico.net
>
>
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From garrett at bgb.cc  Wed Jun 19 14:48:42 2002
From: garrett at bgb.cc (Don Garrett)
Date: Wed, 19 Jun 2002 18:48:42 GMT
Subject: What If.....    Strong Types
References: <mailman.1024498968.2603.python-list@python.org>
Message-ID: <3D10D17B.6070004@bgb.cc>

brueckd at tbye.com wrote:

> One approach is to not even define a weight member variable until I've
> weighed the package, and I can test to see if I've weighed it already by
> checking for the presence of that variable. The alternative is to define a
> weight variable up front and then initialize it to some dummy value, e.g.
> -1. I believe that when I do this, though, it is more because of my
> "upbringing" in other languages rather than it being a good idea. Does the
> package really weigh -1? No.  Worse, to make it be "good code" I'll
> probably define -1 to be NOT_WEIGHED or something, and check against it so
> I'm not using magic numbers. 

   I might suggest the use of None over -1, since it has a real meaning of 
'not really there'. But not defining the member would have the same meaning as 
well. I do get your point here.

> It's sometimes hard for me to model my objects this way because of having
> had to always do it the other (completely statically defined) way, but
> when this approach works the solutions are often quite elegant and
> straightforward (btw - I believe that a good portion of bugs and program
> cruftiness is related to forcing the programmer to clearly define, up
> front, what the complete type of the object is. In many cases this forces
> the programmer to make decisions without enough information, in others it
> is too constricting, and in others it's just too tedious to make the
> programmer worry about all those details, and in all cases general
> hackiness ensues).
> 
> -Dave

   The point about programmers having to make decisions about the data before 
they are ready is a very good one. I've certainly run into that myself, but 
I've never heard it phrased that way, so it never clicked the way it just did.

   Don't you find that you miss the advantages of a clear definition of what a 
class is and/or does? That's more important with group development, but even 
when it's just me.... I need memory aids.

   Regardless, thanks for the reply. I'm learning, and I like that!

-- 
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc



From bokr at oz.net  Sun Jun 30 19:00:47 2002
From: bokr at oz.net (Bengt Richter)
Date: 30 Jun 2002 23:00:47 GMT
Subject: newbie Q: delete list few lines in file
References: <7fe97cc4.0206291519.1c95ce31@posting.google.com> <3D1E5123.FDEFADDF@engcorp.com> <3D1E83C0.8060700@earthlink.net> <3D1E8A08.53DD2272@engcorp.com> <B9451E26.D35B%jonathan@onegoodidea.com>
Message-ID: <afo2iv$lvr$0@216.39.172.122>

On Sun, 30 Jun 2002 20:44:22 +0100, Jonathan Hogg <jonathan at onegoodidea.com> wrote:

>On 30/6/2002 5:33, in article 3D1E8A08.53DD2272 at engcorp.com, "Peter Hansen"
><peter at engcorp.com> wrote:
>
>> Hans Nowak wrote:
>>> 
>>> Peter Hansen wrote:
>>> 
>>>> You cannot write files in place.
>>> 
>>> <nitpick>
>>> 
>>> Sure you can:
>> 
>> [snip proof]
>> 
>> Of course, but we were answering a newbie's questions, right?
>
>I think that it is unclear from Xah's original post whether the files or the
>lines are to be deleted. Certainly, asking how to write to files in place
>suggests that the lines are to be deleted.
> 
>> -trying-not-to-complicate-things-ly yr's,
>
>If you want to say "that's complicated, are you sure you want to do that?",
>go ahead. But it's not helpful to tell someone that something cannot be done
>when it can. That's just misdirection.
>
>-newbies-aren't-children-ly y'rs,
>
Some of them probably are, for which case misdirection is _extra_ pernicious IMO.

Regards,
Bengt Richter


From hesterloli at hotmail.com  Wed Jun 19 04:46:33 2002
From: hesterloli at hotmail.com (George Hester)
Date: Wed, 19 Jun 2002 08:46:33 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com> <mailman.1024460666.24801.python-list@python.org> <B8XP8.47925$GY.15259265@twister.nyroc.rr.com> <3D104142.8948754@alcyone.com>
Message-ID: <JxXP8.47928$GY.15272008@twister.nyroc.rr.com>

Because it is there.  No use what use is it climbing Mt Everest?  Because it is
there.

--
George Hester
_________________________________
"Erik Max Francis" <max at alcyone.com> wrote in message
news:3D104142.8948754 at alcyone.com...
> George Hester wrote:
>
> > I don't want to learn the language.  At least not right now.  This
> > started
> > becuase I read there was a new version of Python.  Then I saw an ASP
> > at
> > Microsoft that used Python.  So I thought I would run an ASP to
> > determine my
> > version of Python.  And then I thought I would ask how to do that
> > here.
>
> If you don't want to learn the language, what possible use could knowing
> what version you're running do you?
>
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ Who'd ever think it / Such a squalid little ending
> \__/ The American and Florence, _Chess_
>     Church / http://www.alcyone.com/pyos/church/
>  A lambda calculus explorer in Python.




From marklists at mceahern.com  Fri Jun 28 16:09:47 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 28 Jun 2002 15:09:47 -0500
Subject: generate all combinations of a list (with variable output length)
In-Reply-To: <3D1CBE6D.30002@huno.net>
Message-ID: <NCBBLFCOHHDIKCAFGCFBCEAMLHAA.marklists@mceahern.com>

> i'm looking for a way to generate all possible combinations of all items
> in a list. list = ["A","D","$","5","R"]

i don't have a solution, but some questions to clarify:

if you think of your list as comprising the digits, then you're asking for
an n-length string composed of all possible digits where the digits include:

  A, D, $, 5, R, and None (the representation of the latter is '')

so, for n = 10, this is a possible value:

  AAAAAAAAAA

as well as:

  [          ]

omit the brackets.

True?

anyway, this might get you started:

  http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/105962

// m

-





From aahz at pythoncraft.com  Sun Jun  2 13:46:46 2002
From: aahz at pythoncraft.com (Aahz)
Date: 2 Jun 2002 13:46:46 -0400
Subject: What is a classmethod for?
References: <addgv30f4d@enews4.newsguy.com>
Message-ID: <addlm6$g44$1@panix1.panix.com>

In article <addgv30f4d at enews4.newsguy.com>,
Dennis Peterson <denpeterson at yahoo.com> wrote:
>
>class C:
>    def foo(cls):
>        print cls
>    foo = classmethod(foo)
>
>class D(C):
>    pass
>
>c = C()
>d = D()
>c.foo() prints __main__.C
>d.foo() prints __main__.D
>
>That's nice, but I can't think of a use for it, given that we have
>inheritance and isinstance(). Can anyone enlighten me?

The point is that you can call class methods *without* creating an
instance:

C.foo()
D.foo()

This allows you to manipulate class variables through methods, rather
than using functions or needing to create an instance.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"In the end, outside of spy agencies, people are far too trusting and
willing to help."  --Ira Winkler


From jjl at pobox.com  Wed Jun  5 10:30:48 2002
From: jjl at pobox.com (John J. Lee)
Date: Wed, 5 Jun 2002 15:30:48 +0100
Subject: urllib, urllib2, httplib -- Begging for consolidation?
In-Reply-To: <mailman.1023226276.32205.python-list@python.org>
References: <mailman.1023226276.32205.python-list@python.org>
Message-ID: <Pine.SOL.4.30.0206051524590.16156-100000@mimosa.csv.warwick.ac.uk>

On Tue, 4 Jun 2002 brueckd at tbye.com wrote:
[...]
> > I suppose the standard answer applies here: it isn't there because
> > nobody has written it yet.
>
> Argh... this misses the whole point of the thread. Never mind, I repent
> for having revived this topic.

I'm trying to understand your posts on this, which are interesting. I just
ported the HTTP cookie handling code from libwww-perl (and integrated it
into urllib2), and since I seem to have got into this HTTP business, I may
(eventually, not for a month or two) experiment with adding features to /
restructuring urllib2 & co.  Having read the original thread (your posts,
at least), I still think we are (mostly) agreeing violently.  If not,
apologies for my slowness.

[quotes pasted from several different postings, all by Dave]
> If there's some division of the feature set (base protocol and higher
> level) it would make sense for the two modules to be httpcore and
> httplib (or something) with urllib built on top.

OK, my main point here: the code is already structured like this; the only
difference between your view of how it should be and how it actually is is
that the conceptual units you would like to label httplib and urllib
happen to live in the same module in the current implementation:

httpcore --> httplib
httplib --> urllib2.AbstractHTTPHandler (which is missing features)
urllib --> urllib2.OpenerDirector, urllib2.build_opener, urllib2.urlopen

But from your last post, it seems I may have completely misunderstood.
Perhaps part of your point is simply that the division into modules should
better reflect the conceptual structure?  The current organisation may (or
may not) have been a mistake, but it's not a significant enough mistake to
warrant reorganisation, is it?

> That's fine, but why would so much useful http-specific knowledge live
> outside of an http-specific module?

Because it's possible to separate out the very basic, minimal HTTP stuff
into httplib, it was the right decision to do so, from an implementation
point of view.  From a use point of view, since HTTP is stateless, it's
always possible to get at the httplib.HTTP interface even when using
urllib2 (I guess that's false for HTTP 1.1, though, which hadn't occurred
to me until you pointed it out).  Are we agreed on this?

A particular example you raised:

> Ok, maybe I'm just not understanding how httplib/urllib/urllib2 work
> together. For example, what is the correct way to do a HTTP HEAD
> request that follows redirects? It's not hard, but it's silly to have
> to code it yourself if somebody already did the work for you. Well,
> httplib doesn't know how to follow an HTTP redirect, but lo and
> behold, urllib does. Unfortunately, there's no way to use it because
> in this case because urllib is for opening URLs (the GET is hardcoded
> and not easily overridable).

We agree on the problem here.  However, I don't see why the basic
architecture of urllib2 is wrong here.  HTTPRedirectHandler can be used
entirely separately from the OpenerDirector machinery, and is documented
in the std. lib. docs.  You are of course right in saying that it fails
completely to handle HEAD requests and cookies, for example.  *However*,
this only has the undesirable effect of punting you back to httplib (which
I think we're agreed is the problem we're discussing here) if you happen
to think that that'll be less work than modifying HTTPRedirectHandler,
which is an almost entirely self-contained class.  Obviously, it would be
better if this job were done once and for all in the proper manner, by
fixing HTTPRedirectHandler to handle HEAD methods (assuming that HEAD
methods are supposed to get redirected), but the architecture is all there
-- we would just have to use the Request object to see if we're doing a
HEAD or a POST, and act accordingly.

To reiterate, the framework is there, the code isn't.

Interested in any more ideas you may have -- the more concrete, the
better!


John



From hesterloli at hotmail.com  Tue Jun 18 04:17:54 2002
From: hesterloli at hotmail.com (George Hester)
Date: Tue, 18 Jun 2002 08:17:54 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com>
Message-ID: <S0CP8.44419$GY.13868719@twister.nyroc.rr.com>

I remember once I tried to explain to my students the theory behind the limit in
calculus.  To 35 students.  Maybe 3 understood it in such a way that they could
present a solution back to me.  But most just didn't want to know.  I was wrong.
I was not teaching Mathematicians but Buisness students that just needed the
credit.  Now I am not someone who necessarily wants to code using Python in ASP.
I just want some code to find my version of Python using ASP.  I'm sorry if I
have not lived up to your expectations.

--
George Hester
_________________________________
"George Hester" <hesterloli at hotmail.com> wrote in message
news:K7PO8.16057$uk2.8406537 at twister.nyroc.rr.com...
> What is a simple script that will tell me the version of python I am currently
> running?  IIS 5.0 Windows 2000.
>
> --
> George Hester
> _________________________________
>
>




From nagarajagadi at yahoo.com  Mon Jun  3 02:08:48 2002
From: nagarajagadi at yahoo.com (nagaraj)
Date: 2 Jun 2002 23:08:48 -0700
Subject: accessing database of other applications
Message-ID: <56334c65.0206022208.363ade63@posting.google.com>

HI 
I am a fresh developer in palmOS. i wanted to know how can I access
the database of some other application (eg:datebook, memopad or any
other user application) through my application. can I use the
datamanager functions (APIs) given in sdk? if so how? if I use
dmOpenDatabaseByTypeCreator how to get the creatorID of a perticular
database?
thanx in advance


From mcfletch at rogers.com  Tue Jun  4 12:08:11 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 04 Jun 2002 12:08:11 -0400
Subject: property in Py2.2
References: <3CFC9814.4761F3CB@gol.ge>
Message-ID: <3CFCE5EB.3040108@rogers.com>

Don't really have time to build an example for you, but there's an 
extensive real-world set of examples in basicproperty (part of the 
wxPython Properties Distribution).

http://wxpypropdist.sf.net/

HTH,
Mike

Giorgi Lekishvili wrote:
> Hi all!
> 
> I wonder if someone finds time to send me a real code of how to use
> properties in Py2.2
> 
> I would be much obliged, if you also send me a version of the code for
> Py2.1, with __setattr__  and __getattr__.
> 
> Thx,
> Giorgi
> 
> PS. Prior to post here, I have read the materials on www.python.org. I
> find them very helpful.
> 


-- 
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From hesterloli at hotmail.com  Tue Jun 18 18:01:28 2002
From: hesterloli at hotmail.com (George Hester)
Date: Tue, 18 Jun 2002 22:01:28 GMT
Subject: python cgi + IIS
References: <3D0F1813.3010304@gmx.de> <x0NP8.165989$Fp1.91529@atlpnn01.usenetserver.com>
Message-ID: <Y4OP8.46694$GY.14566397@twister.nyroc.rr.com>

Have you noticed this guy/gal has errors in their IIS?

http://www.e-coli.net/pyiis_server.html

--
George Hester
_________________________________
"Steve Holden" <sholden at holdenweb.com> wrote in message
news:x0NP8.165989$Fp1.91529 at atlpnn01.usenetserver.com...
> "Axel Grune" <NineOfSix at gmx.de> wrote in message
> news:3D0F1813.3010304 at gmx.de...
> > Hi, I'm trying to get my IIS (Win2k Pro) execute my python scripts.
> > I added the 'App Mapping' info for '.pycgi' files in the properties of
> > IISAdmin (Computermanagement) as I read in the internet. The string
> > for the executable file is 'D:\Python22\python.exe -u %s %s' (the path
> > is valid).
> > But when i try to call it within my browser, it just displays it like a
> > normal text file instead of executing it.
> >
> >
> > the script (its called test.pycgi):
> > #!/usr/bin/python
> >
> > print "Content-Type: text/plain\n\n"
> > print "Hallo Welt"
> >
> >
> > has anyone an idea?
> >
> You have almost certainly made a typo, or you didn't apply the changes to
> the directory you are trying to run the CGI from. I just followed the steps
> at
>
>     http://www.e-coli.net/pyiis.html
>
> to install .cgipy as a Python-executing extension, and everything worked --
> the following program printed out a complete list of environment variables.
> #
> # Python CGI script
> #
> import os
> print """Content-Type: text/html
>
> <html>
> <head>
>   <title>Python Web Page</title>
> </head>
> <body>
> """
>
> for k, v in os.environ.items():
>     print "%s=%s<br>\n" % (k, v)
>
> print """
> </body>
> </html>
> """
>
> Personally I'd rather run CGI than use ASP under IIS, so I hope you can get
> it to work. Clearly at the moment IIS isn't recognising the .cgi extension
> as needing to run the Python interpreter.
>
> regards
> -----------------------------------------------------------------------
> Steve Holden                                 http://www.holdenweb.com/
> Python Web Programming                http://pydish.holdenweb.com/pwp/
> -----------------------------------------------------------------------
>
>
>
>
>




From tim.one at comcast.net  Fri Jun 21 18:32:33 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 21 Jun 2002 18:32:33 -0400
Subject: TimeoutQueue.py
In-Reply-To: <Xns9234F249D7D85cliechtigmxnet@62.2.16.82>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEDJPPAA.tim.one@comcast.net>

[Chris Liechti]
> Anyone ever needed a Queue with timeout?

Yes.  I encourage you to submit a patch to SourceForge, adding optional
timeout arguments to the std Queue module (include the requisite doc changes
and test suite changes too, and I'll secretly ram it in <wink>).  Queue was
written long before threading.py, and, as you noticed, the blocking
operations in threading.py generally do support optional timeouts.  They're
a Good Idea for all blocking operations.





From aahz at pythoncraft.com  Thu Jun 20 01:23:34 2002
From: aahz at pythoncraft.com (Aahz)
Date: 20 Jun 2002 01:23:34 -0400
Subject: Flame fests (was Re: python version?)
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <3D107948.B3FEE4BB@engcorp.com> <aer53s$7is$1@panix1.panix.com> <3D112BA6.4CA5F584@engcorp.com>
Message-ID: <aerosm$p18$1@panix1.panix.com>

In article <3D112BA6.4CA5F584 at engcorp.com>,
Peter Hansen  <peter at engcorp.com> wrote:
>Aahz wrote:
>> 
>> Everybody have fun playing "kick the newbie"?  Maybe it's time to quit?
>
>I think we'd all object strongly to the characterization that we were 
>"kicking a newbie".

Well, then, if you're not kicking a newbie, you're swallowing trolled
bait.  Which makes you feel better?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From mwh at python.net  Thu Jun 13 11:00:28 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 13 Jun 2002 15:00:28 GMT
Subject: filter2
References: <mailman.1023975618.6474.python-list@python.org>
Message-ID: <lkptyvmdfz.fsf@pc150.maths.bris.ac.uk>

Oleg Broytmann <phd at phd.pp.ru> writes:

> Hello!
> 
>    I want to have new python builtin - filter2. It is exactly like filter,
> but it returns also a list of items that didn't pass the test.
> 
>    Of course I can do it with just "for" loop:
> 
> l1 = []
> l2 = []
> for item in alist:
>    if test(item):
>       l1.append(item)
>    else:
>       l2.append(item)
> 
> but this way I am missing filter's loop written in C.

Not much, I suspect.

>    Running filter two times (with test and inverted test) is not an option,
> as test could be a costly function.
> 
>    Opinions?

Write it in Python.  Particularly if test is expensive the loop
overhead will disappear into the noise.

(with-seriousness-level :low
  If you insist on C loops, there's always:

      lt = map(lambda x:(test(x), x), alist)
      l1 = filter(lambda (x,y):x, lt)
      l2 = filter(lambda (x,y):not x, lt)

  But I doubt you need me to tell you why this solution sucks.)

Cheers,
M.

-- 
  I'm not sure that the ability to create routing diagrams 
  similar to pretzels with mad cow disease is actually a 
  marketable skill.                                     -- Steve Levin
               -- http://home.xnet.com/~raven/Sysadmin/ASR.Quotes.html


From peter at engcorp.com  Sun Jun  9 02:52:43 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 09 Jun 2002 02:52:43 -0400
Subject: Correct code/unit testing (was Re: Efficient python programming...)
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com> <834rgd5isv.fsf@panacea.canonical.org>
Message-ID: <3D02FB3B.CF75E925@engcorp.com>

Kragen Sitaker wrote:
> 
> Peter Hansen <peter at engcorp.com> writes:
> > You forgot the even more important first thing for a beginner:
> > get it correct!  The only good way to do this is to write unit
> > tests that prove it.  If you don't have tests, you don't have
> > working code (or how can you prove it?).
> 
> I agree that correctness is important, although I think it is
> sometimes overemphasized; buggy programs can still be useful.  (I hate
> them, though.)  For example, in the ICFP programming contests, any
> entry found to contain a bug, however small, is automatically
> disqualified.
> 
> I also agree that unit tests are very useful for reducing bugginess.
> 
> But I do not think that they are either necessary or sufficient for
> writing correct code.  

Without tests, how much confidence can you have in the correctness
of your code?  It's been said "you don't know it works if you 
don't test it".

> I think (and extensive experimentation has
> shown) that careful reasoning and code reviewing is necessary and
> sufficient to write correct code.

Whoa!  "extensive experimentation has shown"??  I would suggest that
what extensive experience (I change the word deliberately) has shown
many of us is that careful reasoning and code reviewing, no matter
how extensive, will *always* leave bugs in the code.  Have you never
spent a large amount of time carefully designing and coding a wonder
little program, and checked it over very carefully, even with a peer,
and later found a bug in it?  Maybe you're just a better programmer
than I, and many others, but I would say this inevitable outcome is 
the source of the term "egoless programming" -- if you have a big
ego, you'll be devastated by the repeated proof that you will never,
ever get it right the first time, or even after lots of reviewing.

> It is certainly possible to err too far in the direction of believing
> your code correct simply because it passes its test suite, and this is
> a common beginner's error.  I think it is also possible to err too far
> in the direction of writing and running tests instead of desk-checking
> code.

I can only suggest that until you have tried test-first development,
in which the code is written in very, very small steps only to make
small failing tests pass, and discovered the possibilities of 
emergent design, and other aspects of this approach, then we will not
find much common ground in this discussion.  I simply disagree with
your last sentence above, and I believe beginners _never_ write
unit tests anywhere that I've seen, so I disagree with the first
one too. :)

-Peter


From kjetilho at haey.ifi.uio.no  Sun Jun  9 09:40:21 2002
From: kjetilho at haey.ifi.uio.no (Kjetil Torgrim Homme)
Date: 09 Jun 2002 15:40:21 +0200
Subject: Python 2.2.1 vs. gcc 3.1?
References: <yu99vg8v6orb.fsf@europa.research.att.com> <15616.57963.519636.730290@12-248-41-177.client.attbi.com> <mailman.1023468673.31246.python-list@python.org>
Message-ID: <1radq4bmey.fsf@fimm.ifi.uio.no>

[Andrew Koenig]:
>
>   PS: I can't find a "ldconfig" command on my system.

it's called crle (Solaris 8 and newer).  I do not recommend its use
over rpath except in the most desparate situations.

-- 
Kjetil T.		your computer can help researchers understand how
			proteins work.    http://folding.stanford.edu


From emile at fenx.com  Thu Jun 27 09:41:22 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 27 Jun 2002 13:41:22 GMT
Subject: [Newbie]  How to manipulate variable _names_ ?
References: <3d1b0723.6179705@news.texas.net>
Message-ID: <6CES8.325755$352.35447@sccrnsc02>

Ben Fairbank
> I have a program that requires the user to enter a file name, such as
> "othello," which the program will then open, process, and create a
> dictionary based on the contents of the file.  I would like to name
> the dictionary "OthelloCount," and then go on and create more
> dictionaries, each with a name made up of the inputfile name as
> provided by the user, concatenated with "Count," but I do not know how
> to make a variable name based on input.
>

One way is to use getattr:

import sys

def orderspam(): return 'Hmm, spam?'
def ordereggs(): return 'Hmm, eggs?'
for ii in ('spam','eggs'):
    print getattr(sys.modules[__name__], 'order%s' % ii)()


Or, you could create a single dictionary, counts, with keys like othello
and values being the results of your processing.

HTH,

--

Emile van Sebille
emile at fenx.com

---------



From pyth at devel.trillke.net  Thu Jun 13 17:50:28 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 13 Jun 2002 23:50:28 +0200
Subject: Type subclassing: bug or feature
In-Reply-To: <20020613232811.I6609@prim.han.de>; from pyth@devel.trillke.net on Thu, Jun 13, 2002 at 11:28:11PM +0200
References: <aeau48$111$1@panix1.panix.com> <20020613232811.I6609@prim.han.de>
Message-ID: <20020613235028.J6609@prim.han.de>

To Aahz wrote:
> Aahz wrote:
> > Consider the following code:
> > 
> > class MyStr(str):
> >     def contains(self, value):
> >         return self.find(value) >= 0
> > 
> > s = MyStr("hello, world!")
> > s = s.capitalize()
> > if s.contains('Hello'):
> >     print "Found it!"
> > 
> > It fails with an AttributeError when it calls s.contains(), because
> > s.capitalize() returned a str instead of a MyStr.  Anyone want to take a
> > whack at defending this as the correct behavior?
> 
> class Names(str):
>     def __init__(self, namelist):
>         str.__init__(self, ":".join(namelist))
> 
> what would you expect 
> 
>     Names([1,2,3]).capitalize()

oops. sorry. should have checked it better (or at all<self.wink>)...

class Names(str):
    def __new__(cls, namelist):
        return str.__new__(str, ":".join(namelist))

n=Names(['tim','peter','guido'])

But hopefully my point was understandable that it is hard 
to guess a return type for 'n.capitalize()' now ...

> to return? An exception? I think it's dangerous to assume anything
> about the subtyped classes' constructor. You *do* need a constructor
> because strings are immutable so 'str.capitalize' can't modify
> anything in place.
> 
> if-you-subtype-cake-you-gotta-eat-it-all-ly y'rs, holger

cheers,

    holger




From maxm at mxm.dk  Thu Jun 13 03:15:13 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 13 Jun 2002 09:15:13 +0200
Subject: Zope Newbie query
References: <ugefre2980kj09@corp.supernews.com>
Message-ID: <3D084681.9090601@mxm.dk>

Tony B wrote:

> I have been playing with zope and would like to use it properly now the
> problem I'm encountering is: How do I remotely install zope on a raq4 on
> which I am not the main raq4 administrator (i.e I have a subdirectory on
> which my site is hosted).? Do I have to have root priveliges to install and
> run zope?


No Zope can run just fine as any user. You just install Zope and make it 
run on port 8080 or another free port.

Then you get the root to set up proxy for your Zope through Apache. You 
then install VirtualHostMonster on Zope and you are good to go.

Well there are a few steps involved but it's not impossible. The point 
is that it must run on a port that non-root users can use, and then get 
proxied from Apache.

If you don't need Zope to run on port 80 through Apache, you can just 
install Zope an make it run directly on port 8080. No further setup 
nessecary. This will most likely give you problems with firewalls though.

regards Max M



From kragen at pobox.com  Sat Jun  1 15:23:24 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 01 Jun 2002 15:23:24 -0400
Subject: Where can I find .join() in the docs
References: <mailman.1022766950.1396.python-list@python.org> <oIqJ8.118521$L76.198853@rwcrnsc53>
Message-ID: <83y9dyvm4z.fsf@panacea.canonical.org>

"Emile van Sebille" <emile at fenx.com> writes:
> Michael Chermside
> > Specifically, I went to the Python Library Reference (which you will
> > find stashed under your pillow), Chapter 2 (built-in stuff), then went
> > to the section on sequences (which Strings are) and the page on
> Strings.
> 
> ...which illustrates a difficulty for a newbie... knowing it's 'built-in
> stuff' then further knowing that strings are classified under
> sequences...

Yeah, what I did is I read straight through the Language Reference the
first time (didn't understand much, but kind of got the lay of the
land) and eventually learned that the index of the Library Reference
was the place to look for such things.  It took me a while to stop
looking in the Language Reference for things like methods of lists.



From sholden at holdenweb.com  Thu Jun 13 22:46:18 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 13 Jun 2002 22:46:18 -0400
Subject: Socket error problem
References: <mailman.1023661033.29781.python-list@python.org>
Message-ID: <gMcO8.9548$Fp1.7987@atlpnn01.usenetserver.com>

"David LeBlanc" <whisper at oz.net> wrote in message
news:mailman.1023661033.29781.python-list at python.org...
> Ok, that solved the problem. It was caused by the app using it's own
version
> of sockets; overriding the default implementation to create something
called
> a "TimeoutSocket", which don't work on Windows for some reason (something
to
> worry about later if necessary). Fortunately, all that was required to go
> back to using standard sockets was to remove the override. After chasing
> down a few additional bugs (it seems to ignore it's own config options or
> the doc isn't good enough to describe how to set them properly), it now
> actually slurps mail!! :)
>
> The author of the package said it would never work on Windows - hah! :)
> (Actually, I'm not even going to bother with the QT GUI interface,
although
> I'm sure that's doable as well.)
>
> Thanks for the help!
>

Well, that's strange: I have used timeoutsockets on Windows with great
success. Perhaps you are using an old version, some of which were extremely
buggy.

You'll be happy to know that the timeout features are likely to make it into
standard sockets for 2.3.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From shalehperry at attbi.com  Fri Jun 21 12:51:20 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 21 Jun 2002 09:51:20 -0700 (PDT)
Subject: [newbie] Overwriting class definitions
In-Reply-To: <3D13563D.AB80D324@athensgroup.com>
Message-ID: <XFMail.20020621095120.shalehperry@attbi.com>

On 21-Jun-2002 Mike Christie wrote:
> I'm a complete newcomer to Python; I'm teaching myself Python and this is
> just
> about the only Python code I've ever typed in, so my apologies for posting a
> dumb beginner question.  Here's my problem.
> 

tutor at python.org exists exactly to answer new to python (and to programming)
questions.

> 
> Then I did this:
> 
> import circle
> circle.Circle()
> 
> and got
> 
> Hello, version 1
> 
> as expected.  Then I edited the file to say "version 2" and saved it, and
> then
> typed in the import statement and circle.Circle() statement again.  It says
> version 1 again.
> 
> So what's happening?  I assumed that when I imported the file again, it would
> read in the class definition again and overwrite the old definition.  But
> that
> doesn't appear to be happening.  When I exit Pythonwin and reenter, it runs
> just fine.
> 

if you do:

import foo
import foo
import foo

only the first import happens.  python does a check, sees it has done the
work and ignores the rest.  This is what happened to you.

I find that existing the interpreter is the safest thing to do while doing
interative coding like this.




From eddie at holyrood.ed.ac.uk  Fri Jun 14 15:44:44 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Fri, 14 Jun 2002 19:44:44 +0000 (UTC)
Subject: a possibly foolish question about slices
References: <yu99sn3prbqm.fsf@europa.research.att.com> <aedd8k$brf$1@scotsman.ed.ac.uk> <m3znxxu21l.fsf@mira.informatik.hu-berlin.de> <aedf1p$bv1$1@scotsman.ed.ac.uk> <yu99k7p1r77s.fsf@europa.research.att.com>
Message-ID: <aedh3c$c3d$1@scotsman.ed.ac.uk>

Andrew Koenig <ark at research.att.com> writes:

>Eddie> Hmm, I was actually trying to suggest (non-seriously) fixing
>Eddie> zero based indexing! but I realise now that wasn't clear.

>It's hard to imagine how it would be possible to change to 1-based
>indexing without breaking lots of code.

No pain, no gain :)

>For that matter, I had the thought that it might be possible simply
>to define x[n:0] as meaning x[n:len(x)] if n<0, but I imagine that
>even that would break too much code.

No, I was experimenting along those lines but couldn't see a way to do it
without breaking compatability.

Of course that solution from that clever David Eppstein chap works for the
case you asked about.  The case I always seem to want to do is chop n
characters of the end of a string where n sometimes turns out to be 0.  Still
the '[:n or maxint]' solution is good enough.

Eddie


From mgerrans at mindspring.com  Fri Jun  7 05:08:38 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Fri, 7 Jun 2002 02:08:38 -0700
Subject: How do you read unicode files?
References: <adpd7h$6sv$1@slb3.atl.mindspring.net> <m3vg8vaa9u.fsf@mira.informatik.hu-berlin.de>
Message-ID: <adpt72$dds$1@nntp9.atl.mindspring.net>

Thanks Mike, Francis and Martin for the quick replies!

My mistake was using readlines() to read the text file, instead of read().
(I should have posted a snippet of code in the first place and probably you
all would have noticed that immediately).

- Matt




From d_blade8 at hotmail.com  Thu Jun 20 16:23:01 2002
From: d_blade8 at hotmail.com (Lemniscate)
Date: 20 Jun 2002 13:23:01 -0700
Subject: Transforming Clusters into a Color Map
References: <7396d2b2.0205291424.49a9e95f@posting.google.com> <mailman.1022712528.1085.python-list@python.org>
Message-ID: <7396d2b2.0206201223.7e9ff87e@posting.google.com>

Thanks a bunch, I am working on it right now.  Actually, though, I was
wondering if I could ask you another question.  The kind of heatmap I
want is a bit different than the screenshot.  It is more like a
patchgrid (if you go to
http://www.cs.uml.edu/~mtrutsch/research/High-Dimensional_Visualizations-KDD2001-color.pdf
and go to section 3.3, you'll see a definition (and example) of a heat
maps, in the way I am using the word.  It's more like a spreadsheet
with colors instead of numbers.  I don't know if that makes it easier
or harder.  I'm sorry if I seem a bit dense on the subject but I have,
thus far, only used programs as ways in which to process data, not
actually display it in such a format.  I would think it would be
simple, but I know that I have a ton to learn, so it could be simple
or very complex.  Thanks a bunch,

Lem


Tim Churches <tchur at optushome.com.au> wrote in message news:<mailman.1022712528.1085.python-list at python.org>...
> Lemniscate wrote:
> > 
> > Hi all,
> > 
> > I have recently been asked to write a program that will operate on
> > huge arrays of numbers.  As such, I have quickly immersed myself in
> > Numeric.  Things are going well (Numeric, in my experience, is
> > extremely effective and relatively easy to learn).
> > 
> > However, I've been informed that I need to 'color map hotspots' in the
> > results.  For example, my boss wants me to transform the results (of,
> > for example, standard deviations run across each array) and then
> > cluster them by color (I can explain more if anybody needs it).
> > 
> > There are a few programs that do this but they all have propriety
> > formatting of the results (which precludes some of the things we want
> > to do) and limits onto the types of comparisons that can be done.
> > Ideally, we would want our own implementation that we could use in any
> > way we wanted.
> > 
> > There isn't alot of support for this type of programming at my
> > company, so the hope is that I can get a 'minimally' functional
> > prototype so we can see what can be done.
> > 
> > >From the people I've talked with, color-maps like these are actually
> > fairly common, so I was hoping that somebody could point me in the
> > right direction (maybe a module already exists to do this that
> > interfaces with Numeric, etc.)  I'm doing some searches but I'm not
> > having a whole lot of luck (right now, I am going through stuff like
> > "Numeric Color" etc.  Thanks a bunch.
> 
> Take a look at the R package - see http://www.r-project.org
> in particular the screenshots page - there is a nice "heat map" of
> a volcano at the bottom of the page. 
> 
> You can transfer your data to R via delimited text files, but if you 
> are using Unix or Linux, then have a look at Walter Moreira's RPy module
> (see http://rpy.sourceforge.net ) which embeds R in Python and allows
> NumPy arrays and other Python data structures to be transparently
> transferred into the R environment (and acted upon there).
> 
> Tim C


From tonycradwick at clear.net.nz  Mon Jun 24 06:26:49 2002
From: tonycradwick at clear.net.nz (Tony Cradwick)
Date: Mon, 24 Jun 2002 10:26:49 GMT
Subject: converting strings to numbers
Message-ID: <3d16f3e0@clear.net.nz>

I am new to python.

How do you convert (cast?) a string, eg from a user input entry box, to a
number, eg integer?





From eddie at holyrood.ed.ac.uk  Thu Jun 20 10:28:32 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Thu, 20 Jun 2002 14:28:32 +0000 (UTC)
Subject: Advice sought on ps module
References: <6e990e29.0206200325.355f0413@posting.google.com>
Message-ID: <aesoqg$kc9$1@scotsman.ed.ac.uk>

hjwidmaier at web.de (Hans-Joachim Widmaier) writes:

>About 2 years ago I wrote a primitive PostScript module to help me
>create various things like CD inlays or diagrams. It served that
>purpose quite well, despite its primitiveness.
>Since long I wanted to rewrite the whole thing from scratch and do it
>(at least kind of) "right." But I'm still not so sure as which way to
>go. At one time I thought I might employ ghostscript, which would have
>enabled me to, e. g. get the string width for some text, so I can make
>adjustments (not only to the text). Apart from that I couldn't get a
>sensible communication with gs through popen2, the PostScript output
>is created somewhat nonlinear, making it harder to ensure that the gs
>context is correct (in that query case).
>I'm convinced that's the wrong way, but the alternative doesn't look
>too good either. If you want to do some optimizations, you have to
>double some functionality of the ps interpreter.

>Examples:
>  - To calculate the with of a string, you have to read the AFM file,
>add the character widths, aplly kerning, transform it with the CTM and
>the fonts TM.
>  - To know whether a font must be reset or is already the current
>font you have to keep track of all the save/restore, gsave/grestore
>and the like.

>I do not want to write a "do all, end all" PostScript module. I do not
>even want to give it away (I'd be simply too embarrassed).

>Still I hope that some of you might give me some good advice on how to
>best do it.

I'm not sure that I understand completely what you're trying to do so maybe
I'm barking up the wrong gum tree here but ...

Surely you want to use postscript itself to do all the work.  PS already has
an operator called stringwidth and a complete set of programming constructs so
the trick usually is to generate the appropriate PS commands.  As an example,
to draw a border around some arbitrary text I guess you could generate code to
do something like:  save current position, for each string you draw calculate
the bounding box and remember the highest X value and at the end use the new
position along with the saved position and highest X value to define the four
corners of a rectangle.  It does require a fair degree of competence in PS but
you wouldn't tackle such a project without it (or a desire to achieve it)
would you :)

Eddie


From benjl at cse.unsw.edu.au  Fri Jun 14 03:33:10 2002
From: benjl at cse.unsw.edu.au (Benno)
Date: 14 Jun 2002 00:33:10 -0700
Subject: Zip makes sense! was: Re: why not "'in' in 'in'"?
References: <mailman.1024021164.12598.python-list@python.org>
Message-ID: <332c7c11.0206132333.199ba913@posting.google.com>

"David LeBlanc" <whisper at oz.net> wrote in message news:<mailman.1024021164.12598.python-list at python.org>...
> > -----Original Message-----
> > From: python-list-admin at python.org
> > [mailto:python-list-admin at python.org]On Behalf Of Tim Peters
> > Sent: Thursday, June 13, 2002 18:57
> > To: python-list at python.org
> > Cc: Grant Griffin
> > Subject: RE: why not "'in' in 'in'"?
> >

<snip discusion about 'in'>
 
> Hmmm... How about calling it "contains" instead of "in"? "if 'feefifofum'
> contains 'fi'" seems nicer then "if 'fi' in 'feefifofum'" maybe?
> 
> Of course maybe this makes too much sense for a language where zip has zip
> to do with compression and yield nothing to do with threads.


I'm guessing you are mentioning this fairly tounge in cheek, however zip makes
a lot of sense to anyone coming from Haskell. If I was after the zip compression
algorithm i would look in the hypothetical compression package.

it-made-perfect-sense-to-me-ing-ly yours Benno


From kragen at pobox.com  Mon Jun 10 20:52:12 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 20:52:12 -0400
Subject: __imul__ broken for 'objects' in 2.2.1
References: <yfsd6v3qa86.fsf@black132.ex.ac.uk> <3n79guskvoaeqtr4srg11kaqse7jcqutt4@4ax.com> <yfswut76yat.fsf@black132.ex.ac.uk>
Message-ID: <83bsaimybn.fsf@panacea.canonical.org>

Alexander Schmolck <a.schmolck at gmx.net> writes:
> I really wonder where this bizzarre behavior comes from -- it is really none
> of the interpreter's business to take guesses whether my class is a sequence
> or not -- finding out if there is a problem with the supplied arguments should
> be left to the discretion of the method itself.

I think it's a bug, but I can easily imagine how it could happen.

At C level, the sequence protocol and the number protocols are
different protocols.  They both have slots in their "vtables" that are
invoked by *.  Class instances can be either sequences or numbers;
apparently Python is deciding that since instances can be sequences,
it will use the sequence methods (or at least the sequence
implementation of *).



From kragen at pobox.com  Sat Jun  1 17:01:38 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 01 Jun 2002 17:01:38 -0400
Subject: User input question
References: <ad4ovf$tqev9$1@ID-49758.news.dfncis.de> <ad4qbk$sd6$1@wanadoo.fr> <ad4stt$uau73$1@ID-49758.news.dfncis.de> <ad4t64$qli$1@wanadoo.fr> <ad4uq3$trgg5$1@ID-49758.news.dfncis.de> <ad50pc$9s5$1@wanadoo.fr>
Message-ID: <83bsauu30t.fsf@panacea.canonical.org>

Eric Brunel <eric.brunel at pragmadev.com> writes:
> I also feel I have to warn you: debugging a CGI script can be *very* 
> frustrating... It's quite difficult to set the whole thing to work, and 
> when you make errors in your script, you won't get any coherent message 
> from the web server.

I disagree; if you use the cgitb module, you'll get better exception
messages from the web server than you'd get in a terminal, and better
than you'll get from any other language.

> ...: the FieldStorage class does it for you. You'll soon find you
> can't live with this class!

FieldStorage isn't perfect, but it isn't *that* bad --- I don't have
any trouble living with it!



From mwh at python.net  Thu Jun 13 05:20:53 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 13 Jun 2002 09:20:53 GMT
Subject: Extension objects in Python 1.5 and 2.2
References: <20020613071813.GA16764@strakt.com> <15624.19045.700125.215173@12-248-41-177.client.attbi.com> <mailman.1023953777.18976.python-list@python.org>
Message-ID: <lk3cvry1pi.fsf@pc150.maths.bris.ac.uk>

Martin Sj?gren <martin at strakt.com> writes:

> On Thu, Jun 13, 2002 at 02:31:49AM -0500, Skip Montanaro wrote:
> > 
> >     Martin> How do I write extension objects so that they work in both
> >     Martin> Python 1.5 and the 2.x series? 
> > 
> > Michael Hudson has been working on a new include file named pymemcompat.h.
> > Since it's only used for backward compatibility it's in the Misc directory.
> > You can grab it from CVS:
> > 
> >     http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/dist/src/Misc/pymemcompat.h
> > 
> > The idea is that you should be able to include it and program to the 2.3
> > memory api in code that runs under 1.5.2 or newer versions of the
> > interpreter.
> 
> Cool. Thanks a lot!

Pleasure!  Be sure to complain if it doesn't work, won't you?

Cheers,
M.

-- 
  An encyclopedia is about being as accurate as it can, not being
  evenly misinformed.                     -- Coby Beck, comp.lang.lisp


From gorny0 at zonnet.nl  Sat Jun 15 08:43:34 2002
From: gorny0 at zonnet.nl (Gorny)
Date: Sat, 15 Jun 2002 14:43:34 +0200
Subject: socket raw
References: <aedj3v$dk2$1@newsreader.mailgate.org>
Message-ID: <aefcpn$8n7$1@reader05.wxs.nl>

----- Original Message -----
From: "LoGan" <giannivolontosky at TOGLIyahoo.it>

> hi, id like to make a python script that use socket raw,

erhm.. raw sockets are mostly very hard to use and if you
want to spoof packets you can *of course* not receive the
responding packets, unless you're on the same segment
and are sniffing.

> i found the btk
> module but it cant handle received packets,

you can use the pcap() class in the btk module. the parsing
part still sucks but you can have the whole binary packet
date in your hands

> it's really a good module but
> useful only to send,

thanks :)
At the moment I'm rewriting in fact the whole billy the kid
module because it was badly and fast written. It's now more
transparent, portable and so. Support for more protocols
is going to be added including some (if it's even possible??)
socket class merging the pcap() and btk() classes together.
This will create a raw socket implementation but it'll be a very
hard job to make it a good one handling just as normal sockets.

> anyone could help me ?
> bye

Gorny

--
"Spread the sources"
http://gorny.cjb.net




From buzzard at urubu.freeserve.co.uk  Tue Jun 25 17:37:06 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Tue, 25 Jun 2002 22:37:06 +0100
Subject: ? and %s placeholders, help?
References: <af5oq7$sj1$1@news8.svr.pol.co.uk> <m2k7oolcrm.fsf@video.paradise.lost> <3D178542.604@mxm.dk> <af87cj$n5b$1@newsg3.svr.pol.co.uk> <23891c90.0206250257.23d38d30@posting.google.com>
Message-ID: <afao2f$rag$1@newsg2.svr.pol.co.uk>

Thanks all,
                I now have it.  The '%s' placeholder was required.  (But all
the examples I could find used '?' which was a bit unfortunate.)  Cheers.

Duncan




From Uwe.Mayer at ifib.uni-karlsruhe.de  Mon Jun 24 05:20:37 2002
From: Uwe.Mayer at ifib.uni-karlsruhe.de (Uwe Mayer)
Date: Mon, 24 Jun 2002 11:20:37 +0200
Subject: Generic Python
Message-ID: <3D16E465.F0D2509E@ifib.uni-karlsruhe.de>

Hi,

is it possible to to write a class which, f.e. takes an argument in
__init__() and that doesn't return an instance object but a new class
object?

The problem is that I've got a base class and there should be many
subclasses of it. Each subclass just overwriting an output method and
serving as a template: 

class base1:
    def __init__(self):
        ...
    def output(self):
        ...

class template1(base1):
    def output(self):
        ...

class template2(base1):
    def ouptut(self):
        ...

I need many, many of these "template" classes, so a sorter way would be
to accustom the base class to take another argument which then outputs
the right thing:


class base2:
    def __init__(self, text):
       ...
    def output(self):
       print self.text

template1 = base2('some text')
template2 = base2('another example')
...

My problem with this is that this produces instance objects and in the
above example I had class objects. I can hardly use the latter case
because I would always modify the original.
I also cannot create new templates as I need them, because I'm writing
utility classes which are much more complex and take much more
arguments.

Is it somehow possible to have the "base2" class return a class instead
of an instance?
Is there another way to solve this problem?

Thanks in advance
Uwe


From peter at engcorp.com  Wed Jun  5 10:40:34 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 05 Jun 2002 10:40:34 -0400
Subject: Pop a list from beginning ? and memory saving...
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de> <3CFA0676.68F80DE6@engcorp.com> <add289$vud42$1@ID-146704.news.dfncis.de> <Xns9221CE2AF76D0cliechtigmxnet@62.2.16.82> <adhrsl$118kvn$1@ID-146704.news.dfncis.de> <mailman.1023194347.14669.python-list@python.org> <3CFD5B60.90831F21@engcorp.com> <aIeL8.120588$Kp.12235125@bin7.nnrp.aus1.giganews.com> <3CFD86D4.C4F81404@engcorp.com> <3CFD8747.4AA8D29F@engcorp.com> <mailman.1023280688.15830.python-list@python.org>
Message-ID: <3CFE22E2.E88C42C9@engcorp.com>

Jeff Epler wrote:
> 
> On Tue, Jun 04, 2002 at 11:36:39PM -0400, Peter Hansen wrote:
> (speaking about this expression):
> >> l[:] = []
> >
> > Oh, and in my opinion, that is obscure and deserves a
> > comment explaining how it works, and why it is being done.
> > I might not feel that way if I write that idiom a few times,
> > though. :)
> 
> IMO
>     del l[:]
> is clearer than
>     l[:] = []
> I'd accept the first one immediately, and puzzle over the second one for
> a short (or even long) time.

Agreed, although my brain knows that "x = l[:]" is the idiom
for making a copy of a list, so it wants to tell me that "del l[:]"
is just going to make a copy, then delete it right away.

-Peter


From pyth at devel.trillke.net  Sun Jun  2 16:05:45 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 2 Jun 2002 22:05:45 +0200
Subject: List
In-Reply-To: <3cfa7619_1@news.iprimus.com.au>; from occeanlinux@linuxmail.org on Sun, Jun 02, 2002 at 07:56:58PM +0000
References: <3cfa6ed8_1@news.iprimus.com.au> <addrjh$imq$1@news.tudelft.nl> <3cfa7619_1@news.iprimus.com.au>
Message-ID: <20020602220544.C30380@prim.han.de>

Gold Fish wrote:
> Charl P. Botha wrote:
> 
> > In article <3cfa6ed8_1 at news.iprimus.com.au>, Gold Fish wrote:
> >> Assume we got 2 list
> >> list 1 = ['ABC','CDG','DAV','FE','FECA','FEAFA','FEA']
> >> list 2 = []
> >> how can we copy element from list 1 to list 2 say i just want to copy the
> >> 3 element 'CDG','DAV','FE' to list2 only.
> > 
> > list1 = ['ABC','CDG','DAV','FE','FECA','FEAFA','FEA']
> > list2 = list1[1:4]
> > 
> > Is that what you meant?
> > 
> No i mean if i got from list1 i wanna take 3 elements in any position to 
> list2 but i don't know how many item in that list.

if you have a starting index 'i' then you can do

  list1[i:i+3]  

which returns you at most 3 elements, starting from position 'i'.

If this is still not what you want, then please give 
two or three examples of what you want to achieve.

    holger




From usenet at thinkspot.net  Sat Jun  1 16:51:28 2002
From: usenet at thinkspot.net (Sheila King)
Date: Sat, 01 Jun 2002 13:51:28 -0700
Subject: How to open a HTML file when the python cgi program is executing?
References: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de> <ad95l1.i0.1@kserver.org> <ad9vsm$vhjnu$1@ID-49758.news.dfncis.de> <9m6ada.31b.ln@10.0.0.1> <adaeh8$uum7t$1@ID-49758.news.dfncis.de> <ada3k7.lo.1@kserver.org> <83n0uevjvx.fsf@panacea.canonical.org>
Message-ID: <adajh1.13g.1@kserver.org>

On 01 Jun 2002 16:12:02 -0400, Kragen Sitaker <kragen at pobox.com> wrote in
comp.lang.python in article <83n0uevjvx.fsf at panacea.canonical.org>:

> Sheila King <usenet at thinkspot.net> writes:
> > print "Location: http://replacement-url\n"
> > 
> > You should have nothing before this, and there's no point in having
> > anything after it, since the page is being redirected.
> 
> Some browsers (in particular old versions of Opera) have problems with
> redirects every once in a while; that's why redirects traditionally
> contain HTML, in case the browser is broken.

I posted a short, quick, quip of a reply to your message. Although I've
cancelled it now, it will probably still show up on most news servers.

In any case, a Google Groups search does find support for your statements
(which I'm sure is no surprise to you).

As an example:
http://groups.google.com/groups?as_umsgid=DoBG3t.LMv%40world.std.com

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From amuys at shortech.com.au  Tue Jun  4 20:40:41 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 4 Jun 2002 17:40:41 -0700
Subject: 'for every' and 'for any'
References: <20020526135944.A32690@hishome.net> <20020526091742.A987@unpythonic.net> <mailman.1022567685.7803.python-list@python.org> <slrnaf7af3.8kn.quinn@schilling.ugcs.caltech.edu> <mailman.1022658469.27441.python-list@python.org> <7934d084.0206031636.5c29b31f@posting.google.com> <mailman.1023177066.17333.python-list@python.org>
Message-ID: <7934d084.0206041640.69f1974c@posting.google.com>

Gerhard =?unknown-8bit?Q?H=E4ring?= <gh_pythonlist at gmx.de> wrote in message news:<mailman.1023177066.17333.python-list at python.org>...
> * Andrae Muys <amuys at shortech.com.au> [2002-06-03 17:36 -0700]:
> > C dosn't have any 32-bit int types standardised, so what do _you_ use
> > when you need one?
> 
> Do you need 32 bit ints or >= 32 bit ints?
> 
Normally >=32, but in that case anyone using anything other than
standard int or long had better be using a domain specific alias
(size_t, time_t, etc).  However when messing with binary data in it's
various forms (network packets, filesystems, executable formats, just
to name a few personal favourates) there are times when you most
definately want N bits *exactly*.  Naturally sometimes N == 32. :)

True, if this integer type is likely to leak into the specific modules
interface I'll generally alias it myself, but internally I usually
won't bother.

Andrae


From skip at pobox.com  Fri Jun  7 13:51:47 2002
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 7 Jun 2002 12:51:47 -0500
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <200206071649.g57GnmC00708@europa.research.att.com>
References: <yu99vg8v6orb.fsf@europa.research.att.com>
        <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
        <200206071649.g57GnmC00708@europa.research.att.com>
Message-ID: <15616.62131.521803.383860@12-248-41-177.client.attbi.com>

    Andrew> That could be it, but I have been setting LD_RUN_PATH explicitly
    Andrew> when building Python, so I don't understand why it would have
    Andrew> changed.

It's been ages since I used a Solaris system, but shouldn't that be
LD_LIBRARY_PATH?  Perhaps there is no ldconfig command (anymore) on Solaris.

Skip





From whisper at oz.net  Tue Jun 11 10:19:53 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 11 Jun 2002 07:19:53 -0700
Subject: Defining a method final
In-Reply-To: <ae4tvv$rvd$1@wanadoo.fr>
Message-ID: <GCEDKONBLEFPPADDJCOEKECJDIAA.whisper@oz.net>

I think the point of what he is asking for has been missed. Final methods
are not necessisarily private, it means they can't be redefined
("overridden") in derived classes, thus freezing the semantics of the
method.

A plausible scenario for something like this is a class that wraps hardware.
While I might like to create derived classes for text or binary I/O on this
widget, the exact sequence of manipulating the hardware is tricky and can
result in expensive damage if not done exactly right. Therefore, I don't
want derived classes changing the read or write methods of the base class.

In another post, I suggested the idea of pulling the base classes methods
down into the dictionary of the derived classes during __init__ I don't know
if this is a pythonic way to do things (it strikes me as being so), but I
think it would have the desired effect.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Eric Brunel
> Sent: Tuesday, June 11, 2002 8:36
> To: python-list at python.org
> Subject: Re: Defining a method final
>
>
> Emile van Sebille wrote:
> > Accepted practice is to use a single underscore for objects to be
> > considered private.  This does not prevent their use, but strongly
> > signals that re-use ought not to be done.
> >
> >>>> class Test:
> > ...     def _private(self): pass
> > ...
> >
>
> Python now have an "official" means to make attributes private by
> prefixing
> them with a double-underscore. And it *does* prevent their use or
> overloading:
>
> class A:
>   def __private(self): pass
> class B(A):
>   pass
> o = B()
> o.__private()
>
> results in:
>
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> AttributeError: B instance has no attribute '__private'
>
> Regarding the original question (final methods), it has however obvious
> drawbacks:
> - the method can be *redefined*: that doesn't mean it can be overloaded,
> but class B can have its own __private method
> - the method is actually *private*, which means that neither sub-classes,
> nor "outside" classes can call it.
>
> I don't think that actual final methods can be defined in Python...
> --
> - Eric Brunel <eric.brunel at pragmadev.com> -
> PragmaDev : Real Time Software Development Tools -
http://www.pragmadev.com
--
http://mail.python.org/mailman/listinfo/python-list





From jiri at baum.com.au  Sat Jun  1 06:12:25 2002
From: jiri at baum.com.au (Jiri Baum)
Date: Sat, 1 Jun 2002 20:12:25 +1000
Subject: How to open a HTML file when the python cgi program is executing?
References: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de> <ad95l1.i0.1@kserver.org> <ad9vsm$vhjnu$1@ID-49758.news.dfncis.de>
Message-ID: <9m6ada.31b.ln@10.0.0.1>

Ken:
> No, I mean at some point in the script, I want to get out of it and use
> ordinary HTML instead (main.html in this example). I don't mean the
> "print" command.

If you realize this before you've output anything, you can use one of the
redirection / refresh headers, such as "Location".

print """Content-type: text/html
Location: http://replacement-url

<p><a href="http://replacement-url">Click here</a></p>
"""

Theoretically it should be sufficient to just give the Location, but put in
a human-readable redirection just in case.

Otherwise, if you've already started, you must continue. Just open the file
and print it.


As for frames, there's the main document, which decides which frames get
shown; and then the server gets separate requests for each of the frames.
You might want to browse a few pages with telnet to get a feel for how
things work.

Jiri
-- 
Jiri Baum <jiri at baum.com.au>           http://www.csse.monash.edu.au/~jirib
  MAT LinuxPLC project --- http://mat.sf.net --- Machine Automation Tools


From zopestoller at thomas-guettler.de  Thu Jun 27 09:14:14 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 27 Jun 2002 15:14:14 +0200
Subject: Module Help
References: <mailman.1025164967.15063.python-list@python.org>
Message-ID: <3D1B0FA6.9080101@thomas-guettler.de>

Juan M. Casillas wrote:

> Hello!
> 
> 	I want to create an empty module from C, and then add
> 	functions and definitions stored in another file, and
> 	some code that I will create on-the-fly, from C.


I don't understand you. Want to you mean with "from C"?
Why not use python for code generating?



> 	The basic problem I have is that I don't know how to create the
> 	empty module (this module doesn't have any representation
> 	in the filesystem) and then how can I add the code and
> 	save the "environment" for this newly created module, how
> 	can I manage the globals and locals dictionaries ? for
> 	each module ?

Do you want to create a text file containing python code, or
create it at runtime dynamically?

thomas



From neal at metaslash.com  Sun Jun  9 10:48:21 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Sun, 09 Jun 2002 14:48:21 GMT
Subject: Simple pychecker question
References: <mailman.1023482712.8801.python-list@python.org> <pan.2002.06.09.00.07.47.445563.6814@metaslash.com> <3D02FDE9.19198F33@engcorp.com>
Message-ID: <pan.2002.06.09.14.48.16.902484.18868@metaslash.com>

On Sun, 09 Jun 2002 03:04:09 -0400, Peter Hansen wrote:

> Neal Norwitz wrote:
>> 
>> I personally think that superfluous use of named args are bad, but that
>> the feature is good when used properly.  For example, with your code
>> above, calling f(f=2) would be better written f(2).  However for
>> Tkinter, named args can be useful when there are a bunch of parameters
>> and you only want to use the last.
> 
> For a routine named and structured like this:
> 
>   def transmit(data, timeout=1.0):
> 
> which of the following would you say was the better call in some code
> that might be remote from that definition above?
> 
> 1.  transmit('this is a string', 5)
> 
> 2.  transmit('this is a string', timeout=5)
> 
> My point: named arguments can increase readability and maintainability.
> That's not superfluous, to me.

I agree that they can enhance readability.  With the example above, 
it's questionable, but that's my opinion.  Suppose the function was 
transmitWithTimeout()?  Or suppose the code looked like this:

	timeout = 5
	transmit('this is a string', timeout)

There really doesn't seem to be a right way, it just depends.

Neal


From tdelaney at avaya.com  Mon Jun  3 02:51:24 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 3 Jun 2002 16:51:24 +1000 
Subject: Problems with exec scoping (solved)
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A344@natasha.auslabs.avaya.com>

> From: Erik Walthinsen [mailto:omega at temple-baptist.com]
> 
> Well, it turns out that I was being an idiot and not checking 
> the source
> of the returned HTML page.  When repr() is run on an object, 
> it returns a
> string such as:
> 
> <captiveclient.CaptiveClient instance at 0x8189264>
> 
> ...which conveniently doesn't get displayed as text in a 
> browser, looking
> as much like HTML as it does.  I discovered this as I wrote a 
> __repr__()
> function for the class and had large chunks of the page 
> disappear because
> I didnt stick a trailing ">" on the returned string.  
> Switching to "[]"
> solved the problem.
> 
> I will now go and see about writing a HTML-textizer for use in such
> situations.

Here's a very simple one ...

entitydefs = {}

for k, v in htmlentitydefs.entitydefs.items():
    entitydefs[v] = '&' + k + ';'

def htmlEncode (str):
    """"""

    str = string.replace(str, '&', '&amp;')

    for k, v in entitydefs.items():
        if v != '&amp;':
            str = string.replace(str, k, v)

    return str

def htmlDecode (str):
    """"""
    for k, v in entitydefs.items():
        if v != '&amp;':
            str = string.replace(str, v, k)
    
    str = string.replace(str, '&amp;', '&')
    return str




From bt98 at doc.ic.ac.uk  Sun Jun  2 07:10:19 2002
From: bt98 at doc.ic.ac.uk (Benjamin Tai)
Date: Sun, 02 Jun 2002 12:10:19 +0100
Subject: Pop a list from beginning ? and memory saving...
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de> <3CF9EB89.7CA7204B@doc.ic.ac.uk> <adcqon$1012ko$1@ID-146704.news.dfncis.de>
Message-ID: <3CF9FD1B.EFD1FC05@doc.ic.ac.uk>

Shagshag13 wrote:

> > You could try a.pop(0), which pop the first instance of the list, like a FIFO.
> >
> > Thanks
> >
> > Ben
>
> and is it a way to free memory ?
>
> thanks,
>
> s13.

Mr Robert Eanes wrote an email in the list about a month ago:

http://groups.google.com/groups?hl=en&lr=&frame=right&th=5b2e0f9e9c54c69a&seekm=3C6197D0.2D3A7CC8%40pacbell.net#link1

here is another one.

http://groups.google.com/groups?q=free+memory+group:comp.lang.python.*&hl=en&lr=&selm=199609071414.KAA18129%40dolphin.automatrix.com&rnum=3

There are some very good reference when you do a google search for "free memory" in
python group.

I am afriad I have little experience to discuss this issue, however there might be
other articles in the archieve, explaining
1) Python programming techniques in relation to large data structure
2) There might be no direct memory management above Python level. Work is done
inside the interpreter by reference count.

Thanks

Ben




From fgeiger at datec.at  Wed Jun  5 09:26:07 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Wed, 5 Jun 2002 15:26:07 +0200
Subject: python cookbook at oreilly's web site
References: <dCnL8.5$566.1250@news.ecrc.de> <adl2nq$3hrm$1@ID-41653.news.dfncis.de>
Message-ID: <adl3ii$ofq$1@newsreaderg1.core.theplanet.net>

Yes, and I'd really like to know the difference.

Franz


"Tobias Klausmann" <klausman-un040602 at tuts.net> schrieb im Newsbeitrag
news:adl2nq$3hrm$1 at ID-41653.news.dfncis.de...
> Markus Jais <mjais at web.de> wrote:
> > just found
> >
> > http://www.ora.com/catalog/pythoncook/
>
> Does this have any relation to ASPNs Python cookbook?
> http://aspn.activestate.com/ASPN/Cookbook/Python
>
> Wondering,
> Tobias
>




From brueckd at tbye.com  Tue Jun 25 14:52:50 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Tue, 25 Jun 2002 11:52:50 -0700 (PDT)
Subject: Suggestions for good programming practices?
In-Reply-To: <afa7e1$eeq$1@nntp6.u.washington.edu>
Message-ID: <Pine.LNX.4.44.0206251122470.27524-100000@ns0.tbye.com>

On 25 Jun 2002, Donn Cave wrote:

> | So... rather than teaching "avoid W!", let's say "be careful with W 
> | because of X, Y, and Z". I still wouldn't use eval/exec on code posted 
> | through a web form, for example, but there are times when they are very 
> | useful and I can use them in good confidence because I understand their 
> | risks.
> 
> But you weren't going to be deterred by that pronouncement anyway.
> 
> "Avoid exec, execfile, eval, and input" is good advice.  Taken at face
> value, it doesn't necessarily absolutely prohibit their use - if I said
> "avoid walking in the road", you could reasonably reasonably assume I'm
> saying something like "walk on the sidewalk when possible".

If you mean the latter, say the latter. Neither newbies nor veterans have 
to do any reading between the lines. Even better, say "walk on the 
sidewalk when possible because ___".

> Someone whose software engineering skills have been honed by years
> of wrestling with ugly code monsters will apply a different perspective
> to that advice.  A 1-week newbie could do worse than to follow that
> advice religiously.

I understand your point, but the OP is new to Python, not programming.  
This isn't a big deal; I was simply pointing out that just saying "don't
use those!" (Why? Are they broken? Deprecated? What?) isn't as helpful as
explaining the risks.

In the specific case of exec/eval, people are quick to strike them down,
often citing the untrusted input example (what about os.system, .fork,
etc.?), but that's a pretty narrow and uncommon usage scenario. When I
first learned Python I used them in some of the "forbidden" ways because I
hadn't learned getattr/setattr yet. So what? My programs worked fine (a
little slow) and later I learned better ways and now use those instead.

If a person is new to programming altogether then what kind of a teacher
will draw attention to exec/eval in the first place anyway? Besides, the
people who understand the potential for problems with something tend to be
the ones that don't mishandle it (whether it be streets, guns, or Python
functions).

-Dave





From peter at engcorp.com  Sat Jun 29 20:17:35 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 29 Jun 2002 20:17:35 -0400
Subject: Win32/Python serial port event character problem
References: <mailman.1025390438.20684.python-list@python.org>
Message-ID: <3D1E4E1F.AD73DAEA@engcorp.com>

Derek Basch wrote:
> 
> If anyone reads this and can show me a better way to
> work with serial event characters in win32 please feel
> free to :)

I second Chris' comments about using something like PySerial
and just handling this in a thread.  Event characters are
probably non-portable, while you could easily write code
with PySerial that runs under Linux or Windows.

But give it time... you definitely didn't give it enough time
before you reposted. :)

By the way, here's some typical (?) code for handling a termination 
sequence on incoming data for something like this (while inside of
some kind of serial port object):

def read(self, timeout=None):
    buf = self.leftover
    self.leftover = ''

    while true:
        data = FUNCTION_TO_READ_A_RAW_BLOB_OF_DATA()

        # if new data received, add to the buffer and check for
        # a contained input termination sequence
        if data:

            buf = buf + data

            termIndex = string.find(buf, self.inputTerminator)
            if termIndex >= 0:
                # move past the terminator
                termIndex = termIndex + len(termIndex)

                # preserve anything lying after input termination 
                # sequence in the leftover buffer for next time
                buf, self.leftover = buf[:termIndex], buf[termIndex:]

                break

    return buf

Each time you call it, it reads raw data and adds it to a buffer
which it then rescans for the termination sequence.  If the buffer
contains the sequence, it splits it on that and returns everything
up to and including the sequence, preserving everything after it 
for the next call.

-Peter


From Chris.Barker at noaa.gov  Fri Jun 21 15:46:28 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Fri, 21 Jun 2002 12:46:28 -0700
Subject: Windows versions of Python---pros and cons?
References: <mailman.1024613669.14676.python-list@python.org>
Message-ID: <3D138293.85892240@noaa.gov>

Bjorn Pettersen wrote:
> > On that note: does anyone know if there is a way to turn a
> > Python script into something that acts like an application on
> > Windows...without resorting to Py2exe and the like.

> If you use the ActiveState version (not sure about the python.org
> version) this should work out-of-the-box. Just type the name of the
> script, you can even leave out the .py extension.

Actually, I do use the ActiveState version, but *.py is associated with
PythonWin, rather than the python interpreter

Mark Hammond wrote:
> Add the .py extension to your PATHEXT environment variable, and you can
> execute "foo.py" at the command prompt.  Like much of Windows, this is
> based on the extension rather than file attributes.

Which is really annoying, kind of  like the Macintosh file creator
concept: I sometimes want the file to be edited, and sometime run...

Anyway, it looks like my best bet is to re-associate *.py with the
python interpreter, rather than pythonwin. I rarely have a use for
opening pythonwin from the command line, and in the GUI, I can use
openwith from the right button menu if I want to open a *.py file in pythonwin.

Now I have to figure out how to change my PATHEXT environment variable,
which is pretty OT for this group...

-Chris




-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From bit_bucket5 at hotmail.com  Thu Jun  6 18:52:39 2002
From: bit_bucket5 at hotmail.com (Chris Stromberger)
Date: Thu, 06 Jun 2002 22:52:39 GMT
Subject: Using AT on XP to run python script
References: <mchqfu448cnq5q8toh4nq7n1sura1ug72g@4ax.com> <3CFFC694.9000501@bgb.cc>
Message-ID: <q6qvfu0nqcv2sh9ujrakeg839ldo6n8hat@4ax.com>

On Thu, 06 Jun 2002 20:31:58 GMT, Don Garrett <garrett at bgb.cc> wrote:

>
>    Change "c:\python21\python.exe c:\scripts\script.py" to
>"c:\python21\python.exe" "c:\scripts\script.py" otherwise it's trying to turn
>that whole thing into a single file name. However, I would have expected the
>last thing you tried to work.

This appears to be working!  Many thanks Don.

>
>    Try using /interactive just for testing. It may be trying to open a command
>window or something. Maybe name the script .pyw
>
>Chris Stromberger wrote:
>
> > I'm having trouble getting AT on XP to run a python script for me.
> > Have tried:
> >
> > at xx:xx "c:\python21\python.exe c:\scripts\script.py"
> > at xx:xx "start c:\scripts\script.py"
> > at xx:xx "c:\scripts\script.py"
> >
> > and get "the system cannot find the file specified" for the first two.
> > The last version does nothing apparently--no error indication or
> > anything.
> >
> > I imagine it is related to user accounts and such.  Anyone have any
> > pointers?
> >
>



From drew_csillag at yahoo.com  Thu Jun  6 03:30:50 2002
From: drew_csillag at yahoo.com (Drew Csillag)
Date: Thu, 06 Jun 2002 12:30:50 +0500
Subject: [ANN] SkunkWeb 3.2.1 Released
Message-ID: <3CFF0FAA.8030201@yahoo.com>

SkunkWeb 3.2.1 has just been released.

The release is available from the SkunkWeb home page at
    http://skunkweb.sourceforge.net

or directly from
        http://prdownloads.sourceforge.net/skunkweb/skunkweb-3.2.1.tgz


Changes since 3.2
========================================
* Apache 2 support!
* new auth service
* Assorted PyDO fixes, notably support for more PostgreSQL types
* Caching fix for stale
* DateTime objects can now be used as arguments to cacheable components
* rewrite fixes and feature additions
* usual array of general bug fixes
* mod_skunkweb now properly reports its existence in the Server header
* UNIX socket fixes
* server restart fixes

Very best wishes to all,

-Drew









From garrett at bgb.cc  Thu Jun  6 16:31:35 2002
From: garrett at bgb.cc (Don Garrett)
Date: Thu, 06 Jun 2002 20:31:35 GMT
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet> <3CFE23DB.994F8A6C@engcorp.com>
Message-ID: <3CFFC67E.3070609@bgb.cc>

    It's a packaging issue for me. It's worth a lot to be able to just ship a
program to a customer without their needing to understand that I wrote it in
Python (thanks py2exe!). A fact about which they usually don't care. I also
don't have to explain to them how to download/install the Python runtime.

    It's also useful since the runtime matches the version of python I
developed with, and they never have to go update the python runtimes on all
the machines to which they have deployed.

    It would be even more useful if there was an easy way to produce a
standalone RPM or MSI file (depending on the plateform). The MSI file is
probably more useful. The unix people are more capable of dealing with the
extra steps of getting a python program to work.

Peter Hansen wrote:

 >
 > Why do you want to do that?  For presumably increased speed, or because
 > you don't want the hassle of multiple .py files all over the place (in
 > other words, better packaging)?
 >
 > -Peter
 >




From pyth at devel.trillke.net  Tue Jun 18 09:21:38 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 18 Jun 2002 15:21:38 +0200
Subject: what *is* a class?
In-Reply-To: <jp7uguoqjs7t23itb7vihai30cnv3qr7sm@4ax.com>; from op73418@mail.telepac.pt on Tue, Jun 18, 2002 at 12:59:21PM +0100
References: <B43D149A9AB2D411971300B0D03D7E8BF0A396@natasha.auslabs.avaya.com> <mailman.1024398742.13711.python-list@python.org> <jp7uguoqjs7t23itb7vihai30cnv3qr7sm@4ax.com>
Message-ID: <20020618152138.C15079@prim.han.de>

Gon?alo Rodrigues wrote:
> Here goes my personal version of this module. It is largely based off
> Sami Hangaslammi recipe at Active State plus some additions/differences
> of mine.
> ...

thanks. that's a good starting point to implement an iterable module. 
I guess we all agree that iterator versions of the builtin counterparts
should at least accept the same arguments. 

    holger




From sholden at holdenweb.com  Tue Jun  4 12:16:51 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 4 Jun 2002 12:16:51 -0400
Subject: Python command line?
References: <B9224E06.628B%sarmstrong13@mac.com>
Message-ID: <7O5L8.232671$%u2.196351@atlpnn01.usenetserver.com>

"SA" <sarmstrong13 at mac.com> wrote in message
news:B9224E06.628B%sarmstrong13 at mac.com...
> Hi Everyone-
>
> I would like to run python code line by line using:
>
> python -c some line of code
>
> Will this work?
>
No. Sorry. Every time you run python you will get a completely new
environemnt which has no memory of what previous invocations have done. Like
this:

$ python -c "myvar = 21"

$ python -c "print myvar"
Traceback (most recent call last):
  File "<string>", line 1, in ?
NameError: name 'myvar' is not defined

> I have a gui app that lets me type in code in one fiels and then executes
> the code typed into that field line by line and prints the output to
another
> field for evaluation.
>
That's good, b ut you are going to have to go a little further to get what
you want. Of course, IDLE and PythonWin do this already, so at least there's
a body of Python source for you to read.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From kragen at pobox.com  Sun Jun  2 23:34:56 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 02 Jun 2002 23:34:56 -0400
Subject: Speeding up the Canvas Widget
References: <ad8jif$dki$1@netnews.upenn.edu>
Message-ID: <83adqdrq5b.fsf@panacea.canonical.org>

zayats at blue.seas.upenn.edu (Salim Zayat) writes:
> I have been having problems with the Canvas widget in terms of just 
> speed.  I need to draw, and almost constantly redraw, upwards of like 
> 10,000 or so things to the canvas (lines, text, and rectangles, etc.).  

It's possible all these things are visible at once; are they really?

Last time I had a speed problem with Canvas, I had one item, a line
object with hundreds of coord pairs.  I just rewrote my program in C
with GTK+ and just drew the lines myself.



From aahz at pythoncraft.com  Sat Jun  8 20:01:19 2002
From: aahz at pythoncraft.com (Aahz)
Date: 8 Jun 2002 20:01:19 -0400
Subject: Efficient python programming...
References: <mailman.1023513672.27475.python-list@python.org>
Message-ID: <adu5sf$j0o$1@panix1.panix.com>

In article <mailman.1023513672.27475.python-list at python.org>,
Sean 'Shaleh' Perry <shalehperry at attbi.com> wrote:
>
>and now I have walked this sufficiently off topic to bore most of the
>readers of this list.

I disagree that this is off-topic; comp.lang.python has generally been
tolerant of any topic that *can* be applied to Python programming.
Testing is surely an appropriate topic.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From tom at huno.net  Fri Jun 28 08:14:57 2002
From: tom at huno.net (thomas)
Date: Fri, 28 Jun 2002 14:14:57 +0200
Subject: feeding a prompt with python
Message-ID: <3D1C5341.9070001@huno.net>

hello,

is there a way with python to feed a prompt with data? i call sth. with 
system($some_cmd) and the cmd gives me a prompt which of course waits 
for user input. however i want to automate this, is there a way?
i browsed through the essential reference 2nd edt. but didn't find anything.

thx,
thomas



From bokr at oz.net  Wed Jun 26 15:53:44 2002
From: bokr at oz.net (Bengt Richter)
Date: 26 Jun 2002 19:53:44 GMT
Subject: Suggestions for good programming practices?
References: <mailman.1024951532.5616.python-list@python.org> <af8n9i$3at$1@panix1.panix.com> <afafee$fc$0@216.39.172.122> <afaohi$682$1@panix1.panix.com> <afb3a2$skm$0@216.39.172.122> <B93F45F6.CEA6%jonathan@onegoodidea.com>
Message-ID: <afd648$3b4$0@216.39.172.122>

On Wed, 26 Jun 2002 10:20:38 +0100, Jonathan Hogg <jonathan at onegoodidea.com> wrote:

>On 26/6/2002 1:53, in article afb3a2$skm$0 at 216.39.172.122, "Bengt Richter"
><bokr at oz.net> wrote:
>
>> Well, I meant using *args to allow detecting actual "no default-overriding
>> argument passed", in place of x=None. The arg list items (only x here)
>> don't have to remain unnamed long. E.g.,
>> 
>>>>> def foo(*args):
>> ...     if len(args)==1:
>> ...         print "We know we have one non-default arg: %s" % `args[0]`
>> ...         x = args[0]
>> ...     elif len(args)==0:
>> ...         print "We can supply a default arg, even None"
>> ...         x = None
>> ...     else: raise TypeError,'foo() takes 0 or 1 arguments (%d given)' %
>> len(args)
>> ...     print 'Effective arg was: %s' % `x`
>> ...
>
>I find this stuff much easier to write with a helper function:
>
>>>> def shift():
>...     global _0
>...     __, _0 = _0[0], _0[1:]
>...     return __
>... 
>
>Then I can write the function 'foo' much more naturally as:
>
>>>> def foo( *args ):
>...     global _0; _0 = args
>...     try:
>...         x = shift()
>...     except IndexError:
>...         print 'We can supply a default arg, even None'
>...         x = None
>...     else:
>...         try:
>...             shift()
>...         except IndexError:
>...             print 'We know we have one non-default arg: %s' % `x`
>...         else:
>...             raise TypeError( 'foo() takes 0 or 1 arguments' )
>...     print 'Effective arg was: %s' % `x`
>...
>
>This can be further simplified by removing the '*args' nonsense altogether
>and defining another helper function:
>
>>>> def p( f ):
>...     def _p( *args ): global _0; _0 = args; f()
>...     return _p
>...
>>>> def foo():
>...     try:
>...         x = shift()
>[etc]
>...
>>>> p(foo)( 'hello world' )
>
LOL ;-)

>
>How did we get onto this from 'Suggestions for good programming practices'?
>
I guess from where I picked up: x=None usage => default args => alternatives => rambling on ...
=> *args use analogous to C++ void foo(const char* s){...}; and
void foo(const int x, const int y, const char* s){...}; to allow foo("s only") and
foo(123, 456, "s with ints") overloading via Python dynamic recognition instead of statically
via C++ ?foo@@YAXPBD at Z vs ?foo@@YAXHHPBD at Z => non-Python odor? => Bad Thing(tm) or caller-friendly?

ISTM caller-friendly is good, but the *args implementation of this overloading is not as nice
as it could be. Is it better to allow foo(x,y,s) and foo(s) via overloaded foo() than, e.g.,
to use separate names like fooxy(x,y,s) and foo(s), and have foo() call fooxy()?
(That, at least, is an OT question ;-)

Regards,
Bengt Richter


From opengeometry at NOSPAM.yahoo.ca  Tue Jun 25 00:16:44 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 25 Jun 2002 04:16:44 GMT
Subject: Python and remote sensing
References: <3D16E528.5040207@kfunigraz.ac.at>
Message-ID: <af8qrb$ccv2j$2@ID-99293.news.dfncis.de>

Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote:
> Hi,
> 
> 
> a technician of us asked me whether it is possible to do the following:
> 
> 
> Starting a script on Unix at lets say 2 am and calling a remote site 
> (http://www.whatever...) and that site should become passively processed 
> (the site holds 41 observing stations). Is it possible to start a remote 

Not sure what "passively processed" means.  If you are talking about
fetching webpages periodically, then crontab and any script will do.

> script (automatically) connect to a http page and retrieve information 
> from that html-page?
> 
> I answerd Perl or Python can do the job. But I am not sure whether that 
> is correct.
> 
> I am sure the above topic has been asked 1000 times. Or is there a tool 
> or program availabe which can do the job better?; I mean our technican 
> has never heard of Python and I am not interested to make web based (put 
> in the right term here) programming or the like.
> 
> If Python can do it - okay; if not please say so (if Python is dead - 
> okay; if Python lives - okay)!
> 
> 
> Thanks,
> S. Gonzi
> 

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin


From peter at engcorp.com  Sun Jun 23 12:12:52 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 23 Jun 2002 12:12:52 -0400
Subject: The best GUI library in python
References: <8d3f4438.0206220457.b820418@posting.google.com> <3D1481DE.C0331204@engcorp.com> <8d3f4438.0206221422.6f71287a@posting.google.com> <3D14FFB5.90808865@engcorp.com> <8d3f4438.0206230713.535169aa@posting.google.com>
Message-ID: <3D15F384.8E97CD96@engcorp.com>

d2002xx wrote:
> 
> I hope the GUI could work well when I compiling a program in
> background (in my K6-2 400, running program other than compiler in
> this kind of situation is allocated only about 10% to 15% CPU
> resource)

Maybe you can just run the compiler with "nice" to make it
behave better.  Then you don't need to change anything else.
Type "man nice" for details.  This is, of course, assuming you
are using something like Linux.  I don't think you've said 
what your platform is yet.

-Peter


From stagnoNOSPAM at prosa.it  Mon Jun 24 15:08:27 2002
From: stagnoNOSPAM at prosa.it (Marco Stagno)
Date: Mon, 24 Jun 2002 19:08:27 GMT
Subject: [ANN] my_gui: a gui module for MySQL . First public release (0.24)
Message-ID: <mailman.1025147583.13072.clpa-moderators@python.org>

This is the first public release for "my_gui". 

my_gui.py is a simple Grafic User Interface for MySql 

It can be used as Python module to create your own MySQL interface. 

Sorry, no documentation yet

download it at http://www.prosa.it/projects/projects

                                  MAS!




From not.this at seebelow.org  Fri Jun 14 18:14:49 2002
From: not.this at seebelow.org (Grant Griffin)
Date: 14 Jun 2002 15:14:49 -0700
Subject: index() and count() methods
References: <GxproF.1y4K@beaver.cs.washington.edu>
Message-ID: <aedpsp01mh3@drn.newsguy.com>

In article <GxproF.1y4K at beaver.cs.washington.edu>, Douglas says...
>
>Apologies if this is a FAQ, but I haven't run across an answer.
>
>Why are index() and count() methods of only mutable sequences, rather
>than all sequences?  I ran into a situation today where using index()
>on a tuple would have been nice.  It was easy to work around; I'm just
>curious.

This one isn't exactly a FAQ, but IIRC, a patch to this effect was once rejected
by Guido.  (In fact, I think it is even documented in Sourceforge.)  I think the
reasoning might have been that the bloat to Python wasn't worth the narrow
usefulness of the addition.

like-those-who-don't-get-enough-fiber,-python-suffers-from
  -occasional-irregularity-ly y'rs,

=g2

_________________________________________________________________________

Grant R. Griffin                                           g2 at dspguru.com
Publisher of dspGuru                               http://www.dspguru.com
Iowegian International Corporation                http://www.iowegian.com



From skip at pobox.com  Mon Jun 10 11:57:37 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 10 Jun 2002 10:57:37 -0500
Subject: Where is Berkeley DB on your system? Can you live without DB 1.85?
In-Reply-To: <LNBBLJKPBEHFEDALKOLCAELHPLAA.tim.one@comcast.net>
References: <GCEDKONBLEFPPADDJCOECEMFDHAA.whisper@oz.net>
        <LNBBLJKPBEHFEDALKOLCAELHPLAA.tim.one@comcast.net>
Message-ID: <15620.52337.994202.635139@12-248-41-177.client.attbi.com>

    Tim> It wasn't known to be badly damaged at the time it was first
    Tim> shipped, and now it's supplied for backward compatibility.  A
    Tim> better alternative would be most welcome, but PythonLabs has no
    Tim> bandwidth to work on it.

Perhaps a goodhearted member of the community could be persuaded to build a
.pyd file using the 1.86 patch.  It changes the file format, but fixes the
hash implementation bugs.  Check here (bottom of page):

    http://www.sleepycat.com/update/index.html

I'm not sure (and don't really care) about any license implications.

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html




From christophe.delord at free.fr  Sun Jun 23 18:33:46 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Mon, 24 Jun 2002 00:33:46 +0200
Subject: re.rsearch?
References: <slrnahcg6u.382.quinn@baht.ugcs.caltech.edu>
Message-ID: <20020624003346.423dd4de.christophe.delord@free.fr>

Hi,

If you compile your pattern, you can give the pattern object a start position and an end position:

>>> p=re.compile("a")
>>> p.search("1a2a").span()
(1, 2)
>>> p.search("1a2a",2).span()
(3, 4)

This may work as you need:

def rsearch(reg, s):
	start, end = None, None
	i = 0
	p = re.compile(reg)
	while 1:
		m = p.search(s, i)
		if m:
			start, end = m.span()
			i = end
		else:
			break
	return start, end


I found this in the documentation at http://web.pydoc.org/2.2/pre.html#RegexObject-search

--
search(self, string, pos=0, endpos=None)
    search(string[, pos][, endpos]) -> MatchObject or None
     
    Scan through string looking for a location where this regular
    expression produces a match, and return a corresponding
    MatchObject instance. Return None if no position in the string
    matches the pattern; note that this is different from finding
    a zero-length match at some point in the string. The optional
    pos and endpos parameters have the same meaning as for the
    match() method.
--

But of course as suggested by Fredrik Lundth, this solution is shorter:

def rsearch(reg, s):
	return re.search("(?:.*)(%s)"%reg, s).span(1)



Best regards,
Christophe.

On 23 Jun 2002 21:42:54 GMT
quinn at baht.ugcs.caltech.edu (Quinn Dunkan) wrote:

> Occaisionally I find myself wanting an re.search that will find the rightmost
> match.  In those occaisions, I use:
> 
> def rsearch(reg, s):
>     'like re.search only find last match, return start, end instead of match'
>     last_m = None
>     start = end = 0
>     while 1:
>         m = re.search(reg, s[end:])
>         if m is None:
>             if last_m is None:
>                 return None, None
>             else:
>                 return start, end
>         last_m = m
>         end += m.end()
>         start = end - (m.end() - m.start())
> 
> ... of course, it would be nicer to return an actual match object, but I
> haven't figured out how to do that short of faking up my own, and I've only
> needed (start, end) so far.  Unfortunately, AFAIK, re.search also lacks a
> 'start_searching_at' type option.
> 
> Has anyone else solved this probem in a more graceful way?


-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From opengeometry at NOSPAM.yahoo.ca  Tue Jun 25 00:19:13 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 25 Jun 2002 04:19:13 GMT
Subject: Python HTML Data Collection
References: <Pine.GSO.4.33.0206232352240.18910-100000@fissure.scar.utoronto.ca>
Message-ID: <af8r01$ccv2j$3@ID-99293.news.dfncis.de>

Paul Barrett <98barret at utsc.utoronto.ca> wrote:
> Hi,
> 
>   I'm fairly new to Python but I do have a little bit of experience from
> a class I took which taught some Python. I'm working at a job where I
> query a CGI script, collect the data and then post it to another CGI
> script. My main question is whether I could write a python script which
> would automate this process. Both of these scripts are on external
> websites and I only need a select amount of the data. I thought that if
> any language would do this job, then it would be Python, but I haven't
> been able to find any information for something like this. Do you folks
> have any suggestions?

It depends on the structure of data you need to parse out.  But, any script
can do it.

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin


From ulbi at ivs.tu-berlin.de  Mon Jun 24 04:41:14 2002
From: ulbi at ivs.tu-berlin.de (Andreas Ulbrich)
Date: Mon, 24 Jun 2002 10:41:14 +0200
Subject: mail-problem over t-online
References: <mailman.1024859369.26585.python-list@python.org>
Message-ID: <3D16DB2A.7050709@ivs.tu-berlin.de>

ZRZ is simply not allowing sending messages using SMTP from networks 
outside the university network. As long as you connect directly 
university everything is fine. As soon as you move to t-online, zrz 
won't allow sending.

A work around could be to open a ssh tunnel (ssh -L option) and to 
connect through this tunnel in case you have a valid account on a 
university machine. This is what I'm doing.



From prema at prema.co.nz  Mon Jun 17 14:55:01 2002
From: prema at prema.co.nz (Mike MacDonald)
Date: Mon, 17 Jun 2002 18:55:01 GMT
Subject: MS Word -- finding text
References: <mailman.1024247060.12127.python-list@python.org> <3D0D1862.5050708@skippinet.com.au>
Message-ID: <ugsc48godgaja8@news.supernews.com>

Hi Mark !
Thanks for this and also for the superb win32 contribution.
My copy of the book is close to worn out !
Best regards
Mike


From cpbotha at i_triple_e.org  Mon Jun 10 16:05:08 2002
From: cpbotha at i_triple_e.org (Charl P. Botha)
Date: Mon, 10 Jun 2002 20:05:08 +0000 (UTC)
Subject: No such directory?
References: <886c5e4b.0206101158.680c6278@posting.google.com>
Message-ID: <ae30pk$3sv$1@news.tudelft.nl>

In article <886c5e4b.0206101158.680c6278 at posting.google.com>, 
Robin Siebler wrote:
> When I try to change into a certain directory, I get an errorm message
> that states the directory doesn't exist.  The directory does exist,
> however, and the listdir function shows that it exists.  What gives?
> 
>>>> os.chdir('c:\\18_510GMC1\\18_510GMC1')
                                        ^
>>>> os.getcwd()
> 'c:\\18_510GMC1'
>>>> os.listdir(os.getcwd())
> ['18_510GMC']
             ^ no 1 at the end.
             
You made a typo.

Python: 1. You: 0. ;)

-- 
charl p. botha http://cpbotha.net/ http://visualisation.tudelft.nl/


From mcherm at destiny.com  Thu Jun  6 11:42:05 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Thu, 06 Jun 2002 11:42:05 -0400
Subject: array, list, performance...
References: <3CFF5F98.2060406@destiny.com> <87sn40xzou.fsf@twistedmatrix.com>
Message-ID: <3CFF82CD.7000500@destiny.com>

Oops, yes.

What I SHOULD have said was:
    li[n] = x   assignment to the middle is O(1)
    li[n:n] = x   INSERTION in the middle is O(len(li))

-- Michael Chermside


Christopher Armstrong wrote:
>>Michael Chermside <mcherm at destiny.com> writes::
>>  li[n] = x      assignment to the middle is O(len(li))
> 
> 
> This doesn't seem right. Shouldn't that be O(1), just like
> access-time?
> 






From brueckd at tbye.com  Mon Jun  3 15:06:33 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Mon, 3 Jun 2002 12:06:33 -0700 (PDT)
Subject: Question: Importing the same module under different names
In-Reply-To: <013501c20b27$a4e757d0$1717a8c0@suxlap>
Message-ID: <Pine.LNX.4.44.0206031202340.1823-100000@ns0.tbye.com>

On Mon, 3 Jun 2002, Andreas Jung wrote:

> Usually Python imports a module 'foo' only once when one uses
> multiple 'import foo' statements. However Python load the 
> module when one uses packages e.g.
> 
> packages/__init.py
>         /foo.py
> 
> 
> When PYTHONPATH is set to the 'packages' directory then

This defeats the purpose of having packages in the first place.

> import foo
> import packages.foo 
[snip]
> loads the foo module twice. How does Python determine in this case
> to load the module or not?

IMO, setting the PYTHONPATH to include a subtree of your package directory 
is the wrong thing to do. Is there a case when you really need to do this? 
It seems akin to:

ln -s foo.py bar.py

and then noticing that

import foo, bar

causes the module to load twice, i.e. it's too much to hope that Python 
will figure out your trickery! :-)

-Dave





From cliechti at gmx.net  Sun Jun 23 14:40:39 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 23 Jun 2002 20:40:39 +0200
Subject: TimeoutQueue.py
References: <mailman.1024698810.22387.python-list@python.org> <3D1624E8.AF7A9BF1@accessforall.nl>
Message-ID: <Xns9236D2C703635cliechtigmxnet@62.2.16.82>

Ype Kingma <ykingma at accessforall.nl> wrote in
news:3D1624E8.AF7A9BF1 at accessforall.nl: 
> Btw. I would expect Chris' patch to work in jython, too.

i think yes, it only needs time.time and time.sleep
 
> Chris, did you add a timeout to the put() method as well?

in the patch, yes.

chris

-- 
Chris <cliechti at gmx.net>



From larooy at xtar.co.nz  Tue Jun 18 17:19:00 2002
From: larooy at xtar.co.nz (John La Rooy)
Date: Wed, 19 Jun 2002 09:19:00 +1200
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com>
	<aegfck$6tkre$1@ID-114079.news.dfncis.de>
	<wEPO8.16066$uk2.8447171@twister.nyroc.rr.com>
	<3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com>
	<gYgP8.42356$GY.12361719@twister.nyroc.rr.com>
	<aekcug$7js8n$1@ID-114079.news.dfncis.de>
	<9YBP8.44418$GY.13865843@twister.nyroc.rr.com>
	<mailman.1024423525.24359.python-list@python.org>
Message-ID: <20020619091900.70ae2b7f.larooy@xtar.co.nz>

On Tue, 18 Jun 2002 11:02:25 -0700
Cliff Wells <logiplexsoftware at earthlink.net> wrote:


> 
> Perhaps you'd like someone to come out and type it for you as well. 
> 
> Please remember that the people on this list aren't being paid to help you
> write your program.  They expect you to do at least /a little/ of the research
> yourself.  It's understandable if you don't know where to look, but when
> someone points you to the relevant portion of the docs, you should at least
> read it before asking again.  It's clear from your postings that you haven't
> even read the tutorial.
> If you can't do that then you should pay someone to write your program for you.
> 
> -- 
> Cliff Wells, Software Engineer
> Logiplex Corporation (www.logiplex.net)
> (503) 978-6726 x308  (800) 735-0555 x308
> 
> 
I've been utterly astonished reading this thread. I was sure George was flame-baiting
but after his last couple of posts I'm not so sure. Well said Cliff

John


From tjreedy at udel.edu  Tue Jun  4 22:29:26 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 05 Jun 2002 02:29:26 GMT
Subject: Pop a list from beginning ? and memory saving...
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de>	<3CFA0676.68F80DE6@engcorp.com>	<add289$vud42$1@ID-146704.news.dfncis.de>	<Xns9221CE2AF76D0cliechtigmxnet@62.2.16.82>	<adhrsl$118kvn$1@ID-146704.news.dfncis.de> <mailman.1023194347.14669.python-list@python.org> <3CFD5B60.90831F21@engcorp.com>
Message-ID: <aIeL8.120588$Kp.12235125@bin7.nnrp.aus1.giganews.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3CFD5B60.90831F21 at engcorp.com...
> >     self._v[:] = []
> I can't parse that.  What does it do?

Try the wonerful interactive interpreter:

>>> l = range(10)
>>> l
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
>>> l[3:7] = []
>>> l
[0, 1, 2, 7, 8, 9]
>>> l[:] = []
>>> l
[]

Terry J. Reedy





From nessus at mit.edu  Fri Jun 21 18:14:44 2002
From: nessus at mit.edu (Douglas Alan)
Date: 21 Jun 2002 18:14:44 -0400
Subject: Doing a ps in windows.. (Is application XYZZY running?)
References: <m38z8g74hw.fsf@localhost.localdomain> <nnmv9uc6gtkge4at1cfiv0rbijsol31oc4@4ax.com>
Message-ID: <lchejw5lez.fsf@gaffa.mit.edu>

I've been trying to follow your recommendation as you wrote below, but
it isn't working for me.  For instance, if I run the following code:

    while 1:
       junk, instances = win32pdh.EnumObjectItems(
          None, None, "process", win32pdh.PERF_DETAIL_WIZARD)
       print len(instances), instances
       if "winvrut2" in instances: sleep(1)
       else: break

The loop never terminates, and the string printed out never changes,
even after the process "winvrut2" terminates, and many other processes
come and go.  It seems that the list of processes is being cached, and
then EnumObjectItems just returns the value from the cache.  How do I
flush the cache?

This is under Windows2000, and ActivePython-2.1.

|>oug


Wolfgang Strobl <ws at mystrobl.de> writes:

> Mon, 25 Mar 2002 20:54:19 GMT, Andrew Markebo
> <flognat at flognat.myip.org>:
> 
> >*blush* yeah I know.. not really.. here, only almost :-)
> >
> >I have the windows-compiled python 2.2, and I would like to see if
> >Opera or Netscape is running on my computer (poor sod, running Win2k,
> >XP or 98).. How do I do this, I suppose I could use something in the
> >win32api, but what is something.. :-) ??
> >
> >On unix I do "ps -ef | grep netscape" ;-)
> 
> On Windows, I do 
> 
> >>> import win32pdh
> >>> junk, instances = win32pdh.EnumObjectItems(None,None,"process",
> ...     win32pdh.PERF_DETAIL_WIZARD)
> >>> "python" in instances
> 1
> >>> "netscape" in instances
> 0
> >>>
> 
> (straight from an example in the win32pdh documentation)
> 
> -- 
> Wir danken f?r die Beachtung aller Sicherheitsbestimmungen



From peter at engcorp.com  Wed Jun  5 22:38:25 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 05 Jun 2002 22:38:25 -0400
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet> <3CFE23DB.994F8A6C@engcorp.com> <slrnaft0p7.3ac.TuxTrax@fortress.tuxnet>
Message-ID: <3CFECB21.8FA62E80@engcorp.com>

TuxTrax wrote:
> 
> On Wed, 05 Jun 2002 10:44:43 -0400, Peter Hansen <peter at engcorp.com> wrote:
> > TuxTrax wrote:
> >> My question is this; I realize that Python compiles to bytecode automatically
> >> as part of the interpreters tokenization run, but is there a way to permanently
> >> compile python code outputting to an executable binary?
> >
> > Why do you want to do that?  For presumably increased speed, or because
> > you don't want the hassle of multiple .py files all over the place (in
> > other words, better packaging)?

> Packaging is a thought, but speed (so far) isn't an issue, as python is pretty
> fast IMHO.

Okay, no probs.  I was concerned that you might have wanted a speedup,
and with the answers like "py2exe" you would not see that speedup 
because all they really do is bundle everything together with 
"frozen" bytecode and a copy of the interpreter...

-Peter


From phlip_cpp at yahoo.com  Thu Jun 20 10:09:29 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 20 Jun 2002 14:09:29 GMT
Subject: SNIP IT: tk.Canvas Drag Items Around
References: <63604d2.0206031518.2da29d3b@posting.google.com> <wkit4eh3f6.fsf@mail.afm.dtu.dk>
Message-ID: <aesnmp$gvp@dispatch.concentric.net>

Brian Elmegaard wrote:

> I am not very experienced in tkinter, so I would like to understand
> what is the advantage of your solutions compared to what I have in
> Matt Conways example:
> c:/Python22/Demo/tkinter/matt/canvas-moving-w-mouse.py

The advantage is I was unaware of this example.

Reviewing it, I see he uses 'move'; I think I didn't.

        self.draw.move(CURRENT, event.x - self.lastx, event.y - self.lasty)

He correctly adjusts for the mouse location within the object (the 
subtraction) where I think I let the object snick over so its first coord 
is under the mouse point. I have since fixed that in the production code.

However, he does not use canvasx or canvasy. If this example scaled or 
scrolled it would produce funny motions.

-- 
  Phlip
       http://www.greencheese.org/LucidScheming
  --  All analysis and no code makes Jack a dull boy.
      All analysis and no code makes Jack a dull boy.
      All analysis and no code makes Jack a dull boy.  --



From emile at fenx.com  Fri Jun 28 07:31:22 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 28 Jun 2002 11:31:22 GMT
Subject: explain this pickle issue to me...
References: <timo-1418C7.00095228062002@corp.supernews.com>
Message-ID: <eOXS8.346960$cQ3.21460@sccrnsc01>

Timothy O'Malley
[snip]
> And the testme.py file used in execfile
> ---------------------------------------
> import cPickle
> class Simple:
>     def __init__(self):
>         self.a = "a"
>         self.b = 2
> # end class A
>
> def writeSimple():
>     s = Simple()
>     return cPickle.dumps(s)
> # end writeSimple
>
> def readSimple(ss):
>     s = cPickle.loads(ss)
>     assert ininstance(s, A), "Not an instance of Simple"
> # end readSimple
>

The problem is that cPickle expects the name of the module holding the
class to be in the pickle, and that execfile doesn't do module
management, but reads it testme.py as __main__, causing the conflict.

If you add to the top of testme:

__name__ = 'testme'
import testme

and change your assert to:

assert isinstance(s, testme.Simple), "Not an instance of Simple"

that should fake it out.  The first makes the creation of pickles use
testme as the module name, and the second and third allows the
isinstance to work.

--

Emile van Sebille
emile at fenx.com

---------



From robinjim at earthlink.net  Tue Jun  4 20:13:35 2002
From: robinjim at earthlink.net (robin and jim)
Date: Wed, 05 Jun 2002 00:13:35 GMT
Subject: FXPy TextField question
Message-ID: <PIcL8.2382$9H4.1851@newsread2.prod.itd.earthlink.net>

Hello,

I installed the latest MS Windows binary distribution of the FOX GUI library
(FXPy-1.0.5) for Python 2.2, but I am unable to access the data entered in a
FXTextField widgit.

Here is what I have tried.  Any suggestions would be appreciated.

# begin -------------------------------
import sys
from FXPy.fox import *

class Test_Window(FXMainWindow):

    def __init__(self, app):
        FXMainWindow.__init__(self, app, "Text Field Test", opts=DECOR_ALL,
x=20, y=20, w=200, h=150)
        FXMAPFUNC(self, SEL_COMMAND, FXDataTarget.ID_VALUE,
Test_Window.onCmd_Handle_Entry)

        # Initialize the data targets for this example
        self.stringTarget   = FXDataTarget("FOX")

        matrix = FXMatrix(self, 1,
MATRIX_BY_COLUMNS|LAYOUT_SIDE_TOP|LAYOUT_FILL_X|LAYOUT_FILL_Y)

        FXTextField(matrix, 10, self.stringTarget, FXDataTarget.ID_VALUE,

LAYOUT_CENTER_Y|LAYOUT_CENTER_X|FRAME_SUNKEN|FRAME_THICK|LAYOUT_FILL_ROW)

    def create(self):
        # Create the windows
        FXMainWindow.create(self)

        # Show the main window centered on the screen
        self.show(PLACEMENT_SCREEN)


    def onCmd_Handle_Entry(self, sender, sel, ptr):
       # Handle completed data entry message for the starting date-time
field
       print 'handled text field message'

def run():
    # Make application
    application = FXApp("DataTarget", "FoxTest")

    # Open display
    application.init(sys.argv)

    # Create the main window
    window = Test_Window(application)

    # Create application
    application.create()

    # Run
    return application.run()

if __name__ == "__main__":
    run()
# end -------------------------------





From pearu at cens.ioc.ee  Thu Jun 20 06:06:24 2002
From: pearu at cens.ioc.ee (Pearu Peterson)
Date: Thu, 20 Jun 2002 13:06:24 +0300 (EEST)
Subject: Looking for Python tools for extracting information from PDF
 files.
In-Reply-To: <Pine.LNX.4.21.0206201249470.4516-100000@cens.kybi>
Message-ID: <Pine.LNX.4.21.0206201304110.4516-100000@cens.kybi>

Never mind. I found pdfimages that I can use. 
Sorry for bothering.

Regards,
	Pearu





From jadestar at idiom.com  Mon Jun 24 17:52:15 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 24 Jun 2002 21:52:15 GMT
Subject: AF_UNIX + SOCK_DGRAM
References: <aes1he$gv2$1@news.idiom.com> <aet3ai$1l0o$2@nntp6.u.washington.edu>
Message-ID: <af84af$1dkt$1@news.idiom.com>

Donn Cave <donn at u.washington.edu> wrote:
> Quoth "James T. Dennis" <jadestar at idiom.com>:

> |  I was playing with the socket module and Unix domain sockets,
> |  trying to make it work with SOCK_DGRAM.  I could get messages
> |  recvfrom() but any attempt to send responses was failing, 
> |  complaining about the return address.

> |  Now I think I understand the problem.  The recvfrom() tuple
> |  does contain two parts (which WOULD BE address and data in UDP)
> |  but I'm guessing that the address is meaningless for Unix 
> |  domain sockets.  It occurs to me that the "return" address 
> |  must be the same socket (filesystem node of type 's') as the 
> |  one that I did a .bind() to.  

> Actually think you have run into a bug in Python's socket module.
> If it's convenient for you to build your own socket module, change
> it like this and see what happens (I'm looking at version 2.1 here.)

> *** socketmodule.c.dist Sun Apr 15 17:21:33 2001
> --- socketmodule.c      Thu Jun 20 10:10:12 2002
> ***************
> *** 597,602 ****
> --- 597,603 ----
>         case AF_UNIX:
>         {
>                 struct sockaddr_un *a = (struct sockaddr_un *) addr;
> +               a->sun_path[addrlen - (sizeof(*a) - sizeof(a->sun_path))] = 0;
>                 return PyString_FromString(a->sun_path);
>         }
>   #endif


> The address returned from recvfrom is (addr, len), but we've been
> ignoring len on the theory that the string would be NUL-terminated.
> It isn't, so there can be some garbage on the end.

> |  This is academic curiosity, and I'll read through the C sources
> |  to see what I can glean therefrom, eventually.  I'd like to know
> |  how I can tell what process has opened a connection to my 
> |  AF_UNIX socket and I'd love to know how to pass an open file
> |  descriptor through an AF_UNIX socket (in Python).  I've read that
> |  this is possible in C and that it is one of the few features of 
> |  AF_UNIX sockets which is NOT available via other address/protocol 
> |  families.

> Yes, AF_UNIX supports that (SOCK_DGRAM or SOCK_STREAM.)  See man 2
> sendmsg.  That would have to be added to the socket module.  Probably
> easy enough to do, as long as you confine yourself to support for this
> one particular thing (passing file descriptors.)

> |  ... I've also heard that this can be used as a way for a
> |  client to request read-only or append-only access to a file
> |  --- since the UNIX APIs (in C) apparently offer ways for the server
> |  to obtain credential information about the requesting process as
> |  well as passing open file descriptors (which are then roughly akin
> |  to "capabilities" since their access mode can't be changed)

> Don't know, right off hand I can't think what API function you
> might be thinking of there.
> 	Donn Cave, donn at u.washington.edu

 So, should this be reported to the Python dev list?  If so, how?
 Can we get it fixed by 2.3?  How would my code work if it was fixed?
 What would/should be returned for this case?





From peter at engcorp.com  Thu Jun  6 21:24:06 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 06 Jun 2002 21:24:06 -0400
Subject: Medium to Large Scale Python Deployments
References: <mailman.1023405789.30972.python-list@python.org> <aJTL8.21666$ZF5.2505752733@newssvr10.news.prodigy.com>
Message-ID: <3D000B36.5C7BCEBF@engcorp.com>

"Domenic R. Merenda" wrote:
> 
> "Sam Penrose" <spenrose at intersight.com> wrote:
> > I'm also curious about large amounts of Python, period. Who knows of
> > ~50K Python lines-of-code (LOC) systems? 100K LOC? More?
> 
>    This is easily over the 100K LOC mark, although I don't know the actual
> count.  We have 37 lib mods for Accounts Payable alone.

Running pycount (http://starship.python.net/crew/gherman/playground/pycount/ )
as "pycount.py -R *.py" in the top level of your source tree would tell
you something more accurate if you wished.

-Peter


From ftobin at neverending.org  Mon Jun 10 23:16:30 2002
From: ftobin at neverending.org (Frank Tobin)
Date: Mon, 10 Jun 2002 23:16:30 -0400
Subject: What is a classmethod for?
In-Reply-To: <83y9dmokh0.fsf@panacea.canonical.org>
References: <addgv30f4d@enews4.newsguy.com>
 <Pine.LNX.4.44.0206031559001.8474-100000@palanthas.neverending.org>
 <83y9dmokh0.fsf@panacea.canonical.org>
Message-ID: <Pine.LNX.4.44.0206102306060.4174-100000@palanthas.neverending.org>

On 10 Jun 2002, Kragen Sitaker wrote:

> That's so you don't have to write this?
> 
> class HostPortAddress(tuple): pass
> def HostPortAddress_from_str(s): 
>     return HostPortAddress(s.split(':', 1))

Well, it makes sense to throw all the 'knowledge' of how to stringify and 
de-stringify a HostPortAddress into one class.

> Are there times when it's more compelling?

Another example would be anytime where you have something that is close to
a staticmethod which is a helper to the instance methods ,but needs some
information that only the class can provide.

-- 
Frank Tobin			http://www.neverending.org/~ftobin/



From aahz at pythoncraft.com  Mon Jun 24 22:58:56 2002
From: aahz at pythoncraft.com (Aahz)
Date: 24 Jun 2002 22:58:56 -0400
Subject: Suggestions for good programming practices?
References: <WEuR8.2074$5l4.75657@ozemail.com.au>
Message-ID: <af8m9g$s5m$1@panix1.panix.com>

In article <WEuR8.2074$5l4.75657 at ozemail.com.au>,
Dianne van Dulken <dianne at dogmac.com> wrote:
>
>I'm fairly new to python, coming from a perl background, and was wondering
>if anyone has a list of things that they consider as good programming
>practice in Python (I thought this might be an interesting topic for
>discussion, anyway)

In Python 2.1.2 or higher, fire up the interactive interpreter and type
"import this".
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From gerhard.haering at gmx.de  Fri Jun 21 06:03:16 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 21 Jun 2002 12:03:16 +0200
Subject: I'd give up Perl tomorrow if only...
In-Reply-To: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au>
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au>
Message-ID: <20020621100315.GA12204@lilith.my-fqdn.de>

* David Mitchell <djmitchell at NOSPAM.optushome.com.au> [2002-06-21 19:27 +1000]:
> .. there was something equivalent in scope and breadth to CPAN.

You'll have to wait some more time for this. I'm pretty disappointed
that http://sf.net/projects/pythonsiphon didn't progress in the last
months. (I'm subscribed to the mailing lists and contributed a patch
that's not applied, yet).

Also jafo did some great work which (he told me on IRC) he sent to the
pythonsiphon maintainer.

> However, I'm probably equally proficient in Perl, and Perl has this
> incredible resource in CPAN.

I think the main point of CPAN is the automagical installation, that
pythonsiphon tries to bring to Python.

As for a searchable database, there's the Vaults of Parnassus, and the
Python modules category at dmoz.org/directory.google.com/you.name.it.

> For just about any library module I've ever
> wanted, I can download one or more modules from CPAN that gets me 90% of the way
> there.  Perl modules like DBI/DBD, SOAP::Lite, GD::Graph, MQSeries::*,
> Tie::*, Convert::EBCDIC,

No idea what Tie::* is, but EBCDIC sounds easy to implement, should it
not exist already, I haven't heard of an MQSeries binding for Python,
but I've personally used Python libraries for the first three items, and
they worked well.

> If there was only some way to subclass Perl/CPAN modules as Python
> modules

PyPerl exists. That's all I can say to that (never used it myself), as
my brain is incompatible with Perl.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From frithiof.jensen at justremovethis.jensen.tdcadsl.dk  Fri Jun 28 16:16:40 2002
From: frithiof.jensen at justremovethis.jensen.tdcadsl.dk (Frihtiof Andreas Jensen)
Date: Fri, 28 Jun 2002 22:16:40 +0200
Subject: Plugins in Python
References: <afgeol$sgd$1@mulga.cs.mu.OZ.AU> <3D1BD350.8F0C74FF@alcyone.com> <3D1CAD56.6070500@REMOVE-THIS-TO-EMAIL-ME.austin.rr.com>
Message-ID: <3d1cc403$0$127$edfadb0f@dspool01.news.tele.dk>

"Brandon Beck" <bbeck at REMOVE-THIS-TO-EMAIL-ME.austin.rr.com> skrev i en
meddelelse news:3D1CAD56.6070500 at REMOVE-THIS-TO-EMAIL-ME.austin.rr.com...


> Hope this helps some.
>

Hot Damn - Thanks ;-)

I am trying to solve a similar problem with an application, where the user
can add code modules to a larger set of "canned" functionality - well, a
framework for developing stock trading systems. This really helps me along.




From ftobin at neverending.org  Sun Jun 30 16:03:58 2002
From: ftobin at neverending.org (Frank Tobin)
Date: Sun, 30 Jun 2002 16:03:58 -0400
Subject: How to stop a SocketServer?
In-Reply-To: <MPG.177e72971f3f789198968e@news.rz.uni-karlsruhe.de>
References: <7xhejv9af2.fsf_-_@ruckus.brouhaha.com>
 <MPG.177e72971f3f789198968e@news.rz.uni-karlsruhe.de>
Message-ID: <Pine.LNX.4.44.0206301602290.7807-100000@palanthas.neverending.org>

Uwe Mayer, on 2002-06-22, wrote:

> I'll be faceing the same problem in a few weeks. Does sys.exit()
> terminate a thread or the interpreter?

The way I currently do it is by sending a TERM signal to the process.  
Dirty, but SocketServer doesn't really give any other good way.

-- 
Frank Tobin			http://www.neverending.org/~ftobin/



From theller at python.net  Mon Jun 24 09:06:59 2002
From: theller at python.net (Thomas Heller)
Date: Mon, 24 Jun 2002 15:06:59 +0200
Subject: module py2exe
References: <mailman.1024919252.19619.python-list@python.org>
Message-ID: <af75hk$boah0$1@ID-59885.news.dfncis.de>

<jubafre at zipmail.com.br> wrote in message news:mailman.1024919252.19619.python-list at python.org...
> i don?t know how to use a py2exe module, i have made the download but i
> don?t know to install. the direcroty of installation is lib\site-packages.
>
> i have a program using tkinter module and i want to compile to .exe the
> name of the file is lfa.py, i read in http://starship.python.net/crew/theller/py2exe/,
> i have to make a setup.py, so doesn?t work, how i can make this rigth???????????????????
>
> # setup.py
> from distutils.core import setup
> import py2exe
>
> setup(name="lfa",
>       scripts=["lfa.py"],
> )
>
> Juliano Freitas
> www.gebrasil.hpg.com.br
This looks good.
All you now have to do is to create the exe-file by running
'python setup.py py2exe'. If all is well, you can find your program
(together with other files it needs) in the subdirectory dist\lfa.

HTH,

Thomas




From mhammond at skippinet.com.au  Wed Jun 26 22:35:21 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 27 Jun 2002 02:35:21 GMT
Subject: COM server threads/unloading
References: <10d944cb.0206261529.741da082@posting.google.com>
Message-ID: <3D1A7A9A.4050208@skippinet.com.au>

There are a couple of things you should:
* Ensure CoUninitialize() is called for each thread (you would have 
needed to call CoInitialize[Ex] manually for these threads.
* Call CoUninitialize() on the main thread.  You did not need to call
* Ensure pythoncom._GetInterfaceCount() and pythoncom._GetGatewayCount() 
both return zero.  If not, your Python program still has a COM 
reference, and this should be nuked.

As far as I know, if all these conditions are met everytihing will 
unload correctly.

Mark.

Uncle Enzo wrote:
> Hello, 
> 
> I have noticed that my multithreaded COM servers are not unloading
> when their clients have unloaded.
> 
> To reproduce this in a simple case, I am using the py2exe COM server
> recipe. Running the server example with the client code below results
> in the testserver.exe loading after the dispatch call, and unloading
> after the server reference is set to None.
> 
> Then I added the googlelogger import statement and the two lines
> initializing and calling the log method. The logger that has two
> threads. Using the same client code results in the testserver.exe
> loading after the dispatch call, but it does not unload after the
> server reference is set to None, nor does it unload after the client
> terminates.
> 
> What is the proper way to get a multithreaded python COM server to
> unload when there are no external references to it?
> 
> Thanks in advance,
> 
> Enzo
> 
> ps, comments about my logger design are welcome, but I'm mostly
> interested in the threading issues :)
> 
> -----------------------------
> 
> Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on
> win32
> Type "help", "copyright", "credits" or "license" for more information.
> 
>>>>import win32com.client
>>>>server = win32com.client.Dispatch("Python.TestServer")
>>>>print server.Hello('enzo')
>>>
> Hello enzo
> 
>>>>server = None
>>>>^Z
>>>
> 
> ----------------------------
> file: testcomserver.py
> 
> import googlelogger
> import sys
> import pythoncom
> 
> if hasattr(sys, 'importers'):
>     # we are running as py2exe-packed executable
>     pythoncom.frozen = 1
> 
> class HelloWorld:
>     _reg_clsctx_ = pythoncom.CLSCTX_LOCAL_SERVER
>     if hasattr(sys, 'importers'):
>         # In the py2exe-packed version, specify the module.class
>         # to use. In the python script version, python is able
>         # to figure it out itself.
>         _reg_class_spec_ = "__main__.HelloWorld"
>     _reg_clsid_ = "{592C15CE-894D-11D6-9A35-0050DA05A582}"
>     _reg_desc_ = "Python Test COM Server"
>     _reg_progid_ = "Python.TestServer"
>     _public_methods_ = ['Hello']
>     _public_attrs_ = ['softspace', 'noCalls']
>     _readonly_attrs_ = ['noCalls']
>     def __init__(self):
>         self.softspace = 1
>         self.noCalls = 0
>         self.mylogger = googlelogger.getLogger()
>     # __init__()
> 
>     def Hello(self, who):
>         self.noCalls = self.noCalls + 1
>         # insert "softspace" number of spaces
>         self.mylogger.log('caller', str(who))
>         return "Hello" + " " * self.softspace + str(who)
> 
> if __name__ == '__main__':
>     if hasattr(sys, 'importers'):
>         # running as packed executable.
>         if '--register' in sys.argv[1:] or '--unregister' in
> sys.argv[1:]:
>             # --register and --unregister work as usual
>             import win32com.server.register
>             win32com.server.register.UseCommandLine(HelloWorld)
>         else:
>             # start the server.
>             from win32com.server import localserver
>             localserver.main()
>     else:
>         import win32com.server.register
>         win32com.server.register.UseCommandLine(HelloWorld)
> 
> ----------------------
> file: setup.py
> 
> from distutils.core import setup
> import py2exe
> 
> setup(name='testCOMserver',
>       scripts=['testCOMserver.py'])
> 
> ------------------------
> file: googlelogger.py
> 
> import threading
> import Queue
> import sys
> import pavlnchcfg
> import time
> import win32com.client
> 
> # This thread waits for ("name", "value") items to appear in the
> queue.
> # It then checks for a log named "name", creates it if it doesn't
> exist,
> # and appends the message to that log, with a timestamp.
> class Writer(threading.Thread):
>     def setLogger(self, logger):
>         #print "setting self._logger to ", logger
>         self._logger = logger
>         
>     def run(self):
>         print "starting Writer.run thread"        
>         while 1 :
>             # blocks until there are items in the queue
>             event = self._logger._logeventqueue.get(1) 
>             #get returns a tuple, (name, message)            
>             if event[0] == 'QUIT': #magic quit event
>                 print "logger.Writer.run: got magic event: returning"
>                 return
>             
>             logname = event[0] 
>             message = event[1]
>             try:
>                 filename = self._logger._dir + '\\' + logname + ".txt"
>                 logfile = open(filename, 'a')    
>                 logfile.write(time.asctime() + " " + message + '\n')
>                 logfile.flush()
>                 logfile.close()
>                 print "logger.Writer.run: logged ", message, " to ",
> logname
>             except:
>                 print "logger.Writer.run: with to log: ", logname, "
> failed with message: ", message
> 
> class Logger:
>     # class variables
>     _logeventqueue = Queue.Queue() #one thread safe multiple user
> queue
>     _dir = "" 
>     
>     def __init__(self):          
>         try:            
>             Logger._dir="C:\\" # for our example
>             print "logger.Logger.init: got log directory",
> Logger._dir, "from config"
>         except:
>             Logger._dir = "C:\\"
>             print "logger.Logger.init: No log directory specified in
> config, _dir set to  C:\\"
>             
>         writer = Writer()
>         writer.setName("writerthread")
>         writer.setLogger(self)
>         writer.start()
>         
>     def log(self, logname, message):
>         event = logname, message  #making the event tuple
>         # event tuple is tuple[0] = name, tuple[1] = event
>         Logger._logeventqueue.put(event)      
>             
> 
> # factory pattern as module instance
> _loggerInstance = None
>        
> def getLogger():
>     import logger
>     if logger._loggerInstance == None:
>         logger._loggerInstance = Logger()
>         return logger._loggerInstance
>     else:
>         return logger._loggerInstance
> 
> if __name__ == '__main__':  #test main
>     import logger
>     print "starting main"
>     l = getLogger()
>     print l
>     l.log("DEBUG", "l, debug message 1")
>     l.log("USAGE", "l, usagestat 1")
>     l.log("DEBUG", "l, debug message 2")
>     lg = getLogger()
>     print lg
>     lg.log("DEBUG", "lg, debug message 1")
>     lg.log("USAGE", "lg, usagestat 1")
>     lg.log("DEBUG", "lg, debug message 2")    
>     l.log("QUIT", "QUIT")
>     print "done"




From aahz at pythoncraft.com  Mon Jun 10 18:45:35 2002
From: aahz at pythoncraft.com (Aahz)
Date: 10 Jun 2002 18:45:35 -0400
Subject: Iterators vs. Generators
Message-ID: <ae3a6f$9d9$1@panix1.panix.com>

So when would one actually write an iterator instead of a generator?
I've been trying to think of an example and failing.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From ark at research.att.com  Fri Jun  7 15:29:40 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 7 Jun 2002 19:29:40 GMT
Subject: Python 2.2.1 vs. gcc 3.1?
References: <yu99vg8v6orb.fsf@europa.research.att.com> <15616.57963.519636.730290@12-248-41-177.client.attbi.com> <200206071747.g57HlbH01057@europa.research.att.com> <mailman.1023472897.6311.python-list@python.org>
Message-ID: <yu99fzzy7uqj.fsf@europa.research.att.com>

Andrew> How do I go about finding out the cause of the failure?

Skip> Try running the just built python executable and then manually execute
Skip> "import struct".  What error message(s) do you see?

OK... I now understand why I couldn't find struct.so:
the setup.py program deletes it after it finds that the import
attempt fails.

So, in setup.py, I did this, starting at line 164:

#        try:
            __import__(ext.name)
#        except ImportError:
#            self.announce('WARNING: removing "%s" since importing it failed' %
#                          ext.name)
#            assert not self.inplace
#            fullname = self.get_ext_fullname(ext.name)
#            ext_filename = os.path.join(self.build_lib,
#                                        self.get_ext_filename(fullname))
#            os.remove(ext_filename)

That left the various .so libraries around, which allows me to try them out:

$ ./python
Python 2.2.1 (#6, Jun  7 2002, 14:55:06) 
[GCC 3.1] on sunos5
Type "help", "copyright", "credits" or "license" for more information.
>>> import struct
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: ld.so.1: ./python: fatal: relocation error: file /tmp/build-python/Python-2.2.1/build/lib.solaris-2.8-sun4u-2.2/struct.so: symbol PyString_Type: referenced symbol not found

So apparently the shared libraries are not being built properly.

Any more ideas?

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From whisper at oz.net  Mon Jun  3 18:46:16 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 3 Jun 2002 15:46:16 -0700
Subject: Time
In-Reply-To: <3cfbdf0c_1@news.iprimus.com.au>
Message-ID: <GCEDKONBLEFPPADDJCOEMEGIDGAA.whisper@oz.net>

You might want to check out mxDateTime in the mx toolkit at
http://www.egenix.com/files/python/. I don't see a direct +/-day function
but there is access to the elements of a datetime such that you could
add/subtract an integer to the day value

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Gold Fish
> Sent: Monday, June 03, 2002 14:32
> To: python-list at python.org
> Subject: Time
>
>
> Anyone know how to determine the time in python. If the mtime in the file
> filename.py is Mon Jun 3 03:25:57 2002 ,if i want to set the time forward
> or backwards in 5 days time, how can i do that.
> --
> http://mail.python.org/mailman/listinfo/python-list





From ianb at colorstudy.com  Sat Jun  1 02:52:01 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 01 Jun 2002 01:52:01 -0500
Subject: How to open a HTML file when the python cgi program is
	executing?
In-Reply-To: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de>
References: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de>
Message-ID: <1022914321.6150.6848.camel@localhost>

Just do:

print open('main.html').read()

(making sure you print headers, like Content-type, first)

On Sat, 2002-06-01 at 01:43, Ken wrote:
> How do I open a HTML file when the cgi program is executing?
> 
> The only method I can think of, which doesn't work is this:
> <meta HTTP-EQUIV="REFRESH" CONTENT="0; URL=main.html">
> 
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 






From tjreedy at udel.edu  Thu Jun 13 10:41:06 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Thu, 13 Jun 2002 14:41:06 GMT
Subject: Bits and bytes?
References: <mailman.1023896057.21318.python-list@python.org> <3d08839a$0$226$4d4ebb8e@news.nl.uu.net>
Message-ID: <6a2O8.153745$4i.13897852@bin2.nnrp.aus1.giganews.com>

"Guyon Mor?e" <gumuz at looze.net> wrote in message
news:3d08839a$0$226$4d4ebb8e at news.nl.uu.net...
> i am trying to implement my own sort of huffman
encoding/compression.
> converting asci-characters to a compressed bitstream requires me to
write it
> at bit-level. Bill's solution might work in some way, but my
sequence of
> bits aren't allways divideable by eight. a word like 'ape' (which is
3x8=24
> bits) might become 1001101, which is 7 bits.... catch my drift?
strings is
> no option as this would have no positive effect on the 'real size'
;there's
> no compression.

For production use, bit twiddling should be done in assembler, C, or
other suited language.  For algorithm development in Python, you could
do something like the following:

  charmap = <map of tuples of 8 0s and 1s to corresponding char> # (
  work = []
  out = []
  for char in plaintext:
    work.append(cipher(char)) # cipher -> list of 0s and 1s
    if len(work) >= 8:
      out.push(charmap(tuple(work[0:8])))
      work[:] = work[8:]
  ciphertext = ''.join(out)

or make 'out' an array from array module to avoid final .join()

Terry J. Reedy





From shredwheat at attbi.com  Mon Jun  3 11:50:19 2002
From: shredwheat at attbi.com (Pete Shinners)
Date: Mon, 03 Jun 2002 15:50:19 GMT
Subject: thread module under Windows
References: <59d8a5f6.0206030730.1be14ffb@posting.google.com>
Message-ID: <3CFB9030.4030802@attbi.com>

Dmitri wrote:
> I have been searching the conference for the simple question like "
> where the thread module is " and got the answer that it's inside the
> python if the threading support is on.

Thread support defaults to "on" for all python compiles. all windows users 
will have thread support, unless they have specifically compiled there own 
without. an easy way to test is to just try it..


try:
     import thread
     print 'Thread support tastes good'
except ImportError:
     print 'Too bad, no Thread support'



From jknapka at earthlink.net  Sat Jun 29 14:48:22 2002
From: jknapka at earthlink.net (Joseph A Knapka)
Date: Sat, 29 Jun 2002 18:48:22 GMT
Subject: Python needs better error reporting
References: <LNBBLJKPBEHFEDALKOLCEEOKAAAB.tim.one@comcast.net> <mailman.1025351378.7466.python-list@python.org> <W2iT8.45588$n4.10825262@newsc.telia.net>
Message-ID: <3D1E00AD.DDD32BA6@earthlink.net>

Fredrik Lundh wrote:
> 
> Walter D?rwald wrote:
> 
> > This is easy: Use the _G_uido _C_loning _C_ommand! ;)
> 
> the one with a -Wall option?  scary...
> 
> </F>

Let's not let this turn into one of those threads
where everyone parrots one another's puns. I've
become a bit Larry of those.

Cheers,

-- Joe


From bokr at oz.net  Mon Jun 24 20:36:09 2002
From: bokr at oz.net (Bengt Richter)
Date: 25 Jun 2002 00:36:09 GMT
Subject: Web templating/db tool with best designer/coder separation?
References: <af2gh6$2c2$0@216.39.172.122>
Message-ID: <af8dtp$5qp$0@216.39.172.122>

On 22 Jun 2002 18:43:50 GMT, bokr at oz.net (Bengt Richter) wrote:

>E.g., if you were hoping for an improvement over current practice,
>which do you consider to be the tool(s) to beat, and what would
>the killer improvement(s) be? TIA.
>
Thank you all (alphabetically:  Ian Bicking, Paul Boddie, Tim Churches,
Dave Cole, Max M, and Jon Ribbens ;-) for your thoughtful responses and
links to relevant resources.

For the Python aspect of things, Albatross seems very similar to
(though more elaborate than) what I had in mind, but ZPT is more to my
liking in the HTML markup area. 

I looked at stuff at the links offered (and browsed a bit):

    http://jonpy.sourceforge.net/
    http://www.object-craft.com.au
    http://www.object-craft.com.au/projects/albatross/albatross/fig-presimpexec.html
and I Googled for Cheetah:
    http://www.cheetahtemplate.org/

No one mentioned ZPT (Zope Page Templates),
    http://www.zope.org//Wikis/DevSite/Projects/ZPT/FAQ

and from the description, this would seem to have a real advantage over all of
the above, in that it uses attributes of html tags as opposed to special tags
that browsers or design tools won't generally understand and render usefully.

This is touted as enabling the "round trip" from graphic designer to template logic
programmer and back using the same html source. I assume this is the "round trip"
Paul was referring to. To me, it seems a winning point for ZPT, though I don't know
how ZPT works out in practice.

They have a list of design tools that apparently pass through the ZPT markup they
don't understand. According to the FAQ:

   "Q8: What HTML editors work well with ZPT?
        We are still compiling a list, but on the commercial (industrial strength)
        front, GoLive and Dreamweaver 4 work well with ZPT. Recent versions of Amaya
        work well in HTML mode."

Opinions on ZPT?

BTW, which is the most appropriate Python version to assume everyone's servers
will be converging on? 1.5, 2.1, 2.2, or ? Could I reasonably plan on using 2.2?
E.g., I think 2.1 is being included in most newer linux distributions. But more
to the point, what version will most likely be available via server cgi etc.?
I'd like the answer to be 2.2 ;-)

Regards,
Bengt Richter


From BPettersen at NAREX.com  Sat Jun 29 16:48:13 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Sat, 29 Jun 2002 14:48:13 -0600
Subject: private
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E201922151BA@admin56.narex.com>

> From: Dave Reed [mailto:dreed at capital.edu] 
> 
[snip private members]
> 
> Kind of off-topic, but this got me curious. The only way I 
> could think of is to assume (which I think is always true) 
> that the private data members are in order from the starting 
> address so based on the offset from the address of the object 
> you could access each one.

.. and of course #define private public, or just go in and edit the .h
file...

-- bjorn




From anton at vredegoor.doge.nl  Sat Jun 29 19:37:18 2002
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Sun, 30 Jun 2002 01:37:18 +0200
Subject: [MODULE] rubiks cube simulation for visual python
References: <afkrn2$obb$1@news.hccnet.nl>
Message-ID: <aflge6$f4g$1@news.hccnet.nl>

On Sat, 29 Jun 2002 19:43:40 +0200, anton at vredegoor.doge.nl (Anton Vredegoor)
wrote:

>A very basic rubiks cube simulation written for visual python.
>Displays a 3D simulation of a rubiks cube. It is possible to view to cube from
>all sides, the cube can be zoomed, the layers can be rotated
>by typing single keystrokes and there is an undo function.
>
>http://home.hccnet.nl/a.vredegoor/rubik/rubik.py
>
>or (html-version)
>
>http://home.hccnet.nl/a.vredegoor/rubik/rubik.html

I found an error in the code but its fixed now. Download again to get the
fixed script.


From pixie888 at hotmail.com  Wed Jun 19 06:09:20 2002
From: pixie888 at hotmail.com (pixie888 at hotmail.com)
Date: Wed, 19 Jun 2002 10:09:20 GMT
Subject: SafeArray problem
Message-ID: <3d105710.83643750@news.skynet.be>

Hi all,

I am using an Automation object which has a method which takes a
SafeArray as an argument. In my Python module I call this method and
give a list of longs as an argument.

While debugging (using the VC++6 IDE) I notice that the type of the
argument (upon entering my automation object) is VT_ERROR (argument
not found, actually). Now, what is really odd is the following: I
extended my method: it takes two arguments, the first being a
SafeArray and the second a long (which will not be used). When I call
this method again from my Python module (with a list of longs and a
dummy long as arguments) then everything works fine.

Beats me, really.

Any idea where the problem lies?

Thanks,

Henk


From oren-py-l at hishome.net  Thu Jun 27 03:32:29 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Thu, 27 Jun 2002 03:32:29 -0400
Subject: Pythoniac: Thoughts on a hardware Python processor
In-Reply-To: <afc52t$2pj$1@sirius.dur.ac.uk>
References: <afc52t$2pj$1@sirius.dur.ac.uk>
Message-ID: <20020627073229.GA97528@hishome.net>

On Wed, Jun 26, 2002 at 10:29:52AM +0000, Christopher Saunter wrote:
> Dear All,
> 
> I have seen the occasional post about creating a specific hardware 'Python
> Processor', that could then be built and used in programmable logic or the
> like (Valves, relays, core memory - Pythoniac ;-) - in other words, create
> a processor that executes Python bytecode directly.  There would be
> various different reasons for doing this (the 'neat' factor etc...)  

One of the most effective optimizations for such a machine would be to 
support compact objects.  The machine word would be somewhat larger than the 
actual address space. Let's say 40 bits.  If the 4 high bits are 0 this is a
reference to an object allocated somewhere in memory.  If the high bits are 
not zero it means that the actual object is stored inside the register.  You 
have 15 types that can bypass allocation, deallocation, reference counting, 
etc. The most important compact type is int, of course.  Python floats are 
64 bits so if you want them to be compact objects, too, the machine word
would need to be 64+ bits.

Another important hardware feature is accelerated associative lookup.

	Oren





From tim.one at comcast.net  Mon Jun 24 10:26:34 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 24 Jun 2002 10:26:34 -0400
Subject: Python hits the spot
In-Reply-To: <yu99sn3chiy4.fsf@europa.research.att.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCGEKNPPAA.tim.one@comcast.net>

[Andrew Koenig]
> Don't some implementations of malloc offer a way to ask the operating
> system to commit to the memory being available when you allocate it?
> If so, I can imagine testing for that facility at config time and
> using it if available.

SourceForge remains open 24 hours a day for patches.  My bet is that nobody
cares enough to bother -- the barrier for complaining on Usenet is one an
ant can hurdle with ease <wink>.





From thenault at nerim.net  Sun Jun  9 12:58:44 2002
From: thenault at nerim.net (Sylvain Thenault)
Date: 09 Jun 2002 18:58:44 +0200
Subject: problem with isinstance and relative/absolute import
References: <87adq8jsq1.fsf@tux.ntw23.fr> <Xns92265AA1E2FABgmcmhypernetcom@199.171.54.216>
Message-ID: <87k7p8z8vv.fsf@tux.ntw23.fr>

Gordon McMillan <gmcm at hypernet.com> writes:

> Sylvain Thenault wrote:
> 
> > does someone can explain to me the following test case, or should it be
> > considered as a bug ?
> > 
> > [syt at tux syt]$ ls tests/
> > __init__.py  module2.py  module.py
> 
> [snip]
> 
> > [syt at tux syt]$ cd tests
> 
> Bad move. This makes the modules in the tests 
> directory accessable in 2 ways - as a direct
> import (sys.path[0]), or as a package. The
> way you do your imports, you end up getting
> module twice, once as module, once as tests.module.
> 
> Just look at the repr's of a, b and A to see.
> 
> And stay out of package directories.

actually I did understood what was the problem, I was only asking if this
behaviour should be considered as normal...

Since it's current in a package to mix relative (to the python file directory)
and absolute (package) import, isn't there a risk to obtain a strange behaviour
without being in the package directory ?

-- 
Sylvain Th?nault 


From karczma at info.unicaen.fr  Wed Jun  5 07:08:56 2002
From: karczma at info.unicaen.fr (Jerzy Karczmarczuk)
Date: Wed, 05 Jun 2002 13:08:56 +0200
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet>
Message-ID: <3CFDF148.AA2035FF@info.unicaen.fr>

TuxTrax wrote:

> My question is this; I realize that Python compiles to bytecode automatically
> as part of the interpreters tokenization run, but is there a way to permanently
> compile python code outputting to an executable binary?

Once upon a time Guido Van Rossum thought "this seems very difficult,
perhaps impossible".

He changed his mind a little bit now...

See:

http://www.onlamp.com/pub/a/python/2002/06/04/guido.html

and the references cited by The Master:

http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/
http://homepages.ulb.ac.be/~arigo/psyco/

Don't forget also that Jython produces Java code. Not an
"executable binary" directly, but ...

You may also embed a Python kernel in your application, see
the tutorial:

http://www.python.org/doc/current/ext/embedding.html

==

Jerzy Karczmarczuk
Caen, France


From dreed at capital.edu  Sun Jun 23 11:49:15 2002
From: dreed at capital.edu (Dave Reed)
Date: Sun, 23 Jun 2002 11:49:15 -0400
Subject: (Newbie)  Counting Instances ("Hits") with Regular Expressions
In-Reply-To: <3d15e215.43448155@news.texas.net> (baf@texas.net)
References: <3d15e215.43448155@news.texas.net>
Message-ID: <200206231549.g5NFnFV17357@localhost.localdomain>

> From: baf at texas.net (Ben Fairbank)
> Newsgroups: comp.lang.python
> X-Newsreader: Forte Free Agent 1.21/32.243
> Organization: Giganews.Com - Premium News Outsourcing
> X-DMCA-Notifications: http://www.giganews.com/info/dmca.html
> X-Abuse-Info: Please be sure to forward a copy of ALL headers
> X-Abuse-Info: Otherwise we will be unable to process your complaint properly
> Xref: news.baymountain.net comp.lang.python:169353
> Sender: python-list-admin at python.org
> Date: Sun, 23 Jun 2002 15:08:17 GMT
> 
> I am new both to Python and to regular expressions, which may account
> for my difficulty.  I must count the frequenies of certain words in
> files of moderate length (about150 k bytes).  I have been reading
> files and then using count(s,sub), which is fast and easy.  I now have
> to allow for punctuation and eliminate words within words, etc, and so
> am trying to use regular expressions instead of simple words as
> targets.  I do not, however, find a similarly easy to use count
> function in the re module.  Yet this is such common operation it must
> be there, or easy to implement.  What is the usual way of simply
> counting "hits" in the re module?  (And what have I missed in the
> documentation; where is this to be found?  I have looked through Lutz
> and Ascher)
> 
> Thanks for any help.
> 
> BAF
> -- 
> 
> 

There's probably an easier way to do this, but I think this will do
what you want. It will search the file infile.txt for all occurences
of "word" (w/o the quotes). The \s ensures that there is some sort of
space, newline, tab, etc. before and after "word" so it won't match
"words" or anything containing "word". The match.end() will start the
next search at the position after the previous matching search.

import re

infile = open('infile.txt', 'r')
text = infile.read()
word = re.compile('\sword\s')
match = word.search(text)
count = 0
while match != None:
    count = count + 1
    match = word.search(text, match.end())
print 'count', count

HTH,
Dave




From fork at yahoo.corn  Sun Jun 16 19:45:56 2002
From: fork at yahoo.corn (fork)
Date: Mon, 17 Jun 2002 07:45:56 +0800
Subject: Asia =?Big5?B?s8ymbqq6uvSkV6aopEilzqt+sU295qmx?=
Message-ID: <aejc13$qa0$18799@news.ctimail.com>

http://www.ilookasia.com/sextoy/default.asp






From fredrik at pythonware.com  Tue Jun 25 12:01:21 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 25 Jun 2002 16:01:21 GMT
Subject: Parsing strings (\n and \\)
References: <3D188006.7050202@thomas-guettler.de> <mailman.1025017712.20580.python-list@python.org>
Message-ID: <lt0S8.45025$n4.10623651@newsc.telia.net>

Fran?ois Pinard wrote:

> A simple avenue is to get Python itself to evaluate the string as a constant
> (you ensure this by removing evaluation context).  Something like this:
>
>     PARSED = eval(UNPARSED, {}, {})

>>> UNPARSED = "__import__('os').system('echo dream on!')"
>>> PARSED = eval(UNPARSED, {}, {})
dream on!

</F>




From fredrik at pythonware.com  Sat Jun 29 12:33:18 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 29 Jun 2002 16:33:18 GMT
Subject: exception handing
References: <pan.2002.06.29.17.41.28.926443.440@myself.com>
Message-ID: <ijlT8.45602$n4.10832036@newsc.telia.net>

"Rhymes" wrote:

> I'm reading "Learning Python" and in the section "gotchas"
> of the exception handling chapter the book itself states
> that the exception matching is made by identity not equality.
> It also write an example such as this:
>
> >>> ex1 = "spam"
> >>> ex2 = "spam"
> >>>
> >>> ex1 == ex2, ex1 is ex2
> (1, 0)
>
> <--- here i get (1, 1) ---->

since strings cannot be modified in place, Python is free
to reuse the same string object when it notices that you're
assigning the same literal to two different variables.

Python 2.2 does this for literals that looks like Python
identifiers; Python 1.5 obviously didn't.

to get the effect the authors had in mind, you can do
this instead:

>>> ex1 = "sp am"
>>> ex2 = "sp am"

or

>>> ex1 = "spam"
>>> ex2 = "sp" + "am"

</F>




From jadestar at idiom.com  Tue Jun 18 23:40:26 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 19 Jun 2002 03:40:26 GMT
Subject: zlib/gzip + anydbm or ...
Message-ID: <aeoufa$rkk$1@news.idiom.com>

 Is there an obvious and simple way to use the existing standard
 libraries to support transparently compressed dbm files?

 (Any combination of zlib/gzip etc with any of the dbm files would
 be considered).





From egbert at bork.demon.nl  Tue Jun 11 04:57:53 2002
From: egbert at bork.demon.nl (Egbert Bouwman)
Date: Tue, 11 Jun 2002 10:57:53 +0200
Subject: Program to an interface, not an implementation
In-Reply-To: <ae2ad4$s36$1@wanadoo.fr>; from eric.brunel@pragmadev.com on Mon, Jun 10, 2002 at 03:48:55PM +0000
References: <D54BF26DB5A7D511B36B00508B9596A402F8EC63@sdgcaexch2.peregrine.com> <mailman.1023704594.18063.python-list@python.org> <ae2ad4$s36$1@wanadoo.fr>
Message-ID: <20020611105752.A985@bork.demon.nl>

On Mon, Jun 10, 2002 at 03:48:55PM +0000, Eric Brunel wrote:
> Hope it's a bit clearer now...
Yes, much clearer, thanks to you all, your patience and didactic 
qualities. I suppose this thread is essential reading for any 
intermediate python programmer interested in Design Patterns.
egbert
-- 
Egbert Bouwman - Keizersgracht 197 II - 1016 DS  Amsterdam - 020 6257991
========================================================================




From K.Rdt at TU-Berlin.DE  Fri Jun 28 05:35:04 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Fri, 28 Jun 2002 11:35:04 +0200
Subject: menue windowing help needed
Message-ID: <MHJE877496QMKJGCTQLZMJ5ZUPLHSQ.3d1c2dc8@Station25>

---------------------------------------------------------------------
Hi

I tried the samples from wxPython and tkinter
without succes.

I need some scripting help for:
	A menue
	each alternative can open
	one window for print or ..

TiA
		K at Rdt
---------------------------------------------------------------------







From karczma at info.unicaen.fr  Thu Jun 20 04:56:01 2002
From: karczma at info.unicaen.fr (Jerzy Karczmarczuk)
Date: Thu, 20 Jun 2002 10:56:01 +0200
Subject: Pedagogic advice needed
References: <3D104D35.E0932297@info.unicaen.fr> <aeqlll$h4t$1@panix1.panix.com>
Message-ID: <3D1198A1.394105A7@info.unicaen.fr>

In my previous posting I asked about your experience
in teaching scientific programming/visualization in
Python (my case: to biologists, but this is not as relevant).

Thank you very much folks for all your helpful answers,
on this nsgroup and privately. I would just like to point
out that answers such as this and similar, often found
here:

Aahz wrote:

> From http://www.python.org/doc/Newbies.html
> you will find "Bioinformatics course in Python"
> at http://www.pasteur.fr/recherche/unites/sis/formation/python/


sometimes slightly miss the target. Of course I *AM GRATEFUL* to
Aahz Maruch, a known specialist, and I accept all help offered (I
am quite far from the sad hero of the thread named "Python version"),
but, believe me, sometimes people really need concrete, personal
experience, not just links to tutorials. I am not blaming Aahz, of
course, he couldn't guess what did I really need. The course of Katja 
Schuerer and Catherine Letondal (which I recommend to many newcomers) 
is an introductory course treating Python as a programming language, 
with examples oriented towards biology. I needed a dense, shock
material, explicitly application-driven, and with a good deal of
visualization examples, absent from the cited course. Once more:
I asked for your personal experience (mine isn't so bad, about
30 years of teaching...)

OK, I will look closer at SciPy, and the sources kindly mentioned
by Fernando P?rez.

As you know, the "scientific" (in the restricted sense) community
is interested in Python for years, yet in the huge Python FAQ,
especially in the section 'Real World' there is almost nothing
about this, God knows why... [[Anyway, people who are managing
this FAQs: http://www.python.org/doc/FAQ.html perhaps would one
day wish to update some links, eg. the first link pointing to docs:
http://www.pythonlabs.com/tech/python2.0/doc which points nowhere]].



> I imagine you might even be able to find some documentation in French.

Oh, yessir, indeed, thank you for this helpful hint. Merci beaucoup et
dziekuje bardzo in my native tongue as well.


Jerzy Karczmarczuk
Caen, France


From andreas at zope.com  Mon Jun  3 13:52:58 2002
From: andreas at zope.com (Andreas Jung)
Date: Mon, 3 Jun 2002 13:52:58 -0400
Subject: Question: Importing the same module under different names
Message-ID: <013501c20b27$a4e757d0$1717a8c0@suxlap>

Usually Python imports a module 'foo' only once when one uses
multiple 'import foo' statements. However Python load the 
module when one uses packages e.g.

packages/__init.py
        /foo.py


When PYTHONPATH is set to the 'packages' directory then

import foo
import packages.foo 

or

import foo
from packages import foo

loads the foo module twice. How does Python determine in this case
to load the module or not?

Andreas 







From cjw at sympatico.ca  Sat Jun 15 12:56:29 2002
From: cjw at sympatico.ca (Colin J. Williams)
Date: Sat, 15 Jun 2002 12:56:29 -0400
Subject: Integer conversion
References: <3D0B3257.E8572AE5@sympatico.ca> <3D0B3EB7.2020304@wedgetail.com>
Message-ID: <3D0B71BD.68994121@sympatico.ca>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020615/2851c841/attachment.html>

From ffjhenon at club-internet.fr  Mon Jun 17 15:01:58 2002
From: ffjhenon at club-internet.fr (Fabien =?iso-8859-1?Q?H=E9non?=)
Date: Mon, 17 Jun 2002 21:01:58 +0200
Subject: If that interests you...
Message-ID: <3D0E3226.48E6D338@club-internet.fr>

I have written an editor for POV-RAY ( the raytracer) using Tkinter that

features syntax highlighting.
It fires/stops up raytraces, save,....

It can be found at :
http://pyvon.sourceforge.net

Fabien




From garrett at bgb.cc  Wed Jun  5 02:39:13 2002
From: garrett at bgb.cc (Don Garrett)
Date: Wed, 05 Jun 2002 06:39:13 GMT
Subject: Printing under Windows
Message-ID: <3CFDB1B2.7060302@bgb.cc>

   A while back, I posted a request of how to print graphical images on 
Windows from Python. There were no answers suggested that I could use. I have 
recently found that the problem can be solved using the wxPython wrapper for 
wxWindows.

  I just thought I would mention it. After two days of exposure (not a lot, I 
admit), wxPython seems to be a great system to work with and to solve several 
problems I've for a while. I'm able to print, access the system tray, and 
build an interface that looks VERY windows native while remaining cross platform.

http://www.wxwindows.org/
http://wxpython.org/

--
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc



From peter.rams at sap.com  Fri Jun  7 12:07:34 2002
From: peter.rams at sap.com (Peter Rams)
Date: Fri, 7 Jun 2002 18:07:34 +0200
Subject: session managment?
Message-ID: <adqlo6$1da$1@news1.wdf.sap-ag.de>

Hi,

is there a way to have internet session managment with python? I want to
build some pages where someone can log in with user/password...

Regards,
Peter




From pinard at iro.umontreal.ca  Mon Jun 24 00:33:20 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 24 Jun 2002 00:33:20 -0400
Subject: Suggestions for good programming practices?
In-Reply-To: <WEuR8.2074$5l4.75657@ozemail.com.au>
References: <WEuR8.2074$5l4.75657@ozemail.com.au>
Message-ID: <oqhejtz46n.fsf@titan.progiciels-bpi.ca>

[Dianne van Dulken]

> I [...] was wondering if anyone has a list of things that they consider
> as good programming practice in Python (I thought this might be an
> interesting topic for discussion, anyway)

I've seen many ideas expressed on this list.  Being rather maniacal on
details, I surely gave many programming rules to myself which I try to
follow.  Maybe I should take the time and formalise these into some Web page.
On the other hand, style may always be the source of heated debates!

Despite Python is simple, I try to _not_ even use all of it.  For example,
`print' is strictly reserved for debugging purposes, I never use `print'
for production scripts.  I also completely avoided `input' so far.  It is
exceptional that I use `exec' or `eval', and I almost never use `global'.

Python has evolved fastly in recent times.  You might also want to decide
for which Python version you aim your code, and choose to _not_ take
advantage of all the features of the latest Python available.  This is
especially useful if you move between machines, or have a user base.

> For example, in perl you obviously are always encouraged to use strict, and
> we always use eval around our main code to pick up any unexpected errors.

Python is pretty exceptional in that it catches and reports errors
automatically, unless you do take special action to prevent it of doing so.
In C or Perl, this is a constant burden for any programmer who takes style
seriously.  I see this as one of the important reliefs brought by Python.

I also much like `assert' as an expedient way to report consistency
errors, and even usage errors.  I do not mind tracebacks in such cases,
even if they are not much useful in case of usage errors.  With `assert',
the second argument is often a tuple containing all variables which are
meaningfully dumped to understand why the `assert' was triggered.

> Any other tips that you would suggest a new python user should be always
> doing as a good programming practice?

Peter Hansen already suggested the Guido style guide, which hold many
good ideas.

All my executable scripts use similar conventions about how to organise
calling of the main function, decoding options, providing help, etc.
My projects are also structured in similar ways (`Makefile', `setup.py',
etc.) and installable without having to be `root'.  I gave a fair amount
of thought before deciding how those things are best done.  My co-workers
adopted very similar conventions, so we can more easily share development.

The key point of all this is that it pays being consistent between all
one's Python works.  My suggestion is that you should actively develop a
style that suits you, and reasonably be faithful to your own style. :-)

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From fredrik at pythonware.com  Thu Jun 20 04:51:29 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 20 Jun 2002 08:51:29 GMT
Subject: How to find all emails (on the same line)
References: <mailman.1024560806.17188.python-list@python.org>
Message-ID: <lIgQ8.44253$n4.10309460@newsc.telia.net>

"A" <printers at sendme.cz> wrote:

> I need to find ALL emails in a file.

that's about as vague as specifications come.  what kind
of mail?  what kind of file?  given that emails usually con-
sists of a couple of header lines followed by a body, what
kind of file can contain multiple mails on a single line? (etc)

> I can use re module but I do not know how to find all emails
> if there are more emails on the same  line because re module
> (as far as I know ) finds only one/first occurrence on that
> line.

use re.findall (or in recent versions, re.finditer)

</F>




From marklists at mceahern.com  Fri Jun  7 06:43:02 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 7 Jun 2002 05:43:02 -0500
Subject: Program to an interface, not an implementation
In-Reply-To: <20020607104821.A227@bork.demon.nl>
Message-ID: <NCBBLFCOHHDIKCAFGCFBKEHCLDAA.marklists@mceahern.com>

[Egbert Bouwman]
> My next problem is on nearly the same page (17) of the GoF book:
> class versus interface (== type) inheritance, but in the python context.

As the following shows, inheritance in Python is implementation inheritance.
In this particular case, the subclass inherits a method they MUST override,
but you could easily provide an implementation that actually does something
too.  I'm not sure whether this answers your question, though--I haven't
been following this thread.

class PersistentBase:

    def save(self):
        raise NotImplementedError

class Invoice(PersistentBase):

    pass

i = Invoice()

try:
    i.save()
except NotImplementedError:
    pass
else:
    raise

Cheers,

// mark

-





From whisper at oz.net  Thu Jun 27 15:36:00 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 27 Jun 2002 12:36:00 -0700
Subject: Pythoniac: Thoughts on a hardware Python processor
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F06233015DE661@mail.hebmex.com>
Message-ID: <GCEDKONBLEFPPADDJCOEGEIHDMAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Gustavo Cordova
> Sent: Thursday, June 27, 2002 12:18
> To: William Park; python-list at python.org
> Subject: RE: Pythoniac: Thoughts on a hardware Python processor
>
>
> >
> > Christopher Saunter <Christopher.Saunter at durham.ac.uk> wrote:
> > > Dear All,
> > >
> > > I have seen the occasional post about creating a specific
> > > hardware 'Python Processor', that could then be built and
> > > used in programmable logic or the like (Valves, relays,
> > > core memory - Pythoniac ;-) - in other words, create
> > > a processor that executes Python bytecode directly.  There would be
> > > various different reasons for doing this (the 'neat' factor etc...)
> >
> > The only route that has any chance of success is
> >     Python -> Forth -> stack engine
> > which leverage works that has already gone into Forth area.
> > But, my days of hand assembling 68000 or programming HP calculators
> > are long gone... :-)
> >
>
> Ahh! Thanks for posting this; I was thinking of the same thing,
> about Chuck Moore's forth chips and all that, but I didn't want
> to post anything because I've already posted some Forth comments
> and received no replies.
>
> :-)
>
> -gus
>

Ahh! Haven't you heard? Forth came in Fifth and the Forth chip is dead. In
fact I don't think anyone is still offering a stack based computer (only HP
did AFAIK, the HP 3000).

Dave LeBlanc
Seattle, WA USA





From nookieNO_SP_AM at online.no  Tue Jun  4 12:45:37 2002
From: nookieNO_SP_AM at online.no (Erlend J. Leiknes)
Date: Tue, 04 Jun 2002 16:45:37 GMT
Subject: writing a chat program
References: <d4c18195.0206040840.7ece1b8c@posting.google.com>
Message-ID: <R86L8.9676$fG3.329140@news2.ulv.nextra.no>

Here is a little thread for you my friend:

import thread
class MyClass:
    def __init__(self):
        self._running = 0

    def Start(self):
        if self._running = 0:
            self._running = 1
            thread.start_new_thread(_MyThreadFunction,())

    def Stop(self):
        self._running = 0

    def _MyThreadFunction(self):
        while self._running:
            #put your function here
            pass

MyFunc = MyClass()
MyFunc.start()
while 1:
    import time
    time.sleep(1)


"Shaun Whelden" <shaunw at liberateschool.com> wrote in message
news:d4c18195.0206040840.7ece1b8c at posting.google.com...
> Hi all,
> I'm trying to throw together a simple text-based chat program in
> Python.  Here's the problem I have right now: I need a way to allow
> the user to be entering in a message they want to send, while at the
> same time waiting for any data that may come through the socket.  If
> data does come through the socket, I need the user to be able to
> continue typing while the data is printed to the screen.
>
> I'm writing this on a Windows machine, so curses is out of the
> question.  Also, I want this to be a text based program, so I don't
> want to have to mess around with Tkinter or any other forms of GUI
> unless I have to.  I'm pretty sure that I need to do something
> involving thread.py and/or threading.py, but I've never used a thread
> before in my life.  Can anyone help?
>
> Shaun Whelden
> shaunw at liberateschool.com




From jdhunter at nitace.bsd.uchicago.edu  Sat Jun 22 10:45:39 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Sat, 22 Jun 2002 09:45:39 -0500
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at>
Message-ID: <m21yazxtgs.fsf@mother.paradise.lost>

>>>>> "Siegfried" == Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> writes:

    Siegfried> My conclusion: I will rewrite my simulation in Fortran
    Siegfried> 90 and will abadon Python. Python is maybe good for
    Siegfried> some small scripts but not appropiate for serious
    Siegfried> programming.

Well, python *is* used for serious programming:

  http://www.zope.org
  http://www.boost.org/libs/python/doc/
  http://www.scipy.org
  http://pfdubois.com/numpy
  http://www.wxpython.org
  http://www.pythonware.com/products/pil
  http://sketch.sourceforge.net

You are jumping in to the deep water pretty fast, wrapping external
code with SWIG and F2PY.  When I first started wrapping my C++
classes, I encountered some system freezes as well.

If you post some code, you might get some real help, as well as advice
on how to debug where the freezes are occurring and why.  There is
support for debugging mixed language code.

Don't give up too early!

John Hunter


From i.linkweiler at gmx.de  Fri Jun  7 07:29:21 2002
From: i.linkweiler at gmx.de (Ingo Linkweiler)
Date: Fri, 07 Jun 2002 13:29:21 +0200
Subject: Comparison: HTTP with Java, Perl or Python
Message-ID: <3D009911.9000904@gmx.de>

Hello,

for a comparison of languages I need a complete "translation" of this short skript into Perl, Tcl and Java:

 myurl = "http://www.uni-dortmund.de"
 import urllib
 file  = urllib.urlopen(myurl)
 text  = file.read()
 file.close()
 print text

Can you send me some solutions?
The skripts must be able to run, please do not post "partial" solutions.

Ingo




From cben at techunix.technion.ac.il  Mon Jun 10 07:11:31 2002
From: cben at techunix.technion.ac.il (Beni Cherniavksy)
Date: Mon, 10 Jun 2002 14:11:31 +0300 (IDT)
Subject: Does Python need a '>>>' operator?
In-Reply-To: <EFEFIHAINIINNELLHFAIEEMJCIAA.kseehof@neuralintegrator.com>
Message-ID: <Pine.GSO.4.33.0206101315030.4022-100000@techst02.technion.ac.il>

On 2002-06-10, Ken Seehof wrote:

> Beni Cherniavsky <cben at tx.technion.ac.il> wrote:
> > I just got another idea: use 0x1234 for 0-filled numbers and 1xABCD for
> > 1-filled ones.  That way you impose no restrictions on what follows the
> > prefix and keep backward compatibility.  0xFFFFFFFF stays a 2^n-1
> > _positive_ number, as it should be.  The look of 1x is weird at first but
> > it is very logical...
>
> Hey, that's pretty clever.  I like it.  One objection I can forsee is
> that it does add another little thing to learn.  And yes, one must
> consider such a feature as making python more complex if it requires
> a paragraph to be added to the main documentation, even if the feature
> will only be used by "advanced" programmers.  The standard argument
> is that all python programmers have to be able to read your code.
>
> However, I don't see this as making life much more complicated for
> beginners, since we already have 0x1234, which is only familiar to
> experienced programmers.  People who are not familiar with 0x1234 can
> learn about 1xABCD at the same time.  People who are familiar with
> 0x1234 (e.g. from programming in C/C++), can probably handle looking up
> the proposed syntax in the documentation, if they can't guess what it
> means by experimenting in the interactive interpreter.
>
After some thought I discovered that this is actually very consistent:

~1x0123456789ABCDEF == 0xFEDCBA9876543210

One only need to learn a little to about 2's compleemnt to understand:

-1x0123456789ABCDEF == 0xFEDCBA9876543211 # 1 added

if he wishes to...

> In particular, I like:
>
>   1xF == -1
>
> Seems more pythonic than 0xFFFFFFFF, since the latter implies knowledge
> that we are using 32 bits.  I don't know of any current way to express
> 1 filled binary numbers cleanly.
>
Currently you use the ~ or the - unary operators.  So another idea, in the
spirit of repr() is to print -1 as ~0x0.  When you want bit-manipulation,
~ is more natural than -.  However I personally like the 1x prefix more.

> Unfortunately, I don't see an easy way to clean up this blemish:
>
> >>> 0xffff
> 65535

This is correct behaviour.

> >>> 0xffffffff
> -1

While currently it is like this, it cannot be considered standard.  It
depends on the machine's word size so any code relying on it being equal
to -1 is already broken!

> >>> 0xfffffffff
> 68719476735L
>
Correct too.

> I suppose in a perfect world, 0xffffffff would be 4294967295L, but
> we'd have serious compatibility issues with that (note that one
> would use 1xf instead of 0xffffffff to represent -1 in the perfect
> world).
>
> BTW, note that 0xffffffff == 4294967295L would be consistent with
> C/C++ with an unsigned int (32 or more bits).  The idea is that
> all numbers 0x... are non-negative, while number 1x... are negative.
>
Sure.  That's the idea.  To make the 32bit boundary invisible one must
ensure that a given written number represents the same mathematical
integer, on any machine.

> I'm quite certain that any proposed solution (such as issuing a
> warning for 0xXXXXXXXX) will receive flames, so I think I will
> stop here, hoping it's not too late :-)
>
I would propose the 1x for a read syntax.  ~ and - are already accepted
automagically (since python's only read syntax is the eval syntax).  The
only trouble now are non-long constants with msb set.  They should be
warned anyway to help people detect word-size sensitivity bugs...  Maybe a
__future__ should be provided is serious enough compatibility problems
arise.  This point itself is probably not worth a __future__ as the new
semantics of 0x... always positive is compatible with any bug free-code ;)

I do not propose to replace hex() now, one can provide a newhex() [better
name desperately needed, maybe hex1?].  In the long run, it might be
reasonable to support the different combinations in format strings:

We already have (or should have; in C and python1.5 - sadly there is no
up-to-date installation around - %x treats the data as always unsigned;
python1.5 refuses to %x longs but accepts them in hex() - I presume that's
been already fixed):

"%x"  % +2  =>   2
"%x"  % -2  =>  -2
"%+x" % +2  =>  +2
"%+x" % -2  =>  -2
"%04x" % +2 =>   0002
"%04x" % -2 =>  -0002
"%0#4x" %+2 =>   0x02
"%0#4x" %-2 =>  -0x02  # Is this so indeed?

Or maybe it's still always unsigned in '%x' to mirror C?  I son't think
that can be kept because the signed->unsigned coercion exposes the word
length; do (x & 0xffff) when you want to coerce, so that your results will
be predictable...

Proposed:

"%~x" % +2  =>   2
"%~x" % -2  =>  ~1
"%#~x" % +2 =>   0x2
"%#~x" % -2 =>   1xE
# Like the proposal by I-don't-remember-who (sorry) but with a determined
# msb!  No guessing!
"%0~4x" % +2 =>  0x02
"%0~4x" % -2 => ~0x01
"%0#~4x" %+2 =>  0x02
"%0#~4x" %-2 =>  1xFE

Now this is probably overkill.  But one must introduce a new flag symbol,
# is already used for adding the 0x, it can't be 1 (would mean a width)
and it would be logical to do 1xE only when # is also specified, which
brought me to all these combinations.

But the two ideas of ~0x1 and 1xE both serve exactly the same purpose so
only one sould suffice IMHO.  ~0x1 is easy to constuct from the existing
facilities, 1xE needs playing around with the output's first char...

> - Ken Seehof
>

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>





From roelsch at comtrak.com  Mon Jun  3 18:31:32 2002
From: roelsch at comtrak.com (Robert L. Oelschlaeger)
Date: Mon, 3 Jun 2002 17:31:32 -0500
Subject: Version strings in executable created by Py2exe
In-Reply-To: <A7B5C22F20EE3C4E95744E2F03F1887F569B@MAYOR.retzlaff.com>
Message-ID: <000001c20b4e$69e42830$e5091bac@ELPASO>

> -----Original Message-----
> From: Jimmy Retzlaff [mailto:jimmy at retzlaff.com]
> Sent: Monday, June 03, 2002 3:09 PM
> To: python-list at python.org
> Cc: Robert L. Oelschlaeger
> Subject: RE: Version strings in executable created by Py2exe
> 
> 
> Robert L. Oelschlaeger [mailto:roelsch at comtrak.com] wrote:
> > Running with Python 2.2. I've created myAppName.exe using 
> Py2exe, with
> > the "version_companyname" and "version_legalcopyright" specified in
> > the setup .cfg file as follows:
> ...
> > Why are the version_companyname and version_legalcopyright 
> strings not
> > installed for the "Company Name" and "Copyright"?
> 
> "version" should be followed by a dash rather than an underscore. So,
> for example, you should use "version-companyname" rather than
> "version_companyname".
> 
> Jimmy

I tried the name change, but it didn't help: still no 'Company Name' or
'Copyright' strings in the resulting .exe file.

I think there is a documentation problem:

    the example at http://starship.python.net/crew/theller/py2exe/ uses
    hypenated names, but

    the names in py2exe/build_exe.py are the underscored names (e.g.,
    self.version_companyname).

It appears to me that the strings in py2exe/build_exe.py that can come
from either self.<stringname> or md.<stringname>, as in (line 825)

    self.version_filedescription or md.description or ""

are working; strings that are derived from self.<stringname> are empty.
I don't see how the self.<stringname> variables are supposed to get set.

Bob
-------------- next part --------------
A non-text attachment was scrubbed...
Name: winmail.dat
Type: application/ms-tnef
Size: 2368 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020603/1bd69019/attachment.bin>

From tdelaney at avaya.com  Tue Jun 11 20:49:59 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 12 Jun 2002 10:49:59 +1000
Subject: Sorting list (maybe stupid)
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A37A@natasha.auslabs.avaya.com>

> From: Ken Seehof [mailto:kseehof at neuralintegrator.com]
> 
> >>> a = [1,1.0,3.0,3,2.0,2,0,0.0]
> >>> a
> [1, 1.0, 3.0, 3, 2.0, 2, 0, 0.0]
> >>> a.sort()
> >>> a
> [0, 0.0, 1, 1.0, 2.0, 2, 3.0, 3]
> 
> So yes, sort() is stable.

That's a rather small set of data to base that on ...

> I suppose it's theoretically possible for a different
> python implementation to use a different sort algorithm,
> but practically speaking, that is very unlikely.

In fact, the current implementation uses an optimised quicksort IIRC - not
at all stable.

Personally, I think the implementation should be documented to use a stable
sort - presumably mergesort (stable, minimal number of comparisons which is
very important in Python).

Tim Delaney




From gcordova at hebmex.com  Mon Jun  3 13:53:54 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Mon, 3 Jun 2002 12:53:54 -0500 
Subject: encoding and decoding repr'd version of strings
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233010A542B@mail.hebmex.com>

> 
> Hi 
> is there an codec that bundles with python so you can 
> transform strings into the repr'd (escaped) version and
> back? If not, how easy would it be to code? I don't want
> to use eval(...) because I don't trust the source of the 
> string and I might mistakenly eval an expression like "' 
> '*(2**30)" which would allocate about 1 GB of data (eek!)
> 
> Thanks,
> Dan
> 

I might be going out on a limb here, but how about using pickle
or cpickle?

-gustavo




From fredrik at pythonware.com  Thu Jun 13 06:37:57 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 13 Jun 2002 10:37:57 GMT
Subject: PIL _imaging
References: <Xns922AA829A8C5FHotMailcs@130.133.1.4> <3D06E694.4040103@mxm.dk> <Xns922B3150A6B32HotMailcs@130.133.1.4> <3D074529.E848AA6@ipm.fhg.de> <Xns922B561EDB212HotMailcs@130.133.1.4> <3D07BCC1.2000002@mxm.dk> <Xns922BB340B788BHotMailcs@130.133.1.4>
Message-ID: <9C_N8.11930$p56.3660854@newsb.telia.net>

"Cousin Stanley" wrote:

> I only have a single installation of Python :
> ... < K:\Python >
> ... Wed 06-12-2002 16:41:48.86 > python
> ... Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
> ... Type "help", "copyright", "credits" or "license" for more information.
> ... >>>

and you are 100% sure that you're using a version
of PIL built for Python 2.2?

try importing the _imaging module directly, and see
what message you get:

    >>> import _imaging

</F>




From lyle at users.sourceforge.net  Mon Jun 24 13:22:22 2002
From: lyle at users.sourceforge.net (Lyle Johnson)
Date: Mon, 24 Jun 2002 12:22:22 -0500
Subject: File Copying
References: <mailman.1024935031.24788.python-list@python.org>
Message-ID: <3D17554E.9070903@users.sourceforge.net>

Rajat Chopra wrote:
> How do you copy a file from one directory to another. I have looked all over and haven't found any documentation on it. Essentially, I want to copy a file from $SRC_DIRECTORY to $TARGET_DIRECTORY.
> 
> Does anyone know which modules are used? Or does anyone have some sample scripts?

The shutil module should do the trick:

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

Hope this helps,

Lyle



From fredrik at pythonware.com  Thu Jun 13 06:24:18 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 13 Jun 2002 10:24:18 GMT
Subject: Something faster then sgmllib for sucking out URLs
References: <mailman.1023913936.26808.python-list@python.org> <m34rg71xa3.fsf@mira.informatik.hu-berlin.de>
Message-ID: <mp_N8.11926$p56.3660561@newsb.telia.net>

Martin v. Loewis wrote:
>
> > I'm working on a webspider to fit my sick needs. The profiler
> > tells me that about 95% of the time is spent in sgmllib. I use sgmllib
> > solely for extracting URLs. I'm looking for a faster way of doing
> > this. Regular expressions, string searches? What's the way to go? I'm
> > not a python purist. Calling some fast C program with the html as
> > argument and getting back a list of URLs would be fine by me.
>
> I recommend to use sgmlop, which is distributed both as part of PyXML,
> and separately by Fredrik Lundh. It is the fastest SGML/XML parser I
> know of, for use within Python.

the latest version (1.1a3) is available here:

    http://effbot.org/downloads/

here's a code snippet that extracts A HREF anchors
from a webpage:

import sgmlop
import urllib

class AnchorHandler:
    def __init__(self):
        self.anchors = []
    def finish_starttag(self,tag,attrs):
        if tag == "a":
            for k, v in attrs:
                if k == "href":
                    self.anchors.append(v)

def getanchors(page):
    handler = AnchorHandler()
    parser = sgmlop.SGMLParser()
    parser.register(handler)
    parser.feed(urllib.urlopen(page).read())
    parser.close() # we're done
    return handler.anchors

print getanchors("http://www.python.org")

</F>




From emile at fenx.com  Thu Jun 27 10:35:53 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 27 Jun 2002 14:35:53 GMT
Subject: How to get rid the new line
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org> <3D1AD534.A74EC9F4@ipm.fhg.de>
Message-ID: <dpFS8.326047$352.36272@sccrnsc02>

"Markus von Ehr" <markus.vonehr at ipm.fhg.de> wrote in message
news:3D1AD534.A74EC9F4 at ipm.fhg.de...
>         f = open(filename, 'r')
>         lines = f.readlines()
>         line1 = lines[0] # exemplarily for first line
>         line1 = line1[0:len(line1)-1]
>
> or:
>         f = open(filename, 'r')
>         lines = f.readlines()
>         line1 = lines[0][0:len(line1)-1] # exemplarily for first line
>

Easier to strip the last character simply using [:-1]:

>>> s = 'this is a testX'
>>> s[:-1]

--

Emile van Sebille
emile at fenx.com

---------



From Adndrew at novaresponse.com.au  Sun Jun  9 11:22:50 2002
From: Adndrew at novaresponse.com.au (Adndrew at novaresponse.com.au)
Date: Sunday, 09 Jun 2002 09:22:50 -0600
Subject: Urgent News Flash
Message-ID: <09060209.2250@novaresponse.com.au>

URGENT NEWSFLASH

The government has installed BLACK BOXES into ISPs to watch your
Internet surfing.

When you are targeted, the first you will know is the KNOCK AT YOUR
DOOR and the JAIL SENTENCE.

If you think you have cleaned your computer - you are WRONG.

They can recover from your PC every single picture or item you have
ever watched on the Internet.

You need protection. You need an Evidence Eliminator. 

Click Here.

http://www.geocities.com/teste99ee/clean.html




From pinard at iro.umontreal.ca  Tue Jun  4 08:39:29 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 04 Jun 2002 08:39:29 -0400
Subject: Pop a list from beginning ? and memory saving...
In-Reply-To: <adhrsl$118kvn$1@ID-146704.news.dfncis.de>
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de>
	<3CFA0676.68F80DE6@engcorp.com>
	<add289$vud42$1@ID-146704.news.dfncis.de>
	<Xns9221CE2AF76D0cliechtigmxnet@62.2.16.82>
	<adhrsl$118kvn$1@ID-146704.news.dfncis.de>
Message-ID: <oqu1ojw73y.fsf@titan.progiciels-bpi.ca>

[Shagshag13]

>  def reset(self):
>   del self._v
>   self._v = []

Just for the record, one could also write, in one line:

    self._v[:] = []

Probably unneeded in this case, but it is useful when one wants to keep
extra references to the list, and empty it nevertheless.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From dfackrell at DELETETHIS.linuxmail.org  Fri Jun  7 10:14:13 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Fri, 7 Jun 2002 08:14:13 -0600
Subject: Detecting OS and Window Manager
References: <mailman.1023304388.9659.python-list@python.org> <ufvsco8mlhf213@corp.supernews.com>
Message-ID: <3d00bfb5$1_2@hpb10302.boi.hp.com>

> Well, in that case we just choose a default.  The thing is, we want to
> detect if PyKDE is installed.  If it is, then we can do "import kdecore"
> without risk of it throwing an exception.  If it isn't, I'd rather not
> have to try something along the lines of attempting "import kdecore" and
> then having to catch an exception if it's not found.

Correct me if I'm wrong, but isn't this exactly the pythonic way of handling
dynamic loading of modules?

try:
    import kdecore
except ImportError:
    # do something else

Of course, this will probably not do what you want in this case, as kdecore
will probably load if it is installed, without regard to whether the current
user is running KDE.  I don't have much experience with this particular
topic yet.

Daniel Fackrell




From mike at radiomag.com  Wed Jun 19 21:38:32 2002
From: mike at radiomag.com (Michael Steuer)
Date: Thu, 20 Jun 2002 11:38:32 +1000
Subject: Listing all available (not just loaded) modules
References: <aem9om$888fb$1@ID-140151.news.dfncis.de>
Message-ID: <aerc34$98g48$1@ID-140151.news.dfncis.de>

Thanks to all who replied to me post.

I know I didn't mention this in my original post, but I needed this information
to find out which modules are installed on my domains webserver, which means I
couldn't just use one of the IDEs (eg. IDLE)functions to look at the module
tree.

In the .cgi mentioned in an earlier reply I found all I needed.

Cheers,

Mik.e




From SteveC at nospam.innocent.com  Thu Jun 20 15:12:05 2002
From: SteveC at nospam.innocent.com (Steven M. Castellotti)
Date: Thu, 20 Jun 2002 19:12:05 GMT
Subject: Dispatching a Win32 COM in the background?
References: <mailman.1024531044.10120.python-list@python.org>
Message-ID: <9OpQ8.90552$ks6.8015@nwrddc02.gnilink.net>

Thanks a lot guys, your suggestions have worked perfectly!

	Just for the record, here's what the full implementation looks like.

This code allows text to be piped to Microsoft's Text-To-Speech
sythesizer, spoken in the background, in a thread-safe manner:

#####################################################################

import threading

class Background_TTS(threading.Thread):

   def __init__(self, debug_mode):
      threading.Thread.__init__(self)

      self.debug_mode = debug_mode
      self.loop = 1

      import Queue
      self.text_queue = Queue.Queue()


   #####################################################################

   def Speak(self, text):
      self.text_queue.put(text)		

		
   #####################################################################

   def stop(self):

      self.loop = 0
		

   #####################################################################
		
   def run(self):
      from win32com.client import constants
      import win32com.client
      import time

      speech = win32com.client.Dispatch("SAPI.SpVoice")

      while self.loop:

      if not(self.text_queue.empty()):
         text = self.text_queue.get()				
         speech.Speak(text)
         time.sleep(.001)

#####################################################################

initialization looks like this:

try:
   self.speech = Background_TTS(self.gameInformation.debug_mode)
   self.speech.start()
except:
   if (self.gameInformation.debug_mode):
   print "Error Initializing Microsoft Speech API."
else:
   self.text_to_speech_enabled = 1

#####################################################################

calling the synthesizer:

self.speech.Speak(text)

#####################################################################

and finally, stopping the process:

self.speech.stop()



	Once again guys, thanks!

-Steve Castellotti
 SteveC (at) innocent.com
 http://cogengine.sourceforge.net/



On Wed, 19 Jun 2002 18:56:05 -0500, David LeBlanc wrote:

> Hmmm... and the run method of Background could be sitting on one end of
> a Queue (Queue is thread-safe: see your python doc for details) that's
> buffering the asyncronous calls on the speacher. The main task then just
> stuffs a talk request into the queue and the Background processes it
> when it gets to it.
> 
> David LeBlanc
> Seattle, WA USA
> 
>> -----Original Message-----
>> From: python-list-admin at python.org
>> [mailto:python-list-admin at python.org]On Behalf Of Peter Hansen Sent:
>> Wednesday, June 19, 2002 16:47
>> To: python-list at python.org
>> Subject: Re: Dispatching a Win32 COM in the background?
>>
>>
>> "Steven M. Castellotti" wrote:
>> >
>> > 2) Make the call inside a thread which produces the same effect.
>>
>> This might work (I don't know anything about COM though).  Try putting
>> your own code inside the run method of a thread:
>>
>> import threading
>> class Background(threading.Thread):
>>     def __init__(self):
>>         threading.Thread.__init__(self)
>>
>>     def run(self):
>>         print 'Starting'
>>         pass   # your own code goes here
>>         print 'Done'
>>
>> >>> b = Background()
>> >>> b.start()
>> Starting
>> <possibly a pause here?>
>> Done
>>
>> If at this point you get a >>> prompt back right away, but the speech
>> generation is carrying on "in the background" then you're all set. Make
>> sure you do this with code you've already tested outside of the thread,
>> so you'll know that any failure is not the fault of your own code...
>>
>> -Peter
>> --
>> http://mail.python.org/mailman/listinfo/python-list


From chouster at uclink4.berkeley.edu  Mon Jun 24 20:18:41 2002
From: chouster at uclink4.berkeley.edu (Chris)
Date: 24 Jun 2002 17:18:41 -0700
Subject: HTML Error from Python CGI Script
Message-ID: <79b54d28.0206241618.6c78dbde@posting.google.com>

Ok, I'm writing a CGI script and its supposed to generate a temporary
HTML document. By temporary I mean (and assume) that a unique filename
is generated whereby I can refer to that filename, write to it, save,
etc. and will not delete unless I explicity call to delete it. I use
the tempfile module from python. My problem is that the temporary HTML
document cannot be interpreted by IE when I click on a hyperlink that
points to this temp HTML document. Here is the page that is generated:

<HTML>
<!-- Created by /someCGIscript -->
<HEAD>
<TITLE>Submission Request from you</TITLE>
</HEAD>
<BODY BGCOLOR="#F0F0F0"   >
<H2>Rearray Information</H2>
<br>User: me at me.com<br>
Time at Request: Mon Jun 24 17:10:51 2002<br>
<FORM METHOD="POST" ENCTYPE="application/x-www-form-urlencoded"   >
<INPUT TYPE="hidden" NAME="project" value="1">
<INPUT TYPE="hidden" NAME="emailAddress" value="me at me.com">
<INPUT TYPE="hidden" NAME="destplates" value="343">
<INPUT TYPE="hidden" NAME="cloneList" value="394023_3523_32423">
</FORM>
</BODY>
</HTML>

I copy this exact text to my own computer (versus operating over
network) and open it and it looks fine. Exactly what I need. However,
when I put a link to it on another webpage on the network, an internal
server error occurs and complains about the following:

syntax error at line 2: `newline or ;' unexpected

What does this mean??? I've generated hundreds of webpages in my day
and never have I come across something seemingly so simple yet so
elusive. Help me please!
THanks


Chris


From theller at python.net  Wed Jun 19 17:38:06 2002
From: theller at python.net (Thomas Heller)
Date: Wed, 19 Jun 2002 23:38:06 +0200
Subject: How to Store Binary data in a *.py module
References: <3D10E0E9.906DF517@noaa.gov>
Message-ID: <aeqto2$8vpjc$1@ID-59885.news.dfncis.de>

"Chris Barker" <Chris.Barker at noaa.gov> wrote in message news:3D10E0E9.906DF517 at noaa.gov...
> I'd like to be able to store binary data in a python module file.
>

I've done it in distutils.bdist_wininst.py, take a look.

Thomas




From hst at empolis.co.uk  Thu Jun  6 08:57:11 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Thu, 6 Jun 2002 13:57:11 +0100
Subject: check on numerical value?
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220BEB@hendrix.empolisuk.com>

Henk wrote:
> 
> Is there any way to discover when a value is numerical or not in
> Python? I want a check in my code on a given value. If somebody enters
> or gives me something like "akjlvq?irq" this is surely not a numerical
> value.
> 
> Thanks,
> 
> Henk
> -- 

>>> while 1:
...     try:
...         i = int(raw_input('Please enter an integer '))
...         break
...     except ValueError: pass
...
Please enter an integer r
Please enter an integer 78
>>>

Use float instead of int if appropriate

Harvey

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.




From xcenovsk at fi.muni.cz  Sun Jun 23 18:17:59 2002
From: xcenovsk at fi.muni.cz (Lukas Cenovsky)
Date: Sun, 23 Jun 2002 22:17:59 GMT
Subject: Python & Litestep
Message-ID: <3D164917.3000609@fi.muni.cz>

Here is the function how send to Litestep a Bang Command
from Python script. Maybe someone will find it helpful.

def SendBang(bang, arg = ''):
   import win32gui, array, struct, win32api
   try:
     LMBANGCOMMAND = "II64s256s"
     copyDataStruct = "IIP"
     a = struct.pack(LMBANGCOMMAND, struct.calcsize(LMBANGCOMMAND), 0, bang, arg)
     ar = array.array('c', a)
     msg = ar.buffer_info()
     cds = struct.pack(copyDataStruct, 9420, msg[1], msg[0])
     cdsa = array.array('c', cds)
     cdsb = cdsa.buffer_info()
     ls = win32gui.FindWindow("TApplication", "Litestep");
     if (win32api.SendMessage(ls, 0x004A, 0, cdsb[0]) <> 1):
       return 0
     else:
       return 1
   except:
     return 0

Use it like:
SendBang('!alert', '"I am an alert from Python."')
SendBang('!Amp_PlayPause')

-- 
-- Cendicek



From marklists at mceahern.com  Fri Jun 28 12:13:58 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 28 Jun 2002 11:13:58 -0500
Subject: aspect-oriented demo using metaclasses
In-Reply-To: <NCBBLFCOHHDIKCAFGCFBIEPMLGAA.marklists@mceahern.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBKEPPLGAA.marklists@mceahern.com>

I wrote:

> What is a good way to write the rules for enrolling
> in notifications?  (And what's the word for this that AspectJ uses?)

The phrase I was look for was join points.

// mark

-





From mwh at python.net  Thu Jun 13 12:29:08 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 13 Jun 2002 16:29:08 GMT
Subject: filter2
References: <mailman.1023975618.6474.python-list@python.org> <lkptyvmdfz.fsf@pc150.maths.bris.ac.uk> <mailman.1023984463.10662.python-list@python.org>
Message-ID: <lkhek7m9c8.fsf@pc150.maths.bris.ac.uk>

Jeff Epler <jepler at unpythonic.net> writes:

> On Thu, Jun 13, 2002 at 03:00:28PM +0000, Michael Hudson wrote:
> > Write it in Python.  Particularly if test is expensive the loop
> > overhead will disappear into the noise.
> > 
> > (with-seriousness-level :low
> >   If you insist on C loops, there's always:
> > 
> >       lt = map(lambda x:(test(x), x), alist)
> >       l1 = filter(lambda (x,y):x, lt)
> >       l2 = filter(lambda (x,y):not x, lt)
> > 
> >   But I doubt you need me to tell you why this solution sucks.)
> 
> If you're benchmarking, try a solution like this in the mix.  It only
> has one map/filter call, not map+2*filter. (uses nested scopes and the
> new bool() builtin, but this can be avoided if necessary)
> 
> def filter2(test, list):
>     l1 = []
>     l2 = []
>     funcs = [l2.append, l1.append]
>     map(lambda x: funcs[bool(test(x))](x), list)
>     return l1, l2
> 
> >>> print filter2(lambda x: x%3==0, range(16))
> ([0, 3, 6, 9, 12, 15], [1, 2, 4, 5, 7, 8, 10, 11, 13, 14])

Hey, that's pretty neat!  Still three function calls per element,
though.  Well, four including bool, but that could be gotten rid of by
gross hacks (e.g. not not test(x)).

Cheers,
M.

-- 
  Worryingly, DEFUN appears to be a function that removes all the
  fun from something: after using it all your code is converted 
  to C++.                              -- Tim Bradshaw, comp.lang.lisp


From rnd at onego.ru  Tue Jun 25 05:58:14 2002
From: rnd at onego.ru (Roman Suzi)
Date: Tue, 25 Jun 2002 13:58:14 +0400 (MSD)
Subject: Python daemon
In-Reply-To: <3d182dd6$1@news.mt.net.mk>
Message-ID: <Pine.LNX.4.44.0206251355510.22712-100000@rnd.onego.ru>

On Tue, 25 Jun 2002, [iso-8859-5] ?????? ?. wrote:

>Hello, 
>
>I'm writing a Python daemon on Unix and I need the folowing features:
> 1. going to background, diassociate from controling tty (done)
> 2. watchdog, the program should never be stuck

- just have another process which will look after the second one.
This is how many critical daemons works. Most easily it 
can be done by forking.

> 3. restarting of the program if it fails, or the watchdog doesn't trigger

Run the program from the /etc/inittab - if it terminates, 
init will restart it for you.

> 4. Logging
>
>The issue 1. I solved, it's on ActiveState's site. Issue 4. is not that
>hard to do I guess. Any suggestions about 2 and 3 ?

Sincerely yours, Roman Suzi
-- 
rnd at onego.ru =\= My AI powered by Linux RedHat 7.2





From bokr at oz.net  Wed Jun 26 20:24:06 2002
From: bokr at oz.net (Bengt Richter)
Date: 27 Jun 2002 00:24:06 GMT
Subject: Python 2.2 __slots__ Bug?
References: <mailman.1025089474.17780.python-list@python.org> <B93F7990.CECB%jonathan@onegoodidea.com> <mailman.1025105074.31122.python-list@python.org>
Message-ID: <afdlv6$bu6$0@216.39.172.122>

On Wed, 26 Jun 2002 10:23:45 -0500 (CDT), Glyph Lefkowitz <glyph at twistedmatrix.com> wrote:

>----Security_Multipart(Wed_Jun_26_10:23:45_2002_363)--
>Content-Type: Text/Plain; charset=us-ascii
>Content-Transfer-Encoding: 7bit
>
>From: Jonathan Hogg <jonathan at onegoodidea.com>
>Subject: Re: Python 2.2 __slots__ Bug?
>Date: Wed, 26 Jun 2002 14:00:48 +0100
>
>> On 26/6/2002 12:03, in article
>> mailman.1025089474.17780.python-list at python.org, "Glyph Lefkowitz"
>> <glyph at twistedmatrix.com> wrote:
>> 
>> > I get "I'm fine" and then a segmentation fault.  This only happens when
>> > __slots__ is defined on a new-style class.  Seriously, i'm not making those
>> > numbers up -- reliably, 43551 objects will crash the interpreter; fewer will
>> > not.
>> 
>> Interestingly I get the construction working fine, but the 'del' crashes the
>> interpreter with (on Mac OS X):
>
>Hmm.  Sprinkling print statements about shows it dying on the 'del' (this is
>how I encountered the bug originally, I was trying to time the [de]allocation
>of large numbers of objects under 2.1 vs. 2.2), but the stacktrace is still the
>same; is there some difference between the way allocation works on OS X
>vs. Linux? (I assume if you can run OS X we're using the same hardware.)
>
>> This seems to fail for me at 5430 objects. I couldn't get the allocation to
>> crash at all (or at least up to a million objects, after which I got bored).
>
>Yep, same here.  Allocation is no problem.
>
Wondering if there's any 12*n byte chunks involved, fitting
into some 2**n (-overhead) allocation space? Just noticing

 >>> divmod(2**19, 43551)
 (12, 1676)
 >>> divmod(2**16, 5430)
 (12, 376)

Sometimes these things mean something, sometimes they don't ;-)
Maybe it will jog someone's thought. E.g., a last-item index/limit
problem?

Regards,
Bengt Richter


From deltapigz at telocity.com  Sat Jun 29 21:47:32 2002
From: deltapigz at telocity.com (Adonis)
Date: Sat, 29 Jun 2002 21:47:32 -0400
Subject: possibly a dumb question
References: <3d1e5eff$1_7@nopics.sjc> <3D1E6415.8128F3F8@engcorp.com> <3d1e6251$1_3@nopics.sjc>
Message-ID: <3d1e6325$1_1@nopics.sjc>

an addition:
why not use a function blah blah, yes, but this is going to be a modularized
project, essentially i want to load some user made modules which will be
classes and will be loaded into python dynamically. like a memory resident
cgi server skipping the use of pipes.

Adonis

"Adonis" <deltapigz at telocity.com> wrote in message
news:3d1e6251$1_3 at nopics.sjc...
> sure:
>
> class foo:
>     def __init__(self, value):
>         return value
> x = foo(0)
> print x ;yeilds 0
>
> i know the code provided is wrong, but its the general idea.
>
> Adonis
>
>
> "Peter Hansen" <peter at engcorp.com> wrote in message
> news:3D1E6415.8128F3F8 at engcorp.com...
> > Adonis wrote:
> > >
> > > can a class return a value? i have treid to google this, but have
found
> > > myself short of an answer, i know i could place it in a member and
> retrieve
> > > it or through a function, but im trying to simplify my life.. just a
> > > little..
> >
> > You probably don't actually mean a class.  I think you mean an object,
> > which is an instance of a class.  (If I've got that wrong, I think
> > we need clarification of the question.)
> >
> > If that's the case, there's still one more problem left. :-)
> > "Returning a value" is a concept related to calling a method
> > or function.  It doesn't have any particular meaning if you
> > ask "can an object return a value?" unless you mean something
> > like "can a method of an object return a value?".  That is
> > certainly possible, although I'm not this is what you wanted,
> > either:
> >
> > >>> class MyClass:
> > ...    def myMethod(self):
> > ...        return 'spam'
> > ...
> > >>> c = MyClass()
> > >>> print 'Green eggs and %s.' % c.myMethod()
> > Green eggs and spam.
> >
> > Can you ask the question in a different way, or show an example
> > of what you were hoping to do?
> >
> > -Peter
>
>




From max at alcyone.com  Thu Jun 27 23:09:04 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 27 Jun 2002 20:09:04 -0700
Subject: Plugins in Python
References: <afgeol$sgd$1@mulga.cs.mu.OZ.AU>
Message-ID: <3D1BD350.8F0C74FF@alcyone.com>

Lars Yencken wrote:

> I'm trying to do plugins for a program I'm writing in Python, and I'm
> not
> sure how exactly to do it. Basically what I want it to be able to
> write a
> module which loads every python file in a directory and stores an
> instance
> of each one that loaded successfully.

The most obvious implementation is simply to have them each imported as
modules with __import__, and define some kind of simple
initialization/shutdown API that will be called upon import.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ See the son in your bad day / Smell the flowers in the valley
\__/ Chante Moore
    Bosskey.net: Aliens vs. Predator 2 / http://www.bosskey.net/avp2/
 A personal guide to Aliens vs. Predator 2.


From dougz at cs.washington.edu  Mon Jun 24 22:04:45 2002
From: dougz at cs.washington.edu (Douglas Zongker)
Date: Tue, 25 Jun 2002 02:04:45 GMT
Subject: adding attributes to builtin functions
References: <Gy8Fr6.J7D@beaver.cs.washington.edu> <3D17CE49.E3D16ABE@engcorp.com>
Message-ID: <Gy8oFx.qEw@beaver.cs.washington.edu>

Peter Hansen <peter at engcorp.com> wrote:
: Douglas Zongker wrote:
:>
:>    <built-in function myfunction>
:>    Traceback (most recent call last):
:>      File "<stdin>", line 1, in ?
:>    TypeError: 'builtin_function_or_method' object has only read-only
:>      attributes (assign to .special)
:> 
:> Is there a way to define a builtin function whose attributes aren't
:> read-only?
:
: Maybe something like this?
:
:    [wrapper around the builtin]

Well, that works, but what I *really* wanted to do was to set the
attributes from within C.  I tried something like this:

    void initmymodule( void )
    {
        PyObject *m, *f;
    
        m = Py_InitModule( "mymodule", MyModuleMethods );

	// grab the function object
        f = PyObject_GetAttrString( m, "myfunction" );

	// set one of its attributes (to None)
        PyObject_SetAttrString( f, "special", Py_None );

	Py_DECREF( f );
    }

but it gives the same exception as doing it in Python ("builtin
function has read-only attributes").  I was hoping if someone could
tell me how to do it in Python I could translate that back into C...

thanks,
dz



From wesc at deirdre.org  Wed Jun 12 03:40:03 2002
From: wesc at deirdre.org (wesc at deirdre.org)
Date: Wed, 12 Jun 2002 00:40:03 -0700 (PDT)
Subject: ANN: BayPIGgies mailing list *and* meeting
Message-ID: <200206120740.AAA04627@alpha.ece.ucsb.edu>

in addition to this Wednesday's meeting (see below), the
mailing list for BayPIGgies is back up at baypiggies.net
please sign up here:

    http://baypiggies.net/mailman/listinfo/baypiggies

if you had given me your e-mail address or business card
in the past indicating your interest, you have automagically
been added.  if, of course, you *don't* want to be on the
list, you can also unsubscribe at that same page.  you can
click on the "View Subscriber List" to see if you have been
added yet.

here is the main user group website with directions to the
meetings and more info:

    http://deirdre.org/baypiggies

thanks, and see some of you Wednesday night @ Stanford!

-wesley


> From: Wesley J. Chun (wesc at deirdre.org)
> Subject: BayPIGgies meeting Wed 6/12 7:30pm 
> Date: 2002-06-10 16:19:35 PST 
> 
> What:	Python for (Perl) Programmers
> Who:	Silicon Valley-San Francisco Bay Area Python Users Group (BayPIGgies)
> When:	Wednesday, June 12, 2002
> Where:	Stanford University, Palo Alto, CA
> Time:	7:30pm - 9pm
> 
> This is a talk aimed at introducing Python to programmers
> familar with Perl and other languages, and serves as a warm-up
> to the full tutorial that will be presented at the O'Reilly
> Open Source Convention in San Diego near the end of July.
> Although the non-Python examples use Perl, it's aimed at
> experienced programmers of all sorts. 
> 
> Our guest speaker is Aahz, long-time active member of the
> Python community.  He has been kicking around the computer
> industry for more than two decades, doing tech support,
> programming, consulting, tech writing, and training. Aahz
> recently signed a book contract for an intermediate-level
> Python book, which will be published in early 2003. 
> 
> More information + directions:  http://deirdre.org/baypiggies/

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

"Core Python Programming", Prentice Hall PTR, ? 2001
    http://starship.python.net/crew/wesc/cpp/

wesley.j.chun :: wesc at deirdre.org
cyberweb.consulting : henderson, nv : cyberweb at rocketmail.com
http://roadkill.com/~wesc/cyberweb/




From gerhard.haering at gmx.de  Fri Jun 28 15:35:23 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 28 Jun 2002 21:35:23 +0200
Subject: C API for Python's RNG?
In-Reply-To: <3D1CBBED.58EAC61E@astro.cornell.edu>
References: <3D1CBB3A.E1C43D8A@astro.cornell.edu> <3D1CBBED.58EAC61E@astro.cornell.edu>
Message-ID: <20020628193523.GB688@lilith.my-fqdn.de>

* Tom Loredo <loredo at astro.cornell.edu> [2002-06-28 15:41 -0400]:
> Tom Loredo wrote:
> > 
> > I'll dig in the source to try to puzzle
> > it out, but I thought I should ask publicly:  
> 
> Oops!  I should have dug first---the standard lib RNG is pure
> Python.  So I guess it's Numeric or some third-party RNG if
> I want to do this quickly.  Any suggestions?

<clueless>
What about reading from /dev/random?
</clueless>

Isn't rand() even in the C standard library? Ok, the standard
implementation is normally pretty bad, or so I hear.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From christophe.delord at free.fr  Fri Jun 14 12:39:45 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Fri, 14 Jun 2002 18:39:45 +0200
Subject: I feel stoopid ... this code is to verbose
References: <3D0A1791.3060307@mxm.dk>
Message-ID: <20020614183945.17cc73f7.christophe.delord@free.fr>

If you want to split a number you can try this :

def nbSlicer(nb, chunkSize=3):
	if nb==0: return "0"
	l = []
	b = 10**chunkSize
	fmt = "%%0%dd"%chunkSize
	while nb:
		nb, r = divmod(nb, b)
		l.insert(0,(nb and fmt or "%d")%r)
	return '.'.join(l)

Or this if you prefer to split a string :

def stringSlicer(st, chunkSize=3):
	l = []
	while st:
		l.insert(0,st[-3:])
		st = st[:-3]
	return '.'.join(l)

print nbSlicer(42000000)
print stringSlicer('42000000')


On Fri, 14 Jun 2002 18:19:29 +0200
Max M <maxm at mxm.dk> wrote:

> Hmm ... I am working on a problem. In Danish we have a number format 
> that looks like:
> 
> 42.000.000,00 which is 42 millions
> 
> So I need to insert dot's at every three character from end of the 
> string for the integer value of the number.
> 
> Let's discard decimal points and just focus on the meat. I have written 
> a funcion which works::
> 
> def stringSlicer(string, chunkSize=3):
>      chunkList = []
>      reverseString = list(string)
>      reverseString.reverse()
>      for i in range(0, len(string), chunkSize):
>          chunk = reverseString[i:i+chunkSize]
>          chunk.reverse()
>          chunk = ''.join(chunk)
>          chunkList.append(chunk)
>      chunkList.reverse()
>      return '.'.join(chunkList)
> 
> 
> print stringSlicer('42000000')
> 
>  >>> 40.000.000
> 
> I just find that it's a lot of code for such a little function an it 
> annoys my sense of aestetics. I have tried a lot of different approaches 
> including using zip on a list like ['','','.'], and other weird stuff :-)
> 
> I just cannot seem to find the nice 3-liner I expected from the 
> beginning. Has anybody got a better solution ? I thought somebody might 
> find it a fun exercise. Well I have...
> 
> 
> regards Max M
> 


-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From owensmk at earthlink.net  Sun Jun  2 20:36:51 2002
From: owensmk at earthlink.net (Michael Owens)
Date: Sun, 2 Jun 2002 19:36:51 -0500
Subject: ANN: PySQLite 0.2.0 Released
Message-ID: <200206021936.51211.owensmk@earthlink.net>

NAME:
    pysqlite - Extension module for SQLite databases.

DESCRIPTION:
    A Python-DB API 2.0 compliant extension module that connects to
    SQLite databases.

    SQLite is a powerful, embedded relational database in a compact C 
    library. It supports a large subset of SQL92, multiple tables and indices,  
    transactions, and triggers. It has a simple C/C++ interface requiring only        
    three functions to perform queries. It has TCL bindings and an ODBC   
    driver. Sources are uncopyrighted and can be used for any purpose. More 
    information can be found at <http://www.hwaci.com/sw/sqlite/index.html>.

USAGE:
    import sqlite

    conn = sqlite.connect("db")
    cursor = conn.cursor()
    SQL = """ select category, family, genus, species
              from calflora order by genus, species limit 10"""

    cursor.execute(SQL)

    for col in cursor.description:
            print "\t %12s - %3s bytes" % (col[0], repr(col[3]))

    row = cursor.fetchone()

    while row != None:
            print "%14s, %15s, %19s, %8s, %25s" % tuple(row)
            row = cursor.fetchone()

    SQL = "insert into calflora (family,genus,species) values(%s,%s,%s)"
    cursor.execute(SQL, ('itchus', 'ivious', 'posionious')) 

    conn.close()

HOMEPAGE:
    <http://pysqlite.sourceforge.net>

DOWNLOAD:
    Linux(source) and Windows (binary) downloads are available at 
    <http://www.sourceforge.net/projects/pysqlite>

AUTHORS:
    Gerhard H?ring <haering_postgresql at gmx.de>
    Michael Owens <mike at mikesclutter.com>
    William Trenker <wtrenker at shaw.ca>

LICENSE:
    Python

NOTES:
The following are changes and enhancements since the last release:

  - Two new project members whose work has made this release.
  - General code cleanup.
  - Lots of bugfixes.
  - Support for SQL NULL.
  - Changed parameter quoting style from 'format' to nicer 'pyformat' style.
  - Much better DB-API 2.0 compliance.
  - Much improved documentation.
  - A 'smart' result set class that allows the columns in a row to be accessed     
    by name (either dictionary-like or attribute-like) additional to list-like
    indices.
  - Reworked examples using the new features. One additional example for
    mass-importing data from an XML file.
  - Unicode support.
  - SQLite is typeless. We added a possibility to overcome this limitation.  
    The types int, float, str and unicode are supported by default, but you    
    can add user-defined types.
  - User-defined SQL functions and aggregates can be programmed in Python (!)
     and registered to the connection.
  - Better error handling and error messages.
  - An extensive test suite.
  - A homepage with preliminary documentation
  - A users' mailing list.
  - win32 binaries








From greg at cosc.canterbury.ac.nz  Sun Jun 30 03:24:49 2002
From: greg at cosc.canterbury.ac.nz (greg)
Date: Sun, 30 Jun 2002 19:24:49 +1200
Subject: squared functions--most Pythonic way?
References: <mailman.1025207218.19740.python-list@python.org> <3D1C207A.74F24164@gol.ge> <afhu82$15ed$1@news.adamastor.ac.za>
Message-ID: <3D1EB241.FF7ABD7D@cosc.canterbury.ac.nz>

Janto Dreijer wrote:
> 
> def addNumbers(k):
>     def f(x):
>         a = addNumbers(x + k)
>         a.val = x+k
>         return a
>     return f
> 
> >>> addNumbers(9)(5)(2)(4)(6).val
> 26
> 
> Now if only I could figure out how to use __repr__() so I don't 
> need that ".val". It also fails when passed only one number. i.e
> addNumbers(5). Help?

class AddNumbers:

  def __init__(self, x):
    self.val = x

  def __call__(self, k):
    return AddNumbers(self.val + k)
	
  def __repr__(self):
    return repr(self.val)

addNumbers = AddNumbers(0)

>>> print addNumbers(9)(5)(2)(4)(6)
26
>>> print addNumbers(5)
0

--
Greg


From thor__00 at yahoo.com  Thu Jun 13 13:19:48 2002
From: thor__00 at yahoo.com (Thor)
Date: Thu, 13 Jun 2002 19:19:48 +0200
Subject: system command
Message-ID: <aeajuh$5b5eo$1@ID-108351.news.dfncis.de>

Having this code
 
  print "source "+root+".in"
  os.system("source "+root+".in")

the script does not run the root+".in" file, but copying and pasting the 
output form teh 1st line work.. what i'm doing wrong? Thanks in advance
-- 
Thor -- Stockholm -- Sverige


From marklists at mceahern.com  Sun Jun 30 13:46:49 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sun, 30 Jun 2002 12:46:49 -0500
Subject: aspect-oriented demo using metaclasses
In-Reply-To: <pan.2002.06.29.19.23.48.841533.2265@club-internet.fr>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEEGLHAA.marklists@mceahern.com>

Pedro, thanks for the reply.  Comments follow inline below.  Some of my
comments will refer to the latest version of code that I just posted
separately.

[Pedro]
> If you really want something more consistent than my posting, I will
> recommend you to also check:
[snip]
>     - Pythius by Juergen Hermann
>     - Transwarp by Phillip J. Eby
[snip]

I've looked at Transwarp and failed to get it.  Last time I looked there was
no simple demo that showed what problem it was trying to solve and how TW
solves it.  But I'm dense and I should probably give it another try.  I look
forward to checking out Pythius.  Someone else mentioned that
recently--probably in the recent thread on metaclasses.

I wrote:
> (I'm not worried about distinguishing staticmethod and classmethod
> type methods for now):

[Pedro]
> You should <wink>. At least for the purpose of the exercise ;)
> (Honestly I didn't try either, it may end up with a two liner...
> but I wonder if they will take two minutes or two hours to be written ;)

The latest version I posted continues to punt on staticmethod and
classmethod.  They're currently not aspected because my type comparison to
type(lambda x:x) will effectively filter out staticmethod and classmethod.
I would have to change the observer interface slightly since staticmethods
don't have have a self parameter and the first parameter of classmethods is
a reference to the class not the instance.  I don't see this being too
hard--I plan to fix it in the next iteration.

> Even if deprecation of the types module have been discussed on the devel
> list, I think it is preferable.

Yeah, using type(lambda x:x) seems more portable or something.  I just did a
little interactive session to verify that I can do this too:

  type(staticmethod(lambda x:x))
  type(classmethod(lambda x:x))

I can see the outlines of some sort of method wrapper factory even now:

  anon = lambda x:x
  # f, g, and h are make_wrapped_method variants for each type of method.
  wrapper_factory = {type(anon): f,
                     type(staticmethod(anon)): g,
                     type(classmethod(anon)): h}
  for k, v in dict.items():
    wrapper = wrapper_factory[type(v)]
    ...

> Oops... Didn't I make the same mistake...
> ... forgetting about the returned value ;)
>
> Something like could be better:
>      def __call__(self, *args, **kwargs):
>          self.before(*args, **kwargs)
>          ret = self.method(self, *args, **kwargs)
>          self.after(*args, **kwargs)
>          return ret

Great catch!  Thanks.  I fixed that in the latest version.

> This Trace class is an aspect.

Ah, I see that I was calling the metaclass Aspect whereas Trace is really an
Aspect.  I definitely need help with naming and terminology, as you can
tell.  I renamed the metaclass to Aspected.  Does that seem better or am I
still missing the correct term?

> This is not an Aspect. This is your way to implement method call
> interception.

Thank you.

> What about module functions ? Bounded methods ?
> Using metaclass for call interception is too restrictive I believe.

Hmm, you're right, my approach will do NOTHING for methods that aren't
associated with a class.  BTW, I thought a normal method associated with a
class whether new type or old was a bound method?  I'm not too worried about
old style classes--although maybe I should be?  I figure if a requirement of
using this is that it only works with new-style classes, that's fine with
me.  Until I run into a problem, of course.  ;-)

I often use methods that aren't associated with a class and I'd like to
aspect those as well, although, when I think about something like
Persistence, that does seem initimately bound up with classes.  So maybe it
depends on the aspect?

I should say that my motivation for pursuing a metaclass implementation is
partly just to learn metaclasses, but also partly because I want something
that requires no modification to the aspected class and minimal effort,
maximum flexibility for defining join points.  What I like about the
metaclass approach is that I just wire up every single class that's aspected
for all events.  It's then up to the observer to hook into them.  Of course,
I need to try to use this for something real and I'm sure that will expose
its weaknesses/my conceptual gaps quickly.

> This is the trickier part. When you do :
>        setattr(cls, k, wrapped_method(cls, v))
> you substitute a function (actually an unbounded method) by a callable
> object. Unfortunately when you'll invoke this object with a classical
> method call, Python will not pass the instance as the first argument.

Yup, I fixed this.

> - providing a function, and take benifit of nested_scopes to retrieve
>   all the information from your context

That's the approach I took.  And to think, 6 months ago I had no freaking
idea what the point of nested scopes was.  ;-)

> Too intrusive. I don't believe that you can do it dynamically, at least
> not for classes defined at module level. They will be created with the
> metaclass defined at compilation time.

This is a good point.  Here's the sort of dynamism I'm aiming for:

The modules containing classes to be aspected have no reference to the
framework inside them.  Not even a __metaclass__ statement.

Oops, a little testing shows that may simply not work.  Hmm.  I don't want
to have to edit the modules that contain the classes to be aspected, even to
add a __metaclass__ declaration.  Since I can't seem to change that at
runtime, this approach probably won't work.

> Yes. Interception of raised exception is a good (and easy ;) feature.
> Just try to go a step further with 'around' methods.

I avoided adding around because it doesn't seem primitive to me.  In other
words, isn't around just before + after notification?  So I could add
support for around simply by making it so that observers wanting around
notification just got before and after notification?  Not sure what to do if
the method raises an error--skip the after?

> Thanks, for this posting Mark. Reminds me that aop is my task list for
> the moment I will have some spare time (not so far I hope ;)

You're very welcome.  Your comments have been most helpful.

> Just to give some hints on aop, this is what I wrote as a reminder
> 6 months ago along with a more complete implementation of aop :

[snip]

Wow, that's eye opening.  I look forward to seeing more.  In the meantime,
I'm going to reflect on whether to abandon the metaclass approach.  It does
seem, as you say, restrictive.  That is, it doesn't allow sufficient runtime
dynamism and it seems to require modifying the code-to-be-aspected.

Cheers,

// mark

-





From sholden at holdenweb.com  Wed Jun 26 08:48:46 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 26 Jun 2002 08:48:46 -0400
Subject: Pythoniac: Thoughts on a hardware Python processor
References: <afc52t$2pj$1@sirius.dur.ac.uk>
Message-ID: <eKiS8.43021$mF.33425@atlpnn01.usenetserver.com>

"Christopher Saunter" <Christopher.Saunter at durham.ac.uk> wrote in message
news:afc52t$2pj$1 at sirius.dur.ac.uk...
> Dear All,
>
> I have seen the occasional post about creating a specific hardware 'Python
> Processor', that could then be built and used in programmable logic or the
> like (Valves, relays, core memory - Pythoniac ;-) - in other words, create
> a processor that executes Python bytecode directly.  There would be
> various different reasons for doing this (the 'neat' factor etc...)
>
[... Pythoniacity ...]
>
> So, Pythoniac: a hardware (well, FPGA(2)...) processor that directly
> executes Python bytecode.  What do people think about the idea?  Is it
> really feasible?  I have doubts it would be particularly quick (in the
> same way I have my doubts about anti-gravity generators... ;-), but it
> should be fun!
>
You will, of course, have to stop the development team from tinkering with
the bytecode when moving from one major version to another. If you are
prepared to overlook their tendency to do that I would think this would be a
great project. Once you got the hardware core working the major problem
would be C/C++ to P+ compiler.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From jepler at unpythonic.net  Thu Jun 13 12:06:02 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Thu, 13 Jun 2002 11:06:02 -0500
Subject: filter2
In-Reply-To: <lkptyvmdfz.fsf@pc150.maths.bris.ac.uk>
References: <mailman.1023975618.6474.python-list@python.org> <lkptyvmdfz.fsf@pc150.maths.bris.ac.uk>
Message-ID: <20020613160557.GC25416@unpythonic.net>

On Thu, Jun 13, 2002 at 03:00:28PM +0000, Michael Hudson wrote:
> Write it in Python.  Particularly if test is expensive the loop
> overhead will disappear into the noise.
> 
> (with-seriousness-level :low
>   If you insist on C loops, there's always:
> 
>       lt = map(lambda x:(test(x), x), alist)
>       l1 = filter(lambda (x,y):x, lt)
>       l2 = filter(lambda (x,y):not x, lt)
> 
>   But I doubt you need me to tell you why this solution sucks.)

If you're benchmarking, try a solution like this in the mix.  It only
has one map/filter call, not map+2*filter. (uses nested scopes and the
new bool() builtin, but this can be avoided if necessary)

def filter2(test, list):
    l1 = []
    l2 = []
    funcs = [l2.append, l1.append]
    map(lambda x: funcs[bool(test(x))](x), list)
    return l1, l2

>>> print filter2(lambda x: x%3==0, range(16))
([0, 3, 6, 9, 12, 15], [1, 2, 4, 5, 7, 8, 10, 11, 13, 14])

Jeff




From howard at eegsoftware.com  Thu Jun 27 02:51:17 2002
From: howard at eegsoftware.com (Howard Lightstone)
Date: Thu, 27 Jun 2002 06:51:17 GMT
Subject: continuous redirection of stdout & stderr to Tkinter Text
References: <3D1A32B0.CA7D3618@club-internet.fr>
Message-ID: <Xns9239F2B36AC5Ahowardeegsoftwarecom@66.75.162.198>

Fabien H?non <ffjhenon at club-internet.fr> wrote in 
news:3D1A32B0.CA7D3618 at club-internet.fr:

> I am using Tkinter as an editor and a launcher (under Linux) for a
> raytracer (POV-RAY)
> 
> Once started, the raytracer does two things in the same time :
> - it displays the image being raytraced
> - it spits out information to the console (ie : the line which is being
> raytraced). 
> I get the image displayed but not the redirection of stdout
> 
It was always too much for me so I built this simple class (which I use to 
get errors when running with pythonw):
"""
Pseudo debug screens

These capture stderr/std and route them to a Tkinter text widget.
This implementation uses a common window for stdout/stderr with
an asterisk at the beginning of each stderr line.

"""
import Tkinter
import sys,string

class DbgText:
    Dbgtopwin=None
    Dbgwidget=None
    DbgRoot=None
    
    def _kill_topwin(self):
        DbgText.Dbgwidget=None
        if DbgText.Dbgtopwin != None:
            DbgText.Dbgtopwin.destroy()
        DbgText.Dbgtopwin=None
       
    def __init__(self,kind=''):
        self.kind=kind
        self.window=None
        self.widget=None
        self.called=0
        self.hide=0
        self.buffer=''

    def __del__(self):
        "On deletion, wait for user to see the output"
        if DbgText.Dbgtopwin != None:
            See()
        self._kill_topwin()
    
    def write(self,charstr):
        "write text to buffer or window"
        if self.hide:
            self.buffer.append(charstr)
        else:
            if self.window == None:
                if DbgText.Dbgtopwin == None:
                    DbgText.Dbgtopwin=Tkinter.Tk()
                    DbgText.Dbgtopwin.protocol('WM_DELETE_WINDOW',Dbg_kill_
topwin)
                    DbgText.Dbgwidget=Tkinter.Text(DbgText.Dbgtopwin)
                    DbgText.Dbgwidget.pack(expand=1)
                top=DbgText.Dbgtopwin
                wid=DbgText.Dbgwidget
            else:
                if self.widget == None:
                    self.widget=Tkinter.Text(self.window)
                top=self.window
                wid=self.widget
            if self.kind != '':
                ep=wid.index('end')
                sp=string.split(ep,'.')
                # determine length of 'previous' line
                prevl=int(sp[0])
                tx='\n'
                if prevl:
                    pl='%d.0' % (prevl-1)
                    tx=wid.get(pl,ep)
                # if this is start of a new line
                if tx[0] == '\n':
                    wid.insert('end',self.kind)
            wid.insert('end',charstr)     
        self.called=1
        top.update()

def Dbg_kill_topwin():
    f=DbgText()
    f._kill_topwin()
    
def Take_stdout():
    "DIsplay stdout in text widget"
    if not isinstance(sys.stdout,DbgText):
        f=DbgText()
        f.prev=sys.stdout
        sys.stdout=f

def Take_stderr():
    "DIsplay stderr in text widget"
    if not isinstance(sys.stderr,DbgText):
        f=DbgText('*')
        f.prev=sys.stderr
        sys.stderr=f

    
def Restore_stdout():
    f=sys.stdout
    if isinstance(f,DbgText):
        sys.stdout=f.prev
        del f

def Restore_stderr():
    f=sys.stderr
    if isinstance(f,DbgText):
        sys.stderr=f.prev
        del f

def Define_Root():
    root=Tkinter.Tk()
    root.withdraw()
    DbgText.DbgRoot=root


def See():
    db=DbgText()
    if db.Dbgtopwin != None:
        db.Dbgtopwin.mainloop() # loop for me to see

def Take_all():
    "send stderr/stdout to Tkinter text window/widget"
    Take_stdout()
    Take_stderr()

def Restore_all():
    "restore stderr/stdout"
    Restore_stdout()
    Restore_stderr()

        
if __name__ == '__main__':
    print 'stdout is here'
    Take_stdout()
    print 'stdout should now be in window'
    print ' this is the second line'
    raw_input()
    Restore_stdout()
    print 'stdout back to original'
    


From BPettersen at NAREX.com  Mon Jun 24 11:59:05 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 24 Jun 2002 09:59:05 -0600
Subject: converting an array of chars to a string
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158EF5@admin56.narex.com>

> From: JB [mailto:jblazi at hotmail.com] 
> 
> Bengt Richter wrote:
> 
> > Have a look at the array module. It can construct an array
> > of signed or unsigned character elements optionally 
> initialized with 
> > values from a list or string.
> [...]
> 
> Thank you very much. In principle, this was exactly, for 
> what I was looking. But most unforunately if I declare such 
> an array a, then
> 
> a[0]=210
> a[0] += 46
> 
> produces an error instead of the correct result 0. Now I am 
> a bit desperate.

Python helpfully gives you an OverflowError since that's technically
what it is. Having the value magically truncated would surprise even
more people. The simple solution is of course:

  def byteAdd(a, b):
      return (a + b) % 256

  a[0] = byteAdd(a[0], 46)

hth,
-- bjorn




From Chris.Barker at noaa.gov  Tue Jun 18 12:01:42 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 18 Jun 2002 09:01:42 -0700
Subject: NEWbie asks "is Python a good tool for automated file/folder 
 generation?"
References: <mailman.1024073951.24787.python-list@python.org> <3D0E8773.A1BA95ED@engcorp.com>
Message-ID: <3D0F5967.2A9CEE86@noaa.gov>


Peter Hansen wrote:
> Better to learn to use raw-strings, with an "r" in front,
> as in path = r"c:\this is a safe path" or better yet,
> either forward slashes always: path = "c:/this works too".

Or even better, use os.path.join(), you never know when you will
encounter a Mac. (of course, if it's OSX, the above will work as well)

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From dradul at yahoo.com  Fri Jun 21 21:41:05 2002
From: dradul at yahoo.com (P. Alejandro Lopez-Valencia)
Date: Fri, 21 Jun 2002 20:41:05 -0500
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu>
Message-ID: <af0llm$a7ntf$2@ID-99513.news.dfncis.de>

"Tom Loredo" <loredo at astro.cornell.edu> escribi? en el mensaje
news:3D110B55.4D8AD7FA at astro.cornell.edu...
>
> In particular,
> I write many C extensions; can one use distutils and the Borland
> or gcc free compilers to build C extensions in both versions?

Yes, as long as it be possible to convert the link libraries from one
format to the other. Everything is explained in the distutils
documentation (I'm surprised none has mentioned this yet!!!).

In the case of the Borland compilers, you process the MSVC6 import
libraries provided in the ${PYTHONHOME}/libs directory with COFF2OMF and
add the suffix "_bcpp", like so

c:\whatever\libs > coff2omf python22.lib python22_bcpp.lib

and add the resulting libraries to the same directory where the MSVC
import libraries are located originally. To compile your extension with
Borland C++ and distutils, you just type,

c:\foo\bar > python setup.py build --compiler=bcpp

Should know, I just compiled the latest numpy CVS snapshot with BCPP
ten minutes ago ;)

If you want to use mingw32 gcc, the process is *a lot" more complicated.
See the documents I mentioned above (basically you need to extract all
the export symbols from the DLL and then create an import library based
on the exports definition file).



From milos.prudek at tiscali.cz  Mon Jun 10 15:18:56 2002
From: milos.prudek at tiscali.cz (Milos Prudek)
Date: Mon, 10 Jun 2002 21:18:56 +0200
Subject: popen2: Broken pipe for large files
Message-ID: <3D04FBA0.2030703@tiscali.cz>

In Python 2.1, is there a size limit when creating pipes with popen2() ?

The following function fails because of a broken pipe if the size of T 
is larger than 21000:

T=File.read()
cmd='file -b -'
w,r=os.popen2(cmd,'b',1000000)
w.write(T)
w.close()
X = r.read().split()
print X

The error message:
Traceback (most recent call last):
   File "./p26.py", line 33, in ?
     w.close()
IOError: [Errno 32] Broken pipe

I tried to specify buffer size 1000000, but it did not help.

The only possible solution so far is to change the first line:

T=File.read()[:20000]

-- 
Milos Prudek






From giffordj at linkline.com  Wed Jun 19 20:02:12 2002
From: giffordj at linkline.com (Jim Gifford)
Date: Wed, 19 Jun 2002 17:02:12 -0700
Subject: Python Module Install Problem
Message-ID: <uh26s5a851kce7@corp.supernews.com>

I have been trying to get syslogmodule to install. I have followed all the
directions and it still does not install. Here is how I have been compiling
Python and installing the modules.

Compile of Python

./configure --quiet --prefix=/usr --with-ncurses --enable-shared
edit /usr/src/Python*/Modules/Setup
                   uncomment *shared*
                   uncomment _curses
                   uncomment timing
                   uncomment syslog
                   uncomment linuxaudiodev
                   uncomment readline
                   uncomment gdbm
                   uncomment md5
                   change /usr/local to /usr
edit /usr/src/Python*/setup.py
                    change /usr/local to /usr
make
make install
python setup.py install

Thank you in advance for you help.




From gh_pythonlist at gmx.de  Thu Jun 13 22:12:16 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 14 Jun 2002 04:12:16 +0200
Subject: Type subclassing: bug or feature
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F06233012631BC@mail.hebmex.com>
References: <F7DB8D13DB61D511B6FF00B0D0F06233012631BC@mail.hebmex.com>
Message-ID: <20020614021215.GD590@lilith.my-fqdn.de>

* Gustavo Cordova <gcordova at hebmex.com> [2002-06-13 16:18 -0500]:
> maybe all methods of builtin types should do:
> 
> 	return self.__class__(return_value)
> 
> when returning a new object (like str methods).

Looks like a waste of performance in about 99 % of the time. Hopefully,
this can be optimized at the C level.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 17.1 ?C      Wind: 1.3 m/s




From tmagna at online.no  Sun Jun  2 10:22:15 2002
From: tmagna at online.no (Tobias Brandvik)
Date: Sun, 02 Jun 2002 14:22:15 GMT
Subject: Python Turing machine
Message-ID: <d9akfu0c7ro0sebio1vd9c6ik2i0fg59le@4ax.com>

Has anyone seen a model of a Universal Turing Machine coded in Python?
I've seen examples in Scheme, Java and Perl, but I'd really like to
look at one made in Python... Any pointers?

--Tobias Brandvik


From BPettersen at NAREX.com  Sat Jun 29 14:41:29 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Sat, 29 Jun 2002 12:41:29 -0600
Subject: yield equivalent in C/JavaScript?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E201922151B8@admin56.narex.com>

> From: Robin Becker [mailto:robin at jessikat.fsnet.co.uk] 
> 
> Is there any way to do yielding easily in C? I have a bunch 
> of programs which are normally connected together using unix 
> pipes. I would like to merge them and the problem is then how 
> to get them to communicate nicely using a single process.

Using threads, you can simulate yield by having the yielding thread
blocking until the thread calling it tells it to restart. This is rather
obscure however, so I wouldn't suggest that you do it. If you are going
to use threads, I would suggest using a Queue to communicate
work/results (I can send you our C++ version of Python's Queue.Queue()
implementation if that would help...)

-- bjorn




From bcbooker at aol.com  Sun Jun  2 22:14:59 2002
From: bcbooker at aol.com (BCBOOKER)
Date: 03 Jun 2002 02:14:59 GMT
Subject: cursors, etc
Message-ID: <20020602221459.05355.00002083@mb-mm.aol.com>

Is there any way for a user to actively move a cursor or some other marker on a
plane?  That is, if there is a 4x4 square with the cursor at (1,1), can the
user press up, n, 1 or whatever and make the x increase?  I tried making
something along the lines of this, but i had to go type in every situation -
that is, make the map for (2, 3) or (1, 2).

I'm a newbie to programming as a whole, so please excuse any nievete that this
message may exude...

cb


From ponderor at lycos.com  Sat Jun 15 20:53:38 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 15 Jun 2002 17:53:38 -0700
Subject: VBA to Python
References: <HjAO8.4199$iR2.398698@news.xtra.co.nz>
Message-ID: <e81be8b2.0206151653.5d09a77d@posting.google.com>

Keith:
> Is there a way of
> transcribing / translating the VBA code across to Python other than
> literally doing it
> line by line?

I don't think so.
But although it sounds tedious, I would expect a translator to still
require a function by function, most likely line by line (tick and
tie?) analysis.
It may also help you're transition and validation for using Python.
Can code reviews can be a bad thing? (rhetorical.)

-Dean


From ark at research.att.com  Fri Jun 14 15:27:51 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 14 Jun 2002 19:27:51 GMT
Subject: a possibly foolish question about slices
References: <yu99sn3prbqm.fsf@europa.research.att.com> <aedd8k$brf$1@scotsman.ed.ac.uk> <m3znxxu21l.fsf@mira.informatik.hu-berlin.de> <aedf1p$bv1$1@scotsman.ed.ac.uk>
Message-ID: <yu99k7p1r77s.fsf@europa.research.att.com>

Eddie> Hmm, I was actually trying to suggest (non-seriously) fixing
Eddie> zero based indexing! but I realise now that wasn't clear.

It's hard to imagine how it would be possible to change to 1-based
indexing without breaking lots of code.

For that matter, I had the thought that it might be possible simply
to define x[n:0] as meaning x[n:len(x)] if n<0, but I imagine that
even that would break too much code.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From peter at engcorp.com  Thu Jun 27 20:03:47 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 27 Jun 2002 20:03:47 -0400
Subject: Most elegant python program to date.
References: <mailman.1025198615.10254.python-list@python.org>
Message-ID: <3D1BA7E3.522E7697@engcorp.com>

PoulsenL at capecon.com wrote:
> 
> Just curious.  I find it much easier to learn a language by reading lucid
> examples of its prose.  Are there any nominations for the most elegant
> example of a python program to date.  Not sure under what criteria, I leave
> that up to you.  I will say that I use python for matrix manipulation more
> than anything, but that shouldn't prevent you from nominating your favorite
> regardless of purpose.

I think, perhaps, that compared to other languages, the range of 
readability or even elegance in Python programs is rather narrow.
I certainly have not read as much as most, but I don't remember
anything in particular standing out.

-Peter


From dougz at cs.washington.edu  Fri Jun 14 17:00:15 2002
From: dougz at cs.washington.edu (Douglas Zongker)
Date: Fri, 14 Jun 2002 21:00:15 GMT
Subject: index() and count() methods
Message-ID: <GxproF.1y4K@beaver.cs.washington.edu>

Apologies if this is a FAQ, but I haven't run across an answer.

Why are index() and count() methods of only mutable sequences, rather
than all sequences?  I ran into a situation today where using index()
on a tuple would have been nice.  It was easy to work around; I'm just
curious.

thanks,
dz





From donn at u.washington.edu  Mon Jun 10 18:48:06 2002
From: donn at u.washington.edu (Donn Cave)
Date: 10 Jun 2002 22:48:06 GMT
Subject: popen2: Broken pipe for large files
References: <mailman.1023736815.9944.python-list@python.org>
Message-ID: <ae3ab6$8u4$1@nntp6.u.washington.edu>

Quoth Milos Prudek <milos.prudek at tiscali.cz>:
| In Python 2.1, is there a size limit when creating pipes with popen2() ?
|
| The following function fails because of a broken pipe if the size of T 
| is larger than 21000:
|
| T=File.read()
| cmd='file -b -'
| w,r=os.popen2(cmd,'b',1000000)
| w.write(T)
| w.close()
| X = r.read().split()
| print X

I suppose the "file" command only reads so much data (and probably
doesn't use all of that, as it is only interested in the top of the
file.)  When python tries to write more data, the pipe is already
closed and that's an error.

| The only possible solution so far is to change the first line:
|
| T=File.read()[:20000]

Sure, that will work, though it might be more economical to write

  T = File.read(20000)

	Donn Cave, donn at u.washington.edu


From jbublitzNO at SPAMnwinternet.com  Mon Jun 17 14:10:43 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Mon, 17 Jun 2002 18:10:43 GMT
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de> <aeftmj$m29$1@bob.news.rcn.net> <3d0c6d67@nntp.server.uni-frankfurt.de> <3D0CE698.1050806@SPAMnwinternet.com> <aeiu3a$4ff$1@bob.news.rcn.net> <3D0D3223.5070500@SPAMnwinternet.com> <aekrd1$4ao$1@bob.news.rcn.net>
Message-ID: <3D0E2606.5050001@SPAMnwinternet.com>

David Abrahams wrote:
> "Jim" <jbublitzNO at SPAMnwinternet.com> wrote in message
> news:3D0D3223.5070500 at SPAMnwinternet.com...

>>The biggest plus for sip is that the 'IDL' is the h file syntax
>>with a few subtractions (variable names, forward declarations,
>>inline code, but inline methods stay), so that most of the
>>description file creation is easy to automate and still not
>>that hard to do manually for a small number of files. sip
>>supports most C++ h file syntax.

> Hmm, I wonder about that. The C++ type system is pretty sophisticated (to
> the point of being able to represent entire compile-time programs!) and all
> of the parser-based wrapping systems I've seen eventually fail to handle
> some constructs I use, especially if I'm wrapping template-based code.
> That's part of the reason I chose to use the compiler itself in
> Boost.Python. The compiler (in theory) has no problem with C++ syntax ;-)

> On the other hand, I realize there's a large category of simple C++
> interfaces which can be parsed will less-than-complete machinery.

Template-based code does require special treatment with sip.
In most cases where I've wrapped template-based types or
classes, I manually write C++ code for conversion C++ <-> Py
(sip has some functions that make that easier) and implement
it as a sip 'mapped type', which lets me refer to the same
conversion code from any return value or arg. I generally
assign a new name (eg QList<int> -> IntList) but wouldn't
have to (sip will parse QList<int> and look for a mapped
type definition corresponding).

sip won't handle all possible C++ syntax, and it sometimes
takes some creativity to shoehorn something in, but something
well over 80% of h file code translates directly (after
removing variable names), and only 1 or 2% really takes
major effort.

The real problem wouldn't seem to be C++ types as such, but
finding a useful (to the Python programmer) representation
via PyObject, and still maintaining the range of abilities
built into the C++ code (esp things like subclassing).

>>The other features that are helpful on large projects are the
>>built-in support for versioning

> How so? Can you say more about this?

sip uses directives: %If (A - B) ... %End to wrap a block
of code for versioning over a range. It could also be
%If (A - ) (>= A) or %If ( - B) (< B). Acceptable values
for A and B are defined in a %Timeline declaration; actual
values for A and B are passed to sip via the command line,
usually  derived from a version #define in the C++ source
code via configure (at least on Linux). Evaluation is by
lexical ordering of the A and B values. A and B are
strings like 'KDE_3_0_0'.

Not very different from #ifdefs, but maybe a little
easier.

>>and embedded docs (both also fairly
>>easy to automate),

> Do you mean docstring generation?

No - similar, but sip provides %ExportDoc ... %End directives
inside which you could place just about anything you want I
suppose. For PyQt/PyKDE the directives wrap SGML marked-up text.
You can construct an input file which along with the proper sip
command line switch, will collect and concatenate all of the
doc blocks distributed over the description files. The
concatenated file can be processed (in the case of SGML) to
provide HTML docs, or whatever you require.

>>as well as support for 'platforms' and
>>'features'. So for example PyQt (not my project) can support Qt
>>from 1.4.2 through 3.0.4 with different mixes of features (with
>>SQL or not, QTable or not) across Windows, Linux/Unix/BSD, and
>>(soon) Mac, with a single set of source files, and still be
>>manageable by one person.

> Boost.Python works on a similar range of platforms.
> I'm not sure if this configuration stuff is an appropriate job for a
> wrapping system to provide specific support for; what's the argument for
> integrating it with the wrapping system?

PyKDE consists of about 400 description files spread over 10-12
libraries depending on version and not necessarily with the same
description file <-> library correspondence in each version. PyQt
is slightly smaller, but supports more platforms. The alternative
to embedding version/platform/feature info in the description files
is to maintain separate sets of files. In the case of PyKDE (depends
on PyQt) for KDE < 3.0.0, it might build with 2.3.0 <= Qt <= 2.3.2 
depending on what the user has. If the user compiled Qt himself, he 
might have a different feature set than the official distributed Qt
for his platform. The number of permutations is fairly large and that
would be multiplied by the number of files.

On the other hand, changes or bug fixes are usually the same across
versions/platforms/features. So in addition to only supporting one
fileset instead of many, it's only necessary to repair/edit one
fileset instead of many. Similarly, it isn't possible to test every
possible platform - the potential for simple typos crippling a
particular version alone is pretty large, and most code is common.

Again, #ifdefs probably provide similar functionality - this is
just a little more formalized/standardized and supported by
the tools (so error checking is possible, for instance).

>>sip is flexible enough that you can handwrite code to "Pythonize"
>>C++ code (return a Python list or dict, or for example,
>>'void rgb (int&, int&, int&)' returns (int, int, int) and takes
>>no arguments in PyKDE), or create type substitutions/conversions
>>('mapped types') once and re-use them across a project. You can
>>do C++ with anything of course, but sip has it well-integrated
>>(manageable) and has support code to make it easier (arg unpacking,
>>object conversions, etc). IIRC, Boost might not need handwritten
>>C++ code quite as much.

> Not sure about this; we'd need to compare specific functionality.

My guess is that Boost does at least some of this a little more
naturally. I'm not sure the extent that Boost can modify things
like return values or args without resorting to C++, however. sip
has to resort to C++, but has helper functions, so a typical
piece of handwritten code rarely exceeds 10 - 15 lines, including
declarations.

>>I suppose those are a series of small advantages, but the end
>>result seems to me that you write less code

> Are you counting the IDL file? That's code, IMO.

Yes, but I don't write it - I just edit it mostly to correct
mistakes my description file generator creates and to handwrite
code where the generator has flagged that. Crude but effective (TM)
Even without automation, the process is easiest if you start by
saving the h file with a .sip extension and then (mostly) delete
the things that sip doesn't want/need, so it's still mostly
editing (sometimes A LOT of editing).

>>Boost (IIRC) also has a big advantage in being able to bind much
>>more selectively.

> Can you explain what you mean by that?

I should probably go back and review Boost. I had thought that
with Boost it was possible to wrap as much or as little of a
C++ library as you want to. That's more of a problem with sip.

>>sip is easiest when binding everything in a lib
>>and probably takes a little more effort to be selective - you
>>can do it, but it isn't as 'mindless' (ie automatable) as most
>>sip file creation is. And of course Boost is actually
>>documented - no small thing.

> Heh, yeah. Unfortunately, documentation is often harder than coding. Even
> the current official release of Boost.Python is not very well-documented.
> We're working hard to correct that for v2, which is in pre-release.

Well, you have docs ... I found the web site to be pretty good at
explaining how to use Boost (and perhaps haven't remembered it all
correctly).

>>How would you see Boost performing on a very large project -
>>say 10 libs, 20 classes/lib, 10 exposed methods/class (2000
>>methods overall)?

> See http://cctbx.sourceforge.net/
> Also
> http://mambo.peabody.jhu.edu/~karlmac/publications/gamera_python_2001/gamera
> .html

> Of course, these don't prove anything, since wrapping is a nontrivial job
> and people quickly get invested in whatever is working for them. You *can*
> wrap a large project with Boost.Python, but whether it's easier than with
> sip...?

Never having used Boost, I couldn't say either. Bindings do
seem to require a particular mindset. There is a learning curve
certainly, and as mentioned, familiarity with a particular
method introduces a lot of expectations.

>>I haven't looked at Boost in a while -
>>could you incorporate conditional compilation (eg versions,
>>features)

> It's C++; you can use #ifdef in just the same way that the rest of the C++
> code surely does.

sip's only advantage then is that it's able to do some error
checking on versioning/feature conditionals because the
syntax is built into sip.

>>and embed docs easily?

> Docstring support is on the todo list for v2. It will be a matter of passing
> a string literal to the def() function which wraps a function pointer:

> class_<Foo>()
>    .def_init(args<int>(), "Creates a Foo with the specified number of Bars")
>    .def("get_name", &Foo::get_name, "get the Foo's name")
>    .def("set_name", &Foo::set_name, "set the Foo's name")
>    ;

That looks cool - it's a little more cumbersome with
sip.

>>Could you automate code generation to some extent?

> Not sure what you have in mind here. In a sense, the metaprogramming
> techniques used by Boost.Python already *do* code generation.

Yeah - 'code generation' is ambiguous. sip itself is a code generator
(it produces C++/Python files which are the actual bindings or Python
lib). What I meant in this case is generating the description files
directly from the h files. I currently take a set of h files, run them
through a handwritten parser and generate *.sip files which sip then
uses to generate the actual C++ code for the bindings. I had the
impression SWIG was supposed to do most of that in a single step. Just
wondering if I could implement something like that with Boost, or
need to code a line manually for each method/class/whatever.

>>The automation for generating
>>PyKDE is pretty crude at the moment, and in all fairness
>>isn't generally available, so this example would be a lot
>>of work using sip as it comes out of the box. Is Boost
>>really intended for projects of this size (there probably
>>aren't many), or do you have different goals for it?

> Yes, Boost.Python is intended for large projects. One of the largest
> projects I've ever heard is using it at Lawrence Livermore Labs, for
> example.

It's pretty amazing to find out who's using Python and things
like Boost, pyGTK. PyQt, PyKDE, etc.

I left your web site with the impression that I (personally) would
find sip easier for large projects, but Boost would have advantages
on smaller projects (unfortunately I remember the impression better
than the reasons for it). That's especially interesting to me at
the moment, because I have ideas for a lot of smaller projects. I expect
I'll stick with sip, but will have to take another look at Boost, if
only to steal good ideas :)

When I get a chance, I'll check out the links above.

Jim



From steve at lurking.demon.co.uk  Tue Jun  4 19:50:13 2002
From: steve at lurking.demon.co.uk (Stephen Horne)
Date: Wed, 05 Jun 2002 00:50:13 +0100
Subject: Thought on PEP 204 and 276
References: <ckp4fuc4fskjcpilnn2mrt3665fqvld92g@4ax.com> <mailman.1022528744.19859.python-list@python.org> <gm17fu40rh6r675g3a6d1ovl49tb9pd822@4ax.com> <mailman.1022594566.27265.python-list@python.org> <slrnaf7mo4.dfe.philh@comuno.freeserve.co.uk> <6w4L8.229234$%u2.120502@atlpnn01.usenetserver.com> <j36qfu4r84j8320iegrujjf8dpvdvs20bm@4ax.com> <YU9L8.240282$%u2.115403@atlpnn01.usenetserver.com>
Message-ID: <88hqfu8ks2sfq8hkk63jn02dr9tp7mvhdh@4ax.com>

On Tue, 4 Jun 2002 16:57:00 -0400, "Steve Holden"
<sholden at holdenweb.com> wrote:

>"Stephen Horne" <steve at lurking.demon.co.uk> wrote ...
>> On Tue, 4 Jun 2002 10:49:19 -0400, "Steve Holden"
>> <sholden at holdenweb.com> wrote:
>>
>> >"phil hunt" <philh at comuno.freeserve.co.uk> wrote in message
>> >news:slrnaf7mo4.dfe.philh at comuno.freeserve.co.uk...
>>
>> >> I think they are essentially the same. A sequence tpye is just a
>> >> mapping type where the keys are restricted to the integers from
>> >> 0 to however many elements there are minus 1.
>> >>
>> >No matter what you think, they are essentially different. Insertion into
>a
>> >mapping doesn't alter the key to which other elements are mapped, but
>> >insertion into a sequence (potentially) does that.
>>
>> What is normally referred to as insertion in dictionaries is actually
>> more like overwriting a list element. This is why handling both with
>> the 'container [key] = value' syntax is natural.
>>
>aarrgghh. So we can completely ignore the fact that the len() of a
>dictionary will change after an insertion while the len() of a list won't
>after an element is overwritten?

a = {0 : 5}
a [0] = 6
len(a) == 1

a = [5]
a[0] = 6
len(a) == 1

len(a) is unchanged - whether a is a dictionary or a list, an
overwrite is not going to change the length. The special behaviour is
the [] cannot create a new item in a list and therefore raises an
exception, whereas it can create a new item in a dictionary so does.
Nothing that inheritance and overriding cannot handle.

Please compare like with like, and therefore consider what you really
mean by 'insert'.

If you are genuinely creating a new key in the dictionary, the nearest
equivalent for a list is the append() method which creates a new item
without disturbing other items key->value mappings. It is restricted
in the items it can create (the key must be equal to the previous
length of the list) but that is fine - the list is a more specialist
container and should fail if other keys are added. It is only a shame
that 'x[len(x)] = y' is invalid, as this would be a valid list
insertion in the same way that [] for any key on dictionaries is.

>> Insertion with modification of existing keys to make room (ie list
>> insertion) simply does not exist in lists. The usefulness of list
>> insertion derives from the fact that a list is a very specialised type
>> of key-to-value mapping - in this respect, the type 'list' should
>> clearly inheret from and extend the more general key-to-value type
>> called a dictionary.
>>
>It might be clear to you. It certainly isn't to me. I believe you are
>looking to unify two fundamentally different semantics.

They are not fundamentally different semantics - they are both
semantically key-to-value mapping containers and one is semantically a
specialised form of the other. Only the implementation is
fundamentally distinct.

>Neither does it make sense in implementation terms. List insertion's
>usefulness is in fact limited to situations where performance isn't unduly
>degraded by the need to modify the representation too radically. Of course,
>where performance doesn't matter, it doesn't matter.

Of course the usefulness of particular operations on certain types
depends on the implementation of those types. That does not make the
operations fundamentally different, though.

Incidentally, if an insert was applied to a dictionary which is
actually semantically equivalent to the list insert, it would be much
slower as each item to move would need to be removed, it's key
incremented and rehashed, and then the item re-inserted into the
dictionary. That is comparing like with like - fundamentally the same
semantics rather than just the same name.

Such an operation would be absurd so is reserved for the more
specialist list.

>> In essence, you are confusing the operations on the underlying data
>> structures with the operations on the conceptual types. These are two
>> separate concepts. The implementation is distinct from the
>> user-visible abstraction.
>>
>Fine. You write your programs to delete lists element-by-element from the
>front, and don't complain about performance -- that's just an aspect of the
>implementation.

Of course I don't do that. I am perfectly aware of performance issues,
just as you apparently are. However, your argument was that these
types are 'fundamentally different' and I don't believe they are.

>While I admire your attempt to maintain theoretical purity here, Python is a
>pragmatist's language. I don't think I'll be bothering to conceptualise
>dictionary insertion as overwriting a non-existent element, thank you very
>much.

The origin of this thread was the idea of supporting the items()
method for lists in the same way as it is for dictionaries instead of
adding an enumerate function. There is no relative performance hit
whichever notation is used. Being consistent in how key-to-value
mapping types represent their functionality has no performance cost at
all.

If there was a performance hit in what I was suggesting, I'd say that
pragmatics were in your favor. Without that performance hit, perhaps
pragmatics may have more to do with the learning curve.

Someone who has written something like...

  for key, value in l_Dict.items () :

might well try to do the same for a list. Why have two ways to do the
same thing? - apparently, because the implementation of the types
within the interpreter is different. Or perhaps because there are
differences in other operations which have nothing to do with this.

I really don't think that theoretical purity and pragmatics are in
conflict at all. There is no performance issue, but there is a
learning curve issue as two key-to-value mapping containers each have
different syntax to achieve exactly the same objective - despite the
fact that for many other operations (with fundamentally different
implementations) on these types, the syntax IS consist. This
inconsistency serves no pragmatic purpose whatsoever.

Funny thing though - integers and long integers are being unified
DESPITE the performance hit. Pragmatics and performance would suggest
that basic integers should be allowed to just overflow - it should be
the programmers responsibility to ensure it doesn't happen, just as in
many other languages. Maybe sometimes theoretical purity overrides
pragmatics - but that isn't the case in my argument.

Personally, I'm sticking with the argument from much earlier in this
thread (that list.items () intuitively suggests it returns the list
itself) as still the most convincing - though that just suggests to me
that there might be a better name which could be used for the
operation on both lists and dictionaries. Perhaps list.mappings() and
key.mappings(), which to me intuitively suggests that it returns a
list of key-to-value mappings exactly as intended.



From emile at fenx.com  Tue Jun 25 11:35:55 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 25 Jun 2002 15:35:55 GMT
Subject: Python daemon
References: <mailman.1024999953.15765.python-list@python.org> <F733C874CD014E16.148E57261FFD5E7A.AFFA4BD17B2F8D8B@lp.airnews.net>
Message-ID: <v50S8.307703$352.31615@sccrnsc02>

Cameron Laird
> This is one of the great FMMs under Unix--the under-utilization

FMMs ?

--

Emile van Sebille
emile at fenx.com

---------



From eric.brunel at pragmadev.com  Fri Jun 21 10:58:12 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Fri, 21 Jun 2002 14:58:12 +0000
Subject: CR+LF problem
References: <3d11bc33.1129927479@news.mad.ttd.net>
Message-ID: <aev7hp$8oa$1@wanadoo.fr>

Manuel Ruiz wrote:
> 
> Hi,
> 
> ------------------------
> .
> char * txt = "print 123\r\n";
> result = PyRun_Simple( txt, Py_file_input, PyDict_New(), NULL );
> .
> -----------------------
> 
> this code fails to exec txt, I got a SyntaxError exception,
> however I if save to file the statement "print 123\r\n" and try to run
> it ussing python interprete (version 2.2.1) It run fine.
> 
> Can anyone tell me anything about this rare behaviour

BTW, I already noticed this problem when using the "freeze" utility. If 
you're on Unix and you try to freeze a program using modules with 
DOS-style line endings, you end up with a SyntaxError...

I reported the bug (n# 464405), but for the moment, it's still open...
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From irmen at NOSPAMREMOVETHISxs4all.nl  Sat Jun  1 13:36:42 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Sat, 1 Jun 2002 19:36:42 +0200
Subject: Generating unique numbers?
References: <3CF7EBC8.4080802@verio.net> <3CF8FC1C.7060903@lindbergs.org>
Message-ID: <adb0na$c94$1@news1.xs4all.nl>

> Based on what people have said, I was thinking of using some 
> combination of DateTime.gmticks and the id of the obj, 
> making it (I think) extremely unlikely that there would be a 
> collision.


You might want to take a look at my GUID generator that is 
part of Pyro. It is used to generate object IDs to identify objects
within a distributed system. See http://www.sourceforge.net/projects/pyro

If you do download it, it can be found in Pyro/util.py

Irmen



From fredrik at pythonware.com  Thu Jun 20 03:50:24 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 20 Jun 2002 07:50:24 GMT
Subject: Variable Interpolation - status of PEP 215
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com>
Message-ID: <4PfQ8.44245$n4.10307683@newsc.telia.net>

"Ganesan R" wrote:
> I am a relative newbie to Python. I have a reasonably good perl background
> (don't worry I am not about to start a language flame war :-), and one of
> the things I really miss is variable interpolation in strings.

now, given that Python is a real programming language, you can
get all the interpolation you'll ever need with a few lines of code
in a utility module.

just cut and paste (this works with Python 2.0 and newer):

import sre, sys

# match $$ and $var and ${var}
_dosub = sre.compile(r'\$(?:\$|(\w+)|\{([^}]*)\})').sub

def expandvars(string, vars):
    # expand $var and ${var}; leave unknowns as is
    def repl(m, vars=vars):
        if not m.lastindex:
            return "$"
        try:
            return vars[m.group(m.lastindex)]
        except (KeyError, NameError):
            return m.group(0)
    return _dosub(repl, string)

def replacevars(string, vars):
    # same as expandvars, but raises an exception if variable not known
    def repl(m, vars=vars):
        if not m.lastindex:
            return "$"
        return vars[m.group(m.lastindex)]
    return _dosub(repl, string)

def replacevars_from_scope(string):
    # same as replacevars, but gets the variables from caller's current scope
    frame = sys._getframe(1)
    mapping = frame.f_globals.copy()
    mapping.update(frame.f_locals)
    return replacevars(string, mapping)

# (etc)

#
# try it out...

s = '${name} was born in ${country}'
print replacevars(s, {'name': 'Guido', 'country': 'the Netherlands'})

name = 'Barry'
country = 'the USA'
print replacevars_from_scope(s)

# (etc)

# </F>




From bokr at oz.net  Sun Jun 30 01:14:49 2002
From: bokr at oz.net (Bengt Richter)
Date: 30 Jun 2002 05:14:49 GMT
Subject: yield equivalent in C/JavaScript?
References: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk>
Message-ID: <afm449$4vp$0@216.39.172.122>

On Sat, 29 Jun 2002 19:12:46 +0100, Robin Becker <robin at jessikat.fsnet.co.uk> wrote:

>Is there any way to do yielding easily in C? I have a bunch of programs
>which are normally connected together using unix pipes. I would like to
>merge them and the problem is then how to get them to communicate nicely
>using a single process.

But this involves switching back and forth between the processing contexts
writing to and reading from the pipes, so unless you use available threading,
pipe/stream buffering, etc., seems like you are going to reinvent a bunch of it
in one form or another. However, ...

If it's your code and you know what kind of chunks it naturally processes in,
then you could just rework the calling sequence so that you can do one sensible
chunk at a time, using a state parameter to switch to the appropriate internal
functionality (e.g., initialise() vs. fill_input_buffer vs. process_input_buffer).
The output from a chunk-producer can then be the input to a chunk-consumer,
calling both in a loop something like, e.g.,

#define FINI 0
#define INIT 1

int main(){
    int producer_state = INIT;
    int consumer_state = INIT;
    ...
    while(producer_state || consumer_state){   /* exit when both FINI */
        /* maybe feed producer_inbuf here if it doesn't do it*/
        producer_state = producer(producer_state, producer_inbuf, producer_outbuf);
        consumer_state = comsumer(consumer_state, producer_outbuf, consumer_outbuf);
        /* maybe dump consumer_outbuf here if it doesn't do it */
    ...
}
Note that the processes return a state value so they effectively control their internal
sequencing, like a state machine, so it's quite flexible.

Producer and consumer would work something like (just a sketch):

int producer(int state, char* inbuf, char* outbuf){
    switch(state){
    case INIT: {
        /* allocate processing chunk buffers*/
        initialize();
        return CHUNKIN;            /* returns like this yield control. Data goes via buffers */
        }
    case CHUNKIN: 
        add_to_chunk_buffer();     /* while inbuf data is available and chunk is incomplete */
        if( chunk_is_complete() ){
            return BLOCK_ON_OUPUT;
        } else {
            return CHUNKIN;
        }
    case BLOCK_ON_OUPUT: /* poll for space to process inbuf to outbuf */
        if( output_buffer_empty() ){
            return DOCHUNK;
        } else {
            return BLOCK_ON_OUTPUT;
        }
    case DOCHUNK: {
        /* sufficient input exists to do reasonable processing, and output space
           has been checked (assuming you know chunk size is less that buffer size), so do it*/
        ...
        if( end_detected_somehow()){
            return FINI;
        } else {
            return CHUNKIN; /*get more input*/
        }
    ...
}

Doing this conveniently depends on being able to do it at the outermost scope inside the processes,
so that you can return from and resume into the main top state dispatching loop. Otherwise you have
stack state that will need to be saved and restored, and then you're re-inventing stuff again.

Hm, this doesn't look like Python ;-p

>
>the typical process currently looks like
>
>        initialise();
>        while((c=getc(stdin))>=0){
>                process(c);
>                putc(c,stdout);
>                }
>
Is process(c) just eavesdropping on a verbatim copy stream?

>so if the inputs are reasonable I could just run each to completion
>before proceeding with the next by replacing the file streams with
>memory ones. That could get messy if the inputs are large.
>
Can you get the processes to do limited-size chunks at a time, so producer
and consumer can alternate, and you can control that reasonably?

>Is there a neat way to simulate the yield idea? In C I can just about
>consider doing some longjump madness, but that wont wash in the other
>possible environment which has JavaScript for implementation.
You really mean JavaScript, not Java?

Regards,
Bengt Richter


From shalehperry at attbi.com  Fri Jun 28 21:12:41 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 28 Jun 2002 18:12:41 -0700 (PDT)
Subject: I'm an idiot
In-Reply-To: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16>
Message-ID: <XFMail.20020628181241.shalehperry@attbi.com>

> 
> f=open('c:\\temp\\temp.txt', 'r')
> g=open('c:\\temp\\temp1.txt', 'w')
> while 1:
>     try:
>         s=f.readline

should be 's = f.readline() # I am a method, not a variable'

>         g.write(s.split())

split() takes a string and breaks it into pieces, this is not exactly what you
want, for instance 'this is a line' will beceome ['this', 'is', 'a', 'line'].

>     except IOError:
>         break
> 
> g.close
> f.close

these too should have () on the end.




From mgilfix at eecs.tufts.edu  Sun Jun 16 12:31:23 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Sun, 16 Jun 2002 12:31:23 -0400
Subject: What If.....    Strong Types
In-Reply-To: <3D0C0F0D.4030900@bgb.cc>; from garrett@bgb.cc on Sun, Jun 16, 2002 at 04:08:26AM +0000
References: <3D0C0F0D.4030900@bgb.cc>
Message-ID: <20020616123122.B23809@eecs.tufts.edu>

On Sun, Jun 16 @ 04:08, Don Garrett wrote:
>    Primarily, what if classes always had rigidly defined interfaces. I mean 
> that public members had to be declared to exist, and that methods couldn't be 
> modified on an instance. Private members could exist but would be really private.

  Do you think you really gain anything by public/private
interfaces? What's wrong with the way python approaches
intefaces? Information hiding and interface exposure are done by
contract: e.g. When I say: instance.__var, __var is considered private
and cannot be gotten at by conventional means, but can still be gotten
at if a programmer is determined. And sometimes, it's *nice* to be
able to do so. There are times when you have a set, 'rigid' inteface,
which you know won't change, such as in a well-defined API, and you're
using a class which almost does what you need but not exactly. Then
this lack of public/private becomes a blessing because if you
understand how the underlying class works, you can get at information
without breaking the semantics of the class. But you still have the
advantage of private members - no one can use those members without
being very explicit and going out of their way to do so. In fact, I
often believe that the C++ 'protected' can often be what people really
want...  If you're really into a rigid interface, then make use of the
'__'. Same beneift, less hinderance.

                        -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From pyth at devel.trillke.net  Sun Jun 30 03:48:33 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 30 Jun 2002 09:48:33 +0200
Subject: Partial Parsing of Python
In-Reply-To: <20020629141507.GA6507@luminance.reworld.org>; from pde@cs.mu.oz.au on Sun, Jun 30, 2002 at 12:15:08AM +1000
References: <20020629141507.GA6507@luminance.reworld.org>
Message-ID: <20020630094833.B20310@prim.han.de>

Hi Peter,

You wrote:
> I'm currently working on a python-based alternative to a standard *nix
> shell.  I imagine lots of other people have built such animals, but this
> one is supposed to be graphical and include nifty features like smart
> auto-completion and running commentary on one's commands (which should be
> handy for newbies).

That's great! I can offer you code which will help you. 
Moreover some of us had good discussions at EuroPython about *exactly*
this type of project. For a start look at my alpha-version of a new 
command line completer:

    http://home.trillke.net/~hpk/rlcompleter2.py

Michael Hudson moreover has done nice work with replacing readline
with curses. 

    http://starship.python.net/crew/mwh/hacks/pyrepl.html

> Anyway, having implemented some of the completion features the simple 
> way-- using tokenize.tokenize(), string analysis, and the odd call to
> parser.expr() to guess what the user might want to type next, I've
> reached the conclusion that doing too much more will lead to Madness
> and Spaghetti Code.

completly correct. 

> So the alternative is of course to use a proper python parser, which 
> can include hooks suggesting how bits of the grammar can be completed.
> Basically, what this parser needs to do is attempt to accept its input,
> noting along the way when it meets the user's cursor.  If it fails,
> it needs to preserve an incomplete parse tree, record which production
> rule it was trying to use, and which symbol in that rule caused the
> problem.

Actually the problem goes deeper. Have a look at my TryParser class
in rlcompleter2.py for a start. It does handle many cases and should
give you some more ideas :-)

Actually the REAL problem to solve first is to have Multi-Line editing. 
Then all sorts of partial parsers *and* partial evaluation.  Python is
just too damn dynamic to have that one easily.

So what to do? Our idea at EuroPython was to use PyGame (the wrapper
around SDL) to get rid of the uglyness and restrictiveness of 
readline and curses alltogether.  We want to have a REAL playground 
and concentrate on our ideas.

Once i slow down from the mind-blowing EuroPython discussions and talks
i check how easy that is.

Drop me a note if you are interested to stay tuned.

    holger




From rajarshi at presidency.com  Fri Jun 28 14:49:34 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Fri, 28 Jun 2002 14:49:34 -0400
Subject: testing type of an object
Message-ID: <pan.2002.06.28.14.49.33.858327.1787@presidency.com>

Hi,
  I'm writing some code that depends on the type of an argument
passed to it.

Is there a standard trick to do this? My versions don't seem to work:

if type(s) == 'str':
  do something

or

if string.find(type(s), 'str') != -1:
  do something

dont seem to work

TIA,


From max at alcyone.com  Sun Jun 23 15:08:03 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 23 Jun 2002 12:08:03 -0700
Subject: bloody freshman question regarding strings
References: <mailman.1024821149.18077.python-list@python.org> <3D15A173.9010904@mxm.dk>
Message-ID: <3D161C93.568B299@alcyone.com>

Max M wrote:

> A few approaches:
> 
> def mailmsg(severity, message, opts, target, recipient):
>      # first create the message body
>      msgbody = "computer: %s\n" % COMPUTER)
>      "system:   %s\n" % SYSTEMTYPE)
>      "cpu:      %s\n" % CPUTYPE)
	...

Huh?  Syntax errors.

> def mailmsg(severity, message, opts, target, recipient):
>      # first create the message body
>      bodyParts = []
>      a = body.append

You meant a = bodyParts.append.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ See the son in your bad day / Smell the flowers in the valley
\__/ Chante Moore
    Bosskey.net: Aliens vs. Predator 2 / http://www.bosskey.net/avp2/
 A personal guide to Aliens vs. Predator 2.


From pyth at devel.trillke.net  Tue Jun 11 10:34:21 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 11 Jun 2002 16:34:21 +0200
Subject: Filling a space...
In-Reply-To: <ae4v9h$3qg34$9@ID-108351.news.dfncis.de>; from thor__00@yahoo.com on Tue, Jun 11, 2002 at 03:25:35PM +0200
References: <ae4v9h$3qg34$9@ID-108351.news.dfncis.de>
Message-ID: <20020611163421.S6609@prim.han.de>

Thor wrote:
> Having this tupe: (2,3,4), which one would be the msot efficient way to get 
> this (parentheses ommited):
> 
> 0,0,0  0,0,1  0,0,2  0,0,3
> 0,1,0  0,1,1  0,1,2  0,1,3
> 0,2,0  0,2,1  0,2,2  0,2,3
> 1,0,0  1,0,1  1,0,2  1,0,3....
> 
> I mean, to get all the possible integer points in a defined space.

i hope you are going to cope well with all the answers you will certainly get.
i can't resist either to present a >=python2.2 solution :-)

def allpoints_generator(spectuple):
    aslist = list(spectuple)
    if aslist:
        for i in xrange(aslist.pop(0)):
            if aslist:
                for sub in allpoints_generator(aslist):
                     yield (i,)+sub
            else:
                yield (i,)

for i in allpoints_generator( (1,2,3) ):
    print "result", i

result (0, 0, 0)
result (0, 0, 1)
result (0, 0, 2)
result (0, 1, 0)
result (0, 1, 1)
result (0, 1, 2)

have fun,

    holger




From peter at engcorp.com  Sun Jun  9 03:04:09 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 09 Jun 2002 03:04:09 -0400
Subject: Simple pychecker question
References: <mailman.1023482712.8801.python-list@python.org> <pan.2002.06.09.00.07.47.445563.6814@metaslash.com>
Message-ID: <3D02FDE9.19198F33@engcorp.com>

Neal Norwitz wrote:
> 
> I personally think that superfluous use of named args are bad,
> but that the feature is good when used properly.  For example,
> with your code above, calling f(f=2) would be better written
> f(2).  However for Tkinter, named args can be useful when there
> are a bunch of parameters and you only want to use the last.

For a routine named and structured like this:

  def transmit(data, timeout=1.0):

which of the following would you say was the better call
in some code that might be remote from that definition
above?

1.  transmit('this is a string', 5)

2.  transmit('this is a string', timeout=5)

My point: named arguments can increase readability and 
maintainability.  That's not superfluous, to me.

-Peter


From SBrunning at trisystems.co.uk  Tue Jun 25 11:06:45 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 25 Jun 2002 16:06:45 +0100
Subject: Python and Eclipse
Message-ID: <31575A892FF6D1118F5800600846864DCBD438@intrepid>

> From:	Thomas Guettler [SMTP:zopestoller at thomas-guettler.de]
> Sent:	Tuesday, June 25, 2002 4:02 PM
> Can someone tell me what eclipse is about?
> 
> Someone said it is IBM's intend spread the use of
> WebSphere and someone else said it is an IDE.
 
Hmmm. It's basically a framework for building IDEs, and other things. It
ships with a plug in which makes it into a Java IDE, which as I said, is
superb. IBM is using the Eclipse framework for the next generation of its
WebSphere Studio tools, and that is what IBM are in it for. But other tools
vendors are also basing tools on Eclipse.

I cover Eclipse quite a lot in my web log,
<http://www.brunningonline.net/simon/blog/>.
 
Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.




From peter at engcorp.com  Mon Jun 17 21:16:31 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 21:16:31 -0400
Subject: Pronouncing '__init__'
References: <mailman.1024136779.31270.python-list@python.org>
Message-ID: <3D0E89EF.CB6C4CB7@engcorp.com>

Oren Tirosh wrote:
> 
> There has been a thread about this issue in the past. Below are some of the
> proposals people made.  I'd like to have an informal vote - how do *you*
> say Python's magic names when talking with other Pythoneers?
> 
>  (all my Python-related communication is written)
>  double underscore init double underscore
>  under under init under under
>  double bar init double bar
>  bar bar init bar bar
>  unders init unders
>  under init under
>  bar init bar
>  magic init
>  pyinit
> 
> Most of these pronounciations also have a shorter version with the trailing
> part omitted.

I believe, and I'm not going to take the few seconds to check right now :-),
that the documentation refers to __init__ and friends as "special methods",
so I always just say "the init special method" or "the getitem special method"
if there's any possible confusion, and just "init" or "getitem" without
trouble in all other cases.

Rarely I have to haul out "double underscore X double underscore" but
that's usually no more than once per person... after that the above
techniques usually work.

-Peter


From peter at engcorp.com  Sat Jun 29 12:32:50 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 29 Jun 2002 12:32:50 -0400
Subject: private
References: <pan.2002.06.29.17.38.51.789527.440@myself.com>
Message-ID: <3D1DE132.4A890D9D@engcorp.com>

Rhymes wrote:
> 
> Is there any possibility to build up a _real_ private attribute?

Please define what "_real_" means to you.  It's not apparent.

> Such as C++ private data members...

You realize that it's quite possible to get at private data members
in C++, don't you?

> I know that with double-underscore prefix (and the name mangling related)
> I can " hide " data members but I think it's not a so crafty trick...
> Is there any PEP about it? Will Guido insert a private keyword
> in a future release of Python ?

This has been discussed about 479 times in the last decade.  Please
check the archives.

-Peter


From aoeu at go.com  Mon Jun 17 07:21:07 2002
From: aoeu at go.com (Thinkit)
Date: 17 Jun 2002 04:21:07 -0700
Subject: hexadecimal advocacy
Message-ID: <3eeda89d.0206170321.9afc3e5@posting.google.com>

Try to use "0x" whenever you can (that is, integers above nine).  Sexadecimal rules.


From tim at lesher.ws  Thu Jun  6 10:39:35 2002
From: tim at lesher.ws (Tim Lesher)
Date: 6 Jun 2002 07:39:35 -0700
Subject: Redirecting stdout in rexec
Message-ID: <b6562547.0206060639.5bb24474@posting.google.com>

[Python 2.1.1, Linux]

I'm having trouble redirecting output from an rexec call to my own
string.  I want to use commands of the form "print 'foo'" in the
rexec'ed code, and capture them into a string object in the caller. 
This is roughly what I'm doing:
-----
import rexec

class StringCapture:
    def __init__(self):
        self.buf = ""
    def write(self, str):
        self.buf += str

capture = StringCapture()

renv = rexec.RExec()
module = renv.add_module("sys")
module.__dict__["stdout"] = capture

renv.s_exec("print 'Hello World'")

print "Output is: %s" % capture.buf
-----

It doesn't appear that the replacement of stdout is being seen by the
s_exec, although the same procedure works if I replace the local
stdout.  Is there some additional step I need to do to use this with
rexec?

Thanks.

-- 
Tim Lesher
tim at lesher.ws


From peter at engcorp.com  Sat Jun  1 22:03:05 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 01 Jun 2002 22:03:05 -0400
Subject: File Iterator - Print - Double New Lines
References: <mailman.1022978163.23774.python-list@python.org>
Message-ID: <3CF97CD9.5C5BB2DF@engcorp.com>

Mike Brenner wrote:
> 
> Peter wrote:
> > Mike, the print statement adds newlines after every line it prints.
> > If you want to see the real contents of the file, use sys.stdout.write().
> > Alternatively, you might want to strip the newlines from each line
> > as you read them in.  Depends what you are trying to do.

> That is not what happens, though. The first three outputs (from the first call 
> to reader) do not get newlines.  The second three outputs (from the second call 
> to reader) get double newlines.  The explanation that "the print statement adds 
> newlines after every lines it prints" would explain the double newlines the 
> second time, but it does not explain the lack of newlines the first time. 
> Did you run it on a windows machine?

Mike, I'm right, so study this before you respond again. <wink>

> def writer(list):
>     f=open("tmp.tmp","w")
>     for item in list:
>         f.write(item)
>     f.close()

Note that this routine never adds newlines.  Each item in the list
is written straight out as-is.  I'm glad you called them "items"
and not "lines", because they are not lines, as you'll see.

> def reader():
>     f=open("tmp.tmp","r")
>     for line in f:
>         print line
>     f.close()

Note that this routine reads "lines" which are by definition
sequences of bytes terminated with newlines, or the remaining
bytes terminated by EOF (if there's no final newline).

*Each line* printed by the print statement has a newline appended.
But you have only a single line in the first file!  See below for more 
on this.

> print "trying it without the \\n"
> writer(["a","b","c"])
> reader()

Here you are writing out three bytes, without any newlines.
When you read it back in, you get a single line of three bytes, 
again without even a newline, and then you print it (with one
newline appended by print).

> print "trying it with the \\n"
> writer(["a\n","b\n","c\n"])
> reader()

Here you deliberately force a newline after each byte, so
the file has the newlines present.  When you read with 
"for line in f", Python automatically splits the input
up into "lines", each one with a newline attached, because
there are now newlines in the file.

Then you print each of the *three* lines, with their
integral newlines, followed by the newline that print
appends.

To help you understand this, you might try examining
the files that are created with a binary editor.  You
could also change to using sys.stdout.write() which does
nothing funny with the output like print does.  You could
also experiment with repr() around the data, so you can
see the embedded newlines and such.

Cheers,
-Peter


From oren-py-l at hishome.net  Mon Jun 17 02:49:17 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Mon, 17 Jun 2002 02:49:17 -0400
Subject: Pronouncing '__init__' - summary of replies
Message-ID: <20020617064917.GA51668@hishome.net>

1. Implied by context
2. "under under", "double under"
3. "special", "magic"

__oren__




From peter at engcorp.com  Wed Jun 19 21:11:02 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 19 Jun 2002 21:11:02 -0400
Subject: Flame fests (was Re: python version?)
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <3D1010C3.C321F8B0@engcorp.com> <raXP8.47926$GY.15260255@twister.nyroc.rr.com> <3D107948.B3FEE4BB@engcorp.com> <aer53s$7is$1@panix1.panix.com>
Message-ID: <3D112BA6.4CA5F584@engcorp.com>

Aahz wrote:
> 
> Everybody have fun playing "kick the newbie"?  Maybe it's time to quit?

I think we'd all object strongly to the characterization that we were 
"kicking a newbie".

For one thing, any kicking done was entirely unrelated to whether or not 
he was a newbie.

For another thing, he has basically denied being interested in learning
about Python, which I believe makes the label "newbie" inappropriate.

> (Yes, yes, I've read the thread -- I still think it's time to revert
> c.l.py to its usual habits.)

-Peter


From jon+usenet at unequivocal.co.uk  Tue Jun 25 11:35:00 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Tue, 25 Jun 2002 15:35:00 -0000
Subject: HTML Error from Python CGI Script
References: <79b54d28.0206241618.6c78dbde@posting.google.com>
Message-ID: <slrnahh3d3.sar.jon+usenet@snowy.squish.net>

In article <79b54d28.0206241618.6c78dbde at posting.google.com>, Chris wrote:
> I copy this exact text to my own computer (versus operating over
> network) and open it and it looks fine. Exactly what I need. However,
> when I put a link to it on another webpage on the network, an internal
> server error occurs and complains about the following:
> 
> syntax error at line 2: `newline or ;' unexpected

Obviously your server is trying to use your static page as some kind
of script. You haven't given enough information to work out why this
might be though.


From Chris.Barker at noaa.gov  Tue Jun 25 12:41:43 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 25 Jun 2002 09:41:43 -0700
Subject: How to represent the infinite ?
References: <mailman.1024718075.18074.python-list@python.org> <slrnahfl2u.2s7.huaiyu@gauss.almadan.ibm.com>
Message-ID: <3D189CE5.4E05FA10@noaa.gov>

Huaiyu Zhu wrote:
> Before that happens (and I understand that is going to be a long time), I'd
> like to see overflow exception raised only once for each vector, with proper
> information in the exception denoting which element encountered it first.
> This is not likely to upset those who want to see the exceptions, while
> making it much easier to ignore the exceptions if one wants.  I'm not sure
> whether this is a core Python issue or NumPy issue, though.

I think it's a NumPy issue. However, the hope for the future is that
NumPy (actually NumArray) will be part of the standard library someday
(it really should be, a whole lot of people could benefit from it's use
that don't think they re doing "serious" number crunching). If it is,
and even if it is not, the better the computability between NumPy and
core Python the better. Having inf, -inf and NaN is critical for array
operations, but it can be very useful for all numerical work, I really
wish it reliable existed in Python.

I have to say it is very annoying that the vast majority of hardware has
been IEEE 754 compliant for years, and yet there is very little support
in compilers and libraries...aaarrggg!

by the way, The Mathworks seems to be able to provide a reasonable
subset for MATLAB on many platforms..are they just dealing with a
maintenance nightmare?

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From andersjm at inbound.dk  Sat Jun 29 12:11:31 2002
From: andersjm at inbound.dk (Anders J. Munch)
Date: Sat, 29 Jun 2002 18:11:31 +0200
Subject: Dialing out on MODEM (Partially Resolved)
References: <mailman.1025269958.7325.python-list@python.org>
Message-ID: <zalT8.98278$N46.2491998@news010.worldonline.dk>

"David McInnis" <david at dataovation.com> skrev i en meddelelse
news:mailman.1025269958.7325.python-list at python.org...
> Peter,
>
> Thanks for the tip.  I forgot that I am on Windows and needed \r\n
> rather than just \n

Actually this is not a Windows issue, it's a modem issue.  Regardless of
operating system you need \r\n to speak to a Hayes compatible modem.

- Anders





From SBrunning at trisystems.co.uk  Tue Jun 25 08:22:02 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 25 Jun 2002 13:22:02 +0100
Subject: Python and Eclipse
Message-ID: <31575A892FF6D1118F5800600846864DCBD42E@intrepid>

> From:	claird at starbase.neosoft.com [SMTP:claird at starbase.neosoft.com]
> Can you say a few more words about Eclipse?  It ... well, I'm
> wary.  I understand its marketing more than its technology,
> and on hostile days, I think it's *only* marketing.  Maybe you
> can make it more appealing.  "[It] nearly allows me to *enjoy*
> writing Java ..." is certainly a good start.
 
I use Eclipse as my day-to-day Java IDE. It is just superb. If you need to
hack on Java, give it a try. I promise you, it's worthwhile.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.




From pyth at devel.trillke.net  Tue Jun  4 17:02:45 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 4 Jun 2002 23:02:45 +0200
Subject: self
In-Reply-To: <TL7L8.79$4p7.4100181@newnews.cc.stevens-tech.edu>; from vjovanov@stevens-tech.edu on Tue, Jun 04, 2002 at 02:35:24PM -0400
References: <mailman.1023208220.19083.python-list@python.org> <TL7L8.79$4p7.4100181@newnews.cc.stevens-tech.edu>
Message-ID: <20020604230245.B30696@prim.han.de>

Vojin Jovanovic wrote:
> Now let me give you the problem with self which is making my life
> complicated. Consider this class.
> 
> class Foo:
>  def __init__(self):
>   self.__dict__['Equations']={
>     'a':str(5),
>     'b':'self.a+5',
>     'c':'self.b*self.a'}
> 
> (...)

To get rid of 'self' you might like to try this:

class Evaluator:
    def __init__(self, eqs={}): 
        self.__dict__['equations']=eqs
    def __getattr__(self, name):
        begin={}
        while 1:
            try:
                return eval(self.equations[name], begin)
            except NameError, n:
                var=str(n).split("'")[1]
                begin[var]=getattr(self, var)
    def __setattr__(self, name, value):
        self.equations[name]=value

>>> e=Evaluator({'a':'3', 'b':'a*17', 'c': 'b*a'})
>>>
>>> e.c
153
>>> e.f='a*c'
>>> e.f
459

I don't really think that 'execing' is a good idea here
but i don't know your complete usecase.

With a little more care it should be possible to
catch cyclic equations. 

have fun,

    holger




From fredrik at pythonware.com  Thu Jun 13 13:30:01 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 13 Jun 2002 17:30:01 GMT
Subject: For and Dates
References: <mailman.1023983824.29281.python-list@python.org>
Message-ID: <tE4O8.12003$p56.3680034@newsb.telia.net>

"Marcus Vinicius Laranjeira" wrote:

> I need to do a loop that goes from 2002-6-1 to 2002-7-31, but I don't know
> how to do that...
>
> Is there a way to do such range using for !?
>
> Do you have any suggestions !?

if you can live with a while-loop, here's one (rather
simple-minded) way to do it:

    import time

    now = time.mktime((2002, 6, 1) + (0, 0, 0, 0, 0, 0))
    end = time.mktime((2002, 7, 31) + (0, 0, 0, 0, 0, 0))

    while now <= end:
        y, m, d = time.localtime(now)[:3]
        # do something with y, m, d
        print y, m, d
        now = now + 86400

here's another one:

    import calendar

    def tomorrow((year, month, day)):
        weekday, mdays = calendar.monthrange(year, month)
        day = day + 1
        if day > mdays:
            day = 1
            month = month + 1
            if month > 12:
                month = 1
                year = year + 1
        return year, month, day

    now = 2002, 6, 1

    while now <= (2002, 7, 31):
        y, m, d = now
        # do something with y, m, d
        print y, m, d
        now = tomorrow(now)

if you really need to use a for-in loop, and you're using
Python 2.2, it's straightforward to convert either of these
to use a generator instead:

    from __future__ import generators

    ...

    def dayrange(now, end):
        while now <= end:
            yield now
            now = tomorrow(now)

    for y, m, d in dayrange((2002, 6, 1), (2002, 7, 31)):
        # do something with y, m, d
        print y, m, d

</F>




From emile at fenx.com  Mon Jun 24 06:54:17 2002
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 24 Jun 2002 10:54:17 GMT
Subject: mail-problem over t-online
References: <mailman.1024859369.26585.python-list@python.org> <3D16DB2A.7050709@ivs.tu-berlin.de>
Message-ID: <tTCR8.306174$cQ3.16677@sccrnsc01>

"Andreas Ulbrich" <ulbi at ivs.tu-berlin.de> wrote in message
news:3D16DB2A.7050709 at ivs.tu-berlin.de...
> ZRZ is simply not allowing sending messages using SMTP from networks

Ahh... they don't allow relaying...

> outside the university network. As long as you connect directly
> university everything is fine. As soon as you move to t-online, zrz
> won't allow sending.

... so you'll need to use t-online's relay or send directly, except
you'll then have problems sending to ISPs that filter for 'appropriate'
senders.  AOL does this, as do more over time.



--

Emile van Sebille
emile at fenx.com

---------



From max at alcyone.com  Wed Jun 26 14:18:39 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 26 Jun 2002 11:18:39 -0700
Subject: Python 2.2 __slots__ Bug?
References: <mailman.1025089474.17780.python-list@python.org>
Message-ID: <3D1A057F.A81FF84@alcyone.com>

Glyph Lefkowitz wrote:

> I get "I'm fine" and then a segmentation fault.  This only happens
> when
> __slots__ is defined on a new-style class.  Seriously, i'm not making
> those
> numbers up -- reliably, 43551 objects will crash the interpreter;
> fewer will
> not.

Hmm, the script runs fine on my Slackware 8.0 system, even if I crank
the number up to 100000.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ See the son in your bad day / Smell the flowers in the valley
\__/ Chante Moore
    Bosskey.net: Aliens vs. Predator 2 / http://www.bosskey.net/avp2/
 A personal guide to Aliens vs. Predator 2.


From peter at engcorp.com  Sun Jun  9 14:12:03 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 09 Jun 2002 14:12:03 -0400
Subject: How to capture stdout into a string?
References: <2YLM8.905$Yw3.119357@news20.bellglobal.com> <3D039285.72DFC71D@engcorp.com> <moMM8.91$Vr2.130316@news20.bellglobal.com>
Message-ID: <3D039A73.210CDF7F@engcorp.com>

Michael Davis wrote:
> 
> Peter Hansen wrote:
> [...]
> > The best option by far (and really the easiest) to figure out
> > stuff like this is to check the source.
> >
> Thanks again, Peter. I would have accepted the answer RTFM too - I realized
> that I hadn't read the FTP.dir() documentation carefully enough!

:-)  Actually, I assumed you already had RTFM, and my comment above 
was really meant to be "best option by far (assuming you've read the 
docs) ...."  I didn't bother checking the manual to see if it said
the same.

Still helpful sometimes to read the source.  Never know when you'll pick
up a useful turn of phrase or two.

-Peter


From deathtospam43423 at altavista.com  Mon Jun 10 15:13:40 2002
From: deathtospam43423 at altavista.com (netvegetable)
Date: 10 Jun 2002 19:13:40 GMT
Subject: Sydney Australia: Python Interest Group
References: <M9VM8.3341$06.443812@nasal.pacific.net.au>
Message-ID: <Xns922A30DD181E8AusYourStandingInIt@130.133.1.4>

Mary Gardiner <usenet-jun02 at puzzling.org> wrote in 
news:M9VM8.3341$06.443812 at nasal.pacific.net.au:

> Hi,
> 
> A Python Interest Group has just started in Sydney under the banner of
> the Sydney Linux Users Group (SLUG).

Beautty!

Are you starting one in adelaide as aswell? 
> 
> It's not restricted to Python under Linux and it is not restricted to
> SLUG members or people involved in SLUG.
> 
> It meets on the third Monday of every month. Our second meeting is on
> June 17th at 7:00pm at the University of Technology, Sydney, room
> 2.3.16.
> 
> The announcement is at
> http://lists.slug.org.au/archives/announce/2002/06/msg00003.html
> 
> We don't have a webpage or mailing list independent of SLUG at present,
> if you're interested in further announcements you should watch
> http://www.slug.org.au/ or subscribe to SLUG's low volume announce list
> at http://lists.slug.org.au/listinfo/announce
> 
> -Mary.
> 



-- 
netvegetable at excite.com


From skip at pobox.com  Wed Jun 12 12:53:21 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 12 Jun 2002 11:53:21 -0500
Subject: setup.py breaks Modules/Setup.local
In-Reply-To: <ul3bsagsbke.fsf@smaug.fsl.noaa.gov>
References: <ul3bsagsbke.fsf@smaug.fsl.noaa.gov>
Message-ID: <15623.31873.924232.386915@12-248-41-177.client.attbi.com>

    Mike> I am attempting to get Modules/Setup.local working to build python
    Mike> 2.2.1.  But there seems to be a bad interaction with setup.py.  I
    Mike> have a Modules/Setup.local which contains a definition for the
    Mike> _tkinter module.  I would like to use a Tcl/Tk release from a
    Mike> location that setup.py does not find.  It looks like this:

    Mike> # Edit this file for local setup changes
    Mike> *shared*

    Mike> _tkinter _tkinter.c tkappinit.c -DWITH_APPINIT \
    Mike>         -L/usr/local/tcltk-8.3.2/lib \
    Mike>         -I/usr/local/tcltk-8.3.2/include \
    Mike>         -I/usr/X11R6/include \
    Mike>         -ltk8.3 -ltcl8.3 \
    Mike>         -L/usr/X11R6/lib \
    Mike>         -lX11

I've not tried this myself, but I think if you set the LIBDIR and INCLUDEDIR
environment variables you should be good to go:

    export INCLUDEDIR=/usr/local/tcltk-8.3.2/include
    export LIBDIR=/usr/local/tcltk-8.3.2/lib
    ./configure 
    make

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html






From denpeterson at yahoo.com  Sun Jun  2 20:06:27 2002
From: denpeterson at yahoo.com (Dennis Peterson)
Date: Sun, 2 Jun 2002 20:06:27 -0400
Subject: What is a classmethod for?
References: <addgv30f4d@enews4.newsguy.com> <addlm6$g44$1@panix1.panix.com>
Message-ID: <adebub0mdl@enews1.newsguy.com>

> The point is that you can call class methods *without* creating an
> instance:

>C.foo()

But you can do that with a static method:

class C:
    def foo(x):
        print x
    foo = staticmethod(foo)

C.foo(1)

...much like Java static methods. I see the point there, I'm just trying to
figure out the purpose of the implicit class object in classmethod.

-Dennis






From akakakhel at attbi.com  Wed Jun 26 14:55:12 2002
From: akakakhel at attbi.com (Ali K)
Date: Wed, 26 Jun 2002 18:55:12 GMT
Subject: geting python files for use on web
Message-ID: <k6oS8.191997$6m5.159601@rwcrnsc51.ops.asp.att.net>

I would like to know where I can download the nessecary files for usin
python in ASP and/or CGI.  Please reply.  Thank you.




From pinard at iro.umontreal.ca  Wed Jun 19 09:46:55 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 19 Jun 2002 09:46:55 -0400
Subject: procmail replacement in Python
In-Reply-To: <20020619022110.GA6996@lilith.my-fqdn.de>
References: <20020618235901.GA2197@lilith.my-fqdn.de>
	<oqit4grq03.fsf@carouge.sram.qc.ca>
	<20020619022110.GA6996@lilith.my-fqdn.de>
Message-ID: <oqelf3s7n4.fsf@carouge.sram.qc.ca>

[Gerhard H?ring]

> > What is `getmail'?

> A fetchmail replacement, written in Python.

Would you kindly give me an URL for `getmail'?
-------------- next part --------------

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard

From jepler at unpythonic.net  Mon Jun  3 22:32:46 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Mon, 3 Jun 2002 21:32:46 -0500
Subject: ?Tcl's embedded virtual filesystem and Scripted documents
In-Reply-To: <3CFBF133.7DF60BA4@email.sps.mot.com>
References: <3CFBF133.7DF60BA4@email.sps.mot.com>
Message-ID: <20020603213241.A1167@unpythonic.net>

On Mon, Jun 03, 2002 at 03:44:03PM -0700, Norman Shelley wrote:
> Tcl now has an embedded virtual filesystem that looks very useful.
> http://mini.net/tcl/2138.html
> 
> Is there any motivation in the Python camp to have something similiar?

Since files in Python are objects with methods (read, write & friends), all
you need to do is have a factory for file-like objects.  __builtin__.open()
uses only the real filesystem namespace, of course, but it could
potentially be replaced by another sort of factory...

I don't know if this new tcl gimmick supports stuff like 'chdir
ftp://....', but if so then chdir() would have to be replaced as well.

The following assumes posix-style names, and uses "//XXX/path..." to call
the XXX factory with the argument "path...".  Names like "/XXX/path/" still
have their normal meaning.

import __builtin__

builtin_open = __builtin__.open

filetype_registry = {}

def register(name, factory):
    filetype_registry[name] = factory

def open(path):
    initial_slashes = path.startswith('/')
    if (path.startswith('//') and not path.startswith('///')):
        initial_slashes = 2
    if initial_slashes != 2: return builtin_open(path)
    nextslash = path.find("/", 2)
    if nextslash == -1: nextslash = len(path)
    factory_name = path[2:nextslash]
    arg = path[nextslash+1:]
    factory = filetype_registry[factory_name]
    return factory(arg)

def _test():
    import StringIO
    def factory(path):
	print "in test factory"
	return StringIO.StringIO(path)
    register("test", factory)
    f = open("//test/this is a test")
    print "the contents of the file:", `f.read()`
    f.close()
    f = open("/etc/issue")
    print "the contents of the file:", `f.read()`
    f.close()

if __name__ == '__main__': _test()




From sholden at holdenweb.com  Mon Jun  3 20:59:37 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 3 Jun 2002 20:59:37 -0400
Subject: Time
References: <mailman.1023144005.19450.python-list@python.org>
Message-ID: <ZmUK8.205609$%u2.66972@atlpnn01.usenetserver.com>

"David LeBlanc" <whisper at oz.net> wrote in message
news:mailman.1023144005.19450.python-list at python.org...
> You might want to check out mxDateTime in the mx toolkit at
> http://www.egenix.com/files/python/. I don't see a direct +/-day function
> but there is access to the elements of a datetime such that you could
> add/subtract an integer to the day value
>
> David LeBlanc
> Seattle, WA USA
>
> > -----Original Message-----
> > From: python-list-admin at python.org
> > [mailto:python-list-admin at python.org]On Behalf Of Gold Fish
> > Sent: Monday, June 03, 2002 14:32
> > To: python-list at python.org
> > Subject: Time
> >
> >
> > Anyone know how to determine the time in python. If the mtime in the
file
> > filename.py is Mon Jun 3 03:25:57 2002 ,if i want to set the time
forward
> > or backwards in 5 days time, how can i do that.
>

The "mx" way to do what GoldFish wants is to create a
RelativeDateTime(day=5) and subtract it from the current date.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From kragen at pobox.com  Wed Jun 26 14:21:24 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 26 Jun 2002 14:21:24 -0400
Subject: Inter Process Communication with Numeric arrays
References: <wx1yb2uvt3.fsf@velocita.nasoftware>
Message-ID: <83eletlx3v.fsf@panacea.canonical.org>

Ian McConnell <ian at nasoftware.co.uk> writes:
> I'm writing both processes (though one might be in C) so ORB/CORBA seems to
> be overkill and I can't get through to the memory mapped files at
>         http://pobox.com/~kragen/sw/arrayfrombuffer/
> (Internal server error)

My previous message on the subject seems not to have made it to this
thread, but I fixed the error.

Vmaps is more oriented toward IPC than arrayfrombuffer, but
arrayfrombuffer works for IPC, at least on Linux.  arrayfrombuffer
will help you get the data back and forth quickly and conveniently,
but provides no synchronization primitives, so you might have to use
other means (such as pipes, sockets, or filesystem access) for
synchronization.  Vmaps, which I haven't tried, has synchronization
primitives.


From gerhard at bigfoot.de  Fri Jun 14 10:21:23 2002
From: gerhard at bigfoot.de (Gerhard Haering)
Date: 14 Jun 2002 14:21:23 GMT
Subject: Python and ADSI
Message-ID: <slrnagjuud.15g.gerhard@lilith.my-fqdn.de>

Does anybody have an example script using win32com and ADSI to access the users
in an active directory?

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From qigong42 at hotmail.com  Wed Jun  5 22:04:20 2002
From: qigong42 at hotmail.com (qigong42 at hotmail.com)
Date: Thu, 06 Jun 2002 02:04:20 GMT
Subject: NEWBIE - SendKey equivalent in Python
Message-ID: <lrgtfus01fgmc3sd49djos91812qj91hhl@4ax.com>

Is there a way to send keystokes to another application with Python?

If so, can someone post an example?

Thanks for the help.


From emile at fenx.com  Sun Jun 23 21:19:18 2002
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 24 Jun 2002 01:19:18 GMT
Subject: File IO read returning a string -> but I want a byte -> Question: Is there a faster more efficient way?
References: <3f8faafc.0206231627.570b8602@posting.google.com>
Message-ID: <qsuR8.302444$cQ3.16132@sccrnsc01>

Jay
> #Question: Is there a faster more efficient way?
> 
> from time import clock
> 
> len_data = 20*1024*1024
> data = fp.read(len_data)
> 
> start = clock()
> for i in range(len_data): 
>    norm[i] = struct.unpack("B",data[i])
> finish = clock()

map(ord,data)

--

Emile van Sebille
emile at fenx.com

---------



From whisper at oz.net  Thu Jun 27 19:37:27 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 27 Jun 2002 16:37:27 -0700
Subject: Pythoniac: Thoughts on a hardware Python processor
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F06233015DEC3B@mail.hebmex.com>
Message-ID: <GCEDKONBLEFPPADDJCOEMEJMDMAA.whisper@oz.net>

> -----Original Message-----
> From: Gustavo Cordova [mailto:gcordova at hebmex.com]
> Sent: Thursday, June 27, 2002 15:45
> To: David LeBlanc; Python List
> Subject: RE: Pythoniac: Thoughts on a hardware Python processor
>
>
> >
> > Whoever asked if the Intel/AMD FP stack counted: nah - it's
> > not the primary mode of the CPU like it was on the HP 3000
> > and the Forth chip. BTW, speaking of the Forth chip, I think
> > Forth, Inc. is still offering the IP for that if anyone wants
> > to burn a few hunderd k$ on new silicon.
> >
>
> If I recall correctly, Chuck Moore's still churning out
> forth processors, with on-chip fast memory and hardware
> stacks, really really cool stuff. I think there's more info
> in the ColorForth site, or at least it points you in the
> right direction.

Not when I looked 4 months ago, and not just now when I looked at
http://www.colorforth.com/:
"25x Multicomputer Chip
Spectacular chip! 25 microcomputers, each with 512 words of 18-bit memory.
Each capable of 2400 Mips, sustained. Awaiting funding. " Site last updated
7/2001.

> > While I think that creating a Python chip is a great and
> > wonderful idea, I doubt it will or should happen:
> >
> > [... snip ...]
> >
> > For all these reasons and more, I think a good compiler with
> > a retargetable backend is a better expenditure of time/effort
> > and would yield a result far faster then a silicon development effort.
> >
> > Dave LeBlanc
> > Seattle, WA USA
> >
>
> That's what I meant!! This compiler you talk about should generate
> Forth instead of the normal byte code. Hmmm... I think it's doable,
> but I've yet to learn more forth.
>
> Do you know any good resources? Online books? Good archives?
> Anything?
>
> :-)
>
> -gus

You would do better to worry about the front end before you worry about the
backend, but if you insist, then http://www.forth.org/fig.html is the place
to start. http://www.forth.com is less useful.

Dave LeBlanc
Seattle, WA USA





From dubreus at csit.fsu.edu  Tue Jun 11 11:45:59 2002
From: dubreus at csit.fsu.edu (mitsch)
Date: 11 Jun 2002 08:45:59 -0700
Subject: compiling error:   ELF file's phentsize not the expected size
Message-ID: <93523ebb.0206110745.61157ef4@posting.google.com>

I try to wrapp c++ with python using the boost libraries.

but I got some errors, when trying to import the .so created.


Traceback (most recent call last):
  File "./boost.py", line 2, in ?
    import getting_started1
ImportError: ./getting_started1.so: ELF file's phentsize not the
expected size


can somebody help?



here's my compilation command:
g++3 getting_started1.cpp  -c -fPIC -shared  -o getting_started1.so
-I/usr/local/lib/python/boost_1_28_0/ -I /usr/include/python2.1/


here's my getting_started1.cpp file:
#include <boost/python/class_builder.hpp>
namespace python = boost::python;
#include <string>

namespace { // Avoid cluttering the global namespace.

  // A couple of simple C++ functions that we want to expose to
Python.
  std::string greet() { return "hello, world"; }
  int square(int number) { return number * number; }
}

BOOST_PYTHON_MODULE_INIT(getting_started1)
{
    // Create an object representing this extension module.
    python::module_builder this_module("getting_started1");

    // Add regular functions to the module.
    this_module.def(greet, "greet");
    this_module.def(square, "square");
}


From dom at edgereport.put_a_c_o_m_here  Thu Jun  6 21:09:26 2002
From: dom at edgereport.put_a_c_o_m_here (Domenic R. Merenda)
Date: Fri, 07 Jun 2002 01:09:26 GMT
Subject: Medium to Large Scale Python Deployments
References: <mailman.1023405789.30972.python-list@python.org>
Message-ID: <aJTL8.21666$ZF5.2505752733@newssvr10.news.prodigy.com>

"Sam Penrose" <spenrose at intersight.com> wrote in message
news:mailman.1023405789.30972.python-list at python.org...
> [Domenic R. Merenda raises the topic of Python used in large contexts]

   Woah, watch it!  You almost made me sound intellegent.  Or at the very
least, thoughtful.

> I'm also curious about large amounts of Python, period. Who knows of
> ~50K Python lines-of-code (LOC) systems? 100K LOC? More?

   This is easily over the 100K LOC mark, although I don't know the actual
count.  We have 37 lib mods for Accounts Payable alone.


--
Domenic R. Merenda
Editor, The Edge Report
http://www.edgereport.com




From sholden at holdenweb.com  Fri Jun 28 09:28:21 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 28 Jun 2002 09:28:21 -0400
Subject: [Newbie]  How to manipulate variable _names_ ?
References: <3d1b0723.6179705@news.texas.net>
Message-ID: <GuZS8.107176$mF.95384@atlpnn01.usenetserver.com>

"Ben Fairbank" <baf at texas.net> wrote ...
> I have a program that requires the user to enter a file name, such as
> "othello," which the program will then open, process, and create a
> dictionary based on the contents of the file.  I would like to name
> the dictionary "OthelloCount," and then go on and create more
> dictionaries, each with a name made up of the inputfile name as
> provided by the user, concatenated with "Count," but I do not know how
> to make a variable name based on input.
>

Ben:

Please note that it's almost always a bad idea to use data to generate
variable names this way. Why? Because you then end up having to execute
dynamically-constructed code to *use* those generated variables.

Python's dictionary mechanism is almost certainly what you want in this
case: if your file is called "myfile" then create an entry on some
dictionary (say Count["myfile"]) using the filename as a key. Then you don't
*need* to create variables on the fly. So your overall plan might look
something like this (untested):

Count = {}
for filename in "this", "that", "theother":
    Count[filename] = AnalysisOf(file(filename).read())

This way will be easier, trust me.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From jjl at pobox.com  Thu Jun 27 05:59:20 2002
From: jjl at pobox.com (John J. Lee)
Date: Thu, 27 Jun 2002 10:59:20 +0100
Subject: Cookie based authentication
Message-ID: <Pine.SOL.4.44.0206271045250.17148-100000@mimosa.csv.warwick.ac.uk>

On 2002-06-17 11:32:44 PST, Chandan Mathur <cmathur at cis.poly.edu> wrote:

> I am still stuck on how to use client cookie
> (http://wwwsearch.sourceforge.net/ClientCookie/). I am trying to
> write a

As I said last time, you'd be more likely to get a timely answer by
mailing me direct.

> script that automatically logs in to http://www.fool.com/login.asp
> (SSL encryption) and downloads pages that i need. This is what i
> have tried so far: I logged in using internet explorer and opened
> the cookie file from windows explorer. Using set_cookie method in
> Cookies, i tried to set the cookie as best as i could.. (i probably
> did something wrong here).
>
> import ClientCookie
> import urllib2
> import urllib
> request = urllib2.Request("http://www.fool.com/Login.asp")
> result = urllib2.urlopen(request)
> c = ClientCookie.Cookies()
> c.extract_cookies(result, request)
[...]
>
c.set_cookie(None,"Wookie=Ref=","FoolV=5&Uid=123146592&LastVisit=6%2F17%2F20
>
02+1%3A24%3A20+PM&FirstVisit=6%2F17%2F2002+12%3A27%3A22+PM&Username=westlab"
> ,"/","fool.com/",None,None,None,None,None,None)

You're mixing two or three incompatible modes of operation here.  If
you want to do extraction and adding of cookie headers explicitly, use
the extract_cookies and add_cookie_header methods.  Otherwise, just
use urlopen:

import ClientCookie
result = ClientCookie.urlopen("http://www.fool.com/Login.asp")

Note that this is the urlopen from ClientCookie, not urllib2.  If you
use the urlopen from urllib2, it'll do any redirections (without any
cookie handling) before you even get to hear anything back from it at
all.  set_cookie is another level down further still.

All of this is in the docs...

[...]
> it gets the wrong page

'it gets the wrong page' is not all that helpful a description.  Could
you be a little more discursive next time?


John



From pinard at iro.umontreal.ca  Thu Jun 20 14:56:34 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 20 Jun 2002 14:56:34 -0400
Subject: do...while
In-Reply-To: <EzoQ8.12922$p56.4028292@newsb.telia.net>
References: <slrnah40l1.12v.msoulier@pmerd071.ca.nortel.com>
	<EzoQ8.12922$p56.4028292@newsb.telia.net>
Message-ID: <oqlm99lqxp.fsf@titan.progiciels-bpi.ca>

[Fredrik Lundh]

> Michael P. Soulier wrote:

> > How do you people handle the lack of a do...while in Python?  I find
> > that at times, I must initialize variables for a loop with identical
> > code to the loop, which feels like a waste to me.

> the standard pydiom is:

[F.P. - I negated `condition', presuming we ponder `do...while(condition)'.]

>     while 1:
>         do stuff
>         if not condition:
>             break
>         do stuff

Given the above, I would much prefer:

    do stuff
    while condition:
        do stuff

However, neither are satisfactory, when `do stuff' is long, and may not
be factored out into a separate `def', because for example it plays with
local variables.  My usual idiom is then:

     loop = True
     while loop:
         do stuff
         loop = condition

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From gcordova at hebmex.com  Mon Jun  3 13:57:51 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Mon, 3 Jun 2002 12:57:51 -0500 
Subject: string generator
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233010A5453@mail.hebmex.com>

> 
> I like to create a list of file automatical based on series. 
> For example
> user input the number of file say 5. So i want to generate the file 
> according to this range , this mean the generated files could be 
> filename1
> filename2
> filename3
> filename4
> filename5
> I using the for loop like that
> 
> for file in range(5):
>         lis = [file]
>         string.join('filename',lis[1:])
> how can it didn't work. I am confusing how to do it.
> 

Yeah, but where do you wanna put your results?

You could also:

FNs = ["filename%d" % i for i in range(5)]

or also you could:

FNs = map(lambda i:"filename%d" % i, range(5))

or many other variations thereof.

Good luck!

-gustavo




From merkosh at hadiko.de  Sun Jun 30 21:52:10 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Sun, 30 Jun 2002 18:52:10 -0700
Subject: arguments to functions
Message-ID: <MPG.178955a4fc8bfc1c989680@news.rz.uni-karlsruhe.de>

hi,

I wanted to pass one named and a collection of unnamed (an arbitrary 
number) to a function:

def test(expected=None, *misc):
    print(expected)
    print(misc)

Now, unluckyly 'test' interprets the first unnamed parameter to test as 
the value for 'expected':

test(5, expected=10)
TypeError: test() got multiple values for keyword argument 'expected'

on the other side any keyword arguments must follow non-keyword 
arguments:

test(expected=5, 10)
SyntaxError: non-keyword arg after keyword arg

Which is a pitty, because I think this version is clear.

Is that *it*, or am I doing something wrong here?

Thanks
Uwe


From pedro_rodriguez at club-internet.fr  Sat Jun  1 08:38:23 2002
From: pedro_rodriguez at club-internet.fr (Pedro RODRIGUEZ)
Date: Sat, 01 Jun 2002 14:38:23 +0200
Subject: Using nntplib
References: <3cf633df_2@news.newsgroups.com> <04sJ8.51500$%u2.49899@atlpnn01.usenetserver.com> <3cf65536_3@news.newsgroups.com> <IhHJ8.75955$%u2.62874@atlpnn01.usenetserver.com> <3cf74c6f_1@news.newsgroups.com>
Message-ID: <pan.2002.06.01.12.38.22.909201.2808@club-internet.fr>

On Fri, 31 May 2002 12:19:00 +0200, jb wrote:

> The best possibility seems to me, to rewrite nntplib, so that it fits in
> with Qt and as the NNTP protocol is very simple and nntplib is very
> clearly written, this seems to be simple. Maybe I shall try it.
 
Better have some hook/callback mechanism defined somewhere as you already
stated, or some kind of generator you could activate on will - but this
will be less effecient.

Don't know if this can help 
- python2.2 only and use longcmd which is not part of the public interface :

import nntplib

class MyListHook:
    def __init__(self):
        self.reset()

    def reset(self):
        self.count = 0
        self.list = []

    def write(self, line):
        self.count = self.count + 1
        self.list.append(tuple(line.split()))
        # print "GOT : %s" % line

class MyNNTP(nntplib.NNTP):
    def mylist(self, hook):
        resp, list = self.longcmd('LIST', hook)
        return resp, hook.list

n = MyNNTP("localhost")
h = MyListHook()
print n.mylist(h)

Regards,
Pedro

"Il n'y a gu?re dans l'amor qu'une pr?occupation grave: c'est l'envie;" 
(D?sol? ;)


From cliechti at gmx.net  Mon Jun 24 16:27:55 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 24 Jun 2002 22:27:55 +0200
Subject: SOLVED (Re: Python hits the spot)
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at> <af4vrl$ffe$0@216.39.172.122> <3D16DD5C.1050708@kfunigraz.ac.at> <uhdume2n1mhbba@corp.supernews.com> <3D171291.2060905@kfunigraz.ac.at> <mailman.1024950033.16332.python-list@python.org>
Message-ID: <Xns9237E4F0DB10Dcliechtigmxnet@62.2.16.82>

Tim Churches <tchur at optushome.com.au> wrote in 
news:mailman.1024950033.16332.python-list at python.org:

> Siegfried Gonzi wrote:
>> It was really the laptop processor. I tried the calculation on a
>> stationary machine (Pentium II 450 MHz and SuSE 8.0), and it works.
>> 
>> Die Typen k?nnen jetzt was erleben.
> 
> Could someone who understands idiomatic ?sterreichisches Deutsch
> confirm that the above means "And sorry for being such a prat."?

it's more like a threat to the laptop manufacturer.... 

you have to admit that it was a nasty failure... is bad hardware your first 
thought when your program fails? on the other side he realy had a bad start 
with this thread.

chris

-- 
Chris <cliechti at gmx.net>



From garrett at bgb.cc  Tue Jun 18 19:57:40 2002
From: garrett at bgb.cc (Don Garrett)
Date: Tue, 18 Jun 2002 23:57:40 GMT
Subject: What If.....    Strong Types
References: <3D0C0F0D.4030900@bgb.cc>
Message-ID: <3D0FC86D.1030506@bgb.cc>

Don Garrett wrote:
 >   I'm not suggesting any real changes to any, only proposing a thought
 > experiment.
 >
 >   What if Python had been built from the start with strong types? And I
 > mean strong types, not strong typing. The distinction I am trying to
 > draw is that variables would not be typed, but that all types would have
 > rigid interfaces.

    I seem to have drawn a lot of responses that say "Don't mess with Python".
That tells me that I didn't do a very good job of posing the question. To be
clear, I have no intention of writing any code or changing anything. I don't
even think that standard Python would benefit from the changes I'm suggesting.
I'm just tyring to understand what would or would not be a good idea in a
language design and (more importantly) why.

    One thing I was NOT suggesting was type tagging. Part of what led to my 
post was the realization that most verbosity in strongly/statically typed 
languages is the redundancy of type tagging, and that strong interfaces can be 
had without it.

    Here are the two reasons that lead me to think that a Pythonish language
with strong interface definitions would be a good idea (at least in theory).

One:

    I THINK (am not sure) that this is the smallest (major) change
that could be made to the language spec which would produce a language which
was easily compilable.

    This is important because, good as python is, there are jobs for which it's
unsuitable. For example, it can't be used to write an operating system kernel
or device driver. Python modules are often (re)written in C/C++ when
performance is critical. It would be nice if something more Pythonish could be
used instead.

    My biggest doubt here is.... could highly efficient compilation be done 
without continual type tagging? Obviously, careful data flow analysis would be 
needed.

    I'm totally ignoring other properties that such a compiled language would
need to be really useful. An example being that it would need to be able to
link directly against C/C++ libraries.

Two:

    I've had a long term idea about building a really, really smart IDE.
Something that provided high-level programming support on several levels, as
well as detailed analysis of the code. I know this isn't original.

    For example: a programmer should be able to ask which exceptions an method
could raise and which runtime conditions would cause those exceptions to be
raised.

    That kind of analysis is not easy to do, and will often break into results
that are to vague to be useful. It's easier to build such a system with a lot
of facts that are "for sure" not "probably, unless somebody did something
wierd". As a note: I think such a system is almost impossible with general C++
because it simply has too many ways to "do something wierd".

    Python tends do things by convention, not by enforced rules. This is a good
thing for programmers. But it's hard for software to deal with if the software
is going to do more than just run it.

-- 
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc



From Farhad.Fouladi at Dresdner-Bank.com  Wed Jun  5 10:35:20 2002
From: Farhad.Fouladi at Dresdner-Bank.com (Farhad)
Date: Wed, 5 Jun 2002 16:35:20 +0200
Subject: How to configure python to work with apache 2.0 ??
References: <abpta6$us9o$1@engel.scvmaxonline.com.sg>
Message-ID: <adl6ns$fa72@news-1.bank.dresdner.net>

take a look at:

http://www.modpython.org

regards,
Farhad

"Matthew Chong" <arena at pacific.net.sg> schrieb im Newsbeitrag
news:abpta6$us9o$1 at engel.scvmaxonline.com.sg...
> I simply will just to configure python to work with apache 2.0 without
using
> third-party software such as
> mod_python or pyapache. Because I realised that using the latter, I have
to
> deal with terrible ordeal with compiling
> with C language.
>
> Can someone please tell me how can I just simply run python as cgi in
apache
> 2.0 ? without installing any thirdparty software
> other than python and apache 2 ??
>
> thank
>
>




From paul at boddie.net  Wed Jun  5 07:21:37 2002
From: paul at boddie.net (Paul Boddie)
Date: 5 Jun 2002 04:21:37 -0700
Subject: Extracting data from HTML
References: <82096df4.0205311152.5891a17b@posting.google.com> <1022883953.6150.5802.camel@localhost> <mailman.1022908262.30070.python-list@python.org> <82096df4.0206010201.736c4b2c@posting.google.com>
Message-ID: <23891c90.0206050321.d6a64c3@posting.google.com>

lailian98 at hotmail.com (Hazel) wrote in message news:<82096df4.0206010201.736c4b2c at posting.google.com>...
>
> I'm relying onsgmllib to do the work....
> since htmllib requires heavy coding.

First of all, take a look at this document:

  http://www.boddie.org.uk/python/HTML.html

The first section describes the use of sgmllib, although you could
also try various XML package classes instead. If you understand the
document, you barely need to read the rest of this message.

> Here, an instance of what I want to extract....
> the time of the TV programme >> 12:15:00AM
> 
> "<TR>
>    <TD align=right bgColor=#000033><FONT color=#ffffff 
>    face="verdana, arial, helvetica" size=1>12:15:00 
>    AM</FONT></TD>"
> 
> So what do u think?

The most important structural detail is probably the 'TD' element,
even though the data you need is found within the 'FONT' element. What
you could do, therefore, is to set up a handler method for the 'TD'
element called 'start_td' which sets a flag in your parser object
noting that the information inside the element may be interesting; you
could do more checks on the attributes ('align', 'bgColor') if you
believe that they help to distinguish the cells in the table which
contain times from the other cells. You also need an 'end_td' method
which unsets the flag, and you should always beware of nested tables,
too.

  def start_td(self, attributes):
      if <some test with the attributes>:
        self.inside_cell = 1

  def end_td(self):
      self.inside_cell = 0

Once "inside" the 'TD' element, you might then want to check for a
'FONT' element. Again, set up a handler method called 'start_font'
which firstly checks to see if that flag was set, indicating that the
parser is currently "inside" the 'TD' element of interest. Then, you
might want to check for some interesting attributes, but only if you
think you can rely on them - I get suspicious about multiple
presentational attributes (especially when they could have used a
stylesheet), and that's partly why I advocate checking for the
presence of more than one element type (in this case, the 'TD' and the
'FONT' elements) before mining away at the data.

The 'start_font' element will also set a flag indicating that a time
should be extracted from the text inside the element (between the
start and end tags), and again, it's important to implement an
'end_font' element which unsets this new flag.

  def start_font(self, attributes):
      if self.inside_cell and <some test with the attributes>:
          self.ready_to_read = 1

  def end_font(self):
      if self.inside_cell: # Arguably not necessary.
          self.ready_to_read = 0

Finally, you should implement the 'handle_data' element which checks
that this new flag is set before reading the textual data and storing
it somewhere (such as another attribute in your parser class).

  def handle_data(self, data):
      if self.ready_to_read:
          self.my_programme_times.append(data)

There are lots of issues with doing the parsing this way, and having
parsed some pretty complicated pages, I can certainly recommend the
XML approach instead, since it provides much better ways of testing
the structure than setting flags here and there. Unfortunately, you
may well need something like mxTidy to deal with severely broken HTML,
of which there seems to be a lot around.

Paul


From fperez528 at yahoo.com  Thu Jun 20 13:15:13 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 20 Jun 2002 11:15:13 -0600
Subject: logical statements (in Python and Numeric)
References: <3D11F1B2.21587F31@lanl.gov>
Message-ID: <aet2j4$h8c$1@peabody.colorado.edu>

Nathan Given wrote:

> I am graphing piecewise functions, and I have a method of getting the
> proper y values in a vector that works in MATLAB... I was trying to get
> it to work in Python, but it wasn't...
> 
> it has to do with a using an and statement...
> 
> 
> here is my code...
> 
> x = arrayrange(-4.,2.+.001,.001, Float)
> y = zeros(len(x),Float)
> 
> #------ Create Piecewise Function ------#
> y = y + (x < -2)*(((x + 2.)**2) + 4)
> y = y + (x >= -2. and x < 0.) * (x + 6.)
> y = y + (x >=0) * (x ** 2. + 6.)
> 

I agree that the above failure is bloody counterintuitive, to the point where 
I'd call it a bug. It's easy to illustrate with:

In [58]: a=arange(12)

In [59]: a > 5 and a < 8
Out[59]: array([1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0])

What you need to use is logical_and():

In [60]: logical_and(a>5,a<8)
Out[60]: array([0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0])

You may consider posting to the Numeric mailing list about this. The 'and' 
operator should either not work at all or work correctly. The above is a 
heinous behavior.

But at least you can use the logical_* functions to get around the problem. 
Your code would then read:

In [49]: x = arrayrange(-4.,2.+.001,.001, Float)

In [50]: y = zeros(len(x),Float)

In [51]: y += (x < -2)*(((x + 2.)**2) + 4)

In [52]: y += logical_and(x>=-2,x<0) * (x+6)

In [53]: y += (x >=0) * (x ** 2. + 6.)


Note the '+=' operator, if I recall correctly it works in-place for Numeric 
arrays, which has memory benefits (significant for large arrays).

For common range creation with a closed interval, I wrote the following 
function which I use constantly. Feel free to use it if you like it. The npts 
optional argument is especially useful, since you often care more about the 
particular size of the list rather than the specific value of delta.

def frange(xini,xfin=None,delta=None,**kw):
    """frange([start,] stop[, step, keywords]) -> list of floats

    Return a Numeric array() containing a progression of floats. Similar to
    arange(), but defaults to a closed interval.

    frange(x0, x1) returns [x0, x0+1, x0+2, ..., x1]; start defaults to 0, and
    the endpoint *is included*. This behavior is different from that of
    range() and arange(). This is deliberate, since frange will probably be
    more useful for generating lists of points for function evaluation, and
    endpoints are often desired in this use. The usual behavior of range() can
    be obtained by setting the keyword 'closed=0', in this case frange()
    basically becomes arange().

    When step is given, it specifies the increment (or decrement). All
    arguments can be floating point numbers.

    frange(x0,x1,d) returns [x0,x0+d,x0+2d,...,xfin] where xfin<=x1.

    frange can also be called with the keyword 'npts'. This sets the number of
    points the list should contain (and overrides the value 'step' might have
    been given). arange() doesn't offer this option.

    Examples:
    >>> frange(3)
    array([ 0.,  1.,  2.,  3.])
    >>> frange(3,closed=0)
    array([ 0.,  1.,  2.])
    >>> frange(1,6,2)
    array([1, 3, 5])
    >>> frange(1,6.5,npts=5)
    array([ 1.   ,  2.375,  3.75 ,  5.125,  6.5  ])
    """
    #defaults
    kw.setdefault('closed',1)
    endpoint = kw['closed'] != 0

    # funny logic to allow the *first* argument to be optional (like range())
    # This was modified with a simpler version from a similar frange() found
    # at http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66472
    if xfin == None:
        xfin = xini + 0.0
        xini = 0.0

    if delta == None:
        delta = 1.0

    # compute # of points, spacing and return final list
    try:
        npts=kw['npts']
        delta=(xfin-xini)/float(npts-endpoint)
    except KeyError:
        npts=int((xfin-xini)/delta+endpoint)

    return arange(npts)*delta+xini


If you want to be able to plot with gnuplot a single array with g.plot(y), or 
y vs x with g.plot(x,y), ask me for a modified version I have, which also 
fixes the problems with hardcopy(). It's part of ipython 
(http://windom.colorado.edu/~fperez/ipython/) which gives you a python shell 
which feels much like that of IDL or Mathematica. 

Cheers,

f.


From TuxTrax at fortress.tuxnet.net  Thu Jun  6 12:20:19 2002
From: TuxTrax at fortress.tuxnet.net (TuxTrax)
Date: Thu, 06 Jun 2002 16:20:19 -0000
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet> <qkxL8.57201$wd3.9390318@news6-win.server.ntlworld.com>
Message-ID: <slrnafv3d5.3ac.TuxTrax@fortress.tuxnet>

On Thu, 06 Jun 2002 00:44:32 +0100, Graham Ashton <graham at effectif.com> wrote:
> On Wed, 05 Jun 2002 11:31:19 +0100, TuxTrax wrote:
> 
>> My question is this; I realize that Python compiles to bytecode
>> automatically as part of the interpreters tokenization run, but is there
>> a way to permanently compile python code outputting to an executable
>> binary?
> 
> Not exactly, in general. Have a look for py2exe, psyco, and pyrex for info
> on some of the things that you can currently do.
> 
> GvR alludes to compilation being of interest in the future in a recent
> interview:
> 
>    http://www.onlamp.com/pub/a/python/2002/06/04/guido.html
> 

Thanks for the information, Graham.

Cheers,

Mathew

-- 
TuxTrax   (n.) An invincible, all terrain, Linux driven armored assault
vehicle that can take as much fire as it gives ;-)

<Carnivore activation string> 
ASSASINATION ANTHRAX PRESIDENT NUCLEAR TALIBAN AMMONIUM NITRATE
</Carnivore activation string>

Yes, I am a Penguin cult high priest. Fin readings upon request.
ROT13 this email address to mail me: uvtuqrfregzna at lnubb.pbz



From uncleenzo at hushmail.com  Fri Jun 28 17:02:30 2002
From: uncleenzo at hushmail.com (Uncle Enzo)
Date: 28 Jun 2002 14:02:30 -0700
Subject: COM server threads/unloading
References: <10d944cb.0206261529.741da082@posting.google.com> <slrnahmos0.eh5.kzaragoza@localhost.localdomain>
Message-ID: <10d944cb.0206281302.43de821a@posting.google.com>

Thanks Kris, for your reply. I have a follow-up question...

kzaragoza at attbi.com (Kris J. Zaragoza) wrote:
>(snip) Your server doesn't shut down
> because you have threads still running waiting on a queue that will
> never contain more messages.
> 
> As I see it, you have two options.  First, you can build a mechanism
> to explicitly send the logger the message to shut down.

How would you do this? Start another thread that periodically checks
mainthread.GetInterfaceCount(), and if it returns zero put in the quit
event and then return? I'd rather not have a shutdown or release
method exposed to users.

Thanks,

Enzo


From robin at execulink.com  Mon Jun 17 21:47:02 2002
From: robin at execulink.com (robin at execulink.com)
Date: Mon, 17 Jun 2002 21:47:02 -0400
Subject: Creating Dynamic Web Pages (ZOPE)
References: <46ca81a0.0206062016.789d3af6@posting.google.com> <J29M8.358696$%u2.156726@atlpnn01.usenetserver.com> <46ca81a0.0206141201.23db4a8c@posting.google.com>
Message-ID: <kh3tgus5d6hlrt4qjsdg1odr6j9ft0nue2@4ax.com>

Hugh Cowan wrote:

>Also, I thought that Dreamweaver was more of just a HTML Editor, while
>something like Zope provided extra functionality, tools, wizards,
>etc.. to allow you to create Dynamic Web-Sites quicker and easier?  I
>am just trying to figure out where to concentrate my search for
>software and tools.

I would very much advise against DreamWeaver and similar tools that
pretend designing web pages is a WYSIWYG activity, which it certainly
is not.

I prefer a simple text editor with some HTML convenience features (eg:
syntax highlighting) combined with a templating system. The text
editor lets you stay close to the HTML and CSS you'll need to know
inside out. (I use NoteTab Pro from Fookes software.) The templating
system provides a reasonable way of reusing "code" and helps enforce
consistency.

There are so many web frameworks because everyone has different ideas
about what they want to do. Zope is the heaviest solution and requires
buying into a completely different paradigm -- web as object space --
that many find unintuitive.

Other frameworks concentrate on low-level process handling to remove
the overhead of starting a new Python interpreter for each CGI
session. But the truth is that most sites don't benefit from this
optimisation. It depends on how many thousands of hits an hour you
expect.

My Wasp framework sticks close to the single document equals single
web page paradigm and operates on a web server strictly through CGI
with no fancy architectural tricks.

It provides session and cookie handling, error handling, logging,
e-mail notification, and post-mortem debugging. Plus lots of prebuilt
modules for the sorts of web tasks people need time and time again.
But mostly it helps organise your code by providing simple but
flexible templating.

It's just one of many.

-- robin
www.execulink.com/~robin/wasp/readme.html


From dallasm at aiinet.com  Thu Jun 20 14:49:14 2002
From: dallasm at aiinet.com (Mahrt, Dallas)
Date: Thu, 20 Jun 2002 14:49:14 -0400
Subject: Python Documentation generation
Message-ID: <638AA0336D7ED411928700D0B7B0D75B02E1EE98@aimail.aiinet.com>

I am looking for an application which can generate HTML documentation from
Python modules.
(Note: some examples based on JavaDoc syntax)


I have a few desires from such a program:
	- Ability to exclude elements based on naming conventions or
keywords 
		(ex. Exclude names starting with a single underscore '_foo'
or containing the keyword @private)
	- Ability to have inter module linking. (If module A uses module B,
then there should be an '@see B' like link [JavaDoc-ish])
	- The program should also allow for either:
		- generating module documentation from C extension modules
(Should be tricky)
		- allow for other documentation sections to be linked
automatically by registration of module name

I have found several systems for generating documentation of Python modules
(HappyDoc, pythondoc, pydoc) but I don't believe any of them fulfill all of
these desires. Any suggestions from the group?

Thanks.

---
Dallas S. Mahrt               614-923-1194
Software Engineer             Network Management Systems
Applied Innovation Inc.       Dublin, Ohio





From phr-n2002b at NOSPAMnightsong.com  Sat Jun 29 16:26:26 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Jun 2002 13:26:26 -0700
Subject: yield equivalent in C/JavaScript?
References: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk>
Message-ID: <7xy9cxhlvx.fsf@ruckus.brouhaha.com>

There's no native way to do it in C; in some enviroments you might be
able to simulate it with longjmp madness or threads.

In Javascript the only approach I can think of is make a "virtual
machine" that executes javascript statements using JS's eval function,
and implement something like threads or coroutines in the virtual
machine.  I did something like that once--it wasn't as bad as it
might sound, but it may not be worth it for your application.
You can just organize your generator functions like a state machine
instead, saving their locals and yielding location in some structure
returned to the caller.  I haven't examined any Python generator
bytecode but I suspect that's how it works.  By being limited to
"simple" generators (can't yield except to one's direct caller)
it's enough to just save the local variables and "return" address
in the state object, rather than a full continuation including
the call stack.


From cliechti at gmx.net  Sat Jun 22 09:19:37 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 22 Jun 2002 15:19:37 +0200
Subject: The best GUI library in python
References: <8d3f4438.0206220457.b820418@posting.google.com>
Message-ID: <Xns92359C58C8B03cliechtigmxnet@62.2.16.82>

d2002xx at myrealbox.com (d2002xx) wrote in 
news:8d3f4438.0206220457.b820418 at posting.google.com:

> Hello!!
> 
> I want to learn GUI in python, who would tell me which GUI library is
> the best? (don't tell me Tkinter, it's too slow......)

There isn't something like 'the best' for all use(r)es... it depends on 
your situation. do you need multiple platform support? if no, on wich 
platform do you want to run it. do you need it for commercial projects?

- you could learn anygui (http://anygui.sf.net). that way you don't realy 
  need to care about the GUI library, anygui chooses the the best from 
  what you have installed.
- for multiplatform, wxPython is an option.
- for linux wxGTK etc seems widely used.
- for both platforms there is pyQT wich is good too, but you have to buy it 
  for commercial applications.
- there are may others: fox, or SWING with jython but those are not so   
  widely used.

there are different free and commercial GUI desigeners. do you plan onn 
using one of these?

you can browse the archive of this list to find many discusions about this 
issue.

i use wxPython.

chris

-- 
Chris <cliechti at gmx.net>



From sjmachin at lexicon.net  Mon Jun  3 18:23:13 2002
From: sjmachin at lexicon.net (John Machin)
Date: 3 Jun 2002 15:23:13 -0700
Subject: word wrap in regex
References: <mailman.1023127745.29025.python-list@python.org>
Message-ID: <c76ff6fc.0206031423.1884ebe0@posting.google.com>

Sabine Richter <sabine at textraeume.de> wrote in message news:<mailman.1023127745.29025.python-list at python.org>...
> Hello,
> 
> I have a regex which is a bit long (200 characters) and I want to add
> word wraps and comments to make it more clear. I'm sure I read about
> that in the references before but I don't find the passage with the word
> wrap no more. May anybody give me a hint?
> 
> Sabine

It is in the "module contents" section. You need to specify a flag as
an optional argument. You can have not only "word wraps" and comments
but also whitespace between conceptual "words".

"""
X

VERBOSE
This flag allows you to write regular expressions that look nicer.
Whitespace within the pattern is ignored, except when in a character
class or preceded by an unescaped backslash, and, when a line contains
a "#" neither in a character class or preceded by an unescaped
backslash, all characters from the leftmost such "#" through the end
of the line are ignored.
"""

If you are having trouble parsing the second sentence, try
substituting "nor" instead of the second "or".


From shagshag13 at yahoo.fr  Mon Jun  3 13:49:26 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Mon, 3 Jun 2002 19:49:26 +0200
Subject: need Python hacking tips and random access in text file...
Message-ID: <adga6o$10lije$1@ID-146704.news.dfncis.de>

I'm looking for everything which might fasten full (not c) python scripts...

i've found theses pages :
http://manatee.mojam.com/~skip/python/fastpython.html
http://www.python.org/doc/essays/list2str.html

is there others i miss ?

and an efficient random access to line from text files...

is using linecache the best one : http://www.python.org/doc/current/lib/module-linecache.html ?

thanks in advance,

s13.






From shagshag13 at yahoo.fr  Tue Jun  4 06:42:04 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Tue, 4 Jun 2002 12:42:04 +0200
Subject: Looking for a list subclassing example...
References: <adi4dv$118tko$1@ID-146704.news.dfncis.de> <3cfc994e$0$3851$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <adi5hc$10ulbe$1@ID-146704.news.dfncis.de>

i do it like that, it work, but don't know if it's correct :

class Shag(list):
 def __init__(self, emptyValue = None):
  self._emptyValue = emptyValue

 def reset(self):
  del self
  self = []

 def __setitem__(self, key, item):
  try:
   list.__setitem__(self, key, item)
  except IndexError:
   for i in range(len(self), key):
    self.append(self._emptyValue)
   self.append(item)




From syver-en+usenet at online.no  Tue Jun  4 13:14:37 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Tue, 04 Jun 2002 17:14:37 GMT
Subject: python and windows and pydoc and favorite editor == True
References: <un0ugnxa3.fsf@online.no> <adggk2$psv$00$1@news.t-online.com> <uznycc5om.fsf@online.no> <3CFC1B50.E830AC46@engcorp.com>
Message-ID: <uelfn3r1s.fsf@online.no>

Peter Hansen <peter at engcorp.com> writes:

> Syver Enstad wrote:
> > 
> > "J?rgen Hermann" <jh at web.de> writes:
> > 
> > > "Syver Enstad" <syver-en+usenet at online.no> schrieb im Newsbeitrag
> > > news:un0ugnxa3.fsf at online.no...
> > > > Post if there is any interest.
> > >
> > > OK, post. :)
> > >
> > >
> > 
> >
> 
> 
> Looks like this should be:
> 
>   http://home.online.no/~syver-en/files/PyDocEditorStarter.py
> 
> The other 404's me.

That's correct, 
http://home.online.no/~syver-en/files/PyDocEditorStarter.py
is the address, shows what harm copy paste does :-)
-- 

Vennlig hilsen 

Syver Enstad


From kragen at pobox.com  Sun Jun  2 20:44:49 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 02 Jun 2002 20:44:49 -0400
Subject: symbolic python
References: <3CEFAE4C.38A8022B@doc.ic.ac.uk> <3CEFD881.D7992095@doc.ic.ac.uk> <3cf8ce86$1@giga.realtime.net> <83sn46vkgi.fsf@panacea.canonical.org> <3CF942C7.2E4AA4CF@engcorp.com>
Message-ID: <838z5xtcla.fsf@panacea.canonical.org>

Peter Hansen <peter at engcorp.com> writes:
> Kragen Sitaker wrote:
> > "Ira Baxter" <idbaxter at semdesigns.com> writes:
> > > DMS Software Reengineering Toolkit.
> > It sounds very interesting.  How can I get a copy?
> 
> No doubt by sending them a pile of money.  :-)  (Given the customer
> list mentioned, it would probably be a large pile.)

I looked around the web site for a while without finding any
information about how to do that, so I posted the above message.  Ira
Baxter responded in private, saying it "was available commercially",
and suggesting that I look at the web site for more details.

This all leaves a bad taste in my mouth; the Web site makes a point of
explaining that the information they would like to sell me was
developed with my tax dollars.

> no support for Python is (yet) mentioned in the list of supported
> languages.

I didn't even notice.



From jseb at cs.mcgill.ca  Thu Jun 13 18:37:38 2002
From: jseb at cs.mcgill.ca (=?ISO-8859-1?Q?Jean-S=E9bastien_Bolduc?=)
Date: 13 Jun 2002 15:37:38 -0700
Subject: newbie: from .. import *
Message-ID: <56e1eff0.0206131437.4ca8ed9e@posting.google.com>

Hi All

I've been working with Python for a while, but never really felt
confortable with the way namespaces are imported.

I'm working on a simple application that is to be used in the
interactive mode. The idea is that the behavior of a function is
determined, in part, by some global(?) flags. So I can change my
flags, and explicitely call my function. Really simple. Except that
for some reason, it does not work.

In "test.py", I have the following:

  flag = 1
  def foo():
    print flag
  foo()
  flag = 0
  foo()
  
In the interactive mode, if I import the code above as:
>>> from test import *
1
0

The output is as expected. However, if I continue:
>>> flag = 1
>>> foo()
0

(!) Okay... not what _I_ expected. It sounds really silly, but I've
spent a couple of hours on that, without success. I thought at some
point that using "global" would help me. Of course, if I just do
"import test", everything works fine --- but in interactive mode, I
don't want to type "test.foo()"

Some help?
  --JSeb


From aahz at pythoncraft.com  Wed Jun 19 16:25:54 2002
From: aahz at pythoncraft.com (Aahz)
Date: 19 Jun 2002 16:25:54 -0400
Subject: Pop a list from beginning ? and memory saving...
References: <aem8e8$4kq$1@news.idiom.com> <aep4jb$1q8h$1@news.idiom.com> <aeqlgo$gbk$1@panix1.panix.com> <aeqne9$4h1$1@news.idiom.com>
Message-ID: <aeqpci$1ud$1@panix1.panix.com>

In article <aeqne9$4h1$1 at news.idiom.com>,
James T. Dennis <jadestar at idiom.com> wrote:
>Aahz <aahz at pythoncraft.com> wrote:
>> In article <aep4jb$1q8h$1 at news.idiom.com>,
>> James T. Dennis <jadestar at idiom.com> wrote:
>>>
>>> It might be better to have a hack to the implementation of the
>>> list primitives --- so they defer the shift of remaining elements
>>> until GC.  The head of the list could point at the first none 
>>> deleted item.  Then pop(0) could be implemented in O(1) time and
>>> the garbage collector would eventually come along to do it's 
>>> job.
>>
>> While this might be a worthwhile optimization, it has absolutely nothing
>> to do with GC.  Each time an element were popped off the front, its
>> object would be decreffed and possibly deleted.  GC runs completely
>> independently of individual objects.
>
> Sorry, I was speaking abstractly about garbage collection, I know
> nothing about Python's GC implementation.  Perhaps it would be an 
> opportunity to augment Python's GC.  Some containers (and even 
> some classes) could implement their own compaction routines that 
> could be called by the GC. 

That wouldn't work too well.  It's really only vectors that need
compaction, and it's only Python lists that are mutable vectors.
Python's dict structure already has its own compaction algorithm, and
it's not compatible with GC.

(Dict compaction is based on the proportion of free entries; it's a
simple test, and there's no gain to be made by adding GC on top of it.)

Keep in mind that Python's (CPython's, to be precise; Jython does use GC
because that's what Java provides) base memory management semantics is
built on refcounting.  GC only handles cycles.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From BPettersen at NAREX.com  Mon Jun 10 23:48:03 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 10 Jun 2002 21:48:03 -0600
Subject: newbie question - python blocking on int()?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192F1A3F5@admin56.narex.com>

> From: Michael Davis [mailto:michael at damaru.com] 
> 
> Chris Liechti wrote:
> 
> > Michael Davis <michael at damaru.com> wrote in
> > news:4LaN8.2488$Vr2.586019 at news20.bellglobal.com:
> >> I'm writing a specialized ftp client. I'm parsing the 
> output of ftp.dir,
> >> which gives me a string like this (call it str):
> >> 
> >> -rw-r--r--   1 fred       527 Jun  4 22:58 report.php
> >> 
> >> and I'm getting the various parts like this:
> >> 
> >> details = string.split( str )
> >> permissions = details[0]
> >> size = details[3]
> >> name = details[7]
> >> debug( "added remote file size %5d: %s" % (int(size), name) )
> >> 
> >> This works. But when I replace the 3rd line with this:
> >> 
> >> size = int( details[3] )
> >> 
> >> python hangs on that line. Why?
> > 
> > not when i try it... what do you mean by "hangs" does it 
> takes forever or
> > do you see an error message? if its the later, then you might have
> > assigned something to "int" somewhere else, hiding the 
> builtin function.
> 
> Thanks for replying. Actually, it's the former - when I say 
> 'hangs', I mean 
> that the program appears to stop running at that point. If I write:
> print "one"
> size = int( details[3] )
> print "two"
> 
> it prints "one" but not "two".

Try inserting

  print '[%s]' % details[3]

before you convert to an int (that will at least let you know for sure
what int is trying to conver...)

hth,
-- bjorn




From dsavitsk at e-coli.net  Tue Jun  4 18:33:30 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Tue, 04 Jun 2002 22:33:30 GMT
Subject: python interacting with other programs
References: <64eb9145.0206041454.4f70e86e@posting.google.com>
Message-ID: <_ebL8.24085$d7.6575140@newssrv26.news.prodigy.com>

Doing this has less to do with Python, and more to do with COM.  As such,
you should look at the COM interface exposed by Lotus Notes.  A good way to
do this is to use the Python Makepy utility.  You can also use an object
browser (of which there is one included w/ the VBA in Word or whatnot).

-d


"Jakob" <da_woym at hotmail.com> wrote in message
news:64eb9145.0206041454.4f70e86e at posting.google.com...
> now I know python can interact with programs like Microsoft wor and
> access and all but has anyone ever tried to interact with Lotus Notes?
> Is it possible?




From amuys at shortech.com.au  Mon Jun 24 21:30:26 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 24 Jun 2002 18:30:26 -0700
Subject: Generic Python
References: <mailman.1024928731.22453.python-list@python.org>
Message-ID: <7934d084.0206241730.684488bf@posting.google.com>

Michael Chermside <mcherm at destiny.com> wrote in message news:<mailman.1024928731.22453.python-list at python.org>...
> (4) In python, you can change individual methods of an instance, so you 
> may not even NEED to have many separate classes, if they differ only in 
> that one method.
> 
> Here is an illustration of how it would work:
> 
> Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
> Type "help", "copyright", "credits" or "license" for more information.
>  >>> class Base:
> ...     def output_0(self):
> ...         print 'zero'
> ...     def output_1(self):
> ...         print 'one'
> ...     def output_2(self):
> ...         print 'two'
> ...     def __init__(self, n):
> ...         self.output = (
> ...             [self.output_0, self.output_1, self.output_2] [n] )
> ...
>  >>> b = Base(0)
>  >>> b.output()
>  zero
>  >>> b = Base(2)
>  >>> b.output()
> two
> 

And naturally you can extend this slightly with a closure to avoid
having to declare all your methods in the class upfront ie.

class Base:
    def __init__(self, text):
        def closure():
            print text
        self._output = closure
    def output(self):
        self._output()

b = Base("Hello")
c = Base("World")
b.output()
Hello
c.output()
World

> HOWEVER... if your subclasses need more than just the one function 
> defined in them, then you are better off going with solution (1) above.

Agreed.  The original post requested generic python, AFAIK this is
best done by runtime definition of classes inside factory
methods/functions.

Andrae Muys


From jeremy at jdyallop.freeserve.co.uk  Sat Jun 15 18:33:56 2002
From: jeremy at jdyallop.freeserve.co.uk (Jeremy Yallop)
Date: 15 Jun 2002 22:33:56 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com>
Message-ID: <aegfck$6tkre$1@ID-114079.news.dfncis.de>

* George Hester
| What is a simple script that will tell me the version of python I am currently
| running?  IIS 5.0 Windows 2000.

Look up sys.version and sys.version_info.

Jeremy.


From claird at starbase.neosoft.com  Tue Jun 11 08:55:08 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 11 Jun 2002 07:55:08 -0500
Subject: Getting Tkinter to accept cyrillic text
References: <83034362.0206110052.351cb526@posting.google.com>
Message-ID: <9FE568AC2FFCE4C7.5958BD1DC7128B7C.136193D99215D2BB@lp.airnews.net>

In article <83034362.0206110052.351cb526 at posting.google.com>,
DR <ypoi at spray.se> wrote:
>I need Tkinter entry and label widgets to support cyrillic letters
>(koi8-r). I read in some post that you should change
>sys.setdefaultencoding(), but that method doesn't exist in my Python
>(I use version 2.2.1) -Has this feature been changed or is my
>installation bad? Also there is this locale.setlocale() thing. What
>effect does changing it have?
>
>I really would appreciate some kind of checklist, like: Do this, then
>change that, then evereything works! Or a patch or something (Win95)
>
>Is there any comprahensive recource for this type of problem? I've
>only found some scattered posts so far.
			.
			.
			.
Tcl/Tk folks have gotten rather good with Cyrillic
(see <URL: http://wiki.tcl.tk/ruslish > and <URL:
http://wiki.tcl.tk/3426.html >, for example.  I
haven't begun to apply this to Tkinter.  Good luck.
-- 

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


From duncan at NOSPAMrcp.co.uk  Thu Jun  6 06:32:55 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Thu, 6 Jun 2002 10:32:55 +0000 (UTC)
Subject: pygoogle search restricted by date
References: <adfmsk$205i$1@ariadne.rz.tu-clausthal.de>
Message-ID: <Xns92257522D50DCduncanrcpcouk@127.0.0.1>

Lutz Schroeer <ifls at rz.tu-clausthal.de> wrote in 
news:adfmsk$205i$1 at ariadne.rz.tu-clausthal.de:

> restrict_query = 'daterange:%d-%d' % (yesterday, today)
> 
> data = google.doGoogleSearch('python', restrict = restrict_query)

restrict is for searching by country or topic. I think you just want to put 
the daterange in as part of the main query string:

> data = google.doGoogleSearch('python '+restrict_query)


-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From FooWeissBarMike at hotmail.com  Wed Jun 19 08:55:31 2002
From: FooWeissBarMike at hotmail.com (Michael Weiss)
Date: Wed, 19 Jun 2002 08:55:31 -0400
Subject: SafeArray problem
References: <3d105710.83643750@news.skynet.be>
Message-ID: <3d108014$1@news.mhogaming.com>

How is the safearray parameter defined in the IDL on this method? [in, out]
[in] [out, retval], etc...?

<pixie888 at hotmail.com> wrote in message
news:3d105710.83643750 at news.skynet.be...
> Hi all,
>
> I am using an Automation object which has a method which takes a
> SafeArray as an argument. In my Python module I call this method and
> give a list of longs as an argument.
>
> While debugging (using the VC++6 IDE) I notice that the type of the
> argument (upon entering my automation object) is VT_ERROR (argument
> not found, actually). Now, what is really odd is the following: I
> extended my method: it takes two arguments, the first being a
> SafeArray and the second a long (which will not be used). When I call
> this method again from my Python module (with a list of longs and a
> dummy long as arguments) then everything works fine.
>
> Beats me, really.
>
> Any idea where the problem lies?
>
> Thanks,
>
> Henk




From gherron at islandtraining.com  Tue Jun 25 13:34:35 2002
From: gherron at islandtraining.com (Gary Herron)
Date: Tue, 25 Jun 2002 10:34:35 -0700
Subject: How to ImpersonateLoggedOnUser on Windows NT/2000?
Message-ID: <200206251034.35021.gherron@islandtraining.com>

Hi,

I'm trying to write a script for both Windows NT/2000 and Linux which
needs to perform actions on the behalf of several different users.

On Linux I use "os.seteuid" and friends and all is fine.

On NT the comparable thing seems to be LogonUser,
ImpersonateLoggedOnUser, and RevertToSelf.  However for these calls to
be successful, one needs various privileges, such as SE_TCB_NAME, and
SE_CHANGE_NOTIFY_NAME..

At this point I start getting lost, so I hope someone who has been
through this can help me.  How can a Python program running as local
administrator on Windows NT/2000 set those privileges, and then once
set, how does one actually *do* the impersonation?

Thank you,
Dr. Gary Herron






From brian_l at yahoo.com  Wed Jun 19 01:39:43 2002
From: brian_l at yahoo.com (Brian Lenihan)
Date: 18 Jun 2002 22:39:43 -0700
Subject: PythonWorks Pro problem!
References: <q3nrgu019nlnc2bc19jaouhfo0833fknrj@4ax.com> <3lwP8.112157$pw3.6716@sccrnsc03> <8nrtgusniqrs9r3fs7l4nt8qka5fm7imi3@4ax.com>
Message-ID: <afe3b146.0206182139.a01f19@posting.google.com>

Johann <observer at NOSPAM.space.pl> wrote in message news:<8nrtgusniqrs9r3fs7l4nt8qka5fm7imi3 at 4ax.com>...
> On Tue, 18 Jun 2002 01:49:52 GMT, "Emile van Sebille" <emile at fenx.com>
> wrote:
> >> I registered some examples from above link. The problem is, I cannot
> >> find  this "small, coloured icon" , so I cannot cycle through avalable
> >> settings. :-(
> >>
> >> I tried to clink on everything, but without results :( Anybody can
> >> help?
> >
> >Hmmm... that icon appears to be missing from 1.3...
> >
> >It's there in 1.2
> 
> So, it is a bug. :(
> 
> I wrote to authors, but without response. How to get 1.2 if they allow
> to d-l 1.3 only?

Open the "project settings" in the SYSTEM project.  You still have to
right click on project and "reload actions" before the edit prefs
change.


From eric.brunel at pragmadev.com  Mon Jun  3 11:23:47 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 3 Jun 2002 15:23:47 +0000
Subject: HTML links in a GUI tkinter
References: <mailman.1022986323.16196.python-list@python.org> <83elfprs1x.fsf@panacea.canonical.org> <C7JK8.180067$%u2.89342@atlpnn01.usenetserver.com>
Message-ID: <adfqa4$pnf$1@wanadoo.fr>

Steve Holden wrote:
> So, all Aurel needs to know now is how to run wxPython widgets inside
> tkinter programs...

Why? Tkinter has everything to do it:

----------------------------------
from Tkinter import *

root = Tk()
t = Text(root)
t.pack()

def openHLink(event):
  start, end = t.tag_prevrange("hlink",
                               t.index("@%s,%s" % (event.x, event.y)))
  print "Going to %s..." % t.get(start, end)

t.tag_configure("hlink", foreground='blue', underline=1)
t.tag_bind("hlink", "<Control-Button-1>", openHLink)
t.insert(END, "This is a link\n")
t.insert(END, "http://www.python.org", "hlink")
t.insert(END, "\nAnd text goes on...\n")
root.mainloop()
----------------------------------

Control-click on the link and you should see the message "Going to <link 
text>", which may be replaced by a webbrowser.open to actually open the 
URL. To insert a new link, just do:
t.insert(<pos>, <url>, "hlink")

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From david.abrahams at rcn.com  Thu Jun  6 16:15:19 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Thu, 6 Jun 2002 16:15:19 -0400
Subject: numeric limits?
Message-ID: <adofuv$6v$1@bob.news.rcn.net>

Is there some straightforward way to find out the range of a Python Int (in
code)?
The best I could do was to probe it using tricks like:

x = 1
while (x << 1) > 0:
    x <<= 1

TIA,
Dave

--
+---------------------------------------------------------------+
                  David Abrahams
      C++ Booster (http://www.boost.org)               O__  ==
      Pythonista (http://www.python.org)              c/ /'_ ==
  resume: http://users.rcn.com/abrahams/resume.html  (*) \(*) ==
          email: david.abrahams at rcn.com
+---------------------------------------------------------------+




From daniel.dittmar at sap.com  Mon Jun 17 12:28:30 2002
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Mon, 17 Jun 2002 18:28:30 +0200
Subject: tab size
References: <mailman.1024330162.30544.python-list@python.org>
Message-ID: <ael2nf$h3a$1@news1.wdf.sap-ag.de>

tab sizeAlves, Carlos Alberto - Coelce wrote:
> Consider the following code:
>>>> for i in range(10):         print "%d\t"%i,
>
> 0       1       2       3       4       5       6       7       8
> 9
>>>>
> How can I change the default size of tab in formating string \t?!

Not the exact answer to your problem, but how about "%-4d" % i ?

Daniel





From candiazoo at attbi.com  Fri Jun 28 22:42:20 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Sat, 29 Jun 2002 02:42:20 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com>
Message-ID: <3d1d1de6.425769562@netnews.attbi.com>

Again, thank you for all of your input!  I was able to double the speed with
only a few, minor changes.  It seems like the biggest change came when (and I
don't know why) I was selective in importing my os, time and sys routines
(asctime() in particular?!).  I also localized some class/function calls, used
format strings instead of appending strings, etc.

Mike


From bowman at montana.com  Fri Jun 21 09:49:22 2002
From: bowman at montana.com (bowman)
Date: Fri, 21 Jun 2002 07:49:22 -0600
Subject: compile python witout tkinter
References: <mailman.1024645168.28657.python-list@python.org>
Message-ID: <s6GQ8.56$xc4.67404@newsfeed.slurp.net>

kosh at aesaeion.com wrote:
>
> On a machine I need to build python on I need to build it without tkinter.
> How can I do that.

Just moving the tcl and tk headers worked for me. Or I should say, since 
SuSE Personal is light on header files, they weren't there and Python built 
without tkinter.



From catunda at pobox.com  Tue Jun 25 08:37:12 2002
From: catunda at pobox.com (Marco Catunda)
Date: Tue, 25 Jun 2002 09:37:12 -0300
Subject: Thread Memory Leak
Message-ID: <pan.2002.06.25.12.37.12.535288.17223@pobox.com>

Hello,

I've developed a python program with some threads. 
This program is a daemon server. I have seen the
memory of that program growing very fast in a short 
period of time. So I decided to make a little program
to see what's happen.

########################################
import os
import threading

def runthread():
    a = 0
    for i in range(10):
        a += 1

while 1:
    listThread = []

    for i in range(50):
        listThread.append( threading.Thread( None, runthread ) )
    for i in listThread:
        i.start()
    for i in listThread:
        i.join()

    pid = os.getpid()
    print "==================================="
    os.system( "ps -o vsize,rss %d" % pid )
########################################

What's wrong with the above code? I don't know
why this code has memory leak. Take a look at
used memory, after 10 minutes of running the 
used memory grow.

regards

--
Marco Catunda


From marklists at mceahern.com  Thu Jun 13 09:08:32 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 13 Jun 2002 08:08:32 -0500
Subject: using m2crypto to encrypt credit card numbers
In-Reply-To: <20020613194656.04643b54.larooy@xtar.co.nz>
Message-ID: <NCBBLFCOHHDIKCAFGCFBGEGLLEAA.marklists@mceahern.com>

[John La Rooy]
> Whatever you use to identify the credit card number has to be
> guaranteed to be unique for each card number. A hash won't give
> you that. Perhaps you could use public key encryption. Encrypt
> the keys right away on the internet connected machine using a
> key pair. The private key is kept on a separate system and
> can recover the credit card numbers from the cipher.

Thanks for your reply.  If you read my original post:


http://groups.google.com/groups?hl=en&lr=&ie=UTF8&oe=UTF8&selm=mailman.10237
48154.2706.python-list%40python.org

you'll see that's precisely the approach I sketched.  My questions have to
do with:

  o RSA padding
  o key management
  o seeding the encryption library with randomness via /dev/[u]random

> Every purchase made with the card will encrypt the number to the
> same ciphertext which you use to identify that card everywhere in
> your database.

For what it's worth, I have the cards stored in a table keyed by a
uniqueidentifier (guid).  That's what I use (the guid, that is) to refer to
the credit card--not the actual cipher text.  Same thing, in effect, I
suppose.

Cheers,

// mark

-





From logiplexsoftware at earthlink.net  Fri Jun 21 14:38:57 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 21 Jun 2002 11:38:57 -0700
Subject: [newbie] How class works ?
In-Reply-To: <5.1.1.6.0.20020621153612.009ea870@imap.datacraft.com.br>
References: <5.1.1.6.0.20020621153612.009ea870@imap.datacraft.com.br>
Message-ID: <20020621113857.63826c34.logiplexsoftware@earthlink.net>

On Fri, 21 Jun 2002 15:36:14 -0300
Marcus Vinicius Laranjeira wrote:

> 
> 
> I have a problem. I wrote a file with a class. When I try to instantiate 
> this class in another file, I receive an error telling me : "Failed to 
> reload - 'module' object is not callable
> 
> I've tried many things to solve this problem, but I can't find where is the 
> error or how to solve !

It would help us to help you if you posted some code so perhaps we could get an
idea of what you're doing wrong.

Anyway, here's a simple example of creating a class in file1.py and using it in
file2.py:

====== file1.py ========
class foo:
    def __init__(self, arg):
        self.arg = arg

    def printit(self):
        print self.arg

====== file2.py ========
import file1

f = file1.foo("Hello, world")
f.printit()



-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From peter at engcorp.com  Fri Jun  7 19:24:47 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 07 Jun 2002 19:24:47 -0400
Subject: Medium to Large Scale Python Deployments
References: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com> <mailman.1023409393.3361.python-list@python.org> <3D003EC7.C72D27BE@engcorp.com> <3D006320.5040805@web.de> <3D00A209.EC24F671@engcorp.com> <adr2ao$bei$0@216.39.172.122>
Message-ID: <3D0140BF.A239E39E@engcorp.com>

Bengt Richter wrote:
> 
> Is the "comments" number a count of _pure_ comment lines,
> # (like this one) ?

I believe so.  I haven't examined the pycount code yet.

> I think it would be interesting to analyze comments further,
> to see what the relative comment byte count is overall,
> and for lines with both code and comments,
> """like this line:""" # (this has code and comment).

I consider it a mild stylistic no-no to use line-ending
comments, personally.  I believe it's almost always better
to put a block-comment of one or more lines above a block
of code with related purpose.  A good comment there tends to
be more maintainable, and useful, than lots of little comments
on each line.

> Also maybe separate doc-string lines into a separate category,
> since they have a dual nature.

I believe pycount does separate out doc-strings.  There is a
separate column labelled "doc" which, in our code, was relatively
small (we rarely use doc strings, having avoided that aspect of
Python when writing on an embedded system where all the .pyc 
files were stored in our limited 1MB of flash along with the 
interpreter...).

-Peter


From trentm at ActiveState.com  Mon Jun 24 13:51:30 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Mon, 24 Jun 2002 10:51:30 -0700
Subject: Windows versions of Python---pros and cons?
In-Reply-To: <3D138293.85892240@noaa.gov>; from Chris.Barker@noaa.gov on Fri, Jun 21, 2002 at 12:46:28PM -0700
References: <mailman.1024613669.14676.python-list@python.org> <3D138293.85892240@noaa.gov>
Message-ID: <20020624105130.C4637@ActiveState.com>

[Chris Barker wrote]
> Actually, I do use the ActiveState version, but *.py is associated with
> PythonWin, rather than the python interpreter

No, ActivePython registers the "Open" action for .py to run with the
Python interpreter and also registers an "Edit" action to open in
PythonWin. "Open" should be the default action so that double-clicking
on a .py file will run it. You should be a ble to right-click on a .py
file and have "Open" and "Edit" options.


> Now I have to figure out how to change my PATHEXT environment variable,
> which is pretty OT for this group...

Win9x: autoexec.bat
WinME: dunno
WinNT/2000/XP: Start->Settings->Control Panel->System
                    ->Advanced->Environment Variables


Cheers,
Trent


-- 
Trent Mick
TrentM at ActiveState.com




From cliechti at gmx.net  Sun Jun 30 11:39:20 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 30 Jun 2002 17:39:20 +0200
Subject: Win32/Python serial port event character problem
References: <mailman.1025406939.18627.python-list@python.org>
Message-ID: <Xns923DB40D86853cliechtigmxnet@62.2.16.82>

Derek Basch <dbasch at yahoo.com> wrote in 
news:mailman.1025406939.18627.python-list at python.org:

> http://www.geocities.com/dbasch/tempers.txt
> 
> I dont understand what Chris is doing in the middle
> though. Maybe you can answer that Chris?

yeah that's left over from earlier days when there was a loop around the 
reading. it will vanish completely in the future...

chris

-- 
Chris <cliechti at gmx.net>



From gusraan at terra.com.br  Wed Jun 26 13:13:04 2002
From: gusraan at terra.com.br (Raphael Ribeiro)
Date: 26 Jun 2002 10:13:04 -0700
Subject: Problems using samples
Message-ID: <337619fa.0206260913.41e08940@posting.google.com>

1 x = 8
2 y = 4
3 v = x - y
4
5 print v
6 if x = y:
7	print "Eles s?o o mesmo valor"
8	else:
		print "Eles n?o s?o iguais , manja?"

this is my code, it says there's an error in line 6, but i can't find
what **** error is that, can anyone help me?


From peter at engcorp.com  Sat Jun 22 19:04:25 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 22 Jun 2002 19:04:25 -0400
Subject: Python hits the spot
References: <mailman.1024772248.9399.python-list@python.org> <3D14D0FE.9070801@kfunigraz.ac.at>
Message-ID: <3D150279.2E26C067@engcorp.com>

Siegfried Gonzi wrote:
> 
> Roman Suzi wrote:
> 
> > It's no wonder that 1000 lines of uncommented code behave like that, taking
> > into account that you were trying to integrate several (Tcl/Tk, Fortran, C,
> > Python) systems in a hurry, and, I presume, not knowing most of them.
> 
> I am up to the point where you made my day (really). Are you sure do you
> know what you are talking about?

Roman knows exactly what he's talking about, and a large number of 
people here would agree with him, not with you.  Why do you think 
that might be?  Hint: it's not because you're blaming Python and 
we're just being defensive.  There's another reason.

Christian's reply was very to the point here: you are using 
tools that are _very_ well tested and do not exhibit the 
unreliability you attribute to them for 99.999% of other 
users.  The logical explanation is that it _is_ your own
code that is at fault, or the way you have assembled the
components together.  

Would you like us to help you narrow down the source of the problem, 
or have you already made up your mind what it must be, and refuse 
to accept that it might not be Python, or Linux, or even Windows 
which is at fault?  There are people here with decades of
experience solving more difficult problems than this.  Maybe
you'll let them help you?  (If so, you can start by stopping
the comments about "garbage software" and Python "wreaking havoc".
You might note that, in spite of your initial comment, no one 
_else_ is ranting.)

Cheers,
-Peter


From maxm at mxm.dk  Sat Jun 29 15:09:19 2002
From: maxm at mxm.dk (Max M)
Date: Sat, 29 Jun 2002 21:09:19 +0200
Subject: incorrect upper()/lower() of UTF-8
References: <afiecp$oss$1@web1.cup.hp.com> <3phT8.45577$n4.10823047@newsc.telia.net> <m3ofdt9agn.fsf@mira.informatik.hu-berlin.de>
Message-ID: <3D1E05DF.1000608@mxm.dk>

Martin v. Loewis wrote:

> It might be worth noting that the .upper and .lower methods of the
> Unicode string are not locale-aware. Instead, they work on all
> characters independent of the locale (*). 

Ok ... but I have had problems when using the title() method on strings. 
It didn't work correctly with Danish Characters. But When I set the 
locale for other reasons, the title() worked correctly, or rather as I 
wanted ;-)

That's why I said the locale module should be tried.

regards Max M



From tdelaney at avaya.com  Tue Jun 18 04:37:31 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 18 Jun 2002 18:37:31 +1000
Subject: python version?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A39C@natasha.auslabs.avaya.com>

> From: George Hester [mailto:hesterloli at hotmail.com]
> 
> Listen you aren't my daddy.  If you do not want to help 
> constructively then please get off my boat.

Not only have various people given you exactly what you want, but they have
also pointed you to locations where you can understand the solution they
presented.

You have in return heaped abuse on these people.

If you are not willing to take people's suggestions and investigate them
then you will not be a very welcome participant here.

Let me summarise:

---

Jeremy Yallop: Look up sys.version and sys.version_info.

You: Traceback (innermost last): File "<Script Block >", line 5, in ?
...
I think I need a script to do this.

Gerhard: You did not import the sys module.

Dialtone:
>>> import sys
>>> sys.version
'2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)]'
>>> sys.version_info
(2, 2, 1, 'final', 0)

You: Well yours I haven't the slightest idea what you are getting at.

Jeremy: http://www.python.org/doc/current/tut/tut.html

(Note, the above URL will tell you *exactly* what Dialtone's message meant).

Peter: I suggest downloading the tutorial and installing Python on
your own machine and using it interactively from the prompt
before you advance to something like ASP scripting.

(Followed by some examples of using Python).

You: Listen you aren't my daddy.  If you do not want to help constructively
then
please get off my boat.

You: I think you guys don't [know] how to do it.

---

If you want to understand what people tell you, you will need at least a
minimal understanding of the Python language: specifically, importing
modules and calling functions. The easiest way to learn this is to install
Python and use it without the extra complications introduced by ASP. Once
you manage to find the version using Python directly (probably via the
interactive prompt) you have met your stated goal (finding the version). You
may *then* like to try the same thing through ASP - I encourage this.

Tim Delaney




From matt at automagic.org  Sun Jun 23 16:19:36 2002
From: matt at automagic.org (matt)
Date: Mon, 24 Jun 2002 08:19:36 +1200
Subject: Iterating through two lists
References: <mailman.1022234790.26965.python-list@python.org> <7934d084.0206031700.377c691e@posting.google.com> <c.tee-FF2C9D.10574923062002@news.bellatlantic.net>
Message-ID: <3D162D58.7000304@automagic.org>

Sounds like you want to use list comprehension :
http://www.python.org/doc/current/tut/node7.html#SECTION007140000000000000000

eg :

print [(x,y) for x in (1,2,3,4) for y in (10,15,3,22) if x*y > 25]

regards
Matt


Curtis Taylor wrote:
> In article <7934d084.0206031700.377c691e at posting.google.com>,
>  amuys at shortech.com.au (Andrae Muys) wrote:
> 
> 
>>"Denis S. Otkidach" <ods at fep.ru> wrote in message 
>>news:<mailman.1022234790.26965.python-list at python.org>...
>>
>>>On Fri, 24 May 2002, jb wrote:
>>>
>>>j> I have two lists, x and y with the property len(x) = len(y).
>>>j>
>>>j> I should like to achive this (x is a list of class instances)
>>>j>
>>>j>   for (a,b) in (x,y): a.f(b)
>>>j>
>>>j> Is there a fancy way of doing this or have I to introduce an
>>>j> auxillary
>>>j> counter (that is very easy but maybe not very "lispy", that
>>>j> is
>>>j> "python-like").
>>>
>>>If lists are not very lange try this:
>>>
>>>for a, b in zip(x, y): a.f(b)
>>
>>If the lists are large then you probably need something along the lines of 
>>xzip.
>>
>>def xzip(*args):
>>    iters = [iter(a) for a in args]
>>    while 1:
>>        yield tuple([i.next() for i in iters])
>>
>>Andrae Muys
> 
> 
> 
> Hi Andrae,
> 
> I've got a similar situation I'm dealing with, in that I have 2 lists of 
> dictionaries, each of which is of equal length. What's different 
> however, is that I need to do an equality check for the values of 2 of 
> the keys (both of which exist in each dictionary), then set the value of 
> the 2nd dictionary's key to the value of that of the first's, e.g.:
> 
> if dict2[key1] == dict1[key1]:
>    dict2[key2] = dict1[key2]
> 
> I'm concerned with overhead, as the 2 lists could potentially be very 
> large. Can your xzip function be extended to do such a thing?
> 
> Thanks,
> 
> Curtis




From jjl at pobox.com  Tue Jun  4 17:31:51 2002
From: jjl at pobox.com (John J. Lee)
Date: Tue, 4 Jun 2002 22:31:51 +0100
Subject: Problem with urllib.urlopen()
In-Reply-To: <cc14131d.0206032128.5b2efcf6@posting.google.com>
References: <cc14131d.0206031445.5154de31@posting.google.com> <cc14131d.0206032128.5b2efcf6@posting.google.com>
Message-ID: <Pine.SOL.4.30.0206042228330.10073-100000@mimosa.csv.warwick.ac.uk>

On 3 Jun 2002, Christopher wrote:
> There were some problem with posting the whole link so I'll break each
> one up...
[...lengthy URLs snipped...]
> chris_mk at hotmail.com (Christopher) wrote in message news:<cc14131d.0206031445.5154de31 at posting.google.com>...
[...]
> > You should, like I said, get the XML version of the original links
> > (with some extra information, which is why I want to interface with
> > the XML site, btw).  Now, if you do an interface with Python instead
> > of your browser you get a response of an error page instead.  Why is
> > that and what can I do to change it?  If you have any information, I
> > would appreciate it ver much.  Thanks.

What was the error?  Sorry, too lazy to follow the urls.


John



From K.Rdt at TU-Berlin.DE  Wed Jun 26 01:46:03 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Wed, 26 Jun 2002 07:46:03 +0200
Subject: Static py2exe?
Message-ID: <XRYX83FEO1ZVQIE32NHZTD9OLIG3XUO.3d19551b@FRITZweb>

---------------------------------------------------------------------
Hi

Is there a possibility to make a 'static binary',
so that the *.exe hasn't to rely on python2.dll?

		K at Rdt
---------------------------------------------------------------------







From xah at xahlee.org  Wed Jun 12 22:59:47 2002
From: xah at xahlee.org (Xah Lee)
Date: 12 Jun 2002 19:59:47 -0700
Subject: There's more than one way to do it; you bet...
Message-ID: <7fe97cc4.0206121859.749b0c02@posting.google.com>

once i saw a photo on the web where underneath Larry Wall's "there's
more than one way to do it" autograph, Guido van Rossum of Python fame
signed "you bet!". I didn't bookmark the url. Does anyone know where?

I also recall once seeing a photo of some i think Perl monger's
conference in CA which shows a guy who have a big Perl camel tattooed
to his arm. I'd like to save that photo too.

Thanks.

 Xah
 xah at xahlee.org
 http://xahlee.org/PageTwo_dir/more.html


From eddie at holyrood.ed.ac.uk  Fri Jun 21 07:24:36 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Fri, 21 Jun 2002 11:24:36 +0000 (UTC)
Subject: Advice sought on ps module
References: <6e990e29.0206200325.355f0413@posting.google.com> <aesoqg$kc9$1@scotsman.ed.ac.uk> <6e990e29.0206202356.544eb696@posting.google.com>
Message-ID: <aev2dk$pn2$1@scotsman.ed.ac.uk>

hjwidmaier at web.de (Hans-Joachim Widmaier) writes:

>eddie at holyrood.ed.ac.uk (Eddie Corns) wrote in message news:<aesoqg$kc9$1 at scotsman.ed.ac.uk>...

>> I'm not sure that I understand completely what you're trying to do so maybe
>> I'm barking up the wrong gum tree here but ...

>I admit that my writing wasn't very clear ...
> 
>> Surely you want to use postscript itself to do all the work.  PS already has
>> an operator called stringwidth and a complete set of programming constructs so
>> the trick usually is to generate the appropriate PS commands.

>That's how I do it today. Well, I should have given a more concrete
>example. I've written a program that creates good looking timing
>diagrams for documentation purposes. These consist of a drawing part
>and signal names to the left. While the python program knows exactly
>how large the drawing itself is, it has no way of knowing how far the
>signal names extend to the left, making it all but impossible to give
>a valid boundig box for the whole thing -- which is required, as it's
>going to be an EPS file. Sure, I can get at the BBox by piping the
>finished file through one of the ps utils programs, but this is kinda
>cumbersome. And it still doesn't allow me to adjust the right-hand
>drawing width so the combined width is of some fixed size.

OK, that looks like a fairly complex job.  The only other suggestion I can
make is to wonder whether you're getting too hung up on precision.  Does the
BBox really need to be pixel perfect?  I know myself how easy it is, when you
know a perfect solution _should_ be possible, to get fixated on it instead of
just saying, well take the longest string and assume len * em width and see
whether it looks OK (eg would another person notice?).  OK, I'm sure you
wouldn't have gone to the effort of messing about with font metrics for
nothing but don't forget 'good enough' solutions are just that!

>Another example: In my CD cover/inlay generator I have to do insert
>line breaks if the Title+Singer+Time line doesn't fit. I do this in
>PostScript, as I don't know the width in the Python program. This has
>the drawback that the height of the title list changes and my careful
>layout (not to be taken too seriously) is disrupted. And no, I do
>*not* want to write a layout engine in PostScript.

I'm not particularly into PostScript but I have seen a few amazing packages
from time to time, whether you can find something usable though.

>Despite having read most of the PostScript Reference Manual, the AFM
>spec, the DSC specs, etc., ... and having written a bunch of programs
>that output PostScript, I still consider myself a novice (at best!).
>This reverse notation thing with all those stacks gives me the creeps.
>;-)

It's a really elegant model and very suitable in domains where efficiency is
important (like page markup).  I made some effort to learn forth and a little
bit of PS and I thought it was fun but rather too much hard work to do every
day!

>Hope these explanations make the misery I'm in somewhat more clear.

Misery to me is NOT having a challenging programming project :)
Sorry I can't be of any help after all but good luck anyway.

Eddie
P.S.(!!) don't forget comp.lang.postscript

P.P.S. Actually it occurs to me there is another solution, namely creating the
diagram with something else (eg a canvas) and generating (presumably
bitmapped) PS from that.  Though I'm sure you've already rejected that idea as
being unworthy!


From ehab_teima at hotmail.com  Thu Jun  6 18:19:53 2002
From: ehab_teima at hotmail.com (Ehab Teima)
Date: 6 Jun 2002 15:19:53 -0700
Subject: minidom and encoding problem
References: <17aafe08.0206051514.4af054d2@posting.google.com> <m37klcubdu.fsf@mira.informatik.hu-berlin.de>
Message-ID: <17aafe08.0206061419.6366b12b@posting.google.com>

martin at v.loewis.de (Martin v. Loewis) wrote in message news:<m37klcubdu.fsf at mira.informatik.hu-berlin.de>...
> ehab_teima at hotmail.com (Ehab Teima) writes:
> 
> > I'm using Python 2.1. I wrote classes to create xml document from
> > scratch. The code worked fine until I hit an encoding problem. The
> > classes can read text and insert it as is to xml document using
> > creatTextNode. This text had characters > 127, and I got this error.
> 
> This is a bug in your code. You must not insert (byte) string in a DOM
> tree; always use Unicode objects.

I do not have control over the sent text. The issue started when some
bullets were copied from a word document and pasted into a file and
the whole file was passed to my classes. I cound not find a way to
convert this text to UTF-8 or anything else. Is there a way to prevent
this from happening?
> 
> > I know it's not possible to add an enconding attribute using writexml,
> > so the generated document only has <?xml version="1.0"?>. Is there any
> > way to get around this problem. 
> 
> Yes. Use Unicode strings when creating text nodes. When producing the
> serialized document through .toxml, you will find that it produces a
> Unicode string. Since (as you notice) the document has no encoding
> declaration, you need to .encode("UTF-8") that string before saving it
> into a file.

I tried to encode the string using different encodings but I could
not. Here is what I got when I tried .encode("UTF-8"):

UnicodeError: ASCII decoding error: ordinal not in range(128)


From donn at drizzle.com  Wed Jun  5 02:16:13 2002
From: donn at drizzle.com (Donn Cave)
Date: Wed, 05 Jun 2002 06:16:13 -0000
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <3CFCF3E7.AE891EBE@noaa.gov> <adiumj$22g2$1@nntp6.u.washington.edu> <adjduf$92g$0@216.39.172.122> <slrnafqp12.4ea.huaiyu@gauss.almadan.ibm.com>
Message-ID: <1023257772.419542@yasure>

Quoth huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu):
...
| That is one proposal.  My proposal is to disallow += for immutables.
| Another symbol would be used for rebinding operations.  (Or maybe swap the
| role of these symbols.)  Thus for mutables both operations would be
| available.

That suits me, except the mutable/immutable distinction.  That's
how we have been describing the distinction, but it's worth noting
that it isn't really what we mean.

For a silly example, suppose that in 2.3, None supports +=, kind of
like the UNIX /dev/null.  This would not make it mutable in any
reasonable sense of the word, but it's in the "in place" category.
A class instance that implements __iadd__ to return something other
than self, would not be, though it's certainly mutable.  In any
normal case, "in place" and "mutable" go together, but they aren't
the same thing.

	Donn Cave, donn at drizzle.com


From pinard at iro.umontreal.ca  Tue Jun  4 09:11:37 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 04 Jun 2002 09:11:37 -0400
Subject: Python Turing machine
In-Reply-To: <3CFC539B.BCBFC855@replyto.address.invalid>
References: <mailman.1023068284.27231.python-list@python.org>
	<3CFC539B.BCBFC855@replyto.address.invalid>
Message-ID: <oqptz7w5me.fsf@titan.progiciels-bpi.ca>

[Greg Ewing]

> "Delaney, Timothy" wrote:

> > Yeah - the dependence on an infinitely-long tape can be a real bugger ...

> well, any terminating computation only uses a finite amount of tape,
> so you just need to be prepared to buy more tape as needed.  Of course,
> if the universe is closed, you could find yourself in a situation of
> having run out of matter to make new tape from.

This situation is easily met.  Many of you known the Ackerman function:

def ackerman(i, j):
    if i == 0:
        return j + 1
    if j == 0:
        return ackerman(i-1, 1)
    return ackerman(i-1, ackerman(i, j-1))

in which, as you can see, the most "aggressive" operation towwards
bloating the resulting value is adding 1 to `j'.  When I was a teenager,
a few friends and I had a lot of fun at computing it for various small
argument values, but were just unable to climb the arguments to 4 and 4.

A few years later, having to spend a full night waiting, and to kill time,
I revisited the `ackerman(4, 4)' problem, to find out that the estimated
universe is much to small and much too young for having ever computed this
very odd value[1].

If each elementary particle of the universe was a bit, we would not have
enough bits.  If we consider one addition per unit of time, and the fastest
computer clock we can think of, there would not have been enough cycles
since the big bang.  For this computer, I set the clock period to the time
it takes for light to traverse a proton (which, contrarily to many thinks,
is much smaller than an electron).

Strange that such big a number, which is undoubtedly both well defined
and computable theoretically, could be represented by so little Python!

----------
[1] This was a while ago, maybe I do not remember well, but I got that
`ackerman(4, 4)' is `2 ** (2 ** (2 ** (2 ** (2 ** (2 ** 2))))) - 3'.
Because of this, I may dare writing that we get an odd number! :-)

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From bokr at oz.net  Tue Jun 25 15:46:54 2002
From: bokr at oz.net (Bengt Richter)
Date: 25 Jun 2002 19:46:54 GMT
Subject: Suggestions for good programming practices?
References: <mailman.1025012299.3147.python-list@python.org>
Message-ID: <afahbe$6n2$0@216.39.172.122>

On Tue, 25 Jun 2002 14:31:48 +0100, Simon Brunning <SBrunning at trisystems.co.uk> wrote:

>> From:	Emile van Sebille [SMTP:emile at fenx.com]
>> Aahz
>> > In Python 2.1.2 or higher, fire up the interactive interpreter and
>> type
>> > "import this".
>> 
>> Except ActiveState's
>> 
>> F:\Python22as>python
>> Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
>> Type "help", "copyright", "credits" or "license" for more information.
>> >>> import this
>> Traceback (most recent call last):
>>   File "<stdin>", line 1, in ?
>> ImportError: No module named this
> 
>It isn't in the 2.2 series until 2.2.1, IIRC.
>
I think it must be a distribution glitch. My 2.2 banner looks identical:

 Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
 Type "help", "copyright", "credits" or "license" for more information.
 >>> import this
 The Zen of Python, by Tim Peters 

 Beautiful is better than ugly.
 Explicit is better than implicit.
 Simple is better than complex.
 ...

Regards,
Bengt Richter


From peter at engcorp.com  Wed Jun 19 08:36:25 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 19 Jun 2002 08:36:25 -0400
Subject: What If.....    Strong Types
References: <3D0C0F0D.4030900@bgb.cc> <3D0E8C26.7BF0772D@engcorp.com> <3D1011F1.3060205@bgb.cc>
Message-ID: <3D107AC9.680E09EE@engcorp.com>

Don Garrett wrote:
> 
> Peter Hansen wrote:
> > Preventing modifying methods on an instance would make some of my
> > automated testing really difficult, if not impossible.
> 
>    Hum... that's the only specific example I've ever run across where someone
> would need to switch out a method implementation at run time in production
> code. Depickling and several other very nice features probably depend on it,
> but they are core libraries, not general code. I draw a distinction there.
> 
>    I suggest that you could create a testing subclass that overrides the
> methods in questions with testing versions of the methods.
> 
>    Would that work as well? As easily? 

I don't think it would in all cases, although I'm sad to admit I 
didn't actually think of it before.  Python is just too darned accessible
in this way. :-)   

Since we're testing the code sort of "in situ", I cannot always ensure 
that the objects created would be of my subclass without changing other 
parts of the code.  Think of an factory object embedded inside something 
else, and I'm trying to test that one of the items returned responds in 
the proper fashion by dynamically replacing one of its methods which is 
called by another method which I'm going to call.  

Were the code written in a beautiful, elegant, highly decoupled
fashion in the first place I think what you suggest would work.  Since
it's not, the very pragmatic aspects of Python that allow me to do
what I do are very helpful.  I don't believe any highly static language
would let me do this sort of thing without great inconvenience.

-Peter


From hesterloli at hotmail.com  Tue Jun 18 06:53:30 2002
From: hesterloli at hotmail.com (George Hester)
Date: Tue, 18 Jun 2002 10:53:30 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com>
Message-ID: <KiEP8.44424$GY.13974606@twister.nyroc.rr.com>

"If you can't get an answer, please don't take it personally that we don't feel
we can help you. Sometimes the members of the asked group may simply not know
the answer."

--
George Hester
_________________________________
"George Hester" <hesterloli at hotmail.com> wrote in message
news:K7PO8.16057$uk2.8406537 at twister.nyroc.rr.com...
> What is a simple script that will tell me the version of python I am currently
> running?  IIS 5.0 Windows 2000.
>
> --
> George Hester
> _________________________________
>
>




From jtt at hnc.com  Wed Jun  5 13:03:36 2002
From: jtt at hnc.com (Jim)
Date: 5 Jun 2002 10:03:36 -0700
Subject: DCOracle2 - Can't get cursor.execute to return data when SQL query involves datetimes.
References: <be0d725a.0206041139.1f8f6647@posting.google.com>
Message-ID: <be0d725a.0206050903.7afed7ae@posting.google.com>

I have figured out the problem.  It turns out the the TZ environment
variable was not set correctly (off by 1 hour).  This caused the
DCOracle2.Timestamp call to return the hour incorreclty.


From zneptune at zexcite.zcom  Wed Jun 26 13:43:36 2002
From: zneptune at zexcite.zcom (Conrad)
Date: Wed, 26 Jun 2002 17:43:36 GMT
Subject: tomorrow in yyyymmdd format
References: <20020619004409.GA2583@lilith.my-fqdn.de> <XFMail.20020618183810.shalehperry@attbi.com> <mailman.1024451604.17508.python-list@python.org>
Message-ID: <pan.2002.06.26.10.53.26.183469.1298@zexcite.zcom>

Years ago, Nostradamus predicted that on Tue, 18 Jun 2002 20:52:25 -0500,
Gerhard H?ring would write, saying:

> * Sean 'Shaleh' Perry <shalehperry at attbi.com> [2002-06-18 18:38 -0700]:
>> >> 
>> >> Sure there is :-) time.mktime() can take a "malformed" time tuple
>> >> and do something sensible with it. You can therefore get a time
>> >> tuple from gmtime, add one to the day position and call
>> >> time.mktime() on the result:
>> >> 
>> >> >>> x = (2002, 2, 29, 0, 31, 42, 2, 170, 0)
>> >> >>> time.gmtime(time.mktime(x))
>> >> (2002, 3, 1, 7, 31, 42, 4, 60, 0)
>> >> >>>
>> >> >>>
>> > Pure luck that this works. Don't complain if it stops workin in
>> > Python 2.7 and your code breaks ;-)
>> > 
>> > 
>> actually that is a function of the mktime() function in the C library
>> which the python module wraps.  Based on my reading of a man page or
>> two this behaviour seems guaranteed by POSIX.  Now, the assumption made
>> is the mktime() function on YOUR system follows POSIX.
> 
> Fine. I've just read the Python documentation which doesn't make this
> guarantee. I'll leave it to others to check if this works on win32,
> MacOS, as I'm using mxDateTime for this kind of tasks, anyway.
> 
> Gerhard


Hmm  - here's what I've done - a bit klutzy but it seems to work:

from time import *
one_day = 60*60*24                       # number of seconds in day
   # make a julian for tomorrow -
   # note: depending on what you want be careful around midnight!
tomorrow = mktime(localtime()) + one_day
print 'tomorrow is',localtime(tomorrow)
print 'today is',localtime()


tomorrow is (2002, 6, 27, 10, 42, 11, 3, 178, 1)
today is (2002, 6, 26, 10, 42, 11, 3, 178, 1)

Conrad


From phlip_cpp at yahoo.com  Sun Jun  2 13:06:21 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 02 Jun 2002 17:06:21 GMT
Subject: Reading coords from a scrolled Tkinter Canvas widget?
References: <addeiv$df1$1@news6.svr.pol.co.uk>
Message-ID: <addjad$3cp@dispatch.concentric.net>

G. Willoughby wrote:

> I am having a few problems reading the correct coords from a scrolled
> canvas widget! e.g. i have a .jpg about 2000x2000pixels displayed inside a
> canvas widget measuring 564x564pixels when i scroll the canvas using the
> attached scrollbars i want to click the mouse and read the x/y coords of
> the .jpg where my cursor is, not the canvas widget, any ideas? This is the
> code i have at the minute:

Your code appears not to grok canvasx and canvasy. They translate event.x 
style coordinates to canvas model coordinates.

BTW you could be using a Python Megawidgets Pmw.ScrolledCanvas here. More 
features less pain.

-- 
  Phlip
            http://www.greencheese.org/MakeItSo
  --  DARE to resist drug-war profiteering  --



From tjreedy at udel.edu  Tue Jun  4 15:22:18 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 04 Jun 2002 19:22:18 GMT
Subject: self
References: <mailman.1023208220.19083.python-list@python.org> <TL7L8.79$4p7.4100181@newnews.cc.stevens-tech.edu>
Message-ID: <Jr8L8.111119$%y.11617831@bin4.nnrp.aus1.giganews.com>

"Vojin Jovanovic" <vjovanov at stevens-tech.edu> wrote in message
news:TL7L8.79$4p7.4100181 at newnews.cc.stevens-tech.edu...
[snip]
> Why is this all relevant? Well, I am working on embedding an
interpreted
> language in a large engineering application and Python is  one of
few
> candidates.  So I have been looking for its pros and cons.  So far
it passed
> all the tests except the above.  Because, imagine if we required
from the
> user of our system that he needs to type self. or even s. in front
of every
> new variable that he has in his equation, it would be a big turn
off.

Now that you have explained your actual problem, I believe there is a
simple solution that does not require breaking half of all Python code
written.

Add another dict 'Symbols' to your class.  Then eval/exec your
equations 'in' Symbols as the local/global namespace.  No 'self.'
needed then.  This is also a good idea for security reasons, but
others have posted on that.   Plus, you can initialize 'Symbols' with
the contents of the math module and add any other engineering
functions you want available for use in equations.

Terry J. Reedy




From tom at aragne.com  Sat Jun  1 06:02:56 2002
From: tom at aragne.com (Tom Deprez)
Date: Sat, 1 Jun 2002 12:02:56 +0200
Subject: EuroPython2002: Extention of Early-Bird registration
Message-ID: <007901c20953$82fe3ab0$a18d84d5@skullsplitter>

Hi,

Due to the fact that at the end of the early-bird date, lot's of people
waked up and registered for the congress  (and because of the possible
internet problems caused by the bancruptcy of KPNQwest), we've extended
the early bird until 9th of June. We hope that the doubters among us,
will find this interesting now that they are still able to profit from
the early bird registration. But be warned, this is the final date!

At the moment we've reached a total of more than 150 people at the
EuroPython congress. That's already nice for a first congress, but it
should be more.... Don't you think you could make great deals with
people over there? There must be somebody to whom you can speak about a
certain problem. With 150 people, it must be possible to make some
friends or to see some old ones back... It can't be that Europe has only
such a small group of Python and/or Zope developers/users. What about
people outside of Europe? Have you seen the talks? It must be worthwhile
to pay a visit to the conference.... and if not... well enjoy the
Belgian beer, chocolates and food..... that's something I'm sure you
will never forget!

And now the early bird registration has been extended, .... so if you
had doubts earlier, throw them away and just visit EuroPython2002!

Don't forget to visit the http://europython.zope.nl/interviews page,
which will be regulary updated with new interviews of people comming to
EuroPython2002.

Regards,
EuroPython Team.






From jepler at unpythonic.net  Mon Jun 24 12:31:35 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Mon, 24 Jun 2002 11:31:35 -0500
Subject: Embedding in Tkinter (reprise)
In-Reply-To: <1907a5e4.0206240803.6ca0a4c@posting.google.com>
References: <1907a5e4.0206240803.6ca0a4c@posting.google.com>
Message-ID: <20020624163135.GC2142@unpythonic.net>

Search for the "tksteal" package.  This adds a Tk widget that can
"swallow" or "steal" an arbitrary window when its window ID is given.
Once you've compiled this and it is available to your Tcl/Tk
installation, you need merely define a Python class that wraps this new
widget type, like the existing ones in Tkinter.py.

I found a manpage here:
http://www-komo.ise.eng.osaka-u.ac.jp/misc/tcl/tk_steal_man/tksteal.n.html
One copy of the source code for this widget seems to be at
ftp://ftp.neosoft.com/languages/tcl/sorted/packages-7.6/devel/TkSteal4.0c.tar.gz
but there are a lot of dead links related to this piece of software.

Jeff




From spammers.do.not.bother at void.bogus  Mon Jun 24 10:28:12 2002
From: spammers.do.not.bother at void.bogus (Magnus)
Date: Mon, 24 Jun 2002 14:28:12 GMT
Subject: sys.exc_info() into string
References: <KZoR8.44805$n4.10504231@newsc.telia.net> <3D1635B2.99BCD944@engcorp.com>
Message-ID: <00GR8.44905$n4.10556369@newsc.telia.net>

Peter Hansen wrote:

> Magnus wrote:
>> 
>> in my program I am catching exceptions that I might have missed by using
>> something like:
>> 
>> try:
>>         doMyStuff()
>> except:
>>         type = sys.exc_info()[0]
>>         syslog.syslog(syslog.LOG_EMERG, "My program: unexpected error" +
>>         type)
>> 
>> The thing is that I can not figure out how to make "type" into a string
>> above so I can use it for e.g. syslog(). Is there an easy way of doing
>> this?
> 
> try:
>     1/0
> except Exception, ex:
>     print "Error: %s" % ex
> 
> Good enough??
> 
> -Peter


Thanks both of you. I'm still learning Python and issues that are trivial to 
solve for some isn't trivial for me yet. I'm converting from C++ and Java 
and perhaps in a few months I might be a more experienced Python 
programmer.

Thanks,
Magnus



From stephen at xemacs.org  Tue Jun  4 23:31:06 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 05 Jun 2002 12:31:06 +0900
Subject: prototyping good OOdesign in Python?
References: <mailman.1023075244.23766.python-list@python.org>
Message-ID: <873cw2csg5.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Timothy" == Timothy Delaney <Delaney> writes:

    Timothy> Unfortunately, this will only work for 1 or 2 projects
    Timothy> ... after that they will *expect* you to bring everything
    Timothy> in under time and budget ... but by then the time and
    Timothy> budget is massively reduced because of the great results
    Timothy> you've brought in ...

Of course, if your managers figure that out in only one or two
projects, you take your raise in stock.  If they're dumb enough not to
understand that the third project is a little late and over budget
because they've underestimated the needs, you take a short position
... and project 4 is massively over budget because you spent all the
time working on your resume.<wink>

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
 My nostalgia for Icon makes me forget about any of the bad things.  I don't
have much nostalgia for Perl, so its faults I remember.  Scott Gilbert c.l.py


From detlev at die-offenbachs.de  Sat Jun 15 04:52:59 2002
From: detlev at die-offenbachs.de (Detlev Offenbach)
Date: Sat, 15 Jun 2002 10:52:59 +0200
Subject: IDE recommendations??
Message-ID: <b9veea.l85.ln@majestix.gallien.de>

Hi all,

I have googled this list for IDE recommendations, but have found none 
that satisfied me. I am looking for an IDE that allows development of 
Qt/KDE applications using Python. So far I have found BlackAdder. 
However, it looks as if development has stalled. I could use individual 
applications (e.g. qt-designer for gui, scite as an editor, pyuic, ...), 
but this lacks some comfort. I would appreciate any recommendation for a 
suitable IDE

Regards
Detlev
-- 
Detlev Offenbach
detlev at die-offenbachs.de


From mcfletch at rogers.com  Wed Jun 12 05:06:47 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Wed, 12 Jun 2002 05:06:47 -0400
Subject: Python Virtual File System: A spike test
Message-ID: <3D070F27.9050608@rogers.com>

Purpose:

A class for representing file system paths which supports both 
method-based operations, and abstraction through object encapsulation to 
eventually allow for Virtual File System operations (for instance, 
transparently accessing the contents of zip files).

Status:

An initial implementation (spike test) exploring the viability of the 
approach.  Implementation provides:

	equality testing
	parent/child/ancestor/descendent retrieval/testing
	filesystem-role testing (file/dir/root)
	root retrieval and root comparisons
	"walking" (callback-based recursive iteration)
	listing (returning Paths which include their parent directory)
	fragment determination (i.e. c:\\temp\\test -> "c:\\", "temp", "test"
	recursive directory-tree removal (including files)
	open( ) method for path specifiers

The current implementation is a sub-class of str which provides a number 
of new methods (and overrides some built-in methods) to provide the 
above functionality.

Future:

If there is interest, I am planning on creating a new SourceForge 
project with the eventual goal of creating a virtual file system package 
with a consistent interface to most of the major file-and-directory 
systems (zip files, FTP sites, local file systems) and hooks for adding 
new systems.

Of course, it would be considerably better for such a project if the 
standard library returned Path objects, rather than simple strings, but 
I'm not particularly interested in trying to get the project accepted 
into the core at the moment (after all, this is just a test).

Request:

If you're interested in working on such a project (or even using it), 
let me know, so I can decide whether to devote any more resources to it.


You can find the module here:

http://members.rogers.com/mcfletch/programming/filepath.py

Enjoy yourselves,
Mike

_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From irmen at NOSPAMREMOVETHISxs4all.nl  Thu Jun 27 18:33:06 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Fri, 28 Jun 2002 00:33:06 +0200
Subject: M2Crypto: select() behaves weird on SSL sockets
References: <3D04ED17.5070606@NOSPAMREMOVETHISxs4all.nl> <ae2str$b0q$1@nntp6.u.washington.edu> <3D066E77.4070901@NOSPAMREMOVETHISxs4all.nl> <ae6bmb$frc$1@dahlia.singnet.com.sg> <3D0790E1.9080904@NOSPAMREMOVETHISxs4all.nl> <831ybavprx.fsf@panacea.canonical.org> <3D0919AF.7050903@NOSPAMREMOVETHISxs4all.nl> <83it45lxfb.fsf@panacea.canonical.org>
Message-ID: <3D1B92A2.7030605@NOSPAMREMOVETHISxs4all.nl>

Kragen Sitaker wrote:

> There are many pieces of middle ground between complete threading (two
> threads per connection) and complete event-loop-drivenness with just
> one thread --- as you're obviously aware, since you're using select()
> in a threaded program.

Um, yeah. I have 1 loop using select() that waits for new incoming
connections. Every thread that is started for a connection just
recv()s data off the socket.

> What I guess you're not aware of is that asyncore supports the middle
> ground (middle colors?) as well as the event-loop-driven end of the
> spectrum.  For example, Zope has an asyncore thread that handles I/O,
> but actually processes requests and produces results in some child
> threads.

You're probably right. I don't know a lot about the workings of asyncore.
Perhaps if I finally squash the bug that is still in Pyro 3
-- *blushes* it's probably a threading race condition -- I'll spend
some time looking at it.

Doing stuff the Zope way, the worker threads must have some way to
get to the input and output streams that are handled in asyncore, right?


Irmen





From whisper at oz.net  Fri Jun  7 18:44:34 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 7 Jun 2002 15:44:34 -0700
Subject: Python MTA
Message-ID: <GCEDKONBLEFPPADDJCOEAEHDDHAA.whisper@oz.net>

Has anyone developed a simple/stupid Mail Transfer Agent using Python that
does't rely on any external mail agents (and thus would be platform
independent)?

What I have in mind is an app that sits between my mail client (outlook
(dev/null=flames)) and, at most, a few mail hosts. It would act like a
client to those mail hosts and as a host to outlook. The idea is to be able
to insert a spam filter like ASK (or something better?).

Pointers to any such and/or other spam filters appreciated.

TIA,

David LeBlanc
Seattle, WA USA





From theller at python.net  Thu Jun  6 13:24:23 2002
From: theller at python.net (Thomas Heller)
Date: Thu, 6 Jun 2002 19:24:23 +0200
Subject: Emulating classmethod in Python 2.1
References: <915a998f.0206060833.12ad76a8@posting.google.com>
Message-ID: <ado5v8$uih9$1@ID-59885.news.dfncis.de>

"Hamish Lawson" <hamish_lawson at yahoo.co.uk> wrote in message news:915a998f.0206060833.12ad76a8 at posting.google.com...
> Is it possible to define a 'classmethod' class or function for Python
> 2.1 that would effectively provide the functionality of Python 2.2's
> built-in classmethod function and have the same usage pattern? I'd
> like to be able run code like that below against both 2.2 (using its
> builtin classmethod function) and 2.1 (using some added-in wrapper).
>
>     class SomeClass:
>         def hello(cls, name):
>             print "Hi there", cls, name
>         hello = classmethod(hello)
>
>     SomeClass.hello("Peter")
>
> I've come across Thomas Heller's recipe in the online Python Cookbook,
> but this would require SomeClass to be modified, and so fails my
> criteria.
>
Here's the closest I came up with (you must rewrite your code to
derive SomeClass from object, the rest can remain unchanged).

Thomas

-----------

try:
    object
except NameError:
    class classmethod:
        def __init__(self, func):
            self.func = func

    class _ClassMethod:
        def __init__(self, cls, func):
            self.cls = cls
            self.func = func

        def __call__(self, *args, **kw):
            return self.func(self.cls, *args, **kw)

    import ExtensionClass
    object = ExtensionClass.Base
    class object(ExtensionClass.Base):
        def __class_init__(cls):
            d = {}
            for name, value in cls.__dict__.items():
                if isinstance(value, classmethod):
                    d[name] = _ClassMethod(cls, value.func)
            cls.__dict__.update(d)

#######

class SomeClass(object):

    def hello(cls, name):
        print "Hi there", cls, name
    hello = classmethod(hello)

SomeClass.hello("Peter")





From bt98 at doc.ic.ac.uk  Mon Jun 10 06:47:11 2002
From: bt98 at doc.ic.ac.uk (Benjamin Tai)
Date: Mon, 10 Jun 2002 11:47:11 +0100
Subject: PyCObject subtype PyObject
References: <3D022727.2147C50E@doc.ic.ac.uk> <lk1ybflaua.fsf@pc150.maths.bris.ac.uk>
Message-ID: <3D0483AF.5ACB350@doc.ic.ac.uk>

Michael Hudson wrote:

> > 2) For function "PyCObject_FromVoidPtr()", it takes in a "void*" as the
> > first argument.
> > For function "PyCObject_AsVoidPtr()", it returns a "void*".
> > When implementing Python extension, what is the potential danger of
> > casting pointers, of different size, to and from "void*"?
>
> Doesn't C guarantee that you can cast a pointer of type (T*) to
> (void*) and back again for all T?  I was fairly sure it did.
>

I am extending a C++ application with Python. Before attempting other
possibilities, like Boost, I just want to document the work I have done so
far.

Simply speaking I am hiding a C++ pointer, as a PyCObject*, inside a Python
instance.

The scheme works OK. The C++ class hierarchy is reflected by Python classes.
The pair of C++ and Python object is passed around between the two
languages.



> > 4) Apart from exporting function between modules, are there any
> > examples of using PyCObject?
>

The example, in the Python/C API document, is talking about exporting
functions between Python modules. So I am wondering if my method is valid?



Thanks for your feedback.

Ben



From martin at strakt.com  Thu Jun 13 03:18:13 2002
From: martin at strakt.com (Martin =?iso-8859-1?Q?Sj=F6gren?=)
Date: Thu, 13 Jun 2002 09:18:13 +0200
Subject: Extension objects in Python 1.5 and 2.2
Message-ID: <20020613071813.GA16764@strakt.com>

How do I write extension objects so that they work in both Python 1.5 and
the 2.x series? Until now, I've used PyObject_New and PyObject_Del for
2.x, are there an equivalent set for 1.5 and if so, can they be used in
2.x too, or do I have to #ifdef?

What's the correct way to do this? And gee, it would be nice if the
objects were classes in 2.2 too, but I suppose that'd be a LOT of #ifdefs.


Regards,
Martin

-- 
Martin Sj?gren
  martin at strakt.com              ICQ : 41245059
  Phone: +46 (0)31 7710870       Cell: +46 (0)739 169191
  GPG key: http://www.strakt.com/~martin/gpg.html




From kseehof at neuralintegrator.com  Mon Jun 10 05:44:39 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Mon, 10 Jun 2002 02:44:39 -0700
Subject: Does Python need a '>>>' operator?
In-Reply-To: <Pine.GSO.4.33.0206100850550.12342-100000@techst02.technion.ac.il>
Message-ID: <EFEFIHAINIINNELLHFAIEEMJCIAA.kseehof@neuralintegrator.com>

Beni Cherniavsky <cben at tx.technion.ac.il> wrote:
> I just got another idea: use 0x1234 for 0-filled numbers and 1xABCD for
> 1-filled ones.  That way you impose no restrictions on what follows the
> prefix and keep backward compatibility.  0xFFFFFFFF stays a 2^n-1
> _positive_ number, as it should be.  The look of 1x is weird at first but
> it is very logical...

Hey, that's pretty clever.  I like it.  One objection I can forsee is
that it does add another little thing to learn.  And yes, one must
consider such a feature as making python more complex if it requires
a paragraph to be added to the main documentation, even if the feature
will only be used by "advanced" programmers.  The standard argument
is that all python programmers have to be able to read your code.

However, I don't see this as making life much more complicated for
beginners, since we already have 0x1234, which is only familiar to
experienced programmers.  People who are not familiar with 0x1234 can
learn about 1xABCD at the same time.  People who are familiar with
0x1234 (e.g. from programming in C/C++), can probably handle looking up
the proposed syntax in the documentation, if they can't guess what it
means by experimenting in the interactive interpreter.

In particular, I like:

  1xF == -1

Seems more pythonic than 0xFFFFFFFF, since the latter implies knowledge
that we are using 32 bits.  I don't know of any current way to express
1 filled binary numbers cleanly.

Unfortunately, I don't see an easy way to clean up this blemish:

>>> 0xffff
65535
>>> 0xffffffff
-1
>>> 0xfffffffff
68719476735L

I suppose in a perfect world, 0xffffffff would be 4294967295L, but
we'd have serious compatibility issues with that (note that one
would use 1xf instead of 0xffffffff to represent -1 in the perfect
world).

BTW, note that 0xffffffff == 4294967295L would be consistent with
C/C++ with an unsigned int (32 or more bits).  The idea is that
all numbers 0x... are non-negative, while number 1x... are negative.

I'm quite certain that any proposed solution (such as issuing a
warning for 0xXXXXXXXX) will receive flames, so I think I will
stop here, hoping it's not too late :-)

- Ken Seehof





From emile at fenx.com  Wed Jun  5 22:57:21 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 06 Jun 2002 02:57:21 GMT
Subject: NEWBIE - SendKey equivalent in Python
References: <lrgtfus01fgmc3sd49djos91812qj91hhl@4ax.com>
Message-ID: <lcAL8.1965$pw3.42@sccrnsc03>

qigong42 at hotmail.com
> Is there a way to send keystokes to another application with Python?
> 
> If so, can someone post an example?
> 

This gets 68 hits:

http://groups.google.com/groups?q=python+sendkeys&ie=UTF8&oe=UTF8&hl=en

--

Emile van Sebille
emile at fenx.com

---------



From whisper at oz.net  Sat Jun 22 20:06:42 2002
From: whisper at oz.net (David LeBlanc)
Date: Sat, 22 Jun 2002 17:06:42 -0700
Subject: The best GUI library in python
In-Reply-To: <8d3f4438.0206221422.6f71287a@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEIEJCDLAA.whisper@oz.net>

<snip>
> Well, if you doubt what I said, try to use grail. Or write a listbox
> containing 10,000 items, then wait......

Or don't abuse a listbox by insisting that a GUI component also be an
efficient container object. Take the trouble to shift lines in/out of the
listbox as the user scrolls and don't wait around at all. Wow, you could
even create your own BufferedListbox class and be our hero :-)

I've heard of a scheme called Grail, but i've never used it ;-)





From cliechti at gmx.net  Mon Jun 10 19:59:33 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 11 Jun 2002 01:59:33 +0200
Subject: newbie question - python blocking on int()?
References: <4LaN8.2488$Vr2.586019@news20.bellglobal.com>
Message-ID: <Xns922A149F19723cliechtigmxnet@62.2.16.82>

Michael Davis <michael at damaru.com> wrote in 
news:4LaN8.2488$Vr2.586019 at news20.bellglobal.com:
> I'm writing a specialized ftp client. I'm parsing the output of ftp.dir, 
> which gives me a string like this (call it str):
> 
> -rw-r--r--   1 fred       527 Jun  4 22:58 report.php
> 
> and I'm getting the various parts like this:
> 
> details = string.split( str )
> permissions = details[0]
> size = details[3]
> name = details[7]
> debug( "added remote file size %5d: %s" % (int(size), name) )
> 
> This works. But when I replace the 3rd line with this:
> 
> size = int( details[3] )
> 
> python hangs on that line. Why?

not when i try it... what do you mean by "hangs" does it takes forever or 
do you see an error message? if its the later, then you might have assigned 
something to "int" somewhere else, hiding the builtin function.

note that you should not use "str" as a name as that is a builtin 
object/function that can be very handy. e.g. in python 2.2+ you could have
written "details = str.split(line)"
(or even "details = line.split()" which i prefer)

also note that split() might not be safe for every case. e.g when the user 
name contains a space (if possible) and surely when the filename has a 
space in it.

chris

-- 
Chris <cliechti at gmx.net>



From logiplexsoftware at earthlink.net  Sat Jun 15 07:57:51 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: 15 Jun 2002 04:57:51 -0700
Subject: Newbie deluxe
In-Reply-To: <hdDO8.29984$Hn4.969427@news1.east.cox.net>
References: <hdDO8.29984$Hn4.969427@news1.east.cox.net>
Message-ID: <1024142273.5870.26.camel@localhost.localdomain>

On Sat, 2002-06-15 at 01:50, Zach wrote:
> I've just started taking a look at Python (experienced in Delphi) and so far
> it looks very clean and comprehensible.  My question is, what are the
> fundamental differences and similarities between it and Perl?  

IMO, Perl is a scripting language, Python a programming language (the
fact that they are both interpreted being incidental).  Obviously what
qualifies as a "scripting" language versus a "real" lanaguage is
arguable (scripts can be written in C, if you like, and applications in
a combination of sed, awk and sh... oh, wait... that's Perl ;) 

While Python makes an excellent scripting language for doing sysadmin
chores, text parsing and whatnot, its clean design lends itself to large
projects, even more so than languages traditionally used in that
endeavor (C/C++, etc).

I hear so
> much about these two languages, especially in Linux circles and just
> wondered what sets them apart.  Can full fledged graphical applications be
> built with Python or does it lean more towards scripting? 

Check out wxPython: 

http://www.wxpython.org

There are also bindings for Win32, Qt, GTK, Tk and several others.

 I'm aware that it
> has OOP capabilities so I assume it reaches beyond just scripting but like I
> said, I've no experience with it.  I would like to broaden my coding skills
> with the addition of another language so any general opinions with regards
> to its usefulness are appreciated.

Try it and see =)

Regards,
Cliff






From tdelaney at avaya.com  Thu Jun 20 00:28:50 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 20 Jun 2002 14:28:50 +1000
Subject: Variable Interpolation - status of PEP 215
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A3A7@natasha.auslabs.avaya.com>

> From: Gerhard H?ring [mailto:gerhard.haering at gmx.de]
> 
> * Ganesan R <rganesan at myrealbox.com> [2002-06-20 08:50 +0530]:
> > 
> It's currently being discussed on the python-dev mailing list.

He he he - I was about to point out that the time machine strikes again.

http://www.python.org/peps/pep-0292.html

Tim Delaney




From romany at actimize.com  Mon Jun 17 14:17:43 2002
From: romany at actimize.com (Roman Yakovenko)
Date: Mon, 17 Jun 2002 20:17:43 +0200
Subject: Long pathes on windows 
Message-ID: <7647A9F4B298E74DB6793865DA67285004AD86@exchange.adrembi.com>

Good evening. I have big problem Windows 2000 and small one.
The small one is long pathes on windwos. How can I convert long path with ' '
into short path? 

Thanks for the help. Roman.




From SBrunning at trisystems.co.uk  Thu Jun 13 05:16:50 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 13 Jun 2002 10:16:50 +0100
Subject: How to check version of Python under CGI
Message-ID: <31575A892FF6D1118F5800600846864DCBD344@intrepid>

> From:	Johann [SMTP:observer at NOSPAM.space.pl]
> I would like to write a simple CGI script to get as much info about
> installed Python as possible. But I even don't know how to check
> version of Python. :( From shell command  it is easy, but how to do it
> from CGI script? Eg.
> 
> #!/usr/bin/env python
> print "Content-type: text/html\n\n"
> print "Python version: %s" %s (IDoNotKnowWhatToChoose(),)
 
Try Carsten Gaebler's SnakeCharmer from <http://snakefarm.org/>.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.




From syver-en+usenet at online.no  Wed Jun  5 09:53:26 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Wed, 05 Jun 2002 13:53:26 GMT
Subject: manipulating child processes under Windows
References: <lcptz7zw0m.fsf@gaffa.mit.edu> <uadqb3q9v.fsf@online.no> <lcy9duhksg.fsf@gaffa.mit.edu> <d7226884.0206050341.61fa2ab0@posting.google.com> <3CFE0975.8010106@skippinet.com.au>
Message-ID: <uu1ohzvbq.fsf@online.no>

Mark Hammond <mhammond at skippinet.com.au> writes:

> >Syver Enstad <syver-en+usenet at online.no> writes:
> >
> >
> >>I'll have to check the msdn
> >>library to find out how to get the owning process handle from a
> >>window handle.
> 
> FYI, GetWindowThreadProcessId was added in the win32all-147/8
> win32process module.

Good I'll go download the newest version at once.

-- 

Vennlig hilsen 

Syver Enstad


From peter at engcorp.com  Sat Jun 29 21:16:25 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 29 Jun 2002 21:16:25 -0400
Subject: Newbie that don't understand
References: <AIlT8.13914$p56.4478653@newsb.telia.net> <3d1e262f.3755054@news.clara.net> <3D1E4F80.D8EFBB3E@engcorp.com> <3D1E56EC.7080203@SPAMnwinternet.com>
Message-ID: <3D1E5BE9.86189DA9@engcorp.com>

Jim wrote:
> 
> Peter Hansen wrote:
> > Just a dude wrote:
> 
> >>My son is an electronics
> >>design engineer, and I still can't get my head around negative
> >>voltages, how the hell do you get a -12 V supply... to him it's basic
> >>but to me..............
> 
[snip my own reply which rather missed the mark this time,
because it's too embarrassing to leave in :-)]

> While this is way off topic and way pedantic, electrons
> always leave and enter the same terminals (think electro-
> chemistry, like a battery). 

D'oh!  What was I thinking?  And they call me an engineer. :(

Jim is quite right of course.  Grumble grumble.

-Peter


From bokr at oz.net  Wed Jun  5 20:20:43 2002
From: bokr at oz.net (Bengt Richter)
Date: 6 Jun 2002 00:20:43 GMT
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <3CFCF3E7.AE891EBE@noaa.gov> <adiumj$22g2$1@nntp6.u.washington.edu> <adjduf$92g$0@216.39.172.122> <adjklu$18ra$1@nntp6.u.washington.edu>
Message-ID: <adm9sr$7nl$0@216.39.172.122>

On 5 Jun 2002 00:06:22 GMT, "Donn Cave" <donn at u.washington.edu> wrote:

>Quoth bokr at oz.net (Bengt Richter):
>...
>| Here is a contrived case where += lets you use a
>| mutable-lhs-target-expression-with-side-effect and
>| get only one side effect. There are obviously other
>| ways, but at least += is convenient (and presumably
>| faster in a case like this).
>|
>|  >>> a=[0]
>|  >>> def foo():
>|  ...     print 'side effect'
>|  ...     return a
>|  ...
>|  >>> foo()
>|  side effect
>|  [0]
>|  >>> foo()[0]+=1
>|  side effect
>|  >>> a
>|  [1]
>|  >>> foo()[0] = foo()[0]+1
>|  side effect
>|  side effect
>|  >>> a
>|  [2]
>
>I imagine you have some actual application for something like this,
>but the example doesn't seem very compelling - like you say, there
>are obviously other ways.  If they're less convenient, maybe they're
>easier on the eyes.
>
>But it's probably a moot point.  Whether the current behavior
>of += is a wart or not, it's not likely to be removed.
>
Sometimes a <op>= b is explained as though it were a kind of textual abbreviation
for a = a <op> b. You could spec the semantics by saying it should work "as if" that
were true, but that is not true for Python (and I wouldn't like it if it were).

As it says in

    http://www.python.org/doc/ref/augassign.html
 
"An augmented assignment expression like x += 1 can be rewritten as x = x + 1
                         ^^^^^^^^^^-- statement??, BTW ;-)
to achieve a similar, but not exactly equal effect. In the augmented version,
x is only evaluated once. Also, when possible, the actual operation is performed
in-place, meaning that rather than creating a new object and assigning that to
the target, the old object is modified instead."

("Evaluated once" was my main point, which matters most if there are side effects).

I'm not sure how best to translate "assignment" talk to "binding" talk here.
Perhaps we could speak of "composition-bindings" when speaking of the way
composite objects are created or modified to "contain" other objects via
object->object associations, and "name-bindings" for the usual name->object
associations. However, when I think about it, ISTM _all_ bindings are really
ultimately composition-bindings, and a name is just one way of identifying
the origin in one object of a particular binding-arc going to another.

Even ordinary assignment can't guarantee rebinding of a name, since a name
can identify a property or specially managed attribute, etc., so we can't
well expect augmented assignment to be any less complex in its semantics.

<aside>
I would assume you could have a list of properties with no individual names,
but I haven't tried it. If everything works as it should, it should be possible
to write

    property_list[i] += 1

and have it do the right thing. Or are property objects dependent on access
via subversion of attribute name access? (Suspecting so).
</aside>

Anyway, the main subject was += wrt immutables. The key seems to be not whether
a binding is to an immutable target, but whether the binding _itself_ is immutable,
as in an ordinary tuple's bindings of its 'contained' objects.

An ordinary name binding is _itself_ not ordinarily immutable, since we can usually
rebind. I.e., names usually identify binding sites in a mutable name space.

Ordinarily a name doesn't identify an immutable binding-site, so rebinding is not
a problem. The binding sites in a tuple are immutable, so rebinding can't be done
without violating the meaning of whatver is binding to the tuple as such.

However, the components of a tuple (i.e., the objects bound to its binding sites)
can be mutable, and they can be mutated without rebinding. Hence the strange-looking
(at first) value results of Jeff's original post, which our forgiving tuple class
duplicates without complaining:

 >>> t = ([1], [2], [3])
 >>> t[1] += [3]
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
 TypeError: object doesn't support item assignment
 >>> t
 ([1], [2, 3], [3])
 >>>

But:

 >>> t2 = T( ([1], [2], [3]) )
 >>> t2
 ([1], [2], [3])
 >>> t2[1] += [3]
 >>> t2
 ([1], [2, 3], [3])

Regards,
Bengt Richter


From gumuz at looze.net  Thu Jun 27 15:54:51 2002
From: gumuz at looze.net (Guyon Morée)
Date: Thu, 27 Jun 2002 19:54:51 GMT
Subject: self destruction :)
Message-ID: <f4KS8.1326$qq1.14832@castor.casema.net>

i've got a list of dynamicaly instantiated objects of a class of my own. how
can i let one of the objects delete itself?

i've tried: 'del self', but somehow it didn't work. afterwards it should
also _not_ be in the list anymore.

thanx,
guyon




From jepler at unpythonic.net  Fri Jun  7 15:28:57 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Fri, 7 Jun 2002 14:28:57 -0500
Subject: Detect events positively
In-Reply-To: <adqkna$67o$1@pegasus.csx.cam.ac.uk>
References: <adqkna$67o$1@pegasus.csx.cam.ac.uk>
Message-ID: <20020607192856.GG12501@unpythonic.net>

If you don't want to use threads, just call the "update" method from
time to time to let Tk handle events.  Of course, make sure "Calculate"
can't be invoked recursively or you're in for trouble.

On Fri, Jun 07, 2002 at 04:52:52PM +0100, Xiao-Qin Xia wrote:
> Hello,
> 
> The following code display two button:
> 
> from Tkinter import *
> def Calculate(event = None):
>         while 1:
		  root.update() # let tk handle events
>                 pass
> def Exit(event = None):
>         import sys
>         sys.exit(0)
> root = Tk()
> btn_cal = Button(root, text="calculate", command = Calculate)
> btn_cal.pack()
> btn_exit = Button(root, text="exit", command = Exit)
> btn_exit.pack()
> root.mainloop()
> 
> if clicked the "calculate" button, the program will fall in a loop that 
> will never end. So program will not response to clicking to the "exit" 
> button, so it cannot be terminated normally. If we don't use multithread, 
> can we add some code in the loop of the function Calculate to let it exit 
> the loop after the user clicks the "exit" button or press a "Esc" key?
> 
> Many thanks from, 
> Xiao-Qin Xia
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From ballabio at mac.com  Sat Jun 29 07:42:31 2002
From: ballabio at mac.com (Luigi Ballabio)
Date: Sat, 29 Jun 2002 13:42:31 +0200
Subject: squared functions--most Pythonic way?
References: <ballabio-E54F22.21380327062002@newssrv.fastweb.it> <afiqfr$9re$1@panix1.panix.com>
Message-ID: <ballabio-E8E00E.13423129062002@newssrv.fastweb.it>

In article <afiqfr$9re$1 at panix1.panix.com>, aahz at pythoncraft.com (Aahz) 
wrote:

> In article <ballabio-E54F22.21380327062002 at newssrv.fastweb.it>,
> Luigi Ballabio  <ballabio at mac.com> wrote:
> >
> >      suppose we want to define a function which, given f(x), returns 
> >its square. There are a number of ways which come to mind, such as:
> >
> > (snip)
> 
> I'm sensing an ill-defined spec here.  To me, your English sentence
> implies that what's wanted is
> 
>     def fsquare(f, x):
>         return f(x)**2
> 
> but both of your examples define closures of some sort.

The error was in the english. What I meant was:
"a function which takes a function f and returns another function g such 
that g(x) = f(x)**2"

And by the way, thanks to all who answered...

Bye,
      Luigi


From whisper at oz.net  Fri Jun  7 12:37:35 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 7 Jun 2002 09:37:35 -0700
Subject: Newby: How do I strip HTML tags?
In-Reply-To: <Xns92278CC566FEAusYourStandingInIt@130.133.1.4>
Message-ID: <GCEDKONBLEFPPADDJCOECEFHDHAA.whisper@oz.net>

First you take them out for a nice dinner, then back to your place... ;-)

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of netvegetable
> Sent: Friday, June 07, 2002 8:18
> To: python-list at python.org
> Subject: Newby: How do I strip HTML tags?
>
>
>
> I'm mucking around with cgi, and I'm trying to work out a way to
> strip the
> html tags of a string. e.g, I want to convert this...
>
> ><font size = 12><b><big>Really Big String</big></b></font>
>
> to this this ...
>
> >Really Big String
>
> ... and store it as a value.
>
> I worked out a crude, but effective way of doing it (see code
> below), but I
> can't escape the feeling there must be a built in way of doing it
> more. If
> nothing else, I'm sure somebody who knows their regular expressions could
> neaten it up (please?).
>
> def strip_html_tags(it):
> 	left = it[:(len(it)/2)]
> 	right = it[(len(it)/2):]
> 	final = left[left.rfind('>')+1:] + right[:right.find('<')]
> 	return final
>
> --
> netvegetable at excite.com
> --
> http://mail.python.org/mailman/listinfo/python-list





From stefhan at hotmail.com  Thu Jun  6 08:43:29 2002
From: stefhan at hotmail.com (=?ISO-8859-1?Q?Stefhan_Stj=E4rn=E5s?=)
Date: 6 Jun 2002 05:43:29 -0700
Subject: Pipe return mesage from spawnv to string?
Message-ID: <3ebec3ea.0206060443.d2cc5bb@posting.google.com>

Hi!

Anyone who knows how to pipe from spawnv?
I start isql.exe with os.spawnv and must see the result.

run ("isql.exe"," -s ....")

def run (program, *args):
    """Executes program with arguments, using os.spawnv"""
    try:
        return os.spawnv(os.P_WAIT,program, (program,) +args)
    except (os.error), why:
        raise SErr, str(why)

Best regards
Stefhan


From siegfried.gonzi at kfunigraz.ac.at  Sat Jun 22 15:33:18 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Sat, 22 Jun 2002 21:33:18 +0200
Subject: Python hits the spot
References: <mailman.1024772248.9399.python-list@python.org>
Message-ID: <3D14D0FE.9070801@kfunigraz.ac.at>

Roman Suzi wrote:

> It's no wonder that 1000 lines of uncommented code behave like that, taking
> into account that you were trying to integrate several (Tcl/Tk, Fortran, C,
> Python) systems in a hurry, and, I presume, not knowing most of them.

I am up to the point where you made my day (really). Are you sure do you 
know what you are talking about?


Regards,
S. Gonzi



From fperez528 at yahoo.com  Fri Jun 28 18:12:46 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Fri, 28 Jun 2002 16:12:46 -0600
Subject: C API for Python's RNG?
References: <3D1CBB3A.E1C43D8A@astro.cornell.edu> <3D1CBBED.58EAC61E@astro.cornell.edu> <mailman.1025293041.3597.python-list@python.org>
Message-ID: <afin12$sgi$1@peabody.colorado.edu>

Gerhard H?ring wrote:

> <clueless>
> What about reading from /dev/random?
> </clueless>

he said he was doing Monte Carlo, you never want to do that for MonteCarlo, 
since you want to be able to know the state of your simulation for 
calibration purposes. What you need is a RNG with good properties (period, 
correlations, etc) but which you can seed as needed. /dev/random is useless 
for that.

cheeers,

f




From never at mind.info  Sun Jun  2 11:44:17 2002
From: never at mind.info (G. Willoughby)
Date: Sun, 2 Jun 2002 16:44:17 +0100
Subject: Reading coords from a scrolled Tkinter Canvas widget?
Message-ID: <addeiv$df1$1@news6.svr.pol.co.uk>

I am having a few problems reading the correct coords from a scrolled canvas
widget! e.g. i have a .jpg about 2000x2000pixels displayed inside a canvas
widget measuring 564x564pixels when i scroll the canvas using the attached
scrollbars i want to click the mouse and read the x/y coords of the .jpg
where my cursor is, not the canvas widget, any ideas? This is the code i
have at the minute:

[snip]
self.thumbnail=Canvas(self.frame1, bd=1, relief=SUNKEN, width=564,
height=564, scrollregion=(0, 0, 564, 564))
self.thumbnail.scrollX=Scrollbar(self.frame1, orient=HORIZONTAL)
self.thumbnail.scrollY=Scrollbar(self.frame1, orient=VERTICAL)
self.thumbnail['xscrollcommand']=self.thumbnail.scrollX.set
self.thumbnail['yscrollcommand']=self.thumbnail.scrollY.set
self.thumbnail.scrollX['command']=self.thumbnail.xview
self.thumbnail.scrollY['command']=self.thumbnail.yview
self.thumbnail.scrollX.pack(side=BOTTOM, fill=X)
self.thumbnail.scrollY.pack(side=RIGHT, fill=Y)
self.thumbnail.pack(side=LEFT)

self.thumbnail.bind("<Button-1>", self.getCoords)

def getCoords(self):
    xCoord=event.x
    yCoord=event.y
[/snip]

This code was working perfectly with images sized 564pixels but i've only
just noticed this will not work with bigger images  :(

--G. Willoughby




From whisper at oz.net  Sun Jun 23 21:19:08 2002
From: whisper at oz.net (David LeBlanc)
Date: Sun, 23 Jun 2002 18:19:08 -0700
Subject: Problems with Pmw
In-Reply-To: <QUtR8.20$U4.2298@news.uchicago.edu>
Message-ID: <GCEDKONBLEFPPADDJCOECEKPDLAA.whisper@oz.net>

You have to initialize Pmw and that's not the way to create a top level
window.
Try this:

from Tkinter import *
root = Tk()
import Pmw
Pmw.initialise(root)

t2 = Toplevel()
l = Label(t2, text='A label') ; l.pack()
b = Button(t2, text='Press me') ; b.pack()
e = Pmw.EntryField(t2, labelpos=W, label_text='Enter something') ; e.pack()
t2.mainloop()

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Mark Rivers
> Sent: Sunday, June 23, 2002 17:36
> To: python-list at python.org
> Subject: Problems with Pmw
>
>
> Folks,
>
> In a previous post I mistakenly attributed one problem to Pmw.Blt, while
> subsequent testing has shown that the problem has nothing to do
> with Blt, it
> is a problem (?) with Pmw.
>
> The problem is illustrated by the following short script:
> ########################################
> from Tkinter import *
> import Pmw
> t1 = Tk()
> t2 = Tk()
> l = Label(t2, text='A label'); l.pack()
> b = Button(t2, text='Press me'); b.pack()
> e = Pmw.EntryField(t2, labelpos=W, label_text='Enter something'); e.pack()
> ########################################
>
> Executing this script as shown results in the first Tk() window (t1) being
> created, and the second Tk() window (t2) and the Label and Button widgets
> being created, and then the following error when creating the
> Pmw.EntryField:
>
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "testBlt2.py", line 7, in ?
>     e = Pmw.EntryField(t2, labelpos=W, label_text='Enter something');
> e.pack()
>   File "C:\PROGRA~1\Python21\Pmw\Pmw_0_8_5\lib\PmwEntryField.py", line 72,
> in __
> init__
>     sequences = root.bind_class(tag)
>   File "C:\PROGRA~1\Python21\lib\lib-tk\Tkinter.py", line 923, in
> bind_class
>     return self._bind(('bind', className), sequence, func, add, 0)
>   File "C:\PROGRA~1\Python21\lib\lib-tk\Tkinter.py", line 858, in _bind
>     return self.tk.splitlist(self.tk.call(what))
> TclError: bad window path name ".15422476.15422044"
>
> If the line "t1 = Tk()" is commented out then it works fine.
>
> This happens both on Win32 and Linux with Python 2.1, and Pmw 0.8.5.
>
> I think it should be legal to create multiple Tk() windows in an
> application, so this seems like a serious bug?
>
> Thanks,
>
>               Mark Rivers
>               Dept. of Geophysical Sciences, University of Chicago
>               GSECARS, Advanced Photon Source, Argonne National Laboratory
>
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From shagshag13 at yahoo.fr  Tue Jun  4 14:56:34 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Tue, 4 Jun 2002 20:56:34 +0200
Subject: directory
References: <3cfcfd4a_1@news.iprimus.com.au> <Xns9223D3529DC59cliechtigmxnet@62.2.16.82>
Message-ID: <adj2gi$11cl4e$1@ID-146704.news.dfncis.de>

"Chris Liechti" <cliechti at gmx.net> a ?crit dans le message de news: Xns9223D3529DC59cliechtigmxnet at 62.2.16.82...
> Gold Fish <occeanlinux at linuxmail.org> wrote in
> news:3cfcfd4a_1 at news.iprimus.com.au:
>
> > How to read the subdirectories in the directory

> (probably because os.listdir returns only file-/diectorynames and not
> paths)
>
> look at os.path.walk(), it's there to solve your problem ;-)
>
> chris

here uou'll find an example :
http://aspn.activestate.com/ASPN/Mail/Message/541112

and you could aslo look at :
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/52664

s13.




From quinn at upchuck.ugcs.caltech.edu  Thu Jun 27 01:28:56 2002
From: quinn at upchuck.ugcs.caltech.edu (Quinn Dunkan)
Date: 27 Jun 2002 05:28:56 GMT
Subject: OT: autocoder redux (was Re: Reminder of a python Project.)
References: <2218.941T781T12224051threeseas@earthlink.net> <3D193728.663CA774@engcorp.com> <1708.941T2398T14045676threeseas@earthlink.net> <3D1941BA.7AFF5C92@engcorp.com>
Message-ID: <slrnahl8kn.fd3.quinn@upchuck.ugcs.caltech.edu>

On Wed, 26 Jun 2002 00:23:22 -0400, Peter Hansen <peter at engcorp.com> wrote:
>Timothy Rue wrote:
>> 
>> On 25-Jun-02 22:38:16 Peter Hansen <peter at engcorp.com> wrote:
>> >I'm afraid there is no longer any need for an autocoder.
>> 
>> Geee, maybe you should tell that to NIST
>
>No need.  They called me the other day and I let them know, but
>it was too late to get it into the Computerworld article.
>
>Seriously, there's really no issue with software development any
>more.  The problems are solved.  I guess if you've invented the
>world's first autocoder, it kinda sux to be you then, eh?

Damn!  I had an AI-oriented task-solver I was just about to release after a few
more tweaks on the flux capacitor, but I guess I needn't bother because it's
probably obsolete already.  It seemed to work pretty well, but there's no
arguing with that 97% figure.

With the accelerating rate of technical progress, it's hard for all us
ahead-of-our-time geniuses to stay current.


From pyth at devel.trillke.net  Wed Jun 12 13:52:49 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 12 Jun 2002 19:52:49 +0200
Subject: Python Virtual File System: A spike test
In-Reply-To: <3D077EA8.3090004@rogers.com>; from mcfletch@rogers.com on Wed, Jun 12, 2002 at 01:02:32PM -0400
References: <3D070F27.9050608@rogers.com> <20020612142015.X6609@prim.han.de> <3D077EA8.3090004@rogers.com>
Message-ID: <20020612195249.Z6609@prim.han.de>

Mike C. Fletcher wrote:
> Hi Holger,
> 
> I really like the idea of the filter package (I'm guessing it would work 
> without alteration even with the spike-test class).

No, not really. Or maybe it would? em, i don't really know but i don't
think so :-)

> As for providing methods with precisely the same names as the os module; 
> I'm not particularly thrilled with the names used there, based as they 
> are on UNIX commands, rather than descriptions of what the method does; 
> with that said, I'm not pathologically averse to including such method 
> names.

I find the stdlib-names not very convincing, either. But having to learn 
new names is a PITA for users of a filename-module. Leveraging
the knowledge of 'os.path' and 'os' names makes a lot of sense to me.
  
> Creating an abstraction for the various URL designations is definitely 
> attractive.  Being able to pass around typed URL objects, rather than 
> raw strings, with rich methods on each type should make the Python 
> programming experience feel a little more consistent.

I agree! But 'simple' should win anytime about 'featureful' but inconsistent.
Finding a unifyied URL-interaction interface seems quite difficult. Up until now i
only worked towards a scheme for unified URLs which didn't provide direct
interaction with the (local/remote) object they point to. This *is*
desirable, though.

> Is your module available somewhere to take a peek?

Hope to have it downloadable this evening (GMT). (need to re-setup my
webserver and understand more of HTML4.1 first :-) I'll send the link
to this thread as soon as possible. 

    holger




From see at below  Thu Jun 27 22:52:20 2002
From: see at below (Paul Foley)
Date: 28 Jun 2002 14:52:20 +1200
Subject: Pythoniac: Thoughts on a hardware Python processor
References: <afc52t$2pj$1@sirius.dur.ac.uk> <mailman.1025163215.23885.python-list@python.org>
Message-ID: <m28z50xggr.fsf@mycroft.actrix.gen.nz>

On Thu, 27 Jun 2002 03:32:29 -0400, Oren Tirosh wrote:

> On Wed, Jun 26, 2002 at 10:29:52AM +0000, Christopher Saunter wrote:
>> Dear All,
>> 
>> I have seen the occasional post about creating a specific hardware 'Python
>> Processor', that could then be built and used in programmable logic or the
>> like (Valves, relays, core memory - Pythoniac ;-) - in other words, create
>> a processor that executes Python bytecode directly.  There would be
>> various different reasons for doing this (the 'neat' factor etc...)  

> One of the most effective optimizations for such a machine would be to 
> support compact objects.  The machine word would be somewhat larger than the 
> actual address space. Let's say 40 bits.  If the 4 high bits are 0 this is a
> reference to an object allocated somewhere in memory.  If the high bits are 
> not zero it means that the actual object is stored inside the register.  You 

It's usual to use the low bits for this purpose, rather than the high
bits.  You can use at least 2 low bits for free, on ordinary 32 bit
hardware, if pointers are aligned to 32 bit boundaries anyway.

-- 
Majority, n.:
        That quality that distinguishes a crime from a law.

(setq reply-to
  (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))


From ubecher at gmx.net  Fri Jun 21 14:05:37 2002
From: ubecher at gmx.net (Uwe Becher)
Date: Fri, 21 Jun 2002 20:05:37 +0200
Subject: PIL images in CGI scripts
References: <aevld7$bv9$05$1@news.t-online.com>
Message-ID: <3D136AF1.A3417F17@gmx.net>

Well, i should have searched 2 days and 2 hours instead of two days.
I found a solution for my problem with

if sys.platform == "win32":
     import os, msvcrt
     msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)

I didn't know msvcrt yet.

Uwe

Uwe Becher wrote:
> 
> Hi together,
> 
> i try to generate dynamic images in Python CGI scripts on an apache
> server
> on windows. In my first attempts the browser showed miscoloured images.
> Searching the internet i learned, that i have to output the image as
> binary stream to stdout. I succeeded by calling the Python interpreter
> with the option -u via the shebang #!d:/python21/python.exe -u.
> My question:
> Is it possible to define the binary mode in the script without
> using the -u option?
> 
> Thankys in advance
> 
> Uwe


From whisper at oz.net  Wed Jun  5 05:30:41 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 5 Jun 2002 02:30:41 -0700
Subject: Printing under Windows
In-Reply-To: <3CFDB1B2.7060302@bgb.cc>
Message-ID: <GCEDKONBLEFPPADDJCOEAELIDGAA.whisper@oz.net>

I don't recall - does PythonWin do printing?

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Don Garrett
> Sent: Tuesday, June 04, 2002 23:39
> To: python-list at python.org
> Subject: Printing under Windows
>
>
>    A while back, I posted a request of how to print graphical images on
> Windows from Python. There were no answers suggested that I could
> use. I have
> recently found that the problem can be solved using the wxPython
> wrapper for
> wxWindows.
>
>   I just thought I would mention it. After two days of exposure
> (not a lot, I
> admit), wxPython seems to be a great system to work with and to
> solve several
> problems I've for a while. I'm able to print, access the system tray, and
> build an interface that looks VERY windows native while remaining
> cross platform.
>
> http://www.wxwindows.org/
> http://wxpython.org/
>
> --
> Don Garrett                             http://www.bgb.cc/garrett/
> BGB Consulting                                      garrett at bgb.cc
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From rotundo52 at hotmail.com  Thu Jun 20 01:17:00 2002
From: rotundo52 at hotmail.com (Jon J. Morin)
Date: Thu, 20 Jun 2002 05:17:00 GMT
Subject: can I call operator overloader in superclass?
References: <H2dQ8.1805$CZ4.569945@news1.news.adelphia.net> <3D116126.433AF0E4@alcyone.com>
Message-ID: <gzdQ8.1818$CZ4.577235@news1.news.adelphia.net>

Erik Max Francis wrote:

> "Jon J. Morin" wrote:
> 
>> TypeError: repr not string
>> 
>> For one, I don't understand the error.  For seconds, can I call
>> __repr__ in
>> the superclass from __repr__ in the subclass?
> 
> I'm guessing you have an older version of Python.  In Python 2.2 this is
> the error you'll get:
> 
> TypeError: __repr__ returned non-string (type NoneType)
> 
> which I can only presume was changed to be a little more descriptive.
> The calling superclass methods in subclasses is actually leading you
> astray -- all the error is telling you is that your __repr__ method must
> return and yours isn't.  Change the __repr__ method in your subclass so
> that it returns a string and you're off and running.
> 
> To answer your other question (which really wasn't the issue here), you
> can certainly explicitly call a superclass method in a subclass, and
> you'd had the syntax right:  Specify the superclass method is an unbound
> method and then explicitly pass in the self object:
> 
> class Superclass:
> ...
> 
> def method(self):
> ...
> 
> class Subclass(Superclass):
> ...
> 
> def method(self):
> ...
> Superclass.method(self)
> ...
> 
> 
Thanks.  All I forgot was a return statement in the subclass method and it 
works now.  Thanks for the help.

Jon


From emile at fenx.com  Sun Jun 30 12:05:10 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 30 Jun 2002 16:05:10 GMT
Subject: [Newbie] Is the text delimiter changeable?
References: <3d1f26a4.5915886@news.texas.net>
Message-ID: <W_FT8.369208$cQ3.23994@sccrnsc01>

"Ben Fairbank" <baf at texas.net> wrote in message
news:3d1f26a4.5915886 at news.texas.net...
> I have to process large amounts of text and the text includes numerous
> apostrophes and quotation marks.  It would be very handy if I could
> temporarily assign a substitute for Python's use of the caracters "
> and ' to delimit text strings.  If I could use, for example, the @
> character or the # character, neither of which appears in the text I
> have to work with, and if I could disable the " and ' delimiters, then
> my task would be greatly simplified.  I realize I can backslash escape
> these characters in some circumstances, but I do not think that will
> work when a user's input is a word such as "can't" or "don't."
> Anyway, short of doing something truly drastic, such as recompiling
> the system (which I have no intention of doing), is there a workaround
> that will permit this?
>

Python's text indicators (', ") don't appear in the content so you
should have no problems dealing with the content.

>>> s,t = "can't", '''"can't"'''
>>> s
"can't"
>>> t
'"can\'t"'
>>>



--

Emile van Sebille
emile at fenx.com

---------



From phlip_cpp at my-deja.com  Wed Jun 26 20:16:48 2002
From: phlip_cpp at my-deja.com (Phlip)
Date: Thu, 27 Jun 2002 00:16:48 GMT
Subject: continuous redirection of stdout & stderr to Tkinter Text
References: <3D1A32B0.CA7D3618@club-internet.fr>
Message-ID: <QPsS8.9598$jv4.502984967@newssvr21.news.prodigy.com>

"Fabien H?non" wrote:

> I am using Tkinter as an editor and a launcher (under Linux) for a
> raytracer (POV-RAY)
>
> Once started, the raytracer does two things in the same time :
> - it displays the image being raytraced
> - it spits out information to the console (ie : the line which is being
> raytraced).
> I get the image displayed but not the redirection of stdout

Hit this. It's short:

    http://flea.sourceforge.net/browser006.zip

    http://www.c2.com/cgi/wiki?PyUnitTestBrowser

It contains a class called Piper that wraps up all the crap require to pipe
in one reusable package.

Alternately, use os.system and let the stdout and stderr spew out on the
console.

--
  Phlip




From merkosh at hadiko.de  Sat Jun 22 06:40:44 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Sat, 22 Jun 2002 12:40:44 +0200
Subject: How to stop a SocketServer?
References: <7xhejv9af2.fsf_-_@ruckus.brouhaha.com>
Message-ID: <MPG.177e72971f3f789198968e@news.rz.uni-karlsruhe.de>

In article <7xhejv9af2.fsf_-_ at ruckus.brouhaha.com>, phr-
n2002b at NOSPAMnightsong.com says...
[...]
> doing something crude like killing the process?  I'd like to be able
> to connect to the server and enter a "shutdown" command and have the
> server gracefully exit and free the port.  I haven't been able to
> figure out how to do that.
> 
> Suggestions?  Thanks.

I'll be faceing the same problem in a few weeks.
Does sys.exit() terminate a thread or the interpreter?

Uwe 


From dallasm at aiinet.com  Fri Jun 14 16:20:32 2002
From: dallasm at aiinet.com (Mahrt, Dallas)
Date: Fri, 14 Jun 2002 16:20:32 -0400
Subject: Associating a C++ class with a Python class
Message-ID: <638AA0336D7ED411928700D0B7B0D75B682049@aimail.aiinet.com>

I would read the following link
http://www.python.org/doc/PyCPP.html


> -----Original Message-----
> From: jhndnn at yahoo.com [mailto:jhndnn at yahoo.com] 
> Sent: Friday, June 14, 2002 4:13 PM
> To: python-list at python.org
> Subject: Associating a C++ class with a Python class
> 
> 
> I would like to add a new class type to my embedded python 
> engine and I would like this class to wrap an exisiting C++ 
> class instance. All of the examples I have found are more 'c 
> implementation' than 'c wrappers', meaning that the python 
> instance is not associated with an different c++ class. The 
> methods of the python class will just be straight calls 
> through to the C++ class.
> 
> These classes will only be created from C, so I am basically 
> looking for a way to stuff a pointer to my C++ class into the 
> python object. I guess I could have a map of PyObject->C++ 
> class, but I was hoping there was a cleaner way to accompish 
> this task. Is there a standard way to do this? Am I missing 
> something obvious?
> 
> Thanks-
> 
> John
> --
> John Dunn
> Peak Audio, a division of Cirrus Logic, Inc. http://www.peakaudio.com
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 




From cliechti at gmx.net  Sun Jun 16 15:55:15 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 16 Jun 2002 21:55:15 +0200
Subject: HT telnet interact with auto login?
References: <3d0cd198.17917980@news.zrz.tu-berlin.de>
Message-ID: <Xns922FDF67BADDBcliechtigmxnet@62.2.16.82>

K.Rdt at TU-Berlin.DE (Klaus Reinhardt) wrote in 
news:3d0cd198.17917980 at news.zrz.tu-berlin.de:
> I want to combine:
> 
> import getpass
> import sys
> import telnetlib
> 
> HOST = "...de"
> tn = telnetlib.Telnet(HOST)
> 
> user = '...'
> password = '...'
> 
> 1.)
> tn.interact()     # this is working but with manually ..
> 2.)
> tn.read_until("login: ")     # this seams to work
>                # but without connection
> tn.write(user + "\n")
> if password:
>     tn.read_until("Password: ")
>     tn.write(password + "\n")

i don't understand your question... but often a '\r\n' is required
as end of line. maybe this helps :-)

chris

-- 
Chris <cliechti at gmx.net>



From mcfletch at rogers.com  Mon Jun  3 19:04:37 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 03 Jun 2002 19:04:37 -0400
Subject: getting windows processes list
References: <GCEDKONBLEFPPADDJCOEMEGHDGAA.whisper@oz.net>
Message-ID: <3CFBF605.5000006@rogers.com>

If I'm not mistaken, win32pdh should be able to get the information 
using a process counter.  WSH might also be able to get at it.

HTH,
Mike

PS: Grant, your email address bounces.

David LeBlanc wrote:
> I'm not sure that mere mortal programs are allowed access to that
> information due to Window's (NT's) arcane idea of what constitutes security.
> I think you're doing well to get your own PID.
> 
> David LeBlanc
> Seattle, WA USA
...

_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From matt at mondoinfo.com  Sat Jun 29 22:55:22 2002
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Sun, 30 Jun 2002 02:55:22 GMT
Subject: dynamic buttons with tkinter
References: <afl51r$emn$1@lnx107.hrz.tu-darmstadt.de>
Message-ID: <slrnahssor.1bi.matt@happy-hour.mondoinfo.com>

On Sat, 29 Jun 2002 22:17:17 +0200, Mirko Koenig <koenig at v-i-t.de>
wrote:

>  for file in os.listdir( dir ):
>         Button( load_dlg_frame, text=file, relief=GROOVE, 
> 		command=load).pack()

> But how can i find out which button is pressed?  Or another
> question: How can i pass a event funktion more than 1 parameter (the
> event and other things).  I can't give any button a new event
> function, because it's dynamic, because of the number of files in
> the directory.

Dear Mirko,
For the callback function, give Tkinter an object that has a __call__
method. When you create the object, you can give it the filename and
its __call__ method can use the name when the button is clicked. Danny
Yoo has a recipe that does just that at:

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/66521

The general concept of giving a function some or all of its arguments
in advance of calling it is called "currying". There are other recipes
there that do the same thing in a more general way and other
information about its uses around the Net.

Regards,
Matt



From janto_d at hotmail.com  Fri Jun 28 11:31:11 2002
From: janto_d at hotmail.com (Janto Dreijer)
Date: Fri, 28 Jun 2002 17:31:11 +0200
Subject: doctest for java
Message-ID: <afhv70$15mk$1@news.adamastor.ac.za>

Are there any attempts to port the "doctest" module to Java? (All I can find
using google is some testing class for the JavaDoc program calling itself
"DocTest".)

At a glance it looks like a worthwhile effort...extracting Jython "examples"
from /**...*/ comments also seems simple to implement.




From jbublitzNO at SPAMnwinternet.com  Wed Jun 19 12:20:26 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Wed, 19 Jun 2002 16:20:26 GMT
Subject: Embedding Python - time module problem
References: <cIgHcXA4SyD9swmz@edshk.demon.co.uk> <3D0F99EA.7060904@SPAMnwinternet.com> <R5wUQLAkYDE9swMG@edshk.demon.co.uk>
Message-ID: <3D10AF42.5000201@SPAMnwinternet.com>

Richard Townsend wrote:

> Thanks Jim, passing -Wl,-E to the linker does the trick on HP-UX. I can
> import time and math modules OK now.

The thanks should go to Gustavo Niemeyer, who solved
the problem for me. Glad it works.


Jim





From larooy at xtar.co.nz  Fri Jun 28 23:08:49 2002
From: larooy at xtar.co.nz (John La Rooy)
Date: Sat, 29 Jun 2002 15:08:49 +1200
Subject: How to get rid the new line
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org>
	<affljn$e94us$2@ID-99293.news.dfncis.de>
Message-ID: <20020629150849.38d933cb.larooy@xtar.co.nz>

On 27 Jun 2002 18:30:15 GMT
William Park <opengeometry at NOSPAM.yahoo.ca> wrote:

> SiverFish <occeanlinux at linuxmail.org> wrote:
> > Anyone tell me how to skip the empty line when you read through the file
> > and copy the information in to the list(no empty line )
> 
> A bit off topic... but this question just cries out for 
>     sed 's/^$/d'
> 
> -- 
> William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
> 8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin

This is easier to type. I wonder which is faster

grep .

John


From ianb at colorstudy.com  Tue Jun  4 04:47:55 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 04 Jun 2002 03:47:55 -0500
Subject: CGI questions
In-Reply-To: <Pine.LNX.4.21.0206041508270.902-100000@localhost.localdomain>
References: <Pine.LNX.4.21.0206041508270.902-100000@localhost.localdomain>
Message-ID: <1023180476.16917.35.camel@localhost>

On Tue, 2002-06-04 at 12:41, Michael Hall wrote:
> 
> I've been using PHP for a couple of years now, but have begun to look at
> Python as a CGI language. So far so good, I really like writing straight
> OO code and keeping all the HTML spaghetti out of the way in templates.
> Anyway, a couple of questions:
> 
> 1.
> Does Python have the same or a similar problem that PHP had until recently 
> with regard to automatic global registration of GET, POST and COOKIE
> variables within a script's namespace? My guess is that it doesn't, but
> I'd like to know for sure.
> In other words, is all user input in the form of query string or POST
> variables 'filtered' through cgi.Fieldstorage? Is user input ever directly
> available to the script? 
> Let's say I have a variable 'loggedon' in my script. If a user passes
> 'loggedon=yes' in a query string to my script, will it have any effect or
> would it have to be accessed via something like this first:
> 
> form = cgi.Fieldstorage
> loggedon = form["loggedon"].value

No, there's no problem with this -- you will always explicitly say where
you want to get the value from, nothing's ever put in the global
namespace for you.

> 2.
> I'm trying to use the md5 module to create a session id string. It
> produces some output, but it is pretty weird stuff, not the 32 character
> alphanumeric string I'd expect from PHP. Most of it is definitely not
> [a-zA-Z0-9], and it is around 10 characters long. How can I get a PHP
> style md5 result?

Use .hexdigest()

> 3.
> What is the decimal bit on the end of time.time()'s output? Hundredths of
> a second?

Fractions of a second, of course.  The actual precision is
system-dependent (i.e., it's probably not as precise as it looks)

> 4.
> Is the following Python the (best) equivalent to the PHP?
> 
> PHP				Python
> ereg("bit","bigbits")		re.compile("bit").search("bigbits",1)

I'm not sure what the ,1 is in the Python version.  You should just do
re.compile("bit").search("bigbits"), or perhaps 
re.search("bit", "bigbits") (you save your compiled regex to speed later
searches, but if you don't save the compiled regex compiling won't make
a difference).

--
Ian Bicking           Colorstudy Web Design
ianb at colorstudy.com   http://www.colorstudy.com
4869 N Talman Ave, Chicago, IL 60625 / (773) 275-7241





From cbbrowne at acm.org  Mon Jun 10 14:38:30 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: 10 Jun 2002 18:38:30 GMT
Subject: activating a webpage's click command from python
References: <mailman.1023727993.12908.python-list@python.org> <3D04EA96.B3518A37@engcorp.com>
Message-ID: <ae2rn6$37mpf$1@ID-125932.news.dfncis.de>

In an attempt to throw the authorities off his trail, Peter Hansen <peter at engcorp.com> transmitted:
> "Grimes, Jessica C" wrote:
>> 
>> I want to create a  program that will ask the user what stock he/she 
>> wants to look up.  The program will then extract financial data from 
>> yahoo and import it into a spreadsheet.
>> 
>> My problem is that I don't know how to make the application take the 
>> user's input (the stock's ticker symbol) and place the user's input 
>> from python into the space where yahoo accepts input for a ticker 
>> symbol and activate yahoo's "Look up" click command.
>
> You should investigate HTML forms.  The web page contains a form
> which has several fields and a target URL.  If you view the source
> for the page (in Netscape, hit Ctrl-U; in IE use the View->Source
> menu) you can find buried within it the form and the names of the
> fields.  Forms use the <form> tag and contain <input> fields.
>
> Looking at finance.yahoo.com I can see the form calls "/q" when
> executed, and has a field named "s" for the stock symbol.  You
> can easily take such forms and compose a URL which retrieves the
> data.  In this case this:
>
>    http://finance.yahoo.com/q?s=nt.to
>
> does the job for Nortel on the TSE (disclaimer: anyone even 
> thinking of buying Nortel should probably just keep thinking :-).
> The page returned contains the data, but it's embedded deep
> in the HTML and you _really_ don't want to waste much of your
> time trying to figure this out.  There's got to be a better 
> approach, but I have never spent the time to research it.

The SOAPy way would be:

#!/usr/bin/env python
import SOAP
from SOAP import *
server = "http://services.xmethods.com/soap"
namespace= "urn:xmethods-delayed-quotes"
serv = SOAP.SOAPProxy(server, namespace=namespace)
stocks = ["LNUX", "NT", "TSG", "RHAT", "IBM", "MSFT", "AMR"]
print "Stock    Price"
print "================"
stocks.sort
for el in stocks:
    print "%6s %8.3f" % (el, serv.getQuote(symbol=el))
-- 
(reverse (concatenate 'string "moc.enworbbc@" "sirhc"))
http://www.cbbrowne.com/info/lisp.html
"war is an inappropriate analogy; ``flame war'' is a misnomer.
 in any usenet exchange, the only casualty is time.
 there are better uses for regret."
    --thi <ttn at netcom.com>


From barr at cs.cornell.edu  Thu Jun 27 01:26:15 2002
From: barr at cs.cornell.edu (Rimon Barr)
Date: Thu, 27 Jun 2002 01:26:15 -0400 (EDT)
Subject: spyce
Message-ID: <Pine.LNX.4.44.0206270121180.5504-100000@pompom.u.cs.cornell.edu>

SPYCE - Server Python Pages
http://spyce.sourceforge.net

SPYCE is a drop-in Apache plugin that supports simple and efficient
Python-based dynamic HTML scripting. Those who like Python and are
familiar with JSP, or PHP, or ASP, should have a look at this engine. It
allows one to generate dynamic HTML content just as easily, using Python
for the dynamic parts.  Its modular design makes it very flexible and
extensible.  It can also be used as a command-line utility for HTML
pre-processing.

Version 1.1.8 released on 27 Jun 2002.

   WWW: http://spyce.sourceforge.net
   Freshmeat: http://freshmeat.net/projects/spyce/?topic_id=92
   Download:
    tgz: http://freshmeat.net/redir/spyce/30174/url_tgz/spyce-1.1.8-1.tgz?download
    rpm: http://freshmeat.net/redir/spyce/30174/url_rpm/spyce-1.1.8-1.i386.rpm?download
   Changes:
     http://freshmeat.net/redir/spyce/30174/url_changelog/
   Web CVS Tree: http://freshmeat.net/redir/spyce/30174/url_cvs/

Your comments and suggestions are welcome.

Enjoy,
Rimon.
-- 
 * Rimon Barr       Ph.D. candidate, Computer Science, Cornell University
 | barr at cs.cornell.edu - http://www.cs.cornell.edu/barr - Y!IM: batripler
 |
 | Understanding is a kind of ecstasy.
 +----                                   -- Carl Sagan










From amuys at shortech.com.au  Thu Jun 20 20:22:54 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 20 Jun 2002 17:22:54 -0700
Subject: do...while
References: <mailman.1024595068.19382.python-list@python.org>
Message-ID: <7934d084.0206201622.3fd67d91@posting.google.com>

"Sean 'Shaleh' Perry" <shalehperry at attbi.com> wrote in message news:<mailman.1024595068.19382.python-list at python.org>...
> On 20-Jun-2002 Michael P. Soulier wrote:
> >     Greetings.
> > 
> >     How do you people handle the lack of a do...while in Python? I find that
> > at times, I must initialize variables for a loop with identical code to the
> > loop, which feels like a waste to me. 
> > 
> > ie.
> > 
> >     line = filehandle.readline()
> >     while len(line) > 5:
> >         line = filehandle.readline()
> > 
> >     A do...while here would be nice. 
> > 
> >     Just curious. 
> > 
> 
> while 1:
>     line = filehandle.readline()
>     if len(line) <= 5: break
>     ....
>     ....
> 
> is a typical way to implement it.

or if you prefer to avoid the implied infinate loop...

for line in filehandle:
    if len(line) <= 5: break
    ....

I haven't tested for any interaction with this and any
read-ahead-buffering, however the documentation defines the behaviour
as being repeated calls to readline() so it should work fine.

Andrae Muys


From mep at paju.oulu.fi  Mon Jun 17 09:21:10 2002
From: mep at paju.oulu.fi (Marko Poutiainen)
Date: Mon, 17 Jun 2002 13:21:10 GMT
Subject: IDE recommendations??
References: <b9veea.l85.ln@majestix.gallien.de> <3d0b06bd$0$52043$e4fe514c@dreader3.news.xs4all.nl>
Message-ID: <anlP8.7167$ws6.150445@news2.nokia.com>

"Boudewijn Rempt" <boud at valdyas.org> wrote in message
news:3d0b06bd$0$52043$e4fe514c at dreader3.news.xs4all.nl...

> I've tried to use most IDE's out there -- from Pythonwin to Pythonworks.
> None was really comfortable, except for Wing IDE, which is really
powerful.
> It doesn't include integration of a GUI designer, though, which makes it
> little more than a glorified editor/debugger (but _very_ glorified). At
one
> point there was a version of KDevelop that supported Python development,
but
> that has disappeared, it seems.

Is BOA Constructor worth looking at?




From fredrik at pythonware.com  Wed Jun 26 14:17:28 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 26 Jun 2002 18:17:28 GMT
Subject: question od default args
References: <o3vjhuguen6gfm8vnq7rbk5volmi80shl5@4ax.com>
Message-ID: <YynS8.13571$p56.4313990@newsb.telia.net>

Gon?alo Rodrigues wrote:

> When I want default args I usually do
>
> def (arg = None):
>     ...
>
> But what if I want to make None a reasonable argument too? That is, I
> want to know if the user has in fact passed an argument, and if not to
> do something special, with None (or whatever object) being a reasonable
> arg to be passed.

hmm.  wasn't this just discussed in some thread on a
newsgroup near you?

three alternatives:

    __UNDEF__ = [] # guaranteed to have a unique id

    def myfunc(arg=__UNDEF__):
        if arg is __UNDEF__:
            print "no argument"
        else:
            ...

or, quite ugly:

    def myfunc(*args):
        if not args:
            print "no argument"
        else:
            (arg,) = args
            ...

or, preferred:

    come up with a better design; good designs allow people to
    explicitly pass in a "use the default value" argument value...

</F>




From jhndnn at yahoo.com  Wed Jun 12 13:27:43 2002
From: jhndnn at yahoo.com (John Dunn)
Date: 12 Jun 2002 10:27:43 -0700
Subject: Adding and running functions at runtime
Message-ID: <2cc96dc7.0206120927.71e87364@posting.google.com>

Hello-

I am currently embedding Python in my C++ application and would like
to add function 'scriplets' defined by strings and then call them from
C. Let me explain what I am trying to do -

1. load a string that defines a function ( or functions )
2. get the function object for a defined function
3. call function with arguments
4. get return value and modified argument values

so the code might look something like

const char * funcs = 
"def foo( arg1, arg2 )\n"
"    arg2 = arg2\n"
"def foo2( arg1, arg2 )\n"
"    return arg1+arg2\n"

Mythical_PyLoad( funcs );
PyObject* foo = Mythical_PyGet_Function( "foo" );
PyObject* foo2 = Mythical_PyGet_Function( "foo2" );
PyOjbect* args = some code here to create arguments
// this should return the sume of the args
PyObject* result = PyEval_CallObject( foo2, args ); 
// this should modify the second arg
PyObject* result = PyEval_CallObject( foo, args ); 

I have been able to load code using Py_CompileString, but that appears
to run inline code, and I couldn't get it to return a value or modify
something in the dictionary.

I hope that makes sense...

Thanks for any help-

John
--
John Dunn
Peak Audio, Inc. 
johnd at peakaudio.dontspamme.com
http://www.peakaudio.com


From mickey at tm.informatik.uni-frankfurt.de  Fri Jun 14 20:33:50 2002
From: mickey at tm.informatik.uni-frankfurt.de (Michael 'Mickey' Lauer)
Date: 15 Jun 2002 02:33:50 +0200
Subject: SWIG and Callbacks
Message-ID: <3d0a8b6e@nntp.server.uni-frankfurt.de>

Hi,

I have to bind a C-library which does lots of low-level
ioctl and uses c-callback functions. Using SWIG has given
me the possibility to call most of the wrapped functions
without too much hassle, but I just learned that SWIG can't
call Python-functions as callbacks.

I had a quick look at the other wrappers (SILOON, GRAD, CXX, etc.)
but most look either outdated or not appropriate for this task.
Do you know an (automatic) wrapper generator which has this feature?

I know sip can do this for c++ libraries - can sip also be used
to wrap plain c-libs without "handwriting" code?

Yours,

:M:



From morton at dennisinter.com  Tue Jun  4 13:57:43 2002
From: morton at dennisinter.com (damien  morton)
Date: 4 Jun 2002 10:57:43 -0700
Subject: Kalman Filters, Particle Filters and SVM's
References: <671ybne7zm.fsf@mail.uni-oldenburg.de>
Message-ID: <4abd9ce7.0206040957.51093176@posting.google.com>

Heres a filter that will get of nearly 100% of your spam.

if 'not sent only to me' or 'body contains my email address' or 'body
contains "click here"' then delete with prejudice.

BCC is the root of all spam. Servers should reject BCC email. An
efficient implimentation of BCC allows spam at cost O(1). Reject BCC
and the cost becomes O(N).


Johannes Nix <Johannes.Nix at mail.uni-oldenburg.de> wrote in message news:<671ybne7zm.fsf at mail.uni-oldenburg.de>...
> Are there any implementations of Kalman Filters, Particle Filters or
> Support Vector Machines in (Numerical) Python ?
> 
> Especially the latter would be fine, I'm dreaming to implement a SVM
> based script with the capability to distinguish spam from personal
> mails.... but I'm way to lazy to write my own SVM.
> 
> Johannes
> 
> 
> 
> -- 
> --------------------------------------------------
> Johannes Nix
> 
> jnix at medi.physik.uni-oldenburg.de
> 
> AG Medizinische Physik       Fax +441 798-3902
> Universit?t Oldenburg        Tel +441 798-3270
> 26111 Oldenburg
> --------------------------------------------------


From garrett at bgb.cc  Mon Jun 10 12:40:04 2002
From: garrett at bgb.cc (Don Garrett)
Date: Mon, 10 Jun 2002 16:40:04 GMT
Subject: Python 2 Monolithic Executable
References: <mailman.1023724574.1452.python-list@python.org>
Message-ID: <3D04D617.308@bgb.cc>

Anoop P B wrote:
> Hi,
> 
>     I need to write a program which will take user input and mail it to 
> a predesignated email address. The constraint is that it needs to be a 
> single, small executable (without requiring installation and no 
> dependencies) (Windows). Is there a way to create such an executable 
> from a python program?
> 
> Thanks in advance,
> anoop

   Probably the best you can do is to use py2exe, however the output will be 
an EXE and several support files (DLLs and pyd files). The output will also be 
a meg or more in size.

   Unfortunatly, this sounds like Python will not (currently) do what you need.

-- 
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc



From chris at blaue-elise.net  Sun Jun 23 04:35:51 2002
From: chris at blaue-elise.net (Christian Guenther)
Date: 23 Jun 2002 10:35:51 +0200
Subject: bloody freshman question regarding strings
Message-ID: <1024821351.3424.12.camel@dorian.blaue-elise.net>

Hi folks,

I'm absolutely new to python - also I'm not a beginner in programming.
I bought myself a book called "Einf?hrung in python" - should be
"introduction to python" in the original - by Markus Lutz and David
Asher published by O'Reilly. This book is good, nevertheless I'm a bit
in a hurry because I have to write a script that is supposed to populate
an LDAP-directory for user-authentication. 
By the way: Is there a module for python -> ldap? 

My biggest question at the moment is why can't I do something like this:

def mailmsg(severity message opts target recipient):
	# first create the message body
	msgbody="computer: " + COMPUTER + "\n
	         system:   " + SYSTEMTYPE + "\n
		 cpu:      " + CPUTYPE + "\n
		 kernel:   " + KERNELVER + "\n
		 date:     " + DATE + "\n
		 \n
		 severity: " + severity + "\n
		 message:  " + message + "\n\n"


I want to generate messages - to concole and to mail-recipients - that
show the status of user-processing. This message should include all the
information you see above - they are stored in global vars - a severity
and of course a message itself. Let alone, it just doesn't work and I
know I do this in a nodo way, I just don't have alot of time left ;-)

Could someone point me to the right direction????

	chris


-- 
 __    __  __     __  __ ___    ___                      
|  |  |  ||  \   |  ||  |\  \  /  /              chris Guenther
|  |  |  ||   \  |  ||  | \  \/  /               chris at blaue-elise.net
|  |  |  ||  . \ |  ||  |  >    <                Wuppertal / Germany
|  |__|  ||  |\ \|  ||  | /  /\  \  
 \______/ |__| \____||__|/__/  \__\ 

UNIX _IS_ user friendly, it's just selective about who its friends are
  
----------------------------------------------------------------------
  UNIX was not designed to stop you from doing stupid things, 
  because that would also stop you from doing clever things.
                                                         ...Doug Gwyn
----------------------------------------------------------------------





From carbombo at libero.it  Tue Jun  4 05:02:52 2002
From: carbombo at libero.it (CARbomboniere spa)
Date: Tue, 04 Jun 2002 09:02:52 GMT
Subject: Python and openoffice.com
Message-ID: <0n%K8.578$86.11871@twister1.libero.it>

Hi,

Is there a way to access to Openoffice.org 1.0 's spreadsheet (Calc) in
Python ? Maybe through COM ?

Any examples (even very basic ones; e.g. writing into a cell) ?

Thanks in advance,

Fabrizio




From tjreedy at udel.edu  Tue Jun 11 00:40:03 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 11 Jun 2002 04:40:03 GMT
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
References: <Xns9229EE722467Fgustafl@195.100.94.182>
Message-ID: <DafN8.125526$4i.12087144@bin2.nnrp.aus1.giganews.com>

"Gustaf Liljegren" <gustafl at algonet.se> wrote in message
news:Xns9229EE722467Fgustafl at 195.100.94.182...
> I'm making a function that takes one integer and returns a boolean
that
> says if the integer matches a number in a certain group of numbers.
This
> group consists of single numbers aswell as ranges spanning thousands
of
> numbers. I'm looking for an elegant way to write this function.

After reading the other responses, I would try the following
(untested):
  sort ranges and merge contiguous ranges so there is a gap between
each
  express each range (including single ranges) as half-open range
[min,max+1)
  put 2n endpoints in a single list # they *will* be distinct by step
1
  test target number against first and last int in list
  if < first or >= last, no match; (if = first, match # not sure if
need this)
  if within that range, run binsort to get index of largest num <=
target
  # modify binsort or next statement if necessary
  if even(index) match = true else false

Terry J. Reedy







From gashton at cmedltd.com  Mon Jun 17 06:09:08 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 17 Jun 2002 11:09:08 +0100
Subject: Writing a simple linux daemon type thing...
In-Reply-To: <lkznxyko3p.fsf@pc150.maths.bris.ac.uk>
References: <K9eO8.393$LN6.261@newsread1.prod.itd.earthlink.net>
	<rn3cea.rt4.ln@127.0.0.1>
	<yylO8.916$TP4.88596@newsread2.prod.itd.earthlink.net> 
	<lkznxyko3p.fsf@pc150.maths.bris.ac.uk>
Message-ID: <1024308548.843.10.camel@walter>

On Fri, 2002-06-14 at 14:05, Michael Hudson wrote:

> If you run it as a true daemon, you need to be careful about writing
> to stdout/stderr (because they're not there any more).  Think
> tracebacks.

There's an example of how to do this (which I think I took from Python
Standard Library) in the Daemon class in gmalib. It use's gmalib's
Logger class to send all output to either a normal file, or syslog.

You can download gmalib here:

  http://sourceforge.net/project/showfiles.php?group_id=34228

or save yourself the trouble by browsing the CVS archive first:

 
http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/landialler/gmalib/gmalib.py?rev=1.10&content-type=text/vnd.viewcvs-markup

It's GPL'd, so feel free to re-use, etc.

Example usage:

  import gmalib

  class MyApp(gmalib.Daemon):
      def __init__(self):
          gmalib.Daemon.__init__(self, logfile="log.txt", use_syslog=0)
          self.daemonise()

-- 
Graham Ashton





From martin at v.loewis.de  Mon Jun 24 03:36:59 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 24 Jun 2002 09:36:59 +0200
Subject: python, xml, windows, sunos
References: <Pine.SOL.4.44.0206210813510.27754-100000@bast.u.arizona.edu>
Message-ID: <m3znxldt5w.fsf@mira.informatik.hu-berlin.de>

Michael T Hammond <hammond at U.Arizona.EDU> writes:

> The error I get is:
> 
> xml.sax._exceptions.SAXReaderNotAvailable: No parsers found.
> 
> Does anybody have any idea what to do here?

You should first determine whether this analysis is correct. I'd
normally expect that you have both pyexpat and xmlproc available, so
please do

import xml.parsers.expat
import xml.parsers.xmlproc

and report what that does.

Regards,
Martin



From kragen at pobox.com  Mon Jun 10 18:19:44 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 18:19:44 -0400
Subject: question on pattern
References: <3D007E96.3010505@rmb.co.za> <c76ff6fc.0206071440.70ce82f6@posting.google.com>
Message-ID: <83it4qojy7.fsf@panacea.canonical.org>

sjmachin at lexicon.net (John Machin) writes:
> You have got an answer to your problem of how to get DOTALL behaviour
> with .findall(). However you should be aware that your pattern can
> match strings that are not syntactically C or C++ comments. Here is an
> example from page 172 of "Mastering Regular Expressions" by Jeffrey E.
> F. Friedl:
> 
>    const char *cstart = "/*", *cend = "*/";
> 
> I'd agree that it's a pathological case. I'm just pointing out that in
> most work on program source files, to be 100% correct you need to have
> a lexical analyser for the language in question. Regexes won't take
> you the whole distance.

It's true that you need a lexical analyzer for the language, but most
languages (C and C++) can be lexically analyzed with a regular
expression.  Indeed, lex essentially compiles a big regular expression
into C.

It's also possible that a lexical analyzer suitable for parsing will
make a lot of distinctions, such as between { and 3 and strcpy, that
aren't relevant for some particular application.  In particular, a
regular expression to find comments that handles C string and
character constants correctly is relatively simple, although harder
than you'd think.  Friedl brings up the problem on page 172, as you
say, but solves it on page 173.



From derek at wedgetail.com  Fri Jun 21 05:50:54 2002
From: derek at wedgetail.com (Derek Thomson)
Date: Fri, 21 Jun 2002 09:50:54 GMT
Subject: I'd give up Perl tomorrow if only...
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au>
Message-ID: <3D12F6FC.3000608@wedgetail.com>

David Mitchell wrote:
> .. there was something equivalent in scope and breadth to CPAN.

I (almost) agree, and sympathise :)

> 
> Is ANYONE working on this, or thinking about working on it, or knows
> someone who's thinking about working on it, or dreaming about it, or ...?

Among other things, it's one of the reasons I'm tinkering with a Perl ORB.

http://sourceforge.net/projects/perlorb

If you use a Python ORB, such a Fnorb, 
[http://sourceforge.net/projects/fnorb], on the Python side, you should 
be able to write wrappers for your favourite Perl module pretty easily 
and use them from Python.

Yes, it's stop-gap, but when you already have a non-trivial Perl module, 
either on CPAN or in-house, it can save a lot of rewriting. What we 
*really* need to do is dynamically create CORBA interfaces for Perl 
modules, but that might be impossible given that Perl is like Python in 
that module and class interfaces aren't pre-declared :(

Anyway, feel free to email me if you want to try this approach and you 
need guidance.

As for Parrot, maybe someone will port Python to it one day, and then we 
would have something like Jython, but for Perl!

Regards,
Derek.



From jubafre at zipmail.com.br  Tue Jun 18 17:49:37 2002
From: jubafre at zipmail.com.br (jubafre at zipmail.com.br)
Date: Tue, 18 Jun 2002 18:49:37 -0300
Subject: =?iso-8859-1?Q?how=20tranform=20in=20String=3F=3F?=
Message-ID: <3D0F9DAE000002DE@www.zipmail.com.br>

i have s='xavante brasil', i want to transform in

s='xavante', 'brasil' with two indexes

how i make this???????????


Juliano Freitas
www.gebrasil.hpg.com.br



------------------------------------------
Use o melhor sistema de busca da Internet
Radar UOL - http://www.radaruol.com.br







From gerhard.haering at gmx.de  Fri Jun 21 04:32:15 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 21 Jun 2002 10:32:15 +0200
Subject: converting an array of chars to a string
In-Reply-To: <3D12E228.6030501@wedgetail.com>
References: <3d12dda2_6@news.newsgroups.com> <3D12E228.6030501@wedgetail.com>
Message-ID: <20020621083215.GB933@lilith.my-fqdn.de>

* Derek Thomson <derek at wedgetail.com> [2002-06-21 08:22 +0000]:
> JB wrote:
> >How can I convert an array of chars to string?
> >
> >The problem is this: I should like to perform 8 bit 
> >calculations in a buffer. After this has been done, the 
> >buffer has to ba written to a binary file (and file.write 
> >takes strings instead of lists).
> >
> 
> Easy, just use the "join" function to concatenate a sequence of strings 
> into a single string. For example:
> 
> 
> import string
> 
> list = ['a', 'b', 'c', 'd']
> 
> s = string.join(list, '')
> 
> print s

Unless Python 1.5.2 compatibilty is a requirement, just use string
methods:

    lst = ['a', 'b', 'c', 'd']
    s = ''.join(lst)

It's also not a good idea to replace the builtin list function, thus I'm
using 'lst' here.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 25.1 ?C      Wind: 3.7 m/s




From python at tykebsd.net  Mon Jun 17 07:45:20 2002
From: python at tykebsd.net (Dave Moor)
Date: 17 Jun 2002 04:45:20 -0700
Subject: Python and SQL Server 2000
References: <ZCuO8.10006$9b.678135@typhoon.austin.rr.com> <mailman.1024112300.2184.python-list@python.org> <slrnagn75g.263.kzaragoza@localhost.localdomain>
Message-ID: <1bf96e7f.0206170345.515227c1@posting.google.com>

kzaragoza at attbi.com (Kris J. Zaragoza) wrote in message news:<slrnagn75g.263.kzaragoza at localhost.localdomain>...
> 
> My first instinct on seeing this reply was to point out that win32com
> doesn't run on Red Hat Linux. :-) Then I got to thinking a bit more
> about it and realized this may actually be possible.
> 
> So, let me add to the original poster's question: Has anyone ever
> tried running the Windows version of Python (either the main
> distribution or ActiveState's) under Wine to take advantage of the
> database drivers, etc.?

Another solution is to write a simple pair of client/server socket
apps.  The server is passed and executes sql on the windows using COM
or ODBC.  Then pass the results using the struct module back to the
client.

Dave


From logiplexsoftware at earthlink.net  Thu Jun 13 03:52:27 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: 13 Jun 2002 00:52:27 -0700
Subject: How to call a function using apply with keyword args?
In-Reply-To: <ae9fud$u23$00$1@news.t-online.com>
References: <ae9fud$u23$00$1@news.t-online.com>
Message-ID: <1023954749.3299.12.camel@localhost.localdomain>

On Thu, 2002-06-13 at 00:09, Achim Domma wrote:
> Hi,
> 
> I have a function like this:
> 
> def MyFkt(A,B,C): pass
> 
> and some parameters from a config file:
> 
> params = {'A':5,'B':8,'C':9}
> 
> how can I call MyFkt with this parameters? I tried apply(MyFkt,[],params)
> but get this error:
> 
> TypeError: MyFkt() takes exactly 3 non-keyword arguments (0 given)

Works for me:

Python 2.1.1 (#1, Aug 13 2001, 19:37:40) 
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-96)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> def MyFkt(A,B,C): 
...     print A, B, C
... 
>>> params = {'A':5,'B':8,'C':9}
>>> apply(MyFkt,(),params)
5 8 9
>>> 

You can also use:

>>> MyFkt(**params)
5 8 9
>>> 

Regards,
Cliff





From jonathan at onegoodidea.com  Tue Jun 25 07:57:59 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Tue, 25 Jun 2002 12:57:59 +0100
Subject: converting an array of chars to a string
References: <3d12dda2_6@news.newsgroups.com> <3D12E228.6030501@wedgetail.com> <mailman.1024648440.1200.python-list@python.org> <3D1307AF.3070308@wedgetail.com>
Message-ID: <B93E1957.CDC5%jonathan@onegoodidea.com>

On 21/6/2002 12:02, in article 3D1307AF.3070308 at wedgetail.com, "Derek
Thomson" <derek at wedgetail.com> wrote:

> Gerhard H?ring wrote:
>> 
>> 
>> Unless Python 1.5.2 compatibilty is a requirement, just use string
>> methods:
>> 
>>     lst = ['a', 'b', 'c', 'd']
>>     s = ''.join(lst)
> 
> And they say you can't write obfuscated code in Python :)
> 
> I would have expected "join" to be a method on the sequence, if
> anything. As in "s = lst.join('')". In fact, I looked for it, as I
> *knew* I'd seen it used before as a method, I just didn't expect it to
> be in the string class.
> 
> I think I'll just stick with string.join for the time being. As it turns
> out, there's more than one way to do it, even in Python ;)

This one confused me too until I thought about it some more. It follows from
the original location of 'join' in the 'string' module that it be moved to
being a string method.

The 'join' method only operates on a list of strings and always returns a
string. If you look at the list methods, they're all generic with regards to
what types the list contains ('reverse', 'sort', 'append', etc.).

The closest generic list method I could think of would be a 'concat' method
that looks something like:

>>> class list2( list ):
...     def concat( self ):
...         import operator
...         return reduce( operator.add, self )
... 
>>> l = list2( ['hello', 'world'] )
>>> l.concat()
'helloworld'
>>> 

which would work on any list of sequences. [A side effect is that it would
also act as a 'sum' method on lists of numbers.]

Jonathan



From spam at melkor.dnp.fmph.uniba.sk  Fri Jun 14 02:50:35 2002
From: spam at melkor.dnp.fmph.uniba.sk (Radovan Garabik)
Date: Fri, 14 Jun 2002 08:50:35 +0200
Subject: Writing a simple linux daemon type thing...
References: <K9eO8.393$LN6.261@newsread1.prod.itd.earthlink.net>
Message-ID: <rn3cea.rt4.ln@127.0.0.1>

Jones <jones at jones.jones> wrote:
 : Hi everyone,
 : I hope this is a simple question for you guys.  I'm new to Python and Linux.
 : I'm wondering if Python (and this group) can help me.
 : I want a program to always be running in the background that, connects to
 : PostgreSQL every few seconds, queries a table, and for each record it finds,
 : shells out to the system to execute program.

 : My research indicates that "nohup" might do the trick?  So would it be as
 : simple as writing the program in Python and then puting something like
 : "nohup python myprog" in the Linux startup sequence (wherever that may be)?

yes
(python myprog &
 would be probably better)

 : That's my primary goal.  Now... assuming I get past that part, I may want to
 : take it a few steps further later on.  I'd be happy if you guys could take a
 : stab at telling me if these are possible and any tips too...
 : a) Make it multi-threaded...   As I said before, the number of times it
 : executes the external program equals the number of records it finds in the
 : table.  It would be nice if I could execute the program multiple times
 : concurrently in a *controlled* manner (depending on my system resources)
 : instead of having to wait for each one to be done before starting another
 : one.

there is no need for it to be multithreaded. Just use
os.system("program.py &")
and the program will be executed in the background
 
 : b) Make it fail-safe.  I would like to figure out how I could re-launch the
 : program if it was to crash.

look for daemontools at http://cr.yp.to

-- 
 -----------------------------------------------------------
| Radovan Garabik  http://melkor.dnp.fmph.uniba.sk/~garabik |
| __..--^^^--..__         garabik @ fmph . uniba . sk       |
 -----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!


From siegfried.gonzi at kfunigraz.ac.at  Mon Jun 24 04:50:37 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Mon, 24 Jun 2002 10:50:37 +0200
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at> <af4vrl$ffe$0@216.39.172.122>
Message-ID: <3D16DD5C.1050708@kfunigraz.ac.at>

Bengt Richter wrote:
> On Sun, 23 Jun 2002 11:05:02 +0200, Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote:


>>whether it will be a good idea to rewrite it in Fortran 90. Python can 
>>then be used in combination with DISLIN for the plots.
>>
>>
> Still no real diagnosis, though ;-/

Hi,

1. I could not sleep and installed the normal Python version and 
installed it with gcc. But the situation didn't change.

2. Has got anybody any idea how to  write a small script which tests my 
processor for 30 minutes. As I wrote, I wrote a small Python script 
which does claiming memory for 30 minutes. "Top" then tells me something 
like that: processor 5% ;  memory 99%.

Someone guessed that maybe my processor got too hot. Therefore: what 
about a small script of the form: processor 99% ; memory 5%.

That could at least clarify whether the processor becomes to hot after 
beeing used without breaks.

Writing the memory consumption script was easy, but I do not have any 
clue how to write a processor consumption script.


Regards,
S. Gonzi



From gerhard.haering at gmx.de  Fri Jun 28 23:08:09 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 29 Jun 2002 05:08:09 +0200
Subject: Python 2.2 (ActiveState) odbc.odbc vs. mx.ODBC.Windows
In-Reply-To: <3d1d1ef4.426039468@netnews.attbi.com>
References: <3d1d1ef4.426039468@netnews.attbi.com>
Message-ID: <20020629030809.GA1949@lilith.my-fqdn.de>

* candiazoo at attbi.com <candiazoo at attbi.com> [2002-06-29 02:46 +0000]:
> Have any performance tests been done to determine
> advantages/disadvantages between the two?

AFAIR if odbc.odbc works for you, you simply were lucky. It's
unmaintained, which would be reason enough for me to not waste a single
thought on actually using it.

mxODBC is actively developped and even has commercial support available.

As for performance tests, use the profile or hotshot module to find out
yourself. In my experience, the performance of the database interface is
mostly a non-issue.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From brueckd at tbye.com  Tue Jun 25 18:10:40 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Tue, 25 Jun 2002 15:10:40 -0700 (PDT)
Subject: Suggestions for good programming practices?
In-Reply-To: <gfhafa.hb.ln@grey.aerojockey.localdomain>
Message-ID: <Pine.LNX.4.44.0206251500460.28036-100000@ns0.tbye.com>

On Tue, 25 Jun 2002, Carl Banks wrote:

> > Besides, the
> > people who understand the potential for problems with something tend to be
> > the ones that don't mishandle it (whether it be streets, guns, or Python
> > functions).
> 
> Which is why it is important to tell people who don't understand the
> potential for problems to not use eval until they do.
> 
> You seem to have this idea, based on this last paragraph, that a good
> way to keep newbies from misusing eval and exec is to keep them in
> ignorance.  I think that's absurd.  I don't know if you're just
> grasping for arguments, because you previously seemed to think it's
> best to explain WHY, as opposed to simply saying, "Don't do it."

Um, what's absurd is how you reached that conclusion from the paragraph I
wrote. <0.4 wink> I said that people WON'T get hurt if they are NOT
ignorant - I'm advocating an explanation of risk in place of the
categorical "stay away!".

-Dave





From sholden at holdenweb.com  Tue Jun  4 16:32:16 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 4 Jun 2002 16:32:16 -0400
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <3CFCF3E7.AE891EBE@noaa.gov> <adiumj$22g2$1@nntp6.u.washington.edu>
Message-ID: <Mx9L8.239704$%u2.17455@atlpnn01.usenetserver.com>

"Donn Cave" <donn at u.washington.edu> wrote ...
> Quoth Chris Barker <Chris.Barker at noaa.gov>:
> | Huaiyu Zhu wrote:
> |> Great explanation!  The problem stems from the fact that the current +=
> |> operator implements two semantically distinct operations, chosen
according
> |> to some properties of the operands.
> |
> | exactly, and it wa kind of a weird decision. Personally, I feel the only
> | compelling reason to add The += (and friends) operator was to support
> | in-place addition, most notably for Numeric arrays. Frankly it ought to
> | just be disallowed for non-mutable types, but all the folks that want it
> | for integers would scream. to be honest, I've kind of lost track of why
> | it is a good idea for numbers to be immutable, but that's another topic!
>
> That would be almost as cool as the old implementations of FORTRAN
> where you could change the value of 1.
>
Don't you mean "...change the value of True" <wink>?

> |> I'd imagine that at some future time there would be a desire to split
+= to
> |> two distinct operators, one for true in-place operator and one for the
fake
> |> one currently implemented for immutables.  That would remove the
spurious
> |> exceptions for mutable members of immutable structures.
> |
> | Do we really need a new notation? is a = a + b so bad?
>
> No, it's fine.
>
> | ... In fact, are the
> | current semantics of += s bad, or should this bug just be fixed?
>
> Not really - the current semantics are actually better than the fixed
> version, inasmuch as they are at least consistent.  As I interpret the
> proposal, the fixed += would sometimes work without rebinding (if the
> object supports it) and sometimes would be an assignment (falling back
> to +.)  To me, to allow an object to dictate its relationship to its
> computational environment like that is anathema.
>
>>> a = b = (1, 2, 3)
>>> a += ('boo!',)
>>> a, b
((1, 2, 3, 'boo!'), (1, 2, 3))
>>> a = b = [1, 2, 3]
>>> a += ['boo!']
>>> a, b
([1, 2, 3, 'boo!'], [1, 2, 3, 'boo!'])

Doesn't look that consistent to me. Add in the fact that implementors of
mutable objects are free to choose either semantic and what you have is ...
exactly what Guido intended. I regard it as a wart, but not one that worried
me.

> ...
> | By the way, the += and friends operators also cause similar confusion
> | with name scopes in functions, += is assumed to be a re-bind, when it
> | isn't necessarily intended to:
> |
> | >>> l = [1,2,3]
> | >>> def test():
> | ...     l += [4]
> | ...
> | >>> test()
> | Traceback (most recent call last):
> |   File "<stdin>", line 1, in ?
> |   File "<stdin>", line 2, in test
> | UnboundLocalError: local variable 'l' referenced before assignment
> |
> | but:
> |
> | >>> def test2():
> | ...     l.extend([4])
> |
> | >>> test2()
> | >>> l
> | [1, 2, 3, 5]
> |
> | Works just fine. Perhaps  Huaiyu is right, we need different operators.
> | However, then the operator for incrementing a non-mutable would just be
> | a synonym for a = a + b, which looks like unnecessary syntactical sugar
> | to me.
>
> Yes.
>
Warts are warts. If only Guido had listened to me ... :-)

quite-glad-really-it's-someone-else's-fault-ly y'rs  - steve
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From dietmar at bronko.fh-wedel.de  Mon Jun 24 19:40:42 2002
From: dietmar at bronko.fh-wedel.de (Dietmar Lang)
Date: 24 Jun 2002 23:40:42 GMT
Subject: Suggestions for good programming practices?
References: <mailman.1024957832.5633.python-list@python.org>
Message-ID: <slrnahfbph.1au.dietmar@bronko.fh-wedel.de>

Hi,

David LeBlanc wrote:
> The one that always gets me is "lat,lon,timezone = getLocation("seattle")".
> I.E. the ability to return multiple distinct values as opposed to returning
> multiple values in the form of a struct as in C or C++.

I don't consider that a problem, I find it quite useful, because when
you want to use the returned values, you don't have to "rip them out" of
the struct. In an example such as above with the timezones, you might
later want to use the different results in different contexts, where in
my opinion a subscribed object from a struct makes the less readable.

It sometimes, however does make coding or reading code a tad more 
error-prone insofar that at first sight you can't always be sure just
how many objects that call returns.

Greetings,
Dietmar

-- 
I haven't lost my mind -- it's backed up on tape somewhere.


From bokr at oz.net  Fri Jun 28 17:12:58 2002
From: bokr at oz.net (Bengt Richter)
Date: 28 Jun 2002 21:12:58 GMT
Subject: get word base
References: <m24rfn88jp.fsf@mother.paradise.lost>
Message-ID: <afijgq$ph0$0@216.39.172.122>

On Fri, 28 Jun 2002 15:14:50 -0500, John Hunter <jdhunter at nitace.bsd.uchicago.edu> wrote:

>
>I would like to be able to get the root/base of a word by stripping
>off plurals, gerund endings, participle endings etc...  Here is a
>totally naive first attempt that gets it right sometimes:
>
>import re
>
>rgx = re.compile( '(\w+?)(?:ing|ed|es|s)')
>
>def get_base(word):
>
>    m = rgx.match(word)
>    if m:
>        return m.group(1)
>    else:
>        return word
>
>words = ['hello', 'taxes', 'thoughts', 'walked', 'rakes']
>
>for word in words:
>    print word, get_base(word)
>
>Produces the following output
>> python get_baseword.py
>hello hello
>taxes tax
>thoughts thought
>walked walk
>rakes rak
>
>
>I can think of a few things to do to refine this, but before I forge
>ahead, I wanted to solicit advice.
>
Google for python stemmer ;-)

Regards,
Bengt Richter


From edream at tds.net  Thu Jun  6 12:59:03 2002
From: edream at tds.net (Edward K. Ream)
Date: Thu, 06 Jun 2002 16:59:03 GMT
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet> <3CFDF148.AA2035FF@info.unicaen.fr> <3CFE49CA.46A5FEE1@tds.net>
Message-ID: <3CFF94C2.A152B41E@tds.net>

"Edward K. Ream" wrote:

> > http://homepages.ulb.ac.be/~arigo/psyco/
> 
> Thanks for these links.
> 
> Earlier this year I looked into this question in some depth and
> concluded:[snip]

Never mind.  It looks like the psyco project is where the action is.

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


From logiplexsoftware at earthlink.net  Thu Jun 20 14:23:28 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Thu, 20 Jun 2002 11:23:28 -0700
Subject: How to represent the infinite ?
In-Reply-To: <3d11ee9b$1@news.swissonline.ch>
References: <3d11ee9b$1@news.swissonline.ch>
Message-ID: <20020620112328.2dcf3e07.logiplexsoftware@earthlink.net>

On Thu, 20 Jun 2002 17:02:33 +0200
erreur wrote:

> Will somebody have an idea, to represent the infinite one?
> 
> I have variables to initialize with is +inf (or - inf).  To be sure that
> later, all will be smaller (or larger) than my variables.
> I tried to redefine the operators on an object.  It goes for Inf>10 but I do
> not arrive for 10>Inf (because that takes > of Int).
> 
> --------------------
> 
> def __gt__(self, val):
> 
>     return 1
> 
> --------------------

Here's a more complete infinite value (not that you'd need it, judging from
your requirements):

class inf(int):
    def __init__(self):
        int.__init__(self)
        self.sign = 1

    def __str__(self):
        return "<%sinfinity>" % {-1: "-", 1: ""}[self.sign]

    def __repr__(self):
        return str(self)

    def __cmp__(self, n):
        if isinstance(n, inf):
            return cmp(self.sign, n.sign)
        return self.sign
    
    def __neg__(self):
        retval = inf()
        retval.sign = self.sign * -1
        return retval

    def __mul__(self, n):
        retval = inf()
        retval.sign = self.sign * cmp(n, 0)
        return retval

    def __div__(self, n):
        return self * n

    def __add__(self, n):
        if isinstance(n, inf):
            if n.sign != self.sign:
                return 0
        return self

    def __sub__(self, n):
        return self + -n

    
if __name__ == '__main__':
    import sys
    
    m = -inf()
    n = inf()

    print m
    print n
    print m * n
    print n * n
    print m * m

    assert(m < n)
    assert(n > m)
    assert(m == m)
    assert(n == n)
    assert(m != n)
    assert(m < 0)
    assert(n > 0)
    assert(m != 0)
    assert(n
    assert(m < -sys.maxint)
    assert(n > sys.maxint)
    assert(m < -m)
    assert(n == -m)

    assert(n * 1 == n)
    assert(n * 1000 == n)
    assert(n * -1 == m)
    assert(n * -1000 == m)
    assert(n * 0 == 0)
    assert(m * 0 == 0)
    
    assert(n / 1 == n)
    assert(n / -1 == m)
    assert(n / -1000 == m)
    assert(m / 1 == m)
    assert(m / 1000 == m)
    assert(m / -1 == n)
    assert(m / -1000 == n)

    assert(m + 1 == m)
    assert(m - 1 == m)
    assert(n + 1 == n)
    assert(n - 1 == n)
    
    assert(n + m == 0)
    assert(n - m == n)
    assert(n - n == 0)
    assert(m + m == m)
    assert(m - m == 0)
    assert(m - n == m)
    
    # print 1 / m # exception: division by zero
    # print 1 / n # exception: division by zero



-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From piet at cs.uu.nl  Tue Jun 25 07:32:27 2002
From: piet at cs.uu.nl (Piet van Oostrum)
Date: 25 Jun 2002 13:32:27 +0200
Subject: File "in3sun34.py", line 33, in AttributeError: HTTPConnection
References: <mailman.1024982313.3269.python-list@python.org>
Message-ID: <wzd6ufd25w.fsf@nono.cs.uu.nl>

>>>>> Guillermo Fernandez <guillermo.fernandez at epfl.ch> (GF) writes:

GF> Hello!
GF> Please, if you answer me, include my e-mail in the answer as I am not
GF> abonned to this mailing list.

GF> I've written a little program to test python (I'm a beginner). Here it
GF> is:

GF> import httplib

GF> conn = httplib.HTTPConnection("www.python.org")
GF> print "CONNECTED"
GF> conn.request("GET", "/index.html
GF> response=conn.getresponse()
GF> answer=response.read()
GF> conn.close()
GF> print answer

GF> When I execute it, I obtain an error:
GF> [112]host-cfernand% python test.py
GF> Traceback (innermost last):
GF>   File "test.py", line 3, in ?
GF>     conn = httplib.HTTPConnection("www.python.org")
GF> AttributeError: HTTPConnection

GF> Line number 3 is:
GF> conn = httplib.HTTPConnection("www.python.org")

Supposing the ^X^S are not really in your file, this program should work
(I just tried it). Maybe you have a private httplib.py file somewhere in
your Python path.
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From richardkessler at matteicos.com  Tue Jun 18 14:48:57 2002
From: richardkessler at matteicos.com (Richard Kessler)
Date: Tue, 18 Jun 2002 11:48:57 -0700
Subject: Python to MS SQL Server
Message-ID: <3D0F8099.8080106@matteicos.com>

I am interested in using Python in a MS SQL Server environment, but do 
not see any database modules out there, especially for native versus 
using ODBC. Is there such a module?

Thanks in advance.

- Richard



From edcjones at erols.com  Thu Jun  6 21:33:43 2002
From: edcjones at erols.com (Edward C. Jones)
Date: Thu, 06 Jun 2002 21:33:43 -0400
Subject: Problem pickling class instance
Message-ID: <3D000D77.7040008@erols.com>

I think the following problem may be discussed in the Python Library
Reference, section 3.14.4 but I don't understand it.

Here is the file mod.py:

     import pickle

     def Read():
         f = open('saved', 'r')
	return pickle.load(f)

     class A:
         def __init__(self):
             pass

     if __name__ == '__main__':
         a = A()
	f = open('saved', 'w')
         pickle.dump(a, f)

and the file top.py:

     import mod, pickle

     a = mod.Read()

First I run mod.py then top.py. The latter give the message:

     AttributeError: 'module' object has no attribute 'A'

Please explain the problem to me. Is there a standard work-around?

Thanks,
Ed Jones



From james.kew at btinternet.com  Fri Jun  7 16:03:08 2002
From: james.kew at btinternet.com (James Kew)
Date: Fri, 7 Jun 2002 21:03:08 +0100
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com>
Message-ID: <adr3ht$1k37u$1@ID-71831.news.dfncis.de>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3D00A456.4315EDA3 at engcorp.com...
> Eddie Corns wrote:
> >
> > A friend is fond of quoting this to me:
> >
> >   Rules of Optimization:
> >     Rule 1: Don't do it.
> >     Rule 2 (for experts only): Don't do it yet.
> >   -- M.A. Jackson
> >
> You forgot the even more important first thing for a beginner:
> get it correct!

I favour Kent Beck's aphorism: "Make it work, make it right, make it fast."

James





From edream at tds.net  Sun Jun 30 10:49:38 2002
From: edream at tds.net (Edward K. Ream)
Date: Sun, 30 Jun 2002 14:49:38 GMT
Subject: Getting the XP clipboard in tkinter
Message-ID: <3D1F1A81.95602DE0@tds.net>

Hi,

If a clipboard_get() convenience method existed in tkinter, what would
it look like?

I would like to do something that seems trivial using Tkinter:
namely get the contents of the "global" clipboard in Windows XP.  When
the focus in a text widget my app uses virtual Cut, Copy and Paste
events, like this:

w.event_generate(virtual_event_name("Copy"))

This works.  That is, the text copied in this manner can be read by my
app and can be pasted by my app using the similar "Paste" event.

However, there are times when I need to get this same information when
the focus is in a canvas, and I just don't see how to do it.  In such
cases, my app writes to the clipboard using:

app().root.clipboard_clear()
app().root.clipboard_append(s)

where app().root is a top level widget.  Other apps can read what has
been written to the clipboard!  But the obvious code to read the
clipboard:

s = app().root.selection_get()

takes an exception. I must be missing something really simple.  The
following two statements also take exceptions:

s = app().root.selection_get("clipboard")
s = app().root.selection_get(".")

Please help.  This is driving me crazy.  

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


From pyth at devel.trillke.net  Sat Jun 15 08:44:43 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 15 Jun 2002 14:44:43 +0200
Subject: Integer conversion
In-Reply-To: <3D0B3257.E8572AE5@sympatico.ca>; from cjw@sympatico.ca on Sat, Jun 15, 2002 at 08:25:59AM -0400
References: <3D0B3257.E8572AE5@sympatico.ca>
Message-ID: <20020615144443.C15079@prim.han.de>

Colin J. Williams wrote:
> int(string, base) => integer
> 
> For int('127',  8), I had expected this function to deliver 87.

with python2.1/2.2/2.3

>>> int('127',8)
87

> The example below correctly gives errors at the top and bottom of the
> range and
> for int('127', 10) delivers 127.  The other cases are not what I
> expected.

IMO they work very nicely. Please copy the EXACT screen session
so we can help.

    holger




From R.Barrett at ftel.co.uk  Mon Jun 24 06:32:39 2002
From: R.Barrett at ftel.co.uk (Richard Barrett)
Date: Mon, 24 Jun 2002 11:32:39 +0100
Subject: Python and remote sensing
Message-ID: <5.1.0.14.2.20020624113128.00ad9e30@pop.ftel.co.uk>

Repeat correcting error in last line of example quoted:

At 11:23 24/06/2002 +0200, Siegfried Gonzi wrote:
>Hi,
>
>
>a technician of us asked me whether it is possible to do the following:
>
>
>Starting a script on Unix at lets say 2 am and calling a remote site 
>(http://www.whatever...) and that site should become passively processed 
>(the site holds 41 observing stations). Is it possible to start a remote 
>script (automatically) connect to a http page and retrieve information 
>from that html-page?
>I answerd Perl or Python can do the job. But I am not sure whether that is 
>correct.
>
>I am sure the above topic has been asked 1000 times. Or is there a tool or 
>program availabe which can do the job better?; I mean our technican has 
>never heard of Python and I am not interested to make web based (put in 
>the right term here) programming or the like.
>
>If Python can do it - okay; if not please say so (if Python is dead - 
>okay; if Python lives - okay)!
>
>
>Thanks,
>S. Gonzi

You can write a script in Python or Perl to be executed by the UNIX cron 
daemon at the time you specify (see the UNIX crontab command man page). 
There is nothing out of the ordinary about such scripts beyond their being 
listed in an appropriate user's installed crontab and you will not want it 
to prompt for input parameters.

In the case of Python you can simply use the urllib, urllib2 or httplib to 
retrieve data using HTTP; these modules being listed in order of reducing 
simplicity of use.

Try using urllib from the python command line and you'll see how easy it is.

Probably the biggest issue will be if you have to use any form of 
authentication in order to get a response from the HTTP server. But that 
isn't a problem

Perl almost certainly has equivalent modules if you prefer it to Python.

With Python 2.2.1 a basic script like the one that follows will do it if 
you have to use HTTP basic authentication. You'll need to add some extras 
to handle problems with opening the URLs, supplying parameters if the URLs 
need to incorporate them etc. But basically its really quite easy to do.


#! /usr/bin/env python

import urllib

# Hardwired HTTP basic authentication parameters
basic_auth_username = '....'
basic_auth_passwd = '....'
# URL we want data from
url = 'http://.......'
# Destination filename for data retrieved
destination_filename = '.....'

class myURLopener(urllib.FancyURLopener):

     def prompt_user_passwd(self, host, realm):
         return basic_auth_username, basic_auth_passwd

urllib._urlopener = myURLopener()

stuff = urllib.urlretrieve(url)

open(destination_filename, 'w').write(open(stuff[0], 'r').read())




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





From aahz at pythoncraft.com  Wed Jun 19 16:17:55 2002
From: aahz at pythoncraft.com (Aahz)
Date: 19 Jun 2002 16:17:55 -0400
Subject: OS X debugger?
Message-ID: <aeqotj$dm$1@panix1.panix.com>

Anyone know a good visual debugger for OS X?  I'd originally been
planning to just use shell windows for the demos of my OSCON tutorial,
but it occurs to me that a single-stepping debugger with watch variables
would be even better.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From msoulier at nortelnetworks.com  Thu Jun 20 14:12:31 2002
From: msoulier at nortelnetworks.com (Michael P. Soulier)
Date: Thu, 20 Jun 2002 14:12:31 -0400
Subject: do...while
In-Reply-To: <XFMail.20020620104256.shalehperry@attbi.com>
References: <slrnah40l1.12v.msoulier@pmerd071.ca.nortel.com> <XFMail.20020620104256.shalehperry@attbi.com>
Message-ID: <20020620181230.GB2402@nortelnetworks.com>

On Thu, Jun 20, 2002 at 10:42:56AM -0700, Sean 'Shaleh' Perry wrote:
> 
> On 20-Jun-2002 Michael P. Soulier wrote:
> >     Greetings.
> > 
> >     How do you people handle the lack of a do...while in Python? I find that
> > at times, I must initialize variables for a loop with identical code to the
> > loop, which feels like a waste to me. 
> > 
> > ie.
> > 
> >     line = filehandle.readline()
> >     while len(line) > 5:
> >         line = filehandle.readline()
> > 
> >     A do...while here would be nice. 
> > 
> >     Just curious. 
> > 
> 
> while 1:
>     line = filehandle.readline()
>     if len(line) <= 5: break
>     ....
>     ....
> 
> is a typical way to implement it.

    Right, I've used that in the past. It doesn't seem nearly as clean though,
as making the conditions for terminating the loop plain in the loop
declaration. Considering the Python community's leaning towards readable code,
necessitating a break for code factoring goes against that philosophy. 

    Mike

-- 
Michael P. Soulier, QX41, SKY  Tel: 613-765-4699 (ESN: 39-54699)
Optical Networks, Nortel Networks, SDE Pegasus
"...the word HACK is used as a verb to indicate a massive amount
of nerd-like effort."  -Harley Hahn, A Student's Guide to Unix




From mcfletch at rogers.com  Tue Jun  4 12:00:27 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 04 Jun 2002 12:00:27 -0400
Subject: int, float and string...
References: <adhut1$10rtkl$1@ID-146704.news.dfncis.de>
Message-ID: <3CFCE41B.1070104@rogers.com>

When you cast a float to an int you potentially lose information (the 
decimal fraction).  The conversion is also by way of truncation (for 
historic reasons), so 1.99999 -> 1 (not 2), which likely isn't what you 
wanted, (use round( num, 0) instead).  Floats (C doubles) are larger 
than ints (C longs) at the machine level (8 vs 4 bytes on 32-bit 
machines if I recall correctly), but the overhead of Python objects is 
huge compared to the size of the actual data, so don't worry about it 
unless you're using Numeric Python and slogging around huge arrays of 
numbers (where worrying about it might actually save you time).

Store your data in the format appropriate to it (if it's a 
precision-based measurement, store floats, if it's a count, store 
integers.  If you're multiplying an int and a float, store a float 
result (which is the natural result)).

You normally only worry about converting numbers when you're going from 
one working domain to another (I calculate a weighting value (float) 
based on analysis of some data, then want to switch to using that value 
as an index into an array of values, so need to convert to an integer so 
it can be used as a simple index: my data-storage domain has a different 
idea of the data type than my index-calculation domain, so I have to 
translate between them).  Or where calculations require floating-point 
operation (most notably division).

Where possible, use the same data format for all pieces of the same data 
(don't store a distance as an integer just because it happens to equal 
5.00, use a float consistently throughout your code.  Don't store a 
count of people as a float, always keep it as an integer, and convert it 
only when you want to do calculations on it which require non-integral 
operation).

Natural float-to-int conversion (rounded to nearest int):
	integer = int(round(value,0))
Check whether float is an integer value:
	float % 1.0 == 0.0

HTH,
Mike

Shagshag13 wrote:
> Still wondering questions about pythoner's way...
> 
> Why does i have :
> 
> 
>>>>int('0.0123456789')
>>>
> Traceback (most recent call last):
>   File "<pyshell#157>", line 1, in ?
>     int('0.0123456789')
> ValueError: invalid literal for int(): 0.0123456789
> 
> The correct way to avoid is it :
> 
> 
>>>>int(float('0.0123456789'))
>>>
> 0
> 
> ???
> 
> And to check if a number should be use as an int rather than a float, can i use :
> 
> 
>>>>int(float('0.0123456789')) == float('0.0123456789')
>>>
> 0
> 
>>>>int(float('5.00000000000009')) == float('5.00000000000009')
>>>
> 0
> 
>>>>int(float('5.0000000000000')) == float('5.0000000000000')
>>>
> 1
> 
> or i miss a better way  ???
> 
> By doing this kind of stuff do i save memory ?
> (having a huge list of int and float, casting them to float only for math operations)
> 
> thanks,
> 
> s13.
> 
> 


-- 
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From theller at python.net  Fri Jun  7 10:14:19 2002
From: theller at python.net (Thomas Heller)
Date: Fri, 7 Jun 2002 16:14:19 +0200
Subject: Emulating classmethod in Python 2.1
References: <915a998f.0206060833.12ad76a8@posting.google.com> <ado5v8$uih9$1@ID-59885.news.dfncis.de> <915a998f.0206070611.295149d3@posting.google.com>
Message-ID: <adqf6u$1egdj$1@ID-59885.news.dfncis.de>

"Hamish Lawson" <hamish_lawson at yahoo.co.uk> wrote in message news:915a998f.0206070611.295149d3 at posting.google.com...
> Thomas Heller wrote:
>
> > Here's the closest I came up with (you must rewrite your code to
> > derive SomeClass from object, the rest can remain unchanged). ...
>
> While ideally I was hoping not to have change any client code, your
> solution is the next best thing in requiring only a minimal change and
> in working with both Python 2.2's builtin classmethod() and with an
> emulating wrapper. Thanks!
>
You're welcome.
Note that you have to extend my code a little if you are
creating subclasses of SomeClass, but I'll leave that to you ;-).

I was astonished to see that classmethod() can be used in
classic classes, BTW.

Thomas




From jiri at baum.com.au  Sat Jun  1 00:04:40 2002
From: jiri at baum.com.au (Jiri Baum)
Date: Sat, 1 Jun 2002 14:04:40 +1000
Subject: Generating unique numbers?
References: <20020531195255.A2297@unpythonic.net> <EFEFIHAINIINNELLHFAIIEGFCHAA.kseehof@neuralintegrator.com> <mailman.1022899571.29701.python-list@python.org>
Message-ID: <o4h9da.7s9.ln@10.0.0.1>

Jeff (jepler at unpythonic.net) wrote:
> Yeah, but it was fun to try to ballpark the numbers.  One thing I didn't
> mention in my earlier message is that 2**65 unique IDs is right around
> the number necessary to see birthday-paradox collisions in 128-bit hashes
> such as MD5 with p=0.5.  Still very uncommon for a 160-bit hash like SHA,
> though.

Personally, I'd be more comfortable with random numbers than with the
IP/process-id/etc thing, because random numbers eliminate the human factor.
As has been noted, lots of people use 10.0.0.1. *I* use 10.0.0.1. There are
ethernet cards with the same ether addresses. Random numbers can at least
be analyzed mathematically and relied-upon to obey the predictions.


Jiri
-- 
Jiri Baum <jiri at baum.com.au>           http://www.csse.monash.edu.au/~jirib
  MAT LinuxPLC project --- http://mat.sf.net --- Machine Automation Tools


From fgeiger at datec.at  Thu Jun  6 14:48:48 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Thu, 6 Jun 2002 20:48:48 +0200
Subject: Emulating classmethod in Python 2.1
References: <915a998f.0206060833.12ad76a8@posting.google.com>
Message-ID: <adoarc$2ag$1@newsreaderg1.core.theplanet.net>

"Hamish Lawson" <hamish_lawson at yahoo.co.uk> schrieb im Newsbeitrag
news:915a998f.0206060833.12ad76a8 at posting.google.com...
> Is it possible to define a 'classmethod' class or function for Python
> 2.1 that would effectively provide the functionality of Python 2.2's
> built-in classmethod function and have the same usage pattern? I'd
> like to be able run code like that below against both 2.2 (using its
> builtin classmethod function) and 2.1 (using some added-in wrapper).
>
>     class SomeClass:
>         def hello(cls, name):
>             print "Hi there", cls, name
>         hello = classmethod(hello)
>
>     SomeClass.hello("Peter")
>
> I've come across Thomas Heller's recipe in the online Python Cookbook,
> but this would require SomeClass to be modified, and so fails my
> criteria.
>
> I've also come across Alex Martelli's recipe for emulating
> staticmethod, but of course that doesn't deal with passing a class to
> the wrapped method.
>
>
> Hamish Lawson




From lewy0lewy at poczta.onet.pl  Tue Jun 18 14:25:27 2002
From: lewy0lewy at poczta.onet.pl (Pawel Lewicki)
Date: Tue, 18 Jun 2002 20:25:27 +0200
Subject: Caching query resuls
Message-ID: <aentug$m8d$1@news.onet.pl>

Hallo,
Is there any good solution to cache the results of SQL queries? I found the
great cache manager in dmtools (http://csl.anu.edu.au/ml/dm/), but I wonder
if there is any alternative. I will start with MySQL database, but it would
be great to be able to plug it into other RDBMS's. I am interested in both -
file and memory caching. I also know ZSQL Method in Zope.

Pawel Lewicki




From bedge at troikanetworks.com  Thu Jun 20 13:54:14 2002
From: bedge at troikanetworks.com (Bruce Edge)
Date: Thu, 20 Jun 2002 10:54:14 -0700
Subject: simple socket redirector app question
References: <5DrP8.25$OT1.34057@newsfeed.slurp.net>
Message-ID: <bFoQ8.138$vX2.116875@newsfeed.slurp.net>

In <5DrP8.25$OT1.34057 at newsfeed.slurp.net>, Bruce Edge wrote:

> I have an app which talks via sockets to local host clients only. I'd
> like to get at it remotely.
> 
> Basically it would accept remote connections from a remote client, relay
> them the the local app, and forward the replies back to the remote
> client.
> 
> What's the simplest way to write this type of a socket redirector?
> 
> Use SocketServer, or is that overkill?
> 
> Thanks, Bruce.

Thanks to all who responded. You guys basically wrote it for me. 
Let's see M$ support top that.

-Bruce.


From Chris.Barker at noaa.gov  Mon Jun 24 16:39:55 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Mon, 24 Jun 2002 13:39:55 -0700
Subject: Numerical Methods in Python
References: <%32R8.853$xv1.36390126@newssvr15.news.prodigy.com>
Message-ID: <3D178397.6CF931F@noaa.gov>

Russ Dilley wrote:
> My dilemma is that I also have responsibilities in UNIX system
> administration and it seems to me that I might be able to "kill two birds
> with one stone" by learning the Python language. I currently use Perl and
> Shell scripts for system admin tasks and a mixture of FORTRAN, C, MathCAD,
> Matlab, and Excel for numerical work. As you might guess, it tends to get
> confusing constantly jumping from one to another. The syntax and application
> tends to blur and it is somewhat inefficient to "relearn" each to some
> extent whenever I have a new task to address.
> 
> I would like to focus more attention to become proficient with a single
> language and, based on what I've seen, it seems that Python might be
> appropriate for use in system administration AND numerical programming.

Yes, it it good for these tasks. I, and many others, judging by posts to
newsgroups, am using Python for most of the things I used to use MATLAB
for, and also for sys-admin stuff, text file crunching, and developing
GUI programs. One reason I use Python is just what yu stated: I get
confused using too many different languages, I keep getting their syntax confused.

Python with NumPy has many of the advantages and capabilities as MATLAB.
It is weaker in that there are a whole lot fewer bundles libraries of
handy functions, though you can probably fnd much of what you need by
looking around on the web. The performance is similar, if used
similarly, i.e. vectorize!! The other weakness is the lack of well
integrated plotting, but there are a number of packages that will fill
most of your plotting needs, including various GL based 3-d
visualization tools. The primary advantages of Python over MATLAB are:

Much better facilities for non-number chrunching: text file crunching,
GUI development, web development, etc. etc. Python really is a
fabulously general purpose language.

Easier extending with C and Fortran, and with Weave, Pyrex, and
Py_inline, you may not need to.

Cheaper an re-distributable

More platforms supported

NumPy arrays have more data types: sometimes you don't what you use
doubles for everything!

NumPy arrays are N-dimensional, and have "array broadcasting". What this
means is that you can have a (m,1) array for x, and a (1,n) array for y,
and develop functions of x and y that produce an (m,n) array result.
This is a whole lot cleaner an more efficient than all those meshgrid()
cals I used with MATLAB.

Be sure to check out www.scipy.org and good luck!

-Chris
 
-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From neal at metaslash.com  Mon Jun 24 21:46:40 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Tue, 25 Jun 2002 01:46:40 GMT
Subject: Pychecker and /var/tmp/python2-2.2-root/usr/lib/python2.2/.....
References: <d71cc7f6.0206241327.17ebabfa@posting.google.com>
Message-ID: <pan.2002.06.25.01.46.31.138959.7184@metaslash.com>

On Mon, 24 Jun 2002 17:27:30 -0400, Sandeep Gupta wrote:

> I'm using pychecker on my application, which uses BaseHTTPServer.py and
> unittest.py.  Both those files have "No method" warnings that are
> displayed as:
> /var/tmp/python2-2.2-root/usr/lib/python2.2/BaseHTTPServer.py and
> /var/tmp/python2-2.2-root/usr/lib/python2.2/unittest.py
> 
> Does anyone know why /var/tmp/python2-2.2-root gets prepended to the
> file name?
> How can I remove the warnings for these files?

I had this same problem.  From what I remember, Redhat compiles
the .py -> .pyc/.pyo files in the /var/tmp/.... directory.

I didn't find a great way to fix it in pychecker, but to
fix your problem, you should just have to do:

	cd /usr/lib/python2.2
	python compileall.py .

Neal


From donn at drizzle.com  Fri Jun 21 00:44:47 2002
From: donn at drizzle.com (Donn Cave)
Date: Fri, 21 Jun 2002 04:44:47 -0000
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com> <lkznxqjqx2.fsf@pc150.maths.bris.ac.uk> <3D11A21D.9010809@mxm.dk> <uh3e1usajle42c@news.supernews.com> <3D11DA9B.163A7BD0@info.unicaen.fr>
Message-ID: <1024634686.363992@yasure>

Quoth Jerzy Karczmarczuk <karczma at info.unicaen.fr>:
...
| 3. No, in order to speak about genuine gotos *you don't need to think
|    assembler*.
|    What you need is to understand correctly the notion of continuation
|    (something quite standard - again - in the Functional Programming
|    world). This is a high-level model of "flat" branching, but may be
|    used to more exquisite constructions, such as backtracking.

The stackless branch of python supported a very rich set of
continuation options.  I use the past tense, because it seems
to have been a little too powerful for even the mainly feature
hungry python crowd, so current development of stackless has
dropped most of that stuff.  I hope we're talking about the same
thing - these continuations suspend and resume computations like
a kind of serialized multi-threading.

For me, genuine gotos are only a matter of flow control along the
same lines as "while" loops.  Internal to a computation, if you
will - one doesn't goto out of or into a procedure.  I think in
some sense of "ideal", an ideal procedural language would have goto,
because it expresses some problems most directly and the need to
emulate it with "while" or "try" makes worse code.  The branching
in a goto construct is perfectly apparent, it just can be harder
to get an overall sense of the flow if goto is over-used.  Why,
even our beloved python itself had a few judiciously used gotos
last time I looked.

Continuations I'm not so sure about.  I wrote some stackless stuff
that made really interesting (to me, anyway) use of them, but now
we're talking about flow of control that really is easy to misunderstand.

|    I don't believe that there are any technical difficulties in the
|    implementation of gotos at the low level. At the high level this
|    is then a question of the language structure, as seen by Van Rossum,
|    the lack of gotos was a conscious decision.

If Michael Hudson says there are technical difficulties, I would
take that as fact, but a fact about Python's current internals.
If goto had been wanted, I'm sure Python's design could have
accommodated it.

	Donn Cave, donn at drizzle.com


From uwe.becher at web.de  Fri Jun 21 12:49:37 2002
From: uwe.becher at web.de (Uwe Becher)
Date: Fri, 21 Jun 2002 18:49:37 +0200
Subject: PIL images in CGI scripts
Message-ID: <aevld7$bv9$05$1@news.t-online.com>

Hi together,

i try to generate dynamic images in Python CGI scripts on an apache
server
on windows. In my first attempts the browser showed miscoloured images. 
Searching the internet i learned, that i have to output the image as 
binary stream to stdout. I succeeded by calling the Python interpreter 
with the option -u via the shebang #!d:/python21/python.exe -u.
My question:
Is it possible to define the binary mode in the script without
using the -u option?

Thankys in advance

Uwe


From G.H.Lusby at sms.ed.ac.uk  Sat Jun  8 11:58:06 2002
From: G.H.Lusby at sms.ed.ac.uk (Gavin Lusby)
Date: Sat, 08 Jun 2002 16:58:06 +0100 (BST)
Subject: Matrix Multiplications
Message-ID: <1023551886.3d02298e2d69b@sms.ed.ac.uk>

Dear all,

I am an engineering student trying to further my research with python 
programming.  Would someone please tell me how I would do a certain 
matrix multiplication in Pythin/NumPy...

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

I have two large matricies A[n,n] and B[n,n]

I would like to get an outer product of:

C[k+1:,k:] = B[k+1:,k] * A[k,k:]

where,	k = 0,...,n-1

It is in effect, an outer product of two vectors of different length.

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

I keep getting errors and in the end it's probably something 
straightforward but I'm no Computer Scientist having been spoon-fed 
Fortran 77.  

I would be most grateful if someone could tell me how to do this 
very efficiently in Python.

Many thaks in advance,

Gavin Lusby 




*********************************************************
Gavin Lusby
University of Edinburgh
email:    g.h.lusby at sms.ed.ac.uk




From threeseas at earthlink.net  Wed Jun 26 09:57:56 2002
From: threeseas at earthlink.net (Timothy Rue)
Date: Wed, 26 Jun 2002 13:57:56 GMT
Subject: Reminder of a python Project.
References: <2218.941T781T12224051threeseas@earthlink.net> <afc2mn$2kn$1@verence.demon.co.uk>
Message-ID: <2948.942T2959T5796077threeseas@earthlink.net>

On 26-Jun-02 04:49:11 Paul Wright <-$P-W$- at verence.demon.co.uk> wrote:
>In article <2218.941T781T12224051threeseas at earthlink.net>,
>Timothy Rue <threeseas at earthlink.net> wrote:
>>First off, it's ok that I got such a high level of resistance to my
>>initial posting here about this project. Thanks to some research that /.
>>had an article on, regarding the size of GPL project teams.
>>
>>But anyway.... Remember the autocoding project I mentioned?


>                      ______
>                     /      \
>                   .' PLEASE `.
>                   |  DO NOT  |      _____
>                   | FEED THE |    ,'.....`.
>                   `. TROLLS ,'  ,'........ )
>                     \_    _/   |........ ,'
>                       |  |     `. .... _/
>                       |  |      ,'.,'-'
>                       |  |     /../
>                       |  |   ,'.,'
>                       |  |  /../
>                     . |  | /..'
>                   .\_\|  |/_/,
>                   ___ |  | ___
>                     . `--' .
>                      .    .

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


Nice! Mind if I borrow it?

Other words that you can throw around include "spam", hell you don't even
have to know the definition.

Seriously, ifn people want to babel flaming and all that crap, I do
believe Slash dot has a story that links to the 60 billion dollar article,
which links to the 309 page research/report. And we all know that slash
dot have a comment section? Perhaps some of your don't?

You'll have to look unders yesterdays stories but I'm sure you will find
plenty of others just like you, and regardless of where you are in your
perspective about this.

It's pretty clear what intents are of those who want to shoot the
messanger, considering plenty of researched information was linked to.

What's the word to use for such people that ignore or pretend to ignore
such information but insist on flaming anyone mentioning it?

Something that would fit the Matrix Agent metaphor. Or would it be
something far more common? ignorance?

Anyway, the flamers didn't have to respond at all. It was a simple post
and the reference to research indicating that GPL projects have a low
count of developers, with what I believe is 1 as a mean, is as well what
research found (sorry no link - search for it yourself - /. had it linked)

It was an informational post, that all.

Amazing who all something like that get a flaming response from. Good
thing the world is not made up of nothing but, for we wouldn't have any
research then.

---
*3 S.E.A.S - Virtual Interaction Configuration (VIC) - VISION OF VISIONS!*
   *~ ~ ~      Advancing How we Perceive and Use the Tool of Computers!*
Timothy Rue      What's *DONE* in all we do?  *AI PK OI IP OP SF IQ ID KE*
Email @ mailto:timrue at mindspring.com      >INPUT->(Processing)->OUTPUT>v
Web @ http://www.mindspring.com/~timrue/  ^<--------<----9----<--------<



From emile at fenx.com  Fri Jun 28 08:49:12 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 28 Jun 2002 12:49:12 GMT
Subject: Be the first on your block to register your computer.
References: <mailman.1025211444.21440.python-list@python.org> <afg1ks$e6lj5$1@ID-71831.news.dfncis.de> <3D1B6E69.1090204@earthlink.net> <IXWS8.368185$o66.975748@news-server.bigpond.net.au> <0F2CD32D80F7F061.3D8D80DBD1BF70E9.B80FB476A9BA57A6@lp.airnews.net>
Message-ID: <bXYS8.347255$cQ3.21314@sccrnsc01>

Cameron Laird
> soft's marketing machine is sufficiently powerful to
> induce people to believe the people who create problems
> should be trusted to solve them.

And for the past ~20 years, they've similarly convinced the US
government.


--

Emile van Sebille
emile at fenx.com

---------



From siegfried.gonzi at kfunigraz.ac.at  Sat Jun 22 15:44:52 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Sat, 22 Jun 2002 21:44:52 +0200
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122>
Message-ID: <3D14D3B4.7050709@kfunigraz.ac.at>

Bengt Richter wrote:

> Have you never chased a bug that only shows up in secondary effects?
> In C code, the most common way for that to happen is storing something
> through a bad pointer, but instead of getting a memory fault, it corrupts
> some data structure or data value. The value could be used 20 minutes later
> and cause GPF then, or e.g., it could be data that's part of a calculation,
> but maybe it's only overwriting the least significant half of a floating
> point number, so you get the effect systematic bias or random error injected
> into your calculation, but nothing else. If you don't look for explanations
> of last-decimal errors, this kind of thing can persist a long time. OTOH,
> you could be corrupting numbers that are used in an iterative calculation
> that mysteriously starts to take longer to converge, and does so in a systematically
> increasing way, because the corrupting data varies systematically with time.
> Or you could be corrupting data structures and sabotaging memory management
> without causing immediate GPF, etc. etc.

The explanation is good and could explain the bad behavior. But note: I 
use the same C code on Windows and on Linux! On Windows I used SWIG and 
Visual C++. On Linux I used gcc and SWIG.

Theoretically I could do the following: trying to use the Fortran code 
instead of the C code. The C code relies on a Fortran 77 code. I pinched 
the C code from the netlib (or so). But I could try to use F2PY in 
combination with the Fortran 77 code.

The F2PY would have the advantage that I am not forced to write 
explicitely pointer-wrapper functions in Python in order to use the code 
which comes from SWIG.


The above could also explain why under Windows the system sucks down, 
but it is not destroyable!


S. Gonzi



From sholden at holdenweb.com  Wed Jun  5 11:56:28 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 5 Jun 2002 11:56:28 -0400
Subject: DCOracle2 - Can't get cursor.execute to return data when SQL query involves datetimes.
References: <be0d725a.0206041139.1f8f6647@posting.google.com> <be0d725a.0206050557.1592bc3b@posting.google.com>
Message-ID: <ZAqL8.280027$%u2.29861@atlpnn01.usenetserver.com>

"Jim" <jtt at hnc.com> wrote in message
news:be0d725a.0206050557.1592bc3b at posting.google.com...
> Here is better version of the same program.  Sorry for last posting.
>
> ! /usr/bin/env python
> import DCOracle2
> #*************************************************************
> # Define variables and constants
> #*************************************************************
> TRUE = 1
> FALSE = 0
> FalconPwd = 'uid/pwd'
> Results_count = 0
> db = DCOracle2.connect(FalconPwd)
> #
> # Open a new cursor
> #
> dbCursor = db.cursor()
> tran_date = '19970116124248'
> dt = DCOracle2.Timestamp(int(tran_date[0:4]), int(tran_date[4:6]),
> int(tran_date[6:8]), int(tran_date[8:10]), int(tran_date[10:12]),
> int(tran_date[12:14]))
>
> Results_count = DCOracle2.execute(dbCursor,'select * from
> falcon.falt002d where ACCT_NBR = :1 and trn_dt = :2',"2222222222222222
>   ",dt)
>
>
> tabledata = dbCursor.fetchone()
>
> print tabledata

1. Is there any reason why you are using "DCOracle2.execute(dbCursor..."
rather than "dbCursor.execute(..."? You should really be calling teh
cursor's method (assuming the DCOracle2 module follows the DB API
recommendations).

2. Is there any reason why you are using '"2222222222222222",dt)' rather
than '("2222222222222222",dt))' - teh DB PAI suggests execute() should take
two arguments, the first a SQL statement and the second an (optional) tuple
of substitutions.

regards
 Steve
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From huaiyu at gauss.almadan.ibm.com  Tue Jun  4 21:19:44 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 5 Jun 2002 01:19:44 +0000 (UTC)
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <slrnafps1c.4ea.huaiyu@gauss.almadan.ibm.com> <ufqm0voskv86cc@news.supernews.com>
Message-ID: <slrnafqppg.4ea.huaiyu@gauss.almadan.ibm.com>

John Roth <johnroth at ameritech.net> wrote:
>The assignment occurs because the pointer to the list has to
>be changed to point to the (possibly) new location of the list
>in memory. In other words, the second element of the tuple
>has to be rebound.
>
>There is no way of avoiding this, since objects are known
>internally by their location in memory, and if you change the
>object, you may (not necessarilly will) change that object's
>location. Therefore, the pointer to the object must also be
>changed in this instance. That is what causes the exception.

I don't think objects are pointers to memory locations.

>>> a = [1,2,3]
>>> b = a[1] = ['a','b','c']
>>> a[1] += ['d']
>>> b
['a', 'b', 'c', 'd']

This also works when the list is of an appropriate size that appending
actually reallocates memory and do a copy.

Objects are referenced by their ids (or equivalent).  Their memory location
may or may not change when it is modified, but if it is the same object,
every name that was refering to it will remain referencing to it after the
modification, regardless of its memory location.

The operations of a+=b for immutable objects is of a completely different
nature.  If creates a different object (a+b) and rebinds the name 'a' to it.
If any other names were bound to the object that was named 'a', they will
not feel the effect that 'a' is now pointing to a different object.  They
will remain bound to the old object.

Huaiyu 


From FooWeissBarMike at hotmail.com  Wed Jun 19 11:28:36 2002
From: FooWeissBarMike at hotmail.com (Michael Weiss)
Date: Wed, 19 Jun 2002 11:28:36 -0400
Subject: SafeArray problem
References: <3d105710.83643750@news.skynet.be> <3d108014$1@news.mhogaming.com> <3d108545.95472593@news.skynet.be>
Message-ID: <3d10a3fa$1@news.mhogaming.com>

> I have not written an idl-file myself, I just used the classwizard in
> Visual Studio to write my method on the automation object (using MFC).
> What do you really want to know? (I looked for idl-files in my project
> but did not find any)

How about an ODL file?

I was just wondering if the in, in/out, retval, pointer vs. non-pointer
symantics of the parameter were wrong. I was guessing by adding a new
parameter (the long) that the safearray's properties were changed for the
better. But I guess with a pure Dispatch interface (dispinterface) that this
kind of stuff isn't done, or at least its done differently.

Still, I wonder if this still isn't the problem of the function not being
declared correctly before you made the change and then when you added that
long the declaration of the safearray was changed.







From jimmy at retzlaff.com  Mon Jun  3 18:56:47 2002
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Mon, 3 Jun 2002 15:56:47 -0700
Subject: Version strings in executable created by Py2exe
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F569C@MAYOR.retzlaff.com>

Robert L. Oelschlaeger writes:
> > Robert L. Oelschlaeger [mailto:roelsch at comtrak.com] wrote:
> > > Running with Python 2.2. I've created myAppName.exe using 
> > Py2exe, with
> > > the "version_companyname" and "version_legalcopyright" specified
in
> > > the setup .cfg file as follows:
> > ...
> > > Why are the version_companyname and version_legalcopyright 
> > strings not
> > > installed for the "Company Name" and "Copyright"?
> > 
> > "version" should be followed by a dash rather than an underscore.
So,
> > for example, you should use "version-companyname" rather than
> > "version_companyname".
> > 
> > Jimmy
>
> I tried the name change, but it didn't help: still no 'Company Name'
or
> 'Copyright' strings in the resulting .exe file.
>
> I think there is a documentation problem:
>
>     the example at http://starship.python.net/crew/theller/py2exe/
uses
>     hypenated names, but
>
>     the names in py2exe/build_exe.py are the underscored names (e.g.,
>     self.version_companyname).

Probing a little deeper, it appears as if either - or _ should work.
>From the part of distutils which parses setup.cfg:

opt = string.replace(opt, '-', '_')

Here is a part of my setup.cfg (of course the word-wrapping is not in
the file):

[py2exe]
icon=pyramid.ico
version-companyname = Aver Development Corporation
version-legalcopyright = Copyright (C) 2001-2002 Aver Development
Corporation. All Rights Reserved.
version-filedescription = Alchemist
version-fileversion = 0.0.0.1
version-legaltrademarks = Alchemist is a Trademark of Aver Development
Corporation
version-productname = Alchemist
version-productversion = 0.0.0.1

All of those version strings end up in my exe (i.e., I can see them all
if I get properties on the exe from explorer and go to the version tab).
I'm using Python 2.2.1 and py2exe 0.3.3. I know I've been doing this
successfully since Python 2.2 and since the first version of py2exe that
supported Windows version resources. Which version of py2exe are you
using? Anything out of the ordinary? Are you using the standard binary
distributions of Python and py2exe? Maybe you could post your setup.cfg
file verbatim?

Jimmy




From gh_pythonlist at gmx.de  Tue Jun 18 19:59:01 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 19 Jun 2002 01:59:01 +0200
Subject: procmail replacement in Python
Message-ID: <20020618235901.GA2197@lilith.my-fqdn.de>

Does such a thing exist? I've recently looked into the procmail sources
and run away screaming. I'd very much prefer a Python replacement to
such badly written C code.

Does anything exist that I can start from? I've even contemplated not
inventing yet another filter language, but just using plain Python
functions.
 
If I do this, I'd like to replace fetchmail with getmail as well.

What would you suggest as approach?

- writing it as a getmail plug-in
- writing it as a standalone program to deliver to mailboxes a la
  procmail
- let getmail deliver to a temporary folder, then using my procmail
  replacement to filter into the right folders, delete spam, etc.

The third one looks to be the least likely to lose mail, but the
temporary folder approach looks like a kludge.

Gerhard

PS: I've seen something _very_ limited written in Python on Sourceforge
that only does Maildir, but the name is escaping me.
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 23.3 ?C      Wind: 0.9 m/s




From pyth at devel.trillke.net  Sun Jun 30 13:10:06 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 30 Jun 2002 19:10:06 +0200
Subject: arguments to functions
In-Reply-To: <MPG.178955a4fc8bfc1c989680@news.rz.uni-karlsruhe.de>; from merkosh@hadiko.de on Sun, Jun 30, 2002 at 06:52:10PM -0700
References: <MPG.178955a4fc8bfc1c989680@news.rz.uni-karlsruhe.de>
Message-ID: <20020630191006.C10625@prim.han.de>

Uwe Mayer wrote:
> hi,
> 
> I wanted to pass one named and a collection of unnamed (an arbitrary 
> number) to a function:
> 
> def test(expected=None, *misc):
>     print(expected)
>     print(misc)
> 
> Now, unluckyly 'test' interprets the first unnamed parameter to test as 
> the value for 'expected'

which is expected (and obviously you want to have a double **). 
Or what did you expect? 

You can easily do:

    def test(**kwargs):
        expected = kwargs.get('expected',None)

the dictionaries' get method returns None if it doesn't have 'expected'
as a key.

HTH,

    holger




From marklists at mceahern.com  Wed Jun 12 18:32:03 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 12 Jun 2002 17:32:03 -0500
Subject: map -> class instance
In-Reply-To: <m2r8jcp3pv.fsf@video.paradise.lost>
Message-ID: <NCBBLFCOHHDIKCAFGCFBKEFELEAA.marklists@mceahern.com>

> I want to convert dictionary instances to instances of a class that
> has the keys of the map as attributes.

Try this:

class mapper:

    def __str__(self):
        s = "<mapper>"
        for k, v in self.__dict__.items():
            s += "<%s>%s</%s>" % (k, v, k)
        s += "</mapper>"
        return s

    __repr__ = __str__

def map2class(d):
    m = mapper()
    for k, v in d.items():
        setattr(m, k, v)
    return m

d = {'first' : 'John',
     'last' : 'Hunter',
     'age' : 34}

x = map2class(d)

print x

**
Or, slightly different:
**

class mapper:

    def __init__(self, d=None):
        if d:
            for k, v in d.items():
                setattr(self, k, v)

    def __str__(self):
        s = "<mapper>"
        for k, v in self.__dict__.items():
            s += "<%s>%s</%s>" % (k, v, k)
        s += "</mapper>"
        return s

    __repr__ = __str__

d = {'first' : 'John',
     'last' : 'Hunter',
     'age' : 34}

x = mapper(d)

print x


-





From sholden at holdenweb.com  Fri Jun 28 09:53:42 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 28 Jun 2002 09:53:42 -0400
Subject: Is Python growing?
References: <3D1B92FB.59B590A0@pop.ntlworld.com> <slrnahnif7.1c6.matt@happy-hour.mondoinfo.com> <7x66042j50.fsf@ruckus.brouhaha.com>
Message-ID: <pSZS8.108031$mF.55681@atlpnn01.usenetserver.com>

"Paul Rubin" <phr-n2002b at NOSPAMnightsong.com> wrote ...
> Matthew Dixon Cowles <matt at mondoinfo.com> writes:
> > On the other hand, some features make Python much easier to use and
> > read. A while ago, I replaced:
> >
> > def lengthOfLongestStr(list):
> >   l=len(list[0])
> >   for count in range(1,len(list)):
> >     if len(list[count])>l:
> >       l=len(list[count])
> >   return l
> >
> > with
> >
> > def lengthOfLongestStr(myList):
> >   return max([len(s) for s in myList])
> >
> > To my eye, the second one is much easier to read.
>
> Well, you could have said
>
>     def lengthOfLongestStr(myList):
>       return max (map (len, myList))
>
> without needing any new features added.

NOW you tell him.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From uwe.schmitt at procoders.net  Fri Jun  7 17:19:33 2002
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 7 Jun 2002 21:19:33 GMT
Subject: direct output to printer
Message-ID: <adr815$e6m39$2@hades.rz.uni-sb.de>

Hi,

I'd like to print barcodes with a special barcode printer.
Therefore I have to send raw ASCII data to the printer.
How can I do this using Python under Windows ???

Greetings, Uwe.

-- 
Dr. rer. nat. Uwe Schmitt        python, php, c, c++, codes, tips, tutorials at
uwe.schmitt at procoders.net        http://www.procoders.net


From op73418 at mail.telepac.pt  Tue Jun 18 07:59:21 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Tue, 18 Jun 2002 12:59:21 +0100
Subject: what *is* a class?
References: <B43D149A9AB2D411971300B0D03D7E8BF0A396@natasha.auslabs.avaya.com> <mailman.1024398742.13711.python-list@python.org>
Message-ID: <jp7uguoqjs7t23itb7vihai30cnv3qr7sm@4ax.com>

On Tue, 18 Jun 2002 13:11:04 +0200, holger krekel
<pyth at devel.trillke.net> wrote:

>Delaney, Timothy wrote:
>> > From: holger krekel [mailto:pyth at devel.trillke.net]
>> > 
>> > I agree that this is a good thing. But it wouldn't hurt to have
>> > 
>> >     reverse, sort, extend
>> > 
>> > as standalone functions somewhere in the stdlib.  They would return
>> > a copy...
>> 
>> Personally, I think this is a great idea. But I would exclude extend.
>
>I'd even like to have something like extend (see below) ...
>
>> I think there should be a library of "iterable" functions (perhaps in
>> builtins ;) which return an iterator, and accept any iterable (where
>> appropriate). Let's call it builtin module iterable (or maybe iterator ...).
>> Functions defined in this module would be specified to not modify the
>> original iterable (indeed, should all call iter(iterable) before doing
>> anything). Of course, whee appropriate a function passed in could modify the
>> original.
>> 
>> Examples of such functions would be:
>> 
>> iterable.sort(iterable [, func])
>> iterable.reverse(iterable)
>
>that would be quite expensive for common cases (like lists).  reverse and 
>sort should also allow to operate on complete sequence types. 
>iterable.reverse could be cheap again then and iterable.sort would be
>as expensive as it is now (but not requiring any inplace-steps).
>
>> iterable.map(func, iterable, ...)
>> iterable.filter(func, iterable, ...)
>> iterable.reduce(func, iterable, ...)
>> iterable.zip(iterable, ...)
>> iterable.range([start,] stop[, step])
>
>yes, indeed! i'd probably do 'from iterable import *' :-)
>
>seriously, i'd include something like concat as an iterable, too: 
>
>    for item in iterable.concat(iter1,iter2):
>        # items iterates with iter1, then iter2 
>
>All the the iterable functions should accept sequence and 
>possibly mapping types (same semantics as for 'for'-loops now).
>
>> There have been suggestions like this before, but I don't recall one for
>> sort() and reverse(), and these two have fired me :)
>
>I think an iterable module should be done. Although there would 
>inevitably be overlapping functionality with the current builtins.
>
>Maybe something for the train ride to EuroPython :-)
>
>> Tim Delaney
>
>    holger krekel
>
>
>

Here goes my personal version of this module. It is largely based off
Sami Hangaslammi recipe at Active State plus some additions/differences
of mine.

"""
Iterator utilities module.

Acknowledgements: Largely inspired by Iterator utilities by Sami
Hangalasmmi.
"""

#Import generators.
from __future__ import generators


__version__ = 1.0
__author__ = "G. Rodrigues"


def iterzip(*iterators):
    """Iterator version of builtin zip."""
    #Initialize variables.
    iterators = [iter(iterator) for iterator in iterators]
    while 1:
        yield tuple([i.next() for i in iterators])


def iterfilter(iterator, func = lambda x: x):
    """Iterator version of builtin filter."""
    for elem in iterator:
        if func(elem):
            yield elem


def iterwhile(func, iterator):
    """Iterate while func(elem) is true."""
    for elem in iterator:
        if func(elem):
            yield elem
        else:
            raise StopIteration


def itermap(func = lambda *x: x, *iterators):
    """Iterator version of builtin map. No padding is done."""
    for tupleelems in iterzip(iterators):
        yield func(*tupleelems)


def iterconcat(*iterators):
    """Concatenates all the iterators passed as arguments."""
    for iterator in iterators:
        for elem in iterator:
            yield elem


def iterapply(iterator, *args):
    """Applies an iterator of callables iteractively to args."""
    #Initialize variables.
    arg = args
    for func in iterator:
        arg = func(*arg)
        yield arg


def iterreduce(func, iterator, default = None):
    """Lazy iterator version of reduce builtin."""
    #Initialize variables.
    iterator = iter(iterator)

    #Get first value.
    try:
        prev = iterator.next()
    except StopIteration:
        if not default:
            yield default
        raise StopIteration

    #Get second value.
    try:
        next = iterator.next()
    except StopIteration:
        yield func(default, prev)
        raise StopIteration

    #Continue looping.
    for next in iterator:
        prev = func(prev, next)
        yield prev


def iterslice(iterator, slice):
    """Lazy slice of an iterator."""
    #Initialize variables.
    iterator = iter(iterator)

    if isinstance(slice,int):
        stop = slice
    else:
        stop = slice.stop
        start = slice.start or 0
        step = slice.step or 1
    
    #Skip until start if start different from 0.
    if start:
        for i in range(start):
            iterator.next()
    else:
        i = 0

    while i < stop:
        yield iterator.next()
        i += 1
        #Skip step values.
        for i in range(i + 1, i + step):
            iterator.next()
        

#More specialized iterator utilities.
def itergroup(iterator, n):
    """Iterates through iterator, yielding a list of n iterations at a
time."""
    #Initialize variables.
    ret = []

    #Read n elems at a time.
    for elem in iterator:
        if len(ret) == n:
            yield ret
            ret = []
        else:
            ret.append(elem)

    #Yield if something is left.
    if not ret:
        yield ret


def itersum(iterator):
    """Returns the sum of all the elements in the iterator, None if the
iterator has no elements."""
    #Initialize variables.
    iterator = iter(iterator)
    try:
        ret = iterator.next()
    except StopIteration:
        return None
    
    for elem in iterator:
        ret += elem

    #Return result.
    return ret




From maio at stoupa.sh.cvut.cz  Thu Jun 27 10:05:16 2002
From: maio at stoupa.sh.cvut.cz (Marian Schubert)
Date: Thu, 27 Jun 2002 14:05:16 +0000 (UTC)
Subject: [Newbie]  How to manipulate variable _names_ ?
References: <3d1b0723.6179705@news.texas.net>
Message-ID: <slrnahm6qt.6ph.maio@stoupa.sh.cvut.cz>

On Thu, 27 Jun 2002 12:43:01 GMT, Ben Fairbank <baf at texas.net> wrote:
>I have a program that requires the user to enter a file name, such as
>"othello," which the program will then open, process, and create a
>dictionary based on the contents of the file.  I would like to name
>the dictionary "OthelloCount," and then go on and create more
>dictionaries, each with a name made up of the inputfile name as
>provided by the user, concatenated with "Count," but I do not know how
>to make a variable name based on input.

make some temporary dictionary (tmp[dict_name]) fill it with content of file and than you
can use
exec("%sCount = tmp[%s]" % (dict_name, dict_name))

but you should check what is in dict_name (for example check if there are only
[A-z_] chars...) cause it's bit security hole

If it's possible use instead of previsious solution this one:
Count[dict_name] = tmp


cu,
maio


From theller at python.net  Mon Jun 24 15:23:03 2002
From: theller at python.net (Thomas Heller)
Date: Mon, 24 Jun 2002 21:23:03 +0200
Subject: Subclass of array
References: <3d1714c3_7@news.newsgroups.com> <af7kl3$743$0@216.39.172.122>
Message-ID: <af7riq$bdh2q$1@ID-59885.news.dfncis.de>

"Bengt Richter" <bokr at oz.net> wrote in message news:af7kl3$743$0 at 216.39.172.122...
> On Mon, 24 Jun 2002 14:54:21 +0200, JB <jblazi at hotmail.com> wrote:
>
> >Can I derive an own class from array? (I use 2.2.)
>
> Good thought ;-) It doesn't look like it, but if it can be done, I'd like
> to know too. You can make a class that wraps instead of inherits, though.
> There are lots of ways to do that. See one limited way in another reply
> of mine to you in the other thread (which I did because subclassing the
> obvious way didn't work ;-).
>
In 2.2.1, you cannot use derive subclasses of array.
In 2.3 (current CVS version), you can.

Thomas




From spammers.do.not.bother at void.bogus  Fri Jun  7 17:58:26 2002
From: spammers.do.not.bother at void.bogus (Magnus)
Date: Fri, 07 Jun 2002 21:58:26 GMT
Subject: Comparison: HTTP with Java, Perl or Python
References: <3D009911.9000904@gmx.de> <h99M8.11172$p56.3361428@newsb.telia.net> <3D0127D2.2060604@gmx.de>
Message-ID: <60aM8.42170$n4.9585378@newsc.telia.net>

Ingo Linkweiler wrote:

> thanks a lot for your help.
> 
> Do you know SHORTER solutions in Java?


You can try to do some tidy up but all in there is basically needed. Perhaps 
removing the exception handling and expecting things to be okay?

BTW: I am not an educated Java programmer, I just did some programming for 
fun...until I found Python.

Regards,
Magnus



From tdelaney at avaya.com  Wed Jun 19 20:53:45 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 20 Jun 2002 10:53:45 +1000
Subject: python version?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A3A3@natasha.auslabs.avaya.com>

> From: George Hester [mailto:hesterloli at hotmail.com]
> 
> All fixed
> 
[snip questionnaire]

Now I'm confused. What does that mean???

Tim Delaney




From hamish_lawson at yahoo.co.uk  Tue Jun 25 06:42:26 2002
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: 25 Jun 2002 03:42:26 -0700
Subject: ? and %s placeholders, help?
References: <af5oq7$sj1$1@news8.svr.pol.co.uk> <m2k7oolcrm.fsf@video.paradise.lost> <3D178542.604@mxm.dk> <af87cj$n5b$1@newsg3.svr.pol.co.uk>
Message-ID: <915a998f.0206250242.52840ea9@posting.google.com>

Duncan Smith wrote:

> >>> query = 'INSERT INTO %s (%s) VALUES (%s)' % (tblname, string.join(vars,
>  ', '), string.join(['%s']*len(vars), ', '))
> >>> query
>  'INSERT INTO tbl (var1, var2, var3) VALUES (%s, %s, %s)'
> >>> curs.execute(query % ("'a'", "'b'", "'c'"))
> 1L
> 
> #Hurray, but does this avoid parsing the statement on each INSERT?

No, use the two-argument form of the execute method instead:

    curs.execute(query, ("a", "b", "c"))

The execute method will do the interpolation of the values with
appropriate quoting, so it's not necessary to quote them yourself. The
execute method remembers which queries it has seen and only parses a
query the first time it sees it.

Note that Python DB-API drivers differ in which placeholders they
recognise ('%s', '%(fieldname)s', or '?'), usually (I understand)
based on what the native database API itself employs.

See <http://www.python.org/topics/database/DatabaseAPI-2.0.html>.


Hamish Lawson


From r2d2 at acc.umu.se  Thu Jun 27 04:05:46 2002
From: r2d2 at acc.umu.se (Niklas Frykholm)
Date: Thu, 27 Jun 2002 08:05:46 +0000 (UTC)
Subject: Help with an algorithm wanted
References: <rowen-CE3C0F.13020226062002@nntp1.u.washington.edu>
Message-ID: <slrnahlhqq.l3i.r2d2@shaka.acc.umu.se>

[Russell E. Owen]:
> I'd like some help solving a problem in Python.
> 
> The basic idea is I have an item of data that can be represented in any 
> of four ways: A,B,C and D. I also have conversion functions between 
> certain pairs of representations, such that there's always a path 
> between any two data representations. In this case suppose I know A<->C, 
> B<->C and C<->D, or graphically:
> 
> A <--+
>      +--> C <---> D
> B <--+
> 
> I'd like to create objects that have one of these items (A,B,C,D) as a 
> given, then be able to query any such object for any data representation.

Here is one possible solution. Given a set of classes A, B, C... and
known conversion functions A.to_B, B.to_C, etc this code creates all
derived conversion functions (such as A.to_C). It thus solves the problem
statically. (In some cases a dynamic solution is preferrable.)

	def add_conversion(fromm, via, to):
		"Add conversion from class fromm to class to, via class via."
		def convert(self):
			via_object = getattr(self, "to_" + via)()
			return getattr(via_object, "to_" + to)()
		setattr(globals()[fromm], "to_" + to, convert)

	def has_conversion(fromm, to):
		"Is there a conversion function from class fromm to class to?"
		return hasattr(globals()[fromm], "to_" + to)

	def conversions(klass):
		"Returns all the conversion functions of klass."
		l = filter(lambda s: s.startswith('to_'), dir(globals()[klass]))
		return map(lambda s: s[3:len(s)], l)

	def create_quick_conversions(classes):
		"Create all possible derived conversions for the set of classes."
		for via in classes:
			tos = conversions(via)
			for fromm in classes:
				if fromm != via and has_conversion(fromm, via):
					for to in tos:
						if not has_conversion(fromm, to):
							add_conversion(fromm, via, to)
		# end of for

	create_quick_conversions(["A", "B", "C", "D"])

// Niklas


From syver-en+usenet at online.no  Tue Jun  4 13:31:22 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Tue, 04 Jun 2002 17:31:22 GMT
Subject: manipulating child processes under Windows
References: <lcptz7zw0m.fsf@gaffa.mit.edu>
Message-ID: <uadqb3q9v.fsf@online.no>

Douglas Alan <nessus at mit.edu> writes:
> 
> So, I have two questions:
> 
> (1) Is this thing where a stub process fires off the real process a
>     typical thing to do under Windows?  If so, it seems to make
>     synchronizing processes difficult.

Yes, many gui processes do this, it's a way to send cmd line arguments
to a long running process, f.ex explorer which work in this way.

> 
> (2) What's the best way for me to solve the problem.  I.e., what's the
> 
>     best way to get my Python script to wait for the real program to
>     terminate?
> 
> |>oug

Hmm, my first reaction is to think in terms of window handles in some
way, there's a FindWindow function in win32 (and win32all) that let's
you search for a window by name, I'll have to check the msdn library to find
out how to get the owning process handle from a window handle.

.......

This seems to be it:
GetWindowThreadProcessId

It takes a window handle (HWND) as input and returns the thread id and
the process id. Now you can use the process id to get a handle to the
process and wait on that.

Does that help? If you don't already have it get dynwin/calldll so
that you can call api functions that aren't covered by win32all, I
don't seem to find GetWindowThreadProcessId in win32all.

-- 

Vennlig hilsen 

Syver Enstad


From shalehperry at attbi.com  Thu Jun 20 13:42:56 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 20 Jun 2002 10:42:56 -0700 (PDT)
Subject: do...while
In-Reply-To: <slrnah40l1.12v.msoulier@pmerd071.ca.nortel.com>
Message-ID: <XFMail.20020620104256.shalehperry@attbi.com>

On 20-Jun-2002 Michael P. Soulier wrote:
>     Greetings.
> 
>     How do you people handle the lack of a do...while in Python? I find that
> at times, I must initialize variables for a loop with identical code to the
> loop, which feels like a waste to me. 
> 
> ie.
> 
>     line = filehandle.readline()
>     while len(line) > 5:
>         line = filehandle.readline()
> 
>     A do...while here would be nice. 
> 
>     Just curious. 
> 

while 1:
    line = filehandle.readline()
    if len(line) <= 5: break
    ....
    ....

is a typical way to implement it.




From schitz0boy at hotmail.com  Sat Jun  1 21:40:28 2002
From: schitz0boy at hotmail.com (Corey Woodworth)
Date: 1 Jun 2002 18:40:28 -0700
Subject: Does this widget exist?
Message-ID: <24d8090a.0206011740.49d253e3@posting.google.com>

Is there a Tk widget that lists files like you see in napster or
winMX. (you know with the kb/sec and frequency buttons up top to sort
things.)

If I use the search feature in windows and I select the details view I
get a widget very similar (except of course windows isn't written in
python) :)

Corey


From missive at frontiernet.net  Wed Jun 19 18:45:33 2002
From: missive at frontiernet.net (Lee Harr)
Date: Wed, 19 Jun 2002 22:45:33 -0000
Subject: isometric engine
References: <uh1nfitkhs2s36@corp.supernews.com>
Message-ID: <uh22cdp96m2d9f@corp.supernews.com>

In article <uh1nfitkhs2s36 at corp.supernews.com>, Raven wrote:
> Hi, I was wondering if there were any open source iso engines for python. I
> found the old tkclient for worldforge, but its page is gone now and I havent
> found the source anywhere and I really didn't want to have to write one up
> from scratch if their was already one available. Any help wouold be greatly
> appreciated. :)
> 

Take a look at py-game

http://pygame.org

Look on the projects page. I am pretty sure there is a library
there working towards something like this.



From akakakhel at attbi.com  Wed Jun  5 19:18:40 2002
From: akakakhel at attbi.com (Ali K)
Date: Wed, 05 Jun 2002 23:18:40 GMT
Subject: Internat
References: <69QJ8.68711$ux5.82517@rwcrnsc51.ops.asp.att.net> <ad9tkq$g2n$1@agate.berkeley.edu>
Message-ID: <k%wL8.118822$ux5.177487@rwcrnsc51.ops.asp.att.net>

How do I instal the nessecary files?




From huaiyu at gauss.almadan.ibm.com  Fri Jun 21 21:50:14 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Sat, 22 Jun 2002 01:50:14 +0000 (UTC)
Subject: How to represent the infinite ?
References: <mailman.1024698451.7353.python-list@python.org>
Message-ID: <slrnah7lum.27i.huaiyu@gauss.almadan.ibm.com>

Each time we are having this conversation we go around the same circle once
or twice.  Let's see if we can find a way out.  I would like to divide the
issue into three (still somewhat arbitrary) sub-issues, and see if any one
of them can be resolved in easier/simpler/clearer ways:

1. How much 754 support is provided by the underlying platforms.
2. How do such support vary across the platforms.
3. What are the Python designers' wishes.

Tim Peters <tim.one at comcast.net> wrote:
>
>It has nothing to do with Windows specifically; Python's 754 behavior "is
>random" on all platforms, in platform-specific ways, because Python is
>implemented in C, platform C 754 behavior "is (still) random", and the only
>code ever contributed to address this sorry state of affairs is the fpectl
>module (whose purpose is in a direction opposite the one you favor).

I think you are talking about sub-issue 2 here.  Is there some place we can
find out what these variations are, and how severe they are?

I can certainly contribute "the patch to enable inf and nan in Python" :-)
(attached), but I'm sure that's not the reason you are against it.

In any case I'd like to see what problems people report on different
platforms.


>> Since it could be made to work on Linux, I simply commented out
>> the checking in pyport.h.  Voila, instant ieee compliance!
>
>You happened to get one particular 754 behavior you like; full IEEE
>compliance is a large and difficult task; subset compliance is worse than
>none (if we're to believe Kahan's virulent rants about Java <wink>).

I think you are talking about sub-issue 1 here.  Do you actually believe
that subset-compliance is worse than none?  I'd think that would depend on
what subset we are talking about.  Certainly "can't provide full support"
does not imply "avoid as much as possible".  There must be some concrete
criteria that are omitted here.


>> Your example appears to show that Windows can also live with nan and inf.
>
>Virtually all platforms can, given enough platform-specific support code,
>almost none of which Python has.
>
>> So my question is, are there any significant platforms out there
>> that still really need such checkings which make nan and inf almost
>> useless?
>
>If Python were 100% 754-conforming, Guido and I would still favor enabling
>exceptions on overflow, divide-by-zero, and invalid operation, by default.
>As things are, you've got no choice at all, unless you want such exceptions
>enabled and are on a platform where the fpectl module happens to work.

I think you are talking about sub-issue 3 here.  Could you provide a
rationale in the context of numerical computation?

I can offer an example in support of the opposite side.  Suppose x is a
vector containing elements that might be arbitrarily large, then

	   exp(-x**2)

might involve inf at certain elements, yet the result is perfectly regular,
containing only floats between 0 and 1.  How can this be done by vectorized
computation if individual elements may throw exceptions?


Certainly you and Guido have the final say on sub-issue 3.  On the other two
issues, however, I suspect that

1. Most platforms these days support 754 to a remarkably good degree, far
   better than that is allowed in today's Python.

2. Most number crunchers would be more concerned that it works as well as
   possible on the platforms available to them, rather than whether it also
   works on other platforms.

3. The variations across major platforms are minor.


I'd like to hear people confirm or deny these suspicions.  If answering
these questions is easier than answering the "should or shouldn't" question,
then the state of affairs could be advanced a little bit by considering
these questions first.


Huaiyu


PS. Here's the patch:


This patch removes the code that generate an OverflowError for 1e200**2.
It is generated  with:
diff -rcN Python-2.2.1 Python-2.2.1.ieee > python-2.1.1-ieee-fpu.patch

It can be applied as
patch -p1 -d Python-2.2.1 < python-2.1.1-ieee-fpu.patch

Afterwards, it is still necesary to modify Makefile to add -lieee to LIBS.
This itself can be changed by using autoconf, but I've forgotten how to do
it.  The second part of the patch does that.


diff -rcN Python-2.2.1/Include/pyport.h Python-2.2.1.ieee/Include/pyport.h
*** Python-2.2.1/Include/pyport.h	Mon Mar 11 02:16:23 2002
--- Python-2.2.1.ieee/Include/pyport.h	Wed May  8 15:20:30 2002
***************
*** 301,311 ****
   */
  #define Py_ADJUST_ERANGE1(X)						\
  	do {								\
! 		if (errno == 0) {					\
  			if ((X) == Py_HUGE_VAL || (X) == -Py_HUGE_VAL)	\
  				errno = ERANGE;				\
  		}							\
! 		else if (errno == ERANGE && (X) == 0.0)			\
  			errno = 0;					\
  	} while(0)
  
--- 301,311 ----
   */
  #define Py_ADJUST_ERANGE1(X)						\
  	do {								\
! 		/*if (errno == 0) {					\
  			if ((X) == Py_HUGE_VAL || (X) == -Py_HUGE_VAL)	\
  				errno = ERANGE;				\
  		}							\
! 		else */ if (errno == ERANGE && (X) == 0.0)			\
  			errno = 0;					\
  	} while(0)
  
***************
*** 313,320 ****
  	do {								\
  		if ((X) == Py_HUGE_VAL || (X) == -Py_HUGE_VAL ||	\
  		    (Y) == Py_HUGE_VAL || (Y) == -Py_HUGE_VAL) {	\
! 				if (errno == 0)				\
! 					errno = ERANGE;			\
  		}							\
  		else if (errno == ERANGE)				\
  			errno = 0;					\
--- 313,320 ----
  	do {								\
  		if ((X) == Py_HUGE_VAL || (X) == -Py_HUGE_VAL ||	\
  		    (Y) == Py_HUGE_VAL || (Y) == -Py_HUGE_VAL) {	\
! 			/*	if (errno == 0)				\
! 				errno = ERANGE; */			\
  		}							\
  		else if (errno == ERANGE)				\
  			errno = 0;					\
diff -rcN Python-2.2.1/configure.in Python-2.2.1.ieee/configure.in
*** Python-2.2.1/configure.in	Mon Mar 11 02:14:23 2002
--- Python-2.2.1.ieee/configure.in	Wed May  8 15:47:07 2002
***************
*** 1777,1786 ****
  # (none yet)
  
  # Linux requires this for correct f.p. operations
! AC_CHECK_FUNC(__fpu_control,
!   [],
!   [AC_CHECK_LIB(ieee, __fpu_control)
! ])
  
  # Check for --with-fpectl
  AC_MSG_CHECKING(for --with-fpectl)
--- 1777,1787 ----
  # (none yet)
  
  # Linux requires this for correct f.p. operations
! #AC_CHECK_FUNC(__fpu_control,
! #  [],
! #  [AC_CHECK_LIB(ieee, __fpu_control)
! #])
! AC_CHECK_LIB(ieee, __fpu_control)
  
  # Check for --with-fpectl
  AC_MSG_CHECKING(for --with-fpectl)
diff -rcN Python-2.2.1/pyconfig.h.in Python-2.2.1.ieee/pyconfig.h.in
*** Python-2.2.1/pyconfig.h.in	Wed Oct 24 10:10:49 2001
--- Python-2.2.1.ieee/pyconfig.h.in	Wed May  8 15:53:51 2002
***************
*** 748,754 ****
  #undef HAVE_LIBDLD
  
  /* Define if you have the ieee library (-lieee).  */
! #undef HAVE_LIBIEEE
  
  #ifdef __CYGWIN__
  #ifdef USE_DL_IMPORT
--- 748,754 ----
  #undef HAVE_LIBDLD
  
  /* Define if you have the ieee library (-lieee).  */
! #define HAVE_LIBIEEE
  
  #ifdef __CYGWIN__
  #ifdef USE_DL_IMPORT


From anoop79 at myrealbox.com  Thu Jun 13 03:42:24 2002
From: anoop79 at myrealbox.com (Anoop P B)
Date: Thu, 13 Jun 2002 13:12:24 +0530
Subject: How to call a function using apply with keyword args?
References: <ae9fud$u23$00$1@news.t-online.com>
Message-ID: <3D084CE0.4000505@myrealbox.com>

hi Achim,

try this:

def MyFkt(A=0,B=0,C=0): pass
params = {'A':5,'B':8,'C':9}
MyFkt(params)

the reason behind the error is coz u can only pass a dictionary using 
'apply' for keyword arguments. and keyword args are of the form 
keyword=value (see docs)

~anoop

Achim Domma wrote:
> Hi,
> 
> I have a function like this:
> 
> def MyFkt(A,B,C): pass
> 
> and some parameters from a config file:
> 
> params = {'A':5,'B':8,'C':9}
> 
> how can I call MyFkt with this parameters? I tried apply(MyFkt,[],params)
> but get this error:
> 
> TypeError: MyFkt() takes exactly 3 non-keyword arguments (0 given)
> 
> any idea?
> 
> greetings
> Achim
> 
> 






From bokr at oz.net  Sat Jun 15 11:25:10 2002
From: bokr at oz.net (Bengt Richter)
Date: 15 Jun 2002 15:25:10 GMT
Subject: filter2
References: <mailman.1023975618.6474.python-list@python.org> <aeaava$56l8p$1@ID-59885.news.dfncis.de> <mailman.1023979895.13512.python-list@python.org> <slrnagllbc.5mj.quinn@groat.ugcs.caltech.edu>
Message-ID: <aefm8m$3ng$0@216.39.172.122>

On 15 Jun 2002 05:49:33 GMT, quinn at groat.ugcs.caltech.edu (Quinn Dunkan) wrote:

>On Thu, 13 Jun 2002 18:50:23 +0400, Oleg Broytmann <phd at phd.pp.ru> wrote:
>>On Thu, Jun 13, 2002 at 04:39:51PM +0200, Thomas Heller wrote:
>>> 
>>> "Oleg Broytmann" <phd at phd.pp.ru> wrote in message
>news:mailman.1023975618.6474.python-list at python.org...
>>> >    I want to have new python builtin - filter2. It is exactly like filter,
>>> > but it returns also a list of items that didn't pass the test.
>>> 
>>> IIRC, in Smalltalk they were named 'select' (like filter), and 'reject' (like your filter2).
>>> IMO these are much better names then filter and filter2.
>>
>>   It is too late to rename builtin filter, so I modelled fiter2 after it.
>
>In haskell they call it "partition" which I think is better since the
>emphasis is that it returns both selected and rejeted.
>
>I wrote mine in python with a plain for loop.  I'm not sure C would really get
>you that much efficiency, but it would be easy enough to find out.
>
>Not sure it's worth being in the core (maybe there should be a module for
>list utils or something), but I like partition better than filter2.

I wonder how this version whould do:
 >>> def partition(test, alist):
 ...     l2=[]
 ...     return filter(lambda x: test(x) or l2.append(x), alist), l2
 ...
 >>> partition(lambda x: x%3==0, range(16))
 ([0, 3, 6, 9, 12, 15], [1, 2, 4, 5, 7, 8, 10, 11, 13, 14])


I imagine it's faster than my other one (clipped from other post)

 >>> def filter3(test, alist): # not to use keyword
 ...     l2=[]
 ...     return [x for x in alist if test(x) or l2.append(x)], l2
 ...
 >>> print filter3(lambda x: x%3==0,range(16))
 ([0, 3, 6, 9, 12, 15], [1, 2, 4, 5, 7, 8, 10, 11, 13, 14])

Regards,
Bengt Richter


From marklists at mceahern.com  Tue Jun 11 16:12:00 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 11 Jun 2002 15:12:00 -0500
Subject: using m2crypto to encrypt credit card numbers
In-Reply-To: <1023824489.22503.11.camel@walter>
Message-ID: <NCBBLFCOHHDIKCAFGCFBGEBOLEAA.marklists@mceahern.com>

[Graham Ashton]
> I've not followed the thread, but I'm assuming that your problem is that
> you don't want to store the card number in plain text, and that if you
> encrypt and store it in a manner that will allow automatic decryption by
> your software, then a cracker who gains access to your servers will be
> able to decrypt the card numbers with ease.

That's it precisely.

> The only sensible solution to this that I've ever thought of involves
> getting the payment processor (i.e. online transaction processing
> company) to store a hash for each of your customers' credit cards.

I should have mentioned that solutions which involve getting the payment
processor to do anything different are simply not an option.  As far as I
can tell, I either store the credit card number or I can't do recurring
billing.  If you know of any payment processors that support recurring
billing, please share them.

> Consequently you wouldn't need to store the card number at all, just the
> encrypted hash. Job's a good'un; card numbers would be nicked from far
> fewer online web sites with shoddy security.

Also, another reason I need to store the credit card number is in the case
of chargebacks, which don't go through the payment processor--rather, they
go through the bank.  I don't fully understand this part, but I do believe I
need the credit card number in order to link the chargeback to a
transaction.

Thanks for your comments,

// mark

-





From dm_top at hotmail.com  Sat Jun 22 19:01:27 2002
From: dm_top at hotmail.com (Dmitri)
Date: 22 Jun 2002 16:01:27 -0700
Subject: Python+Cygwin+XML SAX build failure of PyXML 0.7.1
References: <59d8a5f6.0206210804.25353aa3@posting.google.com> <mailman.1024711848.1696.python-list@python.org>
Message-ID: <59d8a5f6.0206221501.232d4b3b@posting.google.com>

Jason ,

> 
> Would you like to contribute a pre-built Cygwin PyXML to the Cygwin
> distribution?
> 

That's possible, as long as I will be satisfied by the results.

Dmitri


From brueckd at tbye.com  Wed Jun  5 13:36:31 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 5 Jun 2002 10:36:31 -0700 (PDT)
Subject: urllib, urllib2, httplib -- Begging for consolidation?
In-Reply-To: <Pine.SOL.4.30.0206051524590.16156-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <Pine.LNX.4.44.0206050921510.6788-100000@ns0.tbye.com>

On Wed, 5 Jun 2002, John J. Lee wrote:

> I'm trying to understand your posts on this,

Hi John,

Before I jump in I do want to emphasize that I'm not trying to be too 
critical of httplib/urllib/urllib2 - the OP asked if they should be 
consolidated, and my short response is "yes, and reorganized too". :-)

Anyway, my concerns live on two levels:

1) It's very non-obvious, especially to newcomers, which library should be
used when. A newbie will probably find httplib without too much
difficulty, but if what they need isn't there, why would they look in
urllib? (not to mention urllib2!) To me this is annoying and the cause of
some wheel-reinventing when they don't know about urllib, but not too
great of a problem - it's a naming problem, it's just a matter of
learning, and it helps keep the Python book authors in business. ;-)

2) As richer HTTP functionality is added to the standard library, it is 
getting added to the wrong place, such that it is only narrowly 
accessible. This is my main concern, which I'll try to elaborate on a 
little more.

urllib gives you a very generic facility to retrieve an object, and that
should not change too much: if you have no special needs then it works
great. As your needs require you to descend to lower level APIs (e.g. for
custom functionality), that transition should as much as possible be (1)  
straightforward, (2) obvious, and (3) one that allows you to reuse as much
of the related knowledge and work. My assertion is that right now we don't
have those things.

Again, speaking ideally, facilities for the retrieval of objects could be 
organized into a hierarchy ranging from very high level to low level, 
e.g. (many branches and leaves removed):

urllib
 |
httplib ftplib gopherlib
 |
 ... (possibly several layers here)
 |
socket

So, depending on your specific needs, you'd "plug in" to this hierarchy at
the appropriate level. If all you need is to go fetch some object, you'd
probably go with urllib. If you know that you're using HTTP, you'd
probably use the highest level HTTP layer as your starting point. If you
were building a custom client that worked over a mythical HTTP/1.5 you'd
start even lower, on down until you're handling the sockets themselves.  
The key though is that each level you'd have to reinvent as little as
possible - you'd build on related work from lower levels.

Obviously getting such a hierachy to work right would be tough and not as 
clean (there's not a strict parent-child relationship between all degrees 
of custom features so there'd probably be lots of mix-in class 
functionality instead of enforced layering). The approach we have today is 
*sort of* like this, except that richer and smarter functionality is being 
added at the *top* of the hierarchy. This has two negative side effects: 
(1) it pollutes the higher level, generic interfaces and (2) it makes it 
difficult to reuse functionality if your needs force you to start out at a 
lower level.

Problem #1 is what makes me throw my hands up in frustration when we talk
about, e.g., expanding the urllib APIs to have some way to do a HEAD
request: it doesn't belong on that level of functionality. That generic
interface is for making it trivially easy to fetch the contents associated
with a URL, independent as much as possible from protocol. By crufting it
up in an attempt to make it all things to all people, it is no longer a
clean, generic API. For this specific example, a HEAD request doesn't make
sense in terms of FTP, so you shouldn't be able to do a HEAD request using
the generic API. (an alternative would be to redesign the API to include
GetSize functionality, that would then take appropriate protocol-specific
action under the covers, but the per-protocol smarts ("how do I get the
size of an FTP object?" would certainly not live inside the generic API
module) Again in the example of a HEAD request, in order to do anything
useful with the results, your code would have to realize that it was in
fact an HTTP HEAD response, so being able to call a protocol-agnostic API
to get those results doesn't add much value since the calling code has to 
be aware of the protocol anyway.

Previously in this thread I've focused my attention on problem #2 (reuse
of functionality), as that's the one that I've run into the most. As it
stands today, you benefit from all sorts of nifty HTTP behavior, but only
if what you need to do works through the urllib APIs. If specific needs
force you to descend that protocol/functionaliy tree (or if you simply
know in advance that you'll be dealing with a specific protocol), you
lose, and have to reimplement it all yourself. From a user's point of
view, this just doesn't make sense. Again using the most recent example of
an HTTP HEAD request, it seems odd that the standard library (urllib)  
happily follows redirects, but if I simply want to do a HEAD request (I
simply need slightly more control over what's happening) I have to
implement that functionality myself.

> Perhaps part of your point is simply that the division into modules should
> better reflect the conceptual structure?  The current organisation may (or
> may not) have been a mistake, but it's not a significant enough mistake to
> warrant reorganisation, is it?

Well, like I said before, this was all in response to the query "should
they be consolidated?", so IF we were to invest the time to consolidate,
THEN we should also reorganize. Also, it seems (this is very much my own
interpretation of course) that as these modules have evolved, the model
around which they are built has become fuzzy, or that they have outgrown
their original model altogether and the current implementation is showing
signs of stress. For example, the very existance of a module whose name is
"urllib2" says there's no longer a clear answer to the question "where
should new feature Y or class of features X live?". So, maybe nearly all
of what's there could be reused as is, but the decision of where to put
what and why would be driven by a clear and cleaner big picture (I've
never advocated throwing it all away, for example, as there's lots of
really good stuff in there).

Probably 90% of the problem is naming and/or module organization - urllib, 
a module to retrieve the contents given a generic URL, should be just that 
and little else, and protocol-specific knowledge should be accessible from 
protocol-specific modules.

Have fun,
-Dave





From kzaragoza at attbi.com  Sat Jun 15 15:56:02 2002
From: kzaragoza at attbi.com (Kris J. Zaragoza)
Date: 15 Jun 2002 14:56:02 -0500
Subject: Python and SQL Server 2000
References: <ZCuO8.10006$9b.678135@typhoon.austin.rr.com> <mailman.1024112300.2184.python-list@python.org>
Message-ID: <slrnagn75g.263.kzaragoza@localhost.localdomain>

In article <mailman.1024112300.2184.python-list at python.org>, Alex Polite wrote:
> I tried to do this a couple of years ago. I think MS SQL Server was
> version 6.x 
> 
> We used ADO connections via the win32com module. It worked but it was
> kind of sluggish. Then to speed things up I wrote a wrapper around bpc
> (bulk something something, look it up in the MS SQL manual). I think
> we couldn't use mxODBC due to licensing issues. Then in
> the end we moved to MySql.

My first instinct on seeing this reply was to point out that win32com
doesn't run on Red Hat Linux. :-) Then I got to thinking a bit more
about it and realized this may actually be possible.

So, let me add to the original poster's question: Has anyone ever
tried running the Windows version of Python (either the main
distribution or ActiveState's) under Wine to take advantage of the
database drivers, etc.?

Kris


-- 
Kris J. Zaragoza       | On the face of it, Microsoft complaining about
kzaragoza at attbi.com    | the source license used by Linux is like the
                       | event horizon calling the kettle black.
                       | -- Adam Barr, article on kuro5hin.org


From i.linkweiler at gmx.de  Fri Jun  7 17:38:26 2002
From: i.linkweiler at gmx.de (Ingo Linkweiler)
Date: Fri, 07 Jun 2002 23:38:26 +0200
Subject: Comparison: HTTP with Java, Perl or Python
References: <3D009911.9000904@gmx.de> <h99M8.11172$p56.3361428@newsb.telia.net>
Message-ID: <3D0127D2.2060604@gmx.de>

thanks a lot for your help.

Do you know SHORTER solutions in Java?



From logiplexsoftware at earthlink.net  Fri Jun 28 16:13:05 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 28 Jun 2002 13:13:05 -0700
Subject: generate all combinations of a list (with variable output length)
In-Reply-To: <3D1CBE6D.30002@huno.net>
References: <3D1CBE6D.30002@huno.net>
Message-ID: <20020628131305.318101b4.logiplexsoftware@earthlink.net>

On Fri, 28 Jun 2002 21:52:13 +0200
thomas wrote:

> hi,
> 
> i'm looking for a way to generate all possible combinations of all items 
> in a list. list = ["A","D","$","5","R"]
> 
> but i also want it to use one item several times or not at all and in no 
> particular order. the length of the string is 1-10. so a possible output 
> would be:
> 
> DD$
> A5RD$D555
> A
> 5RR$$A
> DRRR55$5$A
> ...

How about:

import random

l = ["A","D","$","5","R"]
randomList = [l[random.randrange(0, len(l))] for i in range(random.randrange(1,
len(l) + 1))]

This assumes you want at least one item in the list.  Also, please note that
using "list" as a variable name is bad as it is the name of a built-in type.

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From eddie at holyrood.ed.ac.uk  Fri Jun 14 14:39:16 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Fri, 14 Jun 2002 18:39:16 +0000 (UTC)
Subject: a possibly foolish question about slices
References: <yu99sn3prbqm.fsf@europa.research.att.com>
Message-ID: <aedd8k$brf$1@scotsman.ed.ac.uk>

Andrew Koenig <ark at research.att.com> writes:

>Suppose x is a string or list, and I want to refer to
>a subsequence of x with length n starting at position p.
>under the assumption that this subsequence is entirely

>To do so, I can use x[p:p+n] in all circumstances but one,
>namely when p < 0 and p == -n.  Is there a similarly concise
>expression that works even in this special case?

It frequently bugs me that you can't supply any number for b in <string>[a:b]
to indicate spanning to the end of the string.  In other words I'd love to
hear of a solution because I can't think of a way to do it either EXCEPT to
stop using the $%&^%$@@ zero based indexing and go back to using 1 as the
first item in a list (etc.) as nature intended (this is the way Icon works and
I always found it the nicest and most intuitive and frankly just plain
correct!) this way 0 is reserved for the last item (ie -2 -1 0 is a logical
progression).

So if the designers could fix this for version 2.3 I think we'll all be happy
:)

Actually, while you're at it why not also use Icon's <string>[x:+len] syntax
for the special case mentioned above (I've often wondered about that but can't
be bothered searching for previous discussions).

Eddie (the zeroeth amongst equals).


From BPettersen at NAREX.com  Sat Jun 29 15:06:01 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Sat, 29 Jun 2002 13:06:01 -0600
Subject: Most efficient way to write data out to a text file?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E201922151B9@admin56.narex.com>

> From: Dave Kuhlman [mailto:dkuhlman at rexx.com] 
> 
> candiazoo at attbi.com wrote:
> 
> [snip]
> 
> Thank you for posting this.  Very helpful.
> 
> Could we have a bit of evaluation and explanation please.  I'm 
> suspicious about whether some of these changes actually made a 
> difference.  I'd like a little understanding before I start 
> changing my Python coding style.

[snip optimization involving removing free vars, and string concat]

Looking up a name in globals or builtin is very expensive compared to
looking it up in local scope (use the dis module to see the opcodes
involved -- anything that has LOAD_GLOBAL in it is potentially slow).
Most experienced Python programmers will therefore use something like:

  def foo(arg1, arg2, func=MyModule.func):
      ...

after they have profiled the program and found foo to be a bottleneck.
Similarly, looking up an attribute of an object involves searching its
__dict__. If you save the returned value in a local, you basically get a
fast "load from array" operation instead.

When it comes to string concatenation it is an O(n**2) operation, so
it's usually better to do:

  def foo():
      result = []
      append = result.append
      for i in xrange(100000):
          append(`i`)	# `i` is faster than repr(i)
      return ''.join(result)

I would say the string concat optimization is always a good idea, the
global->local optimization is ok as long as it doesn't interfer with
readability -- other people will surely disagree <wink>.

-- bjorn




From imbosol at vt.edu  Fri Jun 21 16:57:10 2002
From: imbosol at vt.edu (Carl Banks)
Date: Fri, 21 Jun 2002 16:57:10 -0400
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com> <lkznxqjqx2.fsf@pc150.maths.bris.ac.uk> <3D11A21D.9010809@mxm.dk> <uh3e1usajle42c@news.supernews.com> <3D11DA9B.163A7BD0@info.unicaen.fr> <uh446elev7af68@news.supernews.com> <3D12EEAA.9C3A3D29@info.unicaen.fr>
Message-ID: <6v30fa.sm1.ln@grey.aerojockey.localdomain>

Jerzy Karczmarczuk wrote:
> (Oh, I adored that wonderful piece of programming madness, an "article"
> where somebody proposed a new super-Fortran where GO TO was replaced by
> CAME FROM statement, and the language had an inspiring non-deterministic
> conditional:   If <<condition>> THEN WHAT?.
> Any references, anybody?)


http://www.fortran.com/fortran/come_from.html


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From mickey at tm.informatik.uni-frankfurt.de  Fri Jun 14 04:07:27 2002
From: mickey at tm.informatik.uni-frankfurt.de (Michael 'Mickey' Lauer)
Date: 14 Jun 2002 10:07:27 +0200
Subject: How can Python bindings for the Qt toolkit?
References: <aec5fa$e3u$1@bowmore.utu.fi>
Message-ID: <3d09a43f@nntp.server.uni-frankfurt.de>

zhao <wangzhao_wz at hotmail.com> wrote:
> Hi,
> I don't understand how the Sip create Python bindings for C++ libraries. How
> does Sip Python module provides support functions to the automatically
> generated code? I have installed Sip, Python2.2.1 and Qt3.0.4 in Windows.
> But if I write python code, how can it be changed to C++ code so that I can
> use it under Qt software. 

Sounds you need a bit more understanding in what a Python language binding does.
A Python language binding gives you access to functions/methods/constants which
reside in a library which is written in a certain language (not Python) by the
means of Python functions/methods/constants... So: The language binding exposes
a Python interface to the library you want to use.

> Can anybody give me the explanation in details and
> how to install it? 

You really should read the instructions which come with PyQt.

> And when I install PyQt and run pyuic.exe, it always says
> that 'the dynamic link library qt-mt230nc.dll could not be found in the
> specified path'. Why?

This sounds like the typical Qt2.3.0-NonCommercial.exe which forgets to copy
the dll to <windows>/system32. Do this manually.

Hope this helps somehow.

:M:



From Taekyon at macchinetta.miadimora  Mon Jun  3 13:37:03 2002
From: Taekyon at macchinetta.miadimora (Tae kyon)
Date: Mon, 03 Jun 2002 17:37:03 GMT
Subject: solarwolf display ?
References: <slrnafl9gv.5h4.Taekyon@macchinetta.miadimora> <3CFB8ED2.2070507@attbi.com>
Message-ID: <slrnafna1k.2de.Taekyon@macchinetta.miadimora>

Nell'articolo <3CFB8ED2.2070507 at attbi.com>, Pete Shinners ha scritto:

> i know it's not documented, but try running solarwolf with the "-window" 
> flag and it will run in a window. strange that your games will run at 1024x 
> but no 800x. Pygame uses SDL, and if SDL cannot change to the desired video 
> mode it will pick a bigger one and 'letterbox' the display.

Thanks a lot ... and sorry if the question was a little OT on this NG.

-- 
Tae Kyon


From altis at semi-retired.com  Sat Jun 15 01:46:03 2002
From: altis at semi-retired.com (Kevin Altis)
Date: Fri, 14 Jun 2002 22:46:03 -0700
Subject: Using Python to script a game?
References: <cdc2589b.0206141718.7dc2ae13@posting.google.com> <mailman.1024106376.21850.python-list@python.org>
Message-ID: <RtAO8.130$tI.120587@news.uswest.net>

Freedom Force is the game you're probably thinking of that uses Python as
the scripting engine.

I blogged about it here:

http://radio.weblogs.com/0102677/2002/04/27.html#a55

ka

"Gerhard H?ring" <gerhard at bigfoot.de> wrote in message
news:mailman.1024106376.21850.python-list at python.org...
> * Ben Blonder <ben at sourcecod.com> [2002-06-14 18:18 -0700]:
> > I'm writing a 3D naval tactics in C++ for the Mac. I'm planning on
> > having all the object behaviors, movements, and AI controlled through
> > a scripting layer. The scripts would be text files the game would
> > parse and then carry out actions. All the scripts would be read at
> > runtime, not compile time, so they'd all have to be interpreted.
> >
> > For example, I might have the C++ code
> >
> > class foo {
> > public:
> > int x;
> > void print() {
> > cout << x << endl;
> > }
> > };
> >
> > In the scripting system, there would have to be some way to create a
> > new foo object, define x, and then call print() while the user was
> > playing.
> >
> > Can I use Python to accomplish this,
>
> Yes. And it would be a pretty good match, too. There's even been a
> commercial game that uses Python for internal scripting, its name
> escapes me, though.
>
> > and if so, can you point me to some tutorials on integratring Python
> > with my existing code?
>
> Python's documentation has (now) good documentation about it's C API.
> Python's API is _C_, though, not C++. There are several C++ wrappers for
> it than can simplify the integration a lot. My personal favourite is
> Boost::Python. http://www.boost.org/ As a C++ developer, you might be
> interested in the other free libraries from Boost, too. What these
> people accomplish is really amazing!
>
> Gerhard
> --
> mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
> web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
> public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
> reduce(lambda x,y:x+y,map(lambda
x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))
>
>




From uwe at rocksport.de  Fri Jun  7 05:23:17 2002
From: uwe at rocksport.de (Uwe Schmitt)
Date: 7 Jun 2002 09:23:17 GMT
Subject: Distributing Packages / os independent installation folder
References: <adpsrg$m02$03$1@news.t-online.com>
Message-ID: <adpu25$e29eb$2@hades.rz.uni-sb.de>

Achim Domma <achim.domma at syynx.de> wrote:
| Hi,

| - where should my package and shared librarie go to? On windows I put the
| dll into python22/DLLs and the rest into a folder
| python22/Lib/site-packages/ImageMagick. What to do on other plattforms? I
| don't expect a DLLs folder on Linux. ;-)

import sys
echo sys.path

... this might help you.

| - In my __init__.py I have to change one environment variable. Currently I
| do it like this:

| import os
| os.environ['MAGICK_HOME'] = r'C:\Python22\Lib\site-packages\ImageMagick'
| from _ImageMagick import *

| Is there an plattform independent way to get the installation folder of my
| package? On windows I can get the python folder from registry, but this will
| also break on linux.

I dont think so... If you use distutils, you have to know where your
Python distribution lives...

Greetings, Uwe


-- 
Dr. rer. nat. Uwe Schmitt      ICQ# 159647634   Uwe.Schmitt at num.uni-sb.de


From peter at engcorp.com  Sun Jun 30 00:33:12 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 30 Jun 2002 00:33:12 -0400
Subject: newbie Q: delete list few lines in file
References: <7fe97cc4.0206291519.1c95ce31@posting.google.com> <3D1E5123.FDEFADDF@engcorp.com> <3D1E83C0.8060700@earthlink.net>
Message-ID: <3D1E8A08.53DD2272@engcorp.com>

Hans Nowak wrote:
> 
> Peter Hansen wrote:
> 
> >>* how to write to file inplace? (or, do i write to a new file, then
> >>delete original, and rename the new file?)
> >
> > You cannot write files in place.
> 
> <nitpick>
> 
> Sure you can:

[snip proof]

Of course, but we were answering a newbie's questions, right?

-trying-not-to-complicate-things-ly yr's,
 Peter


From siegfried.gonzi at kfunigraz.ac.at  Wed Jun 26 03:03:52 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Wed, 26 Jun 2002 09:03:52 +0200
Subject: Integrals with Python
References: <3D188551.6BA00DED@lanl.gov>
Message-ID: <3D196758.3040604@kfunigraz.ac.at>

Nathan Given wrote:
> Hello All,
> 
> I'm somewhat familiar with MATLAB, and I'm trying to move completely
> over to python... and I was wondering if there is a function out there
> that will compute integrals for me...
> 
> here's what I've found so far...
> 
> http://pylab.sourceforge.net/
> 
> but I'm not sure what I need off that page...
> 
> I already have Numeric... what else do I need?


If you have got measurements only and want to numerically integrate that 
you can get a small Simpson function from me (a few lines of python 
code). Drop me a note. I use it heavily in combination with measurements 
(and the precision is okay; you know measurement values...).


There would be the option to pinch a C code from the net an write the 
bining with the help of SWIG; or even better pinch a Fortran code from 
the net and make a shared module with the help of F2PY.


S. Gonzi




From sholden at holdenweb.com  Thu Jun  6 10:08:24 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 6 Jun 2002 10:08:24 -0400
Subject: check on numerical value?
References: <3cff56b4.511834453@news.skynet.be> <l5JL8.123663$cQ3.3519@sccrnsc01> <adnqh7$ri6f$1@ID-146704.news.dfncis.de>
Message-ID: <a6KL8.313639$%u2.238296@atlpnn01.usenetserver.com>

"Shagshag13" <shagshag13 at yahoo.fr> wrote in message
news:adnqh7$ri6f$1 at ID-146704.news.dfncis.de...
> can't you use (after importing types) type(x) == types.FloatType (or
type.IntType)
>

That wouldn't be helpful, since raw_input() always returns a string. The OP
wants to know whether that string is a valid representation of a number.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From python at rcn.com  Sat Jun  1 20:06:01 2002
From: python at rcn.com (Raymond Hettinger)
Date: Sat, 1 Jun 2002 20:06:01 -0400
Subject: Which is fastest dict or list.index() ?
References: <adaih2$vtb8c$1@ID-146704.news.dfncis.de>
Message-ID: <adbnkh$s57$1@bob.news.rcn.net>

There are a lot of ways to implement an OrderedDictionary.
Take a look at
http://home.arcor.de/wolfgang.grafen/Python/Modules/Modules.html
and http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/107747
If you want a skiplist version, let me know and I'll send it.

Raymond Hettinger


"Shagshag13" <shagshag13 at yahoo.fr> wrote in message
news:adaih2$vtb8c$1 at ID-146704.news.dfncis.de...
> I need a structure to hold a correspondence between toons and id like
>
> {'Bugs' : 0, 'Mickey' : 1, 'Babs' : 2, 'Jerry' : 3, 'Tom' : 4, 'Babs' : 5}
>
> I must be able to alpha sort these toons (without loosing their id), and
> access a toon from an id or an id from a toon.
>
> By now a use a dict and a list : the dict gives index from a toon, and the
> list gives me a toon from an index.
>
> {'Bugs' : 0, 'Mickey' : 1, 'Babs' : 2, 'Jerry' : 3, 'Tom : 4, 'Babs' : 5}
> ['Bugs', 'Mickey', 'Babs', 'Jerry', 'Tom', 'Babs']
>
> (the id is used somewhere else a key to a collection of elements)
>
> My question is which is the best way (best = quickest and memory safe) to
> handle this ?
> Should i subclasse from dict or list ?
> Should i consider using only one list with toons (and use their index as
id)
> ?
> Should i use 2 dicts ?
>
> Many thanks to you for helping
>
> s13.
>
>
>




From lljy at students.cs.mu.oz.au  Thu Jun 27 21:39:32 2002
From: lljy at students.cs.mu.oz.au (Lars Yencken)
Date: Fri, 28 Jun 2002 11:39:32 +1000
Subject: Plugins in Python
Message-ID: <afgeol$sgd$1@mulga.cs.mu.OZ.AU>

Hello,

I'm trying to do plugins for a program I'm writing in Python, and I'm not 
sure how exactly to do it. Basically what I want it to be able to write a 
module which loads every python file in a directory and stores an instance 
of each one that loaded successfully.

I tried using exec() and compile() to run each of the files in the 
directory from my Plugins module, and had each of the files import my 
Plugins module and store an instance of themselves there, but at the end of 
it they weren't stored.

Any ideas?

Thanks,
Lars


From jadestar at idiom.com  Wed Jun 19 01:24:59 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 19 Jun 2002 05:24:59 GMT
Subject: Pop a list from beginning ? and memory saving...
References: <aem8e8$4kq$1@news.idiom.com> <BIEJKCLHCIOIHAGOKOLHAELEDEAA.tim.one@comcast.net> <mailman.1024424930.10523.python-list@python.org>
Message-ID: <aep4jb$1q8h$1@news.idiom.com>

holger krekel <pyth at devel.trillke.net> wrote:

> [Tim Peters]
>> [James T. Dennis]
>>> ...
>>>  So, the critical question is:
>>> 	Is the list's reverse() method done in place
>> Yes.
>>> and does it run in linear time
>> Yes.
>>> (or better?
>> No.  It does len(list)/2 swaps.
	O(n/2) is linear, but still slightly better than 
	O(n)

> in another thread we had some discussion about a
> reverse-iterator.  You could write
>     for item in iters.reverse(list):
>         # item iterates over list's items backwards
>         # but the list itself stays unmodified. 
> Do you agree that this would be a good idea?
> please-share-your-brain-ly y'rs, holger

 I'm humbled that you might think that (output from) my brain is 
 worth sharing.

 On the one hand the function should be easy to add (in C)
 IFF there's a pointer to the far end of the list.  On the other
 hand it's hard to imagine an elegant syntax for expressing it
 (the example doesn't qualify) and it's hard to believe that this 
 feature would be used often enought to matter.

 It might be better to have a hack to the implementation of the
 list primitives --- so they defer the shift of remaining elements
 until GC.  The head of the list could point at the first none 
 deleted item.  Then pop(0) could be implemented in O(1) time and
 the garbage collector would eventually come along to do it's 
 job.

 Even if the memory was so tight that the GC was forced to run
 every hundred or so items (assuming they're being pop(0)'d from
 one list into another on a memory constrained system) we've 
 still reduced the work load from O(n**2) (performing n pop(0)s
 n times) by two orders of magnitude O(n/100).  That's because 
 the GC gets to wait for a 100 pops (or so) move each item over
 100 (or so) places in one O(n) pass and let the app get back to
 work.

 In other words, it should be unecessary to add syntax and methods
 when it's probably possible to tweak the implementation.  Then the 
 question becomes:

 Is the tweak worth it?

 (I don't know since I haven't even glanced at the Python sources).



From johnroth at ameritech.net  Thu Jun 20 18:11:12 2002
From: johnroth at ameritech.net (John Roth)
Date: Thu, 20 Jun 2002 18:11:12 -0400
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com> <aetbqv$kgq$1@newshost.accu.uu.nl>
Message-ID: <uh4km1ffrb400c@news.supernews.com>

"Martijn Faassen" <m.faassen at vet.uu.nl> wrote in message
news:aetbqv$kgq$1 at newshost.accu.uu.nl...
> John Roth <johnroth at ameritech.net> wrote:
> [snip]
> > There's no equivalent of a goto command in Python, for good reason,
> > which I won't repeat here. Prof. Djikstra did it very well in 1974,
> > in his letter "Goto Considered Harmful."
>
> The good old Djikstra! You know, some years ago in another newsgroup I
saw
> so many posts referring to some computer scientist called 'Djikstra',
I
> started to doubt myself and had to confirm that he really is called
> 'Dijkstra'.
>
> Now I realize that there is a Djikstra, and a Dijkstra as well.
Djikstra
> is an evil clone of Dijkstra, which the Python Secret Underground is
> keeping on ice in case they need any evil clones.

Given the Professor's reputation as somewhat of a
curmudgeon, wouldn't that be an identity function?

Sorry about the misspelling.

John Roth
>
> Regards,
>
> 'Martjin'
> --
> History of the 20th Century: WW1, WW2, WW3?
> No, WWW -- Could we be going in the right direction?




From sholden at holdenweb.com  Fri Jun 28 10:04:42 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 28 Jun 2002 10:04:42 -0400
Subject: pprinting generators
References: <ad496f8.0206280442.1fdbae42@posting.google.com>
Message-ID: <H0_S8.108273$mF.22455@atlpnn01.usenetserver.com>

"Ville Vainio" <vvainio at tp.spt.fi> wrote ...
> On 2.2.1, pprint.pprint doesn't seem to be able to handle generators:
>
>
> >>> from __future__ import generators
> >>> def f(): yield 0
> ...
> >>> import pprint
> >>> pprint.pprint(f())
> <generator object at 0x007F91D0>
>
> Surely pprint could be more helpful with generators, iterating through
> them fully or just n first elements, or whatever. Or is this a design
> decision with pprint?
>

It wouldn't be very helpful of pprint to exhaust the generator so you
couldn't access the values, would it? What if the generator were infinite?

once-you-call-next()-there's-no-going-back-ly y'rs  - steve
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From candiazoo at attbi.com  Fri Jun 28 22:46:25 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Sat, 29 Jun 2002 02:46:25 GMT
Subject: Python 2.2 (ActiveState) odbc.odbc vs. mx.ODBC.Windows
Message-ID: <3d1d1ef4.426039468@netnews.attbi.com>

Have any performance tests been done to determine advantages/disadvantages
between the two?  (Performance in execution time for simple select, insert and
update statements, not performance in extra goodies and features).

Thanks, 

Mike


From mwh at python.net  Fri Jun 14 09:05:32 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 14 Jun 2002 13:05:32 GMT
Subject: Writing a simple linux daemon type thing...
References: <K9eO8.393$LN6.261@newsread1.prod.itd.earthlink.net> <rn3cea.rt4.ln@127.0.0.1> <yylO8.916$TP4.88596@newsread2.prod.itd.earthlink.net>
Message-ID: <lkznxyko3p.fsf@pc150.maths.bris.ac.uk>

"Jones" <jones at jones.jones> writes:

> Thank you for you reply.  Just one more question.  Doing what you suggest
> was my first thought.  But I thought I read somewhere last night that using
> "&" was not a good idea..... I have no idea why (it sounds good to me), but
> I guess I just want to understand the difference between running it as a
> "daemon" and just running it the background... if there is one.

If you run it as a true daemon, you need to be careful about writing
to stdout/stderr (because they're not there any more).  Think
tracebacks.

Cheers,
M.

-- 
  ARTHUR:  Yes.  It was on display in the bottom of a locked filing
           cabinet stuck in a disused lavatory with a sign on the door
           saying "Beware of the Leopard".
                    -- The Hitch-Hikers Guide to the Galaxy, Episode 1


From roelsch at acm.org  Thu Jun 20 00:56:22 2002
From: roelsch at acm.org (Robert Oelschlaeger)
Date: Thu, 20 Jun 2002 04:56:22 GMT
Subject: Win32com Python script/MS-Access Please help!!
References: <aeo6h1$itf$1@netnews.upenn.edu>
Message-ID: <WfdQ8.36351$I82.3834202285@newssvr12.news.prodigy.com>

Dave:

You've done a good job of re-posting (my) code from
http://aspn.activestate.com/ASPN/Mail/Message/1225515 as
your own work. Yes, you changed two "int"s to "float"s and
added a "c:/" to the "open" call, and removed the "def main():",
but that's about it.

Good luck with your homework. Let's see, that would be "OPIM
101: Introduction to Computers as Analysis Tools" at the Wharton
School, University of Pennsylvania. See also
http://opim-web.wharton.upenn.edu/home/coursedes.html#OPIM101

Next time, you might just ask for help at opim-help at wharton.upenn.edu

One of the stated aims of that course is "Build student skill and
comfort using the computer to solve problems"; I don't
think that they mean for you to use the computer to find
others to do your work for you.

Bob

"Dave" <dave5774 at hotmail.com> wrote in message
news:aeo6h1$itf$1 at netnews.upenn.edu...
> Hey all,
> I'm having trouble writing a Python script relating to Microsoft
> Access.  If
> anyone could help me I would greatly appreciate it.  Here's the
> problem :
>
> I built an MS-Access table with 5 columns(fields) : OrderID, OrderDate,
> PlantID, ProductID, OrderQty, with OrderID as the primary key.  All
> fields
> are numeric, except for OrderDate(which is a date field).
>
> Now I want to populate this table with data from a web page
> (http://opim.wharton.upenn.edu/~opim101/spring02/Plant1.html).  please
> check
> the site.
>
> My object is to read the the data and insert it record by record into my
> Access table.
>
> Here's what I have so far :
>
> import urllib, string
> Plant1 =
> urllib.urlopen
> ('http://opim.wharton.upenn.edu/~opim101/spring02/Plant1.html'
> )
> lines1 = Plant1.readlines()
> import win32com.client
> engine = win32com.client.Dispatch("DAO.DBEngine.35")
> db=engine.OpenDatabase(r'd:\Plants.mdb')
>
> I have several problems :
> Firstly, when I read the lines, I notice that not all of them are
> relevant.
> I need to ignore <html>, <body>, etc. at the beginning of the file, as
> well
> as the column headers, so that the action really begins from line 4 of
> the
> file and ends with </body>
> Also, I need to eliminate <p> from every line in the file in order to
> get to
> the five comma-delimited fields that follow, and then split the rest of
> the
> string to get to each field.  Here's what I think part of the code
> should
> look like :
>
> for line in lines1.readlines():
>     line=line.rstrip()
>     inOrderID,inOrderDate,inRegionID,inProductID,inOrderQty =
> line.split(",")
>
> Something like that, I'm not sure...
> Lastly, now that I've read all the fields, I need to insert them each
> as a
> single record into my table of my access database using Python DAO
> using the
> Insert statement.  Also, the OrderDate field that I read in, looks like
> a
> text type to Access, so I need to pad it with the # sign on either side
> before inserting it into the Access database, so that Access interprets
> it
> as a Date/Time type.
>
> Finally, I need to get the same information from 2 other plants and tag
> them
> onto the same table...from the web page
> http://opim.wharton.upenn.edu/~opim101/spring02/PlantPortal.html
> where plant1's data shows up first, plant2's data shows up below it, and
> plant3's data shows up last...so all plants' data in the same table.
>
> I would like
> to use more standard commands.  I can use SQL, and I know Python -
> Access
> commands like OpenRecordSet(), TableDefs.count(), Execute(), while not
> rs.EOF, Fields().....
> Is there any other way to write the program without SGML?  To give you
> more
> of an idea of the way I'd like the program, I wrote another program for
> doing a similar thing in Excel, not Access, for a file named sales.txt
> and
> the same 5 columns, but put from a file into 5 columns on excel, and
> certain
> parameters to enter at the beginning... :
>
> ProductID = float(raw_input("Please enter the ProductID: "))
> OrderQty = float(raw_input("Please enter the cutoff order quantity: "))
> CSV = open('c:/sales.txt','r')
> import win32com.client
> x1=win32com.client.Dispatch('Excel.Application')
> x1.visible=1
> x1.Workbooks.Add()
> x1.Range("A1").Select()
> x1.Range("A1:B1").Value = ("ProductID:",ProductID)
> x1.Range("A2:B2").Value = ("OrderQty:",OrderQty)
> x1.ActiveCell.Offset(4).Activate()
> x1.ActiveCell.Range("A1:E1").Value=
> ("OrderID","OrderDate","RegionID","Produc
> tID","OrderQty")
> x1.ActiveCell.Offset(2).Activate()
> for line in CSV.readlines():
>     line=line.rstrip()
> inOrderID,inOrderDate,inRegionID,inProductID,inOrderQty = line.split
> (",")
> if float(inProductID) == ProductID:
>        if float(inOrderQty) >= OrderQty:
>            xl.ActiveCell.Range("A1:E1").Value =
> (inOrderID,inOrderDate,inRegionID,inProductID,inOrderQty)
> xl.ActiveCell.Offset(2).Activate()
>
> I just don't know how to adjust this type of thing for Access...
>
> If anyone could help me by writing the script I would greatly appreciate
> it...
> Thank you so much.
> Sincerely,
> Dave
>
>
>
>
>
>




From jjb5 at cornell.edu  Wed Jun 12 16:08:57 2002
From: jjb5 at cornell.edu (Joel Bender)
Date: Wed, 12 Jun 2002 16:08:57 -0400
Subject: dict = urllib.urldecode(string) ?
References: <jjb5-743571.14235312062002@newsstand.cit.cornell.edu>
Message-ID: <jjb5-390295.16085712062002@newsstand.cit.cornell.edu>

Well, after a little more digging I found cgi.parse_qsl() which does 
what I wanted to do and more, and the resulting list is can be passed 
into urlencode.  Thanks for the help Oleg.


From cbbrowne at acm.org  Tue Jun 18 23:35:43 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: 19 Jun 2002 03:35:43 GMT
Subject: procmail replacement in Python
References: <20020618235901.GA2197@lilith.my-fqdn.de> <mailman.1024451876.25700.python-list@python.org>
Message-ID: <aeou6e$8n2bu$1@ID-125932.news.dfncis.de>

In an attempt to throw the authorities off his trail, pinard at iro.umontreal.ca (Fran?ois Pinard) transmitted:
> [Gerhard H?ring]
>
>> I've recently looked into the procmail sources and run away screaming.
>
> ROTFL!  If I remember well, the original author knew he had a peculiar
> writing style, and was writing in some prominent file (either an FAQ or
> the README, or something like) that his style was not to be discussed! :-)
> I wondered if the new maintainer revised the sources or not, I did not check.
>
>> I'd very much prefer a Python replacement to such badly written C code.
>
> Is it bad style?  I never went as far as asserting that it was bad style,
> because tastes may differ.  Yet in that case, I knew that tastes may differ
> very widely. :-).
>
> Surely, `procmail' successfully addressed difficult problems, and did it
> rather solidly, with almost no bugs.  So, despite the writing style is rather
> alien to mine, it surely had mysterious virtues.  I could not condemn it!

The thing that would be more interesting would be a translator that
would transform procmail _scripts_ into (say) Sieve scripts.

Who cares what the procmail code base looks like if you can take the
sets of rules and transform _them_ into a form that proved more
useful?
-- 
(concatenate 'string "cbbrowne" "@cbbrowne.com")
http://www.ntlug.org/~cbbrowne/ifilter.html
All generalizations are false, including this one. 


From zopestoller at thomas-guettler.de  Thu Jun 20 06:34:31 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 20 Jun 2002 12:34:31 +0200
Subject: Writing to Excel
References: <79b54d28.0206191151.52ee78a3@posting.google.com>
Message-ID: <3D11AFB7.9070803@thomas-guettler.de>

Chris wrote:

> Hi I'm using a Python cgi script to write the contents of a table to
> an Excel sheet. I retrieve all the information fine but it seems like
> Excel doesnt want to print more than one line for me. It prints
> 'headerLine' but nothing more. Why???
> 
> ... 
> headerLine = "some crap"
> 
> sys.stdout.write( Html.header( 'application/x-msexcel' ) )
> print string.strip(headerLine)
> allText = ""


I do it with a different content-type. The code is attached, maybe
it helps

  thomas

+++
         if REQUEST:
             REQUEST.RESPONSE.setHeader('Content-Type',
                                        'application/vnd.ms-excel')

         output=[]
         output.append("<html><body><table border='1'>\n")
         headers=list_of_all_properties.keys()
         headers.sort()
         output.append('<tr>')
         for header in headers:
             output.append('<th>%s</th>' % cgi.escape(header))
         output.append('</tr>\n\n')
         for line in lines:
             output.append('<tr>')
             for col in headers:
                 value=line.get(col, '')
                 if type(value) in [type([]), type(())]:
                     new_value=[]
                     for keyword in value:
                         new_value.append(cgi.escape(keyword))
                     value=string.join(new_value, '<br>')
                 else:
                     value=cgi.escape(value)
                 output.append("<td>%s</td>" % value)
             output.append('</tr>\n\n')
         output.append("</table></body></html>")
         return string.join(output)
+++



From NKluge at optonline.net  Wed Jun  5 02:14:20 2002
From: NKluge at optonline.net (Nicola Kluge)
Date: Wed, 05 Jun 2002 06:14:20 GMT
Subject: self
References: <mailman.1023225848.26315.python-list@python.org>
Message-ID: <0%hL8.22575$tK.5490038@news02.optonline.net>

Interesting approach.  However, once you do that.  The symbolic form of the
equations is lost.  For example,
>>> c=Eval()
>>> c.y='3'
>>> c.y
3
>>> c.a='4'
>>> c.a
4
>>> c.f='y*a'
>>> c.f
12

but
 >>> c.__dict__['dict']['f']
12

doesn't give you the equation back.

Your, approach goes under a lazy evaluation business.  However, I need to
have all dependent attributes which are nested deeper to be evaluated every
time.  That is why the equations should be retained in the  dictionary.  I
don't want to worry whether the previously evaluated attribute holds the
updated value or not.  Evaluations have to propagate all the way to the end,
for example in the above case, while evaluating c.f the evaluation should
evaluate 'y' so that we are sure that 'y' is still 3.  In you approach the
evaluation stops at 'f'=12 but it can't see whether 'y' changed even though
'f' once depended on 'y'.

Vojin Jovanovic


"Bjorn Pettersen" <BPettersen at NAREX.com> wrote in message
news:mailman.1023225848.26315.python-list at python.org...
> From: Vojin Jovanovic [mailto:vjovanov at stevens-tech.edu]
>
> I don't see how that is going to work.  In fact I tried
> things like that before without success.  Just sticking the
> dictionary into eval/exec is missing the point.  Every time
> parsing is done flow needs to go through __getattr__ special
> method to evaluate nested equations.  Without self. you can't
> achieve that effect.  Try the approach you suggested and
> you'll get '#U'.

Perhaps something like this?

-- bjorn


class Eval:
    def __init__(self):
        self.__dict__['dict'] = {}
        exec "from math import *" in self.dict

    def __call__(self, expr):
        try:
            return eval(expr, self.dict, self.dict)
        except SyntaxError:
            exec expr in self.dict, self.dict

    def __getattr__(self, attr):
        return self.dict[attr]

    def __setattr__(self, attr, val):
        self.dict[attr] = eval(val, self.dict, self.dict)

e = Eval()

print e('pi * 3')
e('x = sin(0.5)')
print e.x
e.y = 'sin(0.5)'
print e.y






From fredrik at pythonware.com  Sat Jun 29 08:51:00 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 29 Jun 2002 12:51:00 GMT
Subject: Python needs better error reporting
References: <mailman.1025350057.14486.python-list@python.org>
Message-ID: <U2iT8.45587$n4.10825231@newsc.telia.net>

Tim Peters wrote:

> > I really DO have a point!
>
> Error-reporting in parsers has been the subject of intense
> research.  Do a little digging and come back when you can
> make a case that your point can be addressed effectively.

fwiw, I've seen Lisp advocates argue that this may be
one of Lisp's greatest advantages...

</F>




From opengeometry at NOSPAM.yahoo.ca  Thu Jun 20 11:52:54 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 20 Jun 2002 15:52:54 GMT
Subject: latex-tex-packages
References: <mailman.1024570406.30170.python-list@python.org>
Message-ID: <aestom$9q4al$1@ID-99293.news.dfncis.de>

Alves, Carlos Alberto - Coelce <calves at coelce.com.br> wrote:
> Hi all,
> You forgive me, c'os I know this issue it's not to this list; But, anybody
> can tell me where can I get tex/latex packages to install. I just browse
> some url but can't find'em.
> 
> Thanks, in advance.

www.tug.org

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin


From tdelaney at avaya.com  Thu Jun 13 19:57:12 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 14 Jun 2002 09:57:12 +1000
Subject: Why does Python mix OO concepts and non OO concepts for opera
	tion s  on basic types?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A387@natasha.auslabs.avaya.com>

> From: holger krekel [mailto:pyth at devel.trillke.net]
> 
> probably. With python 1.6 many methods of the string *module* were
> included included directly with the *type* string ('str'). Most 
> python developers think that the string module should be deprecated 

I think "most" is a little bit strong. The general feeling that I get seems
to be pretty much split down the middle.

I personally *don't* favour getting rid of the string module - there are
many things which lend themselves better to a non-object-oriented style -
and many of the string functions are included.

Now, if the str type included *all* of this functionality, *and* was aliased
to string, then you may have a case ... although 'import string' would still
need to work (for backwards compatibility) until 3.0.

Whilst I like str as a type name, I don't think it is expressive enough for
the string functions.

Tim Delaney




From tim.one at comcast.net  Thu Jun 13 22:09:16 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 13 Jun 2002 22:09:16 -0400
Subject: Type subclassing: bug or feature
In-Reply-To: <aeau48$111$1@panix1.panix.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEPFPMAA.tim.one@comcast.net>

[Aahz]
> Consider the following code:
>
> class MyStr(str):
>     def contains(self, value):
>         return self.find(value) >= 0
>
> s = MyStr("hello, world!")
> s = s.capitalize()
> if s.contains('Hello'):
>     print "Found it!"
>
> It fails with an AttributeError when it calls s.contains(), because
> s.capitalize() returned a str instead of a MyStr.  Anyone want to take a
> whack at defending this as the correct behavior?

Short course:

    http://www.python.org/sf/468887

Long course:

    http://www.python.org/sf/460020

Making sure this pisses you off in all cases required one of the biggest
patch sets that went into Python 2.2 <wink>.





From hesterloli at hotmail.com  Tue Jun 18 16:45:25 2002
From: hesterloli at hotmail.com (George Hester)
Date: Tue, 18 Jun 2002 20:45:25 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com> <gYgP8.42356$GY.12361719@twister.nyroc.rr.com> <aekcug$7js8n$1@ID-114079.news.dfncis.de> <9YBP8.44418$GY.13865843@twister.nyroc.rr.com> <Xns92317F85D48EDduncanrcpcouk@127.0.0.1>
Message-ID: <FZMP8.46299$GY.14478971@twister.nyroc.rr.com>

Jesus if that is all it is why the runaround?  You may think I'm impolite but to
ask what looks as though a simple question and have it take 21 responses and
demands that I learn the language NOW something is very wrong with you people
that have answered.  Thanks for peoviding what I asked for.  Let's see if it
works.  If it don't then I suspect the runaround continues.

--
George Hester
_________________________________
"Duncan Booth" <duncan at NOSPAMrcp.co.uk> wrote in message
news:Xns92317F85D48EDduncanrcpcouk at 127.0.0.1...
> "George Hester" <hesterloli at hotmail.com> wrote in
> news:9YBP8.44418$GY.13865843 at twister.nyroc.rr.com:
>
> > I think you guys don't how to do it.
>
> And I think you are one of the most impolite posters we have seen around
> here.
>
> Try this:
> <html>
> <head>
>   <title>Python Version</title>
> </head
> <body>
> <SCRIPT Language="Python" RUNAT=Server>
> import sys
> Response.Write(sys.version)
> </SCRIPT>
> </body>
> </html>
>
>
> --
> Duncan Booth                                             duncan at rcp.co.uk
> int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
> "\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?




From tdelaney at avaya.com  Tue Jun 18 20:13:18 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 19 Jun 2002 10:13:18 +1000
Subject: If you use Python -OO is it a permanent condition??
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A39F@natasha.auslabs.avaya.com>

For some reason, I parse the subject line as "If you use Python in an
object-oriented way, is it a permanent condition?"

I'm not sure if that's a philosophical question, but it confused the hell
out of me ;)

Tim Delaney




From duncan at NOSPAMrcp.co.uk  Fri Jun 14 04:24:04 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Fri, 14 Jun 2002 08:24:04 +0000 (UTC)
Subject: Date formats
References: <3D09A1B1.6040705@mxm.dk>
Message-ID: <Xns922D5E1FF60FEduncanrcpcouk@127.0.0.1>

Max M <maxm at mxm.dk> wrote in news:3D09A1B1.6040705 at mxm.dk:

> 
>      gDateFormat = '%d-%m-%Y'
> 
> This give dates of the type:
> 
>      01-01-2002
> 
> For some reason my customer wants dates of the type:
> 
>      1-1-2002
> 
> Without the leading zero in single digit dates. But when I look in the 
> docs for date time formats there is no way to specify days and months in 
> that format.

The documentation says that on some platforms an optional field width and 
precision specification can follow the initial '%'. This is obviously 
highly non-portable, but might provide a suitable solution if you are not 
desperate for portability. For example, if you are running under Windows 
using "%#d-%#M-%Y" as the format string gives the format you require.

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From vvainio at tp.spt.fi  Thu Jun 20 05:07:36 2002
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 20 Jun 2002 02:07:36 -0700
Subject: 'else' clause in 'for' and 'while'
Message-ID: <ad496f8.0206200107.291c5a40@posting.google.com>

How do you people use the else clause in for and while constructs? Are
they used widely for some specific applications? Do they offer
elegant, idiomatic solutions to some problems?

I haven't found any use for them yet myself, but knowing that they
exist and still not using them troubles me, core python being quite
'small' after all.

-- Ville


From jiscmail-support at jiscmail.ac.uk  Tue Jun 11 11:47:46 2002
From: jiscmail-support at jiscmail.ac.uk (jiscmail-support at jiscmail.ac.uk)
Date: Tue, 11 Jun 2002 16:47:46 +0100
Subject: Your mail sent to mailbase.ac.uk with subject  Hello,introduction on ADSL
Message-ID: <200206111547.g5BFlkl10314@ori.rl.ac.uk>

This Mailbase list has moved to JISCmail.

For further information see http://www.jiscmail.ac.uk/docs/transition.htm.

The email message that you sent has not been delivered, if you
want your email to be delivered you must re-address it as described on the
pages referenced above.

If you have any queries about this message then please contact

jiscmail-helpline at jiscmail.ac.uk






From rotundo52 at hotmail.com  Thu Jun 20 00:42:15 2002
From: rotundo52 at hotmail.com (Jon J. Morin)
Date: Thu, 20 Jun 2002 04:42:15 GMT
Subject: can I call operator overloader in superclass?
Message-ID: <H2dQ8.1805$CZ4.569945@news1.news.adelphia.net>

Hi all.  I have a particularly trivial bit of code I'm working with to learn 
Python and I have a question about operator overloading.  Say I have class 
AList which is partially defined:

class AList:
        def __init__(self, other=[]):
                self.data = []
                for x in other:
                        self.data.append(x)
        def __repr__(self):
                return "List: " + `self.data`
...
...
...

I have a subclass AListSub which is partially defined as follows:

classAListSub(AList):
        repr_calls = 0
        def __init__(self, list=[]):
                AList.__init__(self, list)
        def __repr__(self):
                AListSub.repr_calls = AListSub.repr_calls + 1
                print "Calling __repr__ in superclass"
                AList.__repr__(self)
                print "Method called %d times." % (AListSub.repr_calls)
...
...
...
I can create instances on AListSub by passing in a string, or a list as in
>>> x = AListSub('sasquatch')
>>> x
Calling __repr__ in superclass
Method called 1 times.
Traceback (innermost last):
  File "<stdin>", line 1, in ?
TypeError: repr not string

For one, I don't understand the error.  For seconds, can I call __repr__ in 
the superclass from __repr__ in the subclass?

Jon J. Morin


From marklists at mceahern.com  Sat Jun 29 11:44:24 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sat, 29 Jun 2002 10:44:24 -0500
Subject: private
In-Reply-To: <pan.2002.06.29.17.38.51.789527.440@myself.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBKECGLHAA.marklists@mceahern.com>

> Is there any possibility to build up a _real_ private attribute?
> Such as C++ private data members...
> I know that with double-underscore prefix (and the name mangling related)
> I can " hide " data members but I think it's not a so crafty trick...
> Is there any PEP about it? Will Guido insert a private keyword
> in a future release of Python ?

What problem are you trying to solve?

Cheers,

// m
-





From tismer at tismer.com  Wed Jun 19 11:28:50 2002
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 19 Jun 2002 17:28:50 +0200
Subject: [Python-Dev] Tcl adept wanted for Stackless problem
References: <3D108966.9050402@tismer.com> <200206191507.g5JF7ag02088@pcp02138704pcs.reston01.va.comcast.net>
Message-ID: <3D10A332.7010909@tismer.com>

Guido van Rossum wrote:
>>My big question is:
>>When does Tcl use C stack entries as globals, which
>>are passed as function arguments to interpreter calls?
> 
> 
> It's a performance hack, just as stackless :-).

Where the effect of my hack is slightly bigger.
We can fight that out in Charleroi. :-)

> Tcl's interpreter data structure has a return value field which can
> receive a string of arbitrary length.  In order to make this
> efficient, this is initialized with a pointer to a limited-size array
> on the stack of the caller; when the return value is longer, a
> malloc()'ed buffer is used.  There is a little dance you have to do to
> free the malloc()'ed buffer.  The big win is that most calls return
> short strings and hence you save a call to malloc() and one to free()
> per invocation.  This is used *all over* the Tcl source, so good luck
> getting rid of it.

Thank you! I should better not try this. Instead, I'd like
not to touch it at all.
I have patched tkinter in a way that it does not slice the stack
while some Tcl stuff is running (maybe I didn't catch all).
That should mean that the small stack stings are all alive.
That is, in the context of Tcl, I dispensed with the
"stackless" concept.

The remaining problem is switching of tasklets which contain
Tcl invocations. I thought so far that this is no problem,
since these are disjoint contexts, but Jeff Senn reported
problems as well.
I fear I have the problem that Tcl thinks it is still using
the same interp, or it creates a nested one, while the
tasklets are not nested, but seen as independent. Somehow
I need to create a new Tcl frame chain for every tasklet
that uses Tcl.
Can this be the problem?

Still no clue how to do it but thanks - ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From kfw at woeltje.org  Thu Jun 13 19:59:53 2002
From: kfw at woeltje.org (Keith Woeltje)
Date: Thu, 13 Jun 2002 19:59:53 -0400
Subject: Python 2.2.1
References: <II7O8.622$e27.275217555@twister1.starband.net>
Message-ID: <3D0931F9.4030103@woeltje.org>

Dave,
I concur with the above--good choice in language. If you get frustrated 
with that particular tutorial (it didn't look that great on first 
glance), be sure to look at all of the resources on the Python web site at:
http://www.python.org/doc/Newbies.html

In particular, if you are new to programming, you owe it to yourself to 
check out Alan Guald's site:
http://www.freenetpages.co.uk/hp/alan.gauld/
if you can spring for it, the print version is an excellent starting 
point (I'm fond of real books personally).

Best of luck.
 >K

Dave wrote:
>     Hi, I am 12 years old and would like to begin learning Python over 
> the summer. I have Windows 98 SE. I have python 2.2.1 and all of its 
> utilities on my computer. I am taking a tutorial 
> from  http://softwaredev.earthweb.com/sdopen/article/0,,12382_626311,00.html  
> If you go to it and scroll down to the paragraph entitled *Getting 
> Started* there is a paragraph that says, *"You will need to cause the 
> directory containing the file named python.exe to be listed in your 
> system environment variable named path. If you already know how to set 
> the path, go ahead and do it.  If you don't already know how, you may 
> need to get some help." *
> ** 
>  
>     How do I set the path and what should the name of the path be? All 
> help and information would be appreciated. Thanks.:)




From gleki at gol.ge  Fri Jun  7 06:51:24 2002
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Fri, 07 Jun 2002 12:51:24 +0200
Subject: map
References: <3D007666.2B6D3582@gol.ge>
Message-ID: <3D00902C.DF2CC70@gol.ge>

Ok, there is a function in Numeric, and Numeric.power(a, y) works as
expected.

Anyway, I still would like to know wether there's an opportunity to employ
map function as stated.

Giorgi Lekishvili wrote:

> Hi all!
>
> Is there a way to map a function with several arguments to a list? E.g.,
>
> >>>import math
> >>>import MLab
> >>>a=MLab.rand(3,5)
> >>>b=map(math.pow,a)
>
> of course, this doesn't work.
>
> What to do? My task is to optimize the power (i.e., y in pow(x,y)) to
> achieve maximal performance.
>
> thanx,
> Giorgi
>
> PS: Perhaps I have missed the needed functionality, that's presented in
> NumPy?



From mwarden22 at hotmail.com  Tue Jun 11 14:36:18 2002
From: mwarden22 at hotmail.com (nealj)
Date: 11 Jun 2002 11:36:18 -0700
Subject: Embedded Python script debugger???
References: <qjAE8.494039$K52.83658610@typhoon.tampabay.rr.com> <dwCE8.135970$o66.404796@news-server.bigpond.net.au> <m3bsbe94k3.fsf@daystar.control.com> <d3768569.0205221233.8079de2@posting.google.com> <3ceceec6$1_3@goliath.newsgroups.com>
Message-ID: <d3768569.0206111036.50190748@posting.google.com>

That is exactly what I hope to do with Hap - you could start debugging
by importing the hapclient module.  The dll would still need to be
ported to CE but that shouldn't be too big of a hurdle, we are
planning a mac port in the near future.

"Brad Clements" <bkc at Murkworks.com> wrote in message news:<3ceceec6$1_3 at goliath.newsgroups.com>...
> I'm keeping an eye on HAP because I think it will be useful for debugging
> Python CE applications.
> 
> The holdup is that I need a seperate .exe on the client to do this.
> 
> What would be very interesting .. is a HAP client DLL that is also a Python
> module.
> 
> So, from python I could instantiate a HAP client debug session (the HAP
> client module would create a new interpreter to do this).  Just need some
> options for hooking up debug output text so it could be sent back to the
> original interpreter if needed (rather than a console, because CE doesn't
> have a console).
> 
> --
> Novell DeveloperNet Sysop #5
> 
> _
> "nealj" <mwarden22 at hotmail.com> wrote in message
> news:d3768569.0205221233.8079de2 at posting.google.com...
> > Hap could be adapted to do this quite easily (in fact I will be doing
> > it in the next few weeks).  Currently, Hap uses an app that takes the
> > place of python.exe to execute the python scripts.  This app handles
> > the socket and the debugging stuff - instead of "hapclient.exe", it
> > could just as easily be your application that gets run.
> >
> > My plan is to repackage the code in the app to make this easier - then
> > all you would have to do is to link to a dll and you're off.  Keep an
> > eye on the Hap site on sourceforge - there will be important and
> > interesting updates coming soon.
> >
> > If you are intersted in trying it out, get the source from SourceForge
> > (its currently only there in .zip format) and take a look at
> > ConsoleEmbed.cpp in the HapClient directory - you'll see it doesn't
> > really do much other than setup the debugging and run the script.  Let
> > me know if you have any questions - the best way to get in contact is
> > through the sourceforge email https://sourceforge.net/users/nealj/
> >
> > Good Luck!
> > Neal
> 
> 
> 
> 
> -----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
>    http://www.newsfeed.com       The #1 Newsgroup Service in the World!
> -----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From steve at ferg.org  Fri Jun 21 09:10:30 2002
From: steve at ferg.org (Stephen Ferg)
Date: 21 Jun 2002 06:10:30 -0700
Subject: a Tree data-structure class?
Message-ID: <b16e4ef7.0206210510.77a956b6@posting.google.com>

I do a lot of processing of tree-structured data, so I'm thinking it
would be useful to develop a generic Tree class that I could adapt to
various applications.  Or perhaps a Tree class and a generic Node
class.

But before I start trying to re-invent the wheel, I thought I'd ask if
anybody knows of something like this that has already been developed.

Note that I'm not talking about a Tree GUI widget, but about a tree
data structure, like a list, a queue, etc.

-- Steve Ferg (steve at ferg.org)


From mlh at vier.idi.ntnu.no  Thu Jun  6 20:01:10 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Fri, 7 Jun 2002 00:01:10 +0000 (UTC)
Subject: email: msg.add_payload not synonymous with msg.attach?
Message-ID: <slrnafvtu6.7q0.mlh@vier.idi.ntnu.no>

I've been fiddling with the email package, and tried to construct a
Message object, using add_payload to add the email body (It's a
text/plain message). This worked fine, except for a deprecation
warning telling me to use attach instead. But when I did, my program
crashed with the message "TypeError: string payload expected: <type
'list'>" (from Generator.py). Well... My payload was indeed simply a
string, so I don't understand the error message. Also, I don't
understand how this discrepancy could have occurred if, as the docs
claim, add_payload and attach are synonymous.

(The code is very simple -- basically just instantiating Message and
adding a payload.)

I'm using the email package from CVS (downloaded everything a day or
two ago).

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From thenemesisjaguar at aol.com  Thu Jun  6 20:43:30 2002
From: thenemesisjaguar at aol.com (TheNemesisJaguar)
Date: 07 Jun 2002 00:43:30 GMT
Subject: Movie on Mac
Message-ID: <20020606204330.08563.00000041@mb-fr.aol.com>

How do I get a movie or .bik file to open and play in a python program if I am
using a mac.
Tell me exactly (if it is on my desktop)


From db3l at fitlinxx.com  Thu Jun  6 21:07:33 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 06 Jun 2002 21:07:33 -0400
Subject: Installer vs py2exe: Windows DLLs
References: <u1ooxxn8.fsf@morpheus.demon.co.uk> <Xns9220730326E0Fgmcmhypernetcom@199.171.54.215> <u1ogoiqu.fsf@morpheus.demon.co.uk> <Xns922557D7E23EFgmcmhypernetcom@199.171.54.213>
Message-ID: <uk7pbuca2.fsf@ctwd0143.fitlinxx.com>

Gordon McMillan <gmcm at hypernet.com> writes:

> _winreg.pyd didn't exist as of Py 1.5, but your patch looks
> good.

Perhaps I'm one of the only guys left, but I'm still successfully
using installer 5 with Python 1.5.2 scripts, so if this would be the
only reason to break that compatibility, it would be great to make it
dynamic in some way so as to remain compatible while removing the
dependency for later Python releases.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From pinard at iro.umontreal.ca  Sun Jun 30 10:36:46 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 30 Jun 2002 10:36:46 -0400
Subject: Preforked server in Python
In-Reply-To: <3D1EB82C.1000501@awal.net.sa>
References: <3D1EB82C.1000501@awal.net.sa>
Message-ID: <oqpty896kh.fsf@titan.progiciels-bpi.ca>

[rayed]

> I am looking for module similar to SocketServer that prefork chlidren instead
> of forking on each request.

I wonder.  Is there a way to transmit opened sockets to pre-forked children?

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From evan at 4-am.com  Mon Jun 24 21:35:05 2002
From: evan at 4-am.com (Evan)
Date: Tue, 25 Jun 2002 01:35:05 GMT
Subject: Web templating/db tool with best designer/coder separation?
References: <af2gh6$2c2$0@216.39.172.122> <af8dtp$5qp$0@216.39.172.122>
Message-ID: <3D17C8D4.6090704@4-am.com>

Bengt Richter wrote:
> No one mentioned ZPT (Zope Page Templates),
>     http://www.zope.org//Wikis/DevSite/Projects/ZPT/FAQ
> 
> and from the description, this would seem to have a real advantage over all of
> the above, in that it uses attributes of html tags as opposed to special tags
> that browsers or design tools won't generally understand and render usefully.
> 
> This is touted as enabling the "round trip" from graphic designer to template logic
> programmer and back using the same html source. I assume this is the "round trip"
> Paul was referring to. To me, it seems a winning point for ZPT, though I don't know
> how ZPT works out in practice.

It works out quite well, if you're using Zope.  I haven't seen much 
feedback from people using it independently of Zope, so I can't say one 
way or another in that case.

> Opinions on ZPT?

I like it, but then I would :-)  My wife (graphic designer) and I were 
major contributors to the design of ZPT, so it suits our tastes.

There are also plans in the works for a stylesheet-like extension to ZPT 
that would allow the logic (even the presentation logic) to be further 
separated from the HTML, without plunging the programmer into the dark 
mists of DOM-walking.

Cheers,

Evan @ 4-am



From mgilfix at eecs.tufts.edu  Fri Jun 14 10:05:47 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Fri, 14 Jun 2002 10:05:47 -0400
Subject: Writing a simple linux daemon type thing...
In-Reply-To: <aecrfe$a4l$1@minotaurus.cip.informatik.uni-muenchen.de>; from dysfate@gmx.de on Fri, Jun 14, 2002 at 03:35:42PM +0200
References: <K9eO8.393$LN6.261@newsread1.prod.itd.earthlink.net> <rn3cea.rt4.ln@127.0.0.1> <yylO8.916$TP4.88596@newsread2.prod.itd.earthlink.net> <aecrfe$a4l$1@minotaurus.cip.informatik.uni-muenchen.de>
Message-ID: <20020614100547.D4109@eecs.tufts.edu>

  Wouldn't it be nice if some daemonizing code and basic log support
was built into SocketServer? SocketServer aims at simplifying
the infrastructure for people who don't need a full-blown server
app. Might also be a nice addition for smaller projects who don't want
to deal with this sort of thing... Might make an interesting patch
IMHO.

                   -- Mike

On Fri, Jun 14 @ 15:35, Johannes Stiehler wrote:
> I suppose you would have to redirect all output to stdout and log directly
> to a specified log file. That's usually in the code for all daemonizing
> processes.
> The point is to completely unconnect the process from the console where it
> was started.

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From gerhard.haering at gmx.de  Wed Jun 19 18:58:00 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Thu, 20 Jun 2002 00:58:00 +0200
Subject: Windows versions of Python---pros and cons?
In-Reply-To: <20020619154654.C619@ActiveState.com>
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <20020619154654.C619@ActiveState.com>
Message-ID: <20020619225800.GB12907@lilith.my-fqdn.de>

* Trent Mick <trentm at ActiveState.com> [2002-06-19 15:46 -0700]:
> My understanding is that for C extensions to work they must be built
> with the same compiler as that used to build the Python with which they
> will run. (I may be wrong. Someone please correct me if I am.)

You are wrong. As long as you link to the python{major}{minor}.dll you
are binary compatible, no matter which compiler you use. It's just a
matter of providing an additional compiler option to distutils, like in:

    python setup.py build --compiler=mingw32

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From whisper at oz.net  Mon Jun 24 22:40:11 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 24 Jun 2002 19:40:11 -0700
Subject: Windows versions of Python---pros and cons?
In-Reply-To: <7934d084.0206241750.221bbff2@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEAEPADLAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Andrae Muys
> Sent: Monday, June 24, 2002 18:51
> To: python-list at python.org
> Subject: Re: Windows versions of Python---pros and cons?
>
>
> Tom Loredo <loredo at astro.cornell.edu> wrote in message
> news:<3D124575.2A1393FF at astro.cornell.edu>...
> > Fernando P?rez wrote:
> > >
> > > Don't know if there's a misunderstanding here:
> >
> > Very likely!  (On my part!)
> >
> > > cygwin is a unix-like
> > > environment under Windows.
> >
> > Okay, my concern was that it wouldn't work with or be able to produce
> > Windows binaries---like if I had to produce binary extension modules
> > to ship to others, or a McMillan-ized install that used PythonWin
> > and Tk or wxWindows.  Cygwin is so huge that I got the impression
> > it was a self-contained "subenvironment" so to speak, so I didn't
> > realize I could use it to produce stuff that non-cygwin Windows users
> > could use.
> >

If redistribution of executables that use the cygwin.dll is important, it
should be noted that you can only do this if it's a GPL app (or maybe even
that requires permission) - commercial redistribution definately requires a
licen$e from RedHat. At one time it was $100 per developer.

Dave LeBlanc
Seattle, WA USA





From olc at ninti.com  Sat Jun  8 21:27:56 2002
From: olc at ninti.com (Michael Hall)
Date: Sun, 9 Jun 2002 10:57:56 +0930 (CST)
Subject: Removing ^M
In-Reply-To: <004501c20efb$386ad1e0$0101010a@local>
Message-ID: <Pine.LNX.4.21.0206091054490.7509-100000@localhost.localdomain>

Python ... I love it! I knew about dos2unix but it's not on my box for
some reason, and downloading and installing it seemed like overkill.
This script worked like a charm.

Thanks 
Mick



On Sat, 8 Jun 2002, Chris Gonnerman wrote:

> ----- Original Message ----- 
> From: "Michael Hall" <olc at ninti.com>
> 
> 
> > I am trying remove ^M characters (some kind of newline character) from an
> > HTML file. I've tried all sorts of string.replace and sed possibilities
> > but the
> > things just won't go away. Does anyone have a way of removing such
> > characters?
> 
> Unixoid OS, right? 
> 
> ------------------------------------
> 
> import sys
> 
> data = sys.stdin.read()
> 
> for ch in data:
>     if ch != "\r":
>         sys.stdout.write(ch)
> 
> ------------------------------------
> 
> Run it like this:
> 
>     python cleaner.py <source.html >destination.html
> 
> Yeah, it could have a bunch more nice stuff...  This is the Q&D 
> version.
> 
> If the file is really large this won't be real efficient, but for 
> small files it's probably the best way.
> 
> 
> Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
> http://newcenturycomputers.net
> 
> 
> 
> 

-- 
--------------------------------
n   i   n   t   i  .   c   o   m
php-python-perl-mysql-postgresql
--------------------------------
Michael Hall     ninti at ninti.com
--------------------------------






From syn_uw at NOSPAM_hotmail.com  Sun Jun 23 17:29:43 2002
From: syn_uw at NOSPAM_hotmail.com (Marc Bigler)
Date: Sun, 23 Jun 2002 23:29:43 +0200
Subject: Zope
Message-ID: <3D163DC7.739A83FD@NOSPAM_hotmail.com>

Hello,

I would like to know how I can generate a navigation bar in zope ?

Something like that: when you are at the home page you get:

Home

When you are at a new page let say flowers you get:

Home > Flowers

and when you are at a subpage from Flowers called Orchidee you get:

Home > Flowers > Orchidee

Is that somehow possible examining the BASE0, BASE1,... and so on
variables or is there a better method ?

Any suggestions are welcomed !

Many thanks...
Regards




From ak at silmarill.org  Sat Jun  1 10:28:23 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Sat, 01 Jun 2002 14:28:23 GMT
Subject: What does Python offer?
References: <ad216r$lol$1@news8.svr.pol.co.uk>
Message-ID: <slrnafbntq.j99.ak@ak.silmarill.org>

In article <ad216r$lol$1 at news8.svr.pol.co.uk>, Solosnake wrote:
> Hello
> 
> I have beome curious about Python after playing a game called 'Severance -
> Blade of Darkness', which uses python scripts. I know very little about the
> language. I am a C++ programmer, and would like to ask the Python community
>:
> 
> What is unique or special about Python?
>

On one hand, it's very high-level (like perl or basic), so that
it's very easy to quickly write a prototype; on the other hand,
it is very scalable in the sense that you can write huge programs
that perl and basic would be too "small" for. As someone said..
it bridges the gap between scripting and programming languages.

> Why would it be useful to a games programmer?
>
You can quickly prototype a game in python and then decide
whether you want to re-write it in C++ and most likely you'll
find python is fast enough to use for most of the game except for
the most expensive graphics code.

> Could it be used for allowing users to customize games, eg reprogramming AI
> bots etc?
>
Certainly!

 - Andrei

-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From logiplexsoftware at earthlink.net  Mon Jun 10 17:59:30 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 10 Jun 2002 14:59:30 -0700
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
In-Reply-To: <Xns9229EE722467Fgustafl@195.100.94.182>
References: <Xns9229EE722467Fgustafl@195.100.94.182>
Message-ID: <20020610145930.5201949e.logiplexsoftware@earthlink.net>

On Mon, 10 Jun 2002 21:25:06 +0000 (UTC)
Gustaf Liljegren wrote:

> I'm making a function that takes one integer and returns a boolean that 
> says if the integer matches a number in a certain group of numbers. This 
> group consists of single numbers aswell as ranges spanning thousands of 
> numbers. I'm looking for an elegant way to write this function.

def isin(n, ranges):
    for r in ranges:
        try: 
            s, e = r
            if s <= n <= e:
                return 1
        except TypeError:
            if n == r:
                return 1
    return 0

ranges = [(1, 999), 1004, 1006, (1050, 2000)]

print isin(0, ranges)
print isin(1, ranges)
print isin(2, ranges)
print isin(1000, ranges)
print isin(1999, ranges)

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From db3l at fitlinxx.com  Fri Jun  7 00:53:39 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 07 Jun 2002 00:53:39 -0400
Subject: Problem pickling class instance
References: <3D000D77.7040008@erols.com>
Message-ID: <uelfju1t8.fsf@ctwd0143.fitlinxx.com>

"Edward C. Jones" <edcjones at erols.com> writes:

> First I run mod.py then top.py. The latter give the message:
> 
>      AttributeError: 'module' object has no attribute 'A'
> 
> Please explain the problem to me. Is there a standard work-around?

Pickle only pickles object instance data, not the code representing
the object (class) itself.  Instead it includes a reference to the
name of the class which the object data represents.

When you pickled your instance of class A, that's the name used in the
resulting pickle.  It tries to instantiate a new instance of that
object class when de-pickling in order to populate that instance with
the instance data from the pickle.

But when you try to load it, you've imported the mod module, so that
same class is now known in your namespace as "mod.A", and thus pickle
can't find the normal "A" class it expects for the data it is
de-pickling.

In this specific case, you could either:

  * Do a "from mod import A" (whether or not you're importing mod itself)
  * After doing an "import mod" somewhere before the pickle do "A=mod.A"

Both of these will add a local module-level binding for the name "A"
to the actual class in the module "mod".  Note that you need to have
these statements at module level since pickle expects to find the
class definition at that level.

Whether or not that's an appropriate overall solution depends on the
problem.  If a class is something that is going to be used in a pickle
that is common between two parts of the system, it probably makes
sense to isolate it to its own module (or package), and have all users
of it be in a separate module and import that module to access the
class.  Either that or include the pickling/de-pickling operations
both within the same module as the class so they are always local.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From jepler at unpythonic.net  Tue Jun 25 16:24:43 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 25 Jun 2002 15:24:43 -0500
Subject: bug in os.getgroups?
In-Reply-To: <1025033152.17806.TMDA@nightshade.la.mastaler.com>
References: <1025033152.17806.TMDA@nightshade.la.mastaler.com>
Message-ID: <20020625202442.GG4435@unpythonic.net>

This is liely to be because /tmp is "sticky":
       The  `sticky'  bit  (S_ISVTX)  on a directory means that a
       file in that directory can be renamed or deleted  only  by
       the  owner of the file, by the owner of the
       directory, and by root.
(from my system's manpage for stat(2))

You aren't root, so you can't delete this file in /tmp, which is
typically a "sticky" directory.  It doesn't matter that the group
matches.

I can observe this behavior on my own system from the shell:
$ id 
uid=229(jepler) gid=15(man) groups=15(man),20(games),22(utmp),23(document)
$ ll blah
-rw-rw-r--    1 root     man             0 Jun 25 15:23 blah
$ rm blah
rm: cannot unlink `blah': Operation not permitted

Jeff




From gscott at peregrine.com  Fri Jun  7 11:43:23 2002
From: gscott at peregrine.com (Gordon Scott)
Date: Fri, 7 Jun 2002 08:43:23 -0700 
Subject: Program to an interface, not an implementation
Message-ID: <D54BF26DB5A7D511B36B00508B9596A402F8EC63@sdgcaexch2.peregrine.com>

"My next problem is on nearly the same page (17) of the GoF book:
 class versus interface (== type) inheritance, but in the python context.

 In C++ and Eiffel inheritance means both interface and implementation
 inheritance and in Smalltalk inheritance means just implementation
inheritance.
 I don't know any of these languages."

   To understand this comment that they make you have to make sure you've
got the 
previous distinction they make between an objects class and an objects type.
An objects CLASS is an objects IMPLEMENTATION, while an objects TYPE is the 
INTERFACE or functions it has.  Note that an object has ONE implementation
but can
have MANY types.
   Unfortunately the best example I have is the C++/Java vs Smalltalk/Python
example.
Perhaps if you said what languages you were familiar with we might be able
to provide
more relevant examples.  But I hope you'll get enough to make sense of
this...

   In C++ (and perhaps to some extent) Java, a class is simply a template
for instantiating
objects.  Because of the static typing of C++/Java an objects class and type
are directly
tied together.  The class directly defines an objects type  You can't have
one w/o the other.
Meaning you can't try and call an operation on an object unless it was
explicitly defined when
the code was compiled.
  Example, if you want to have a piece of client code that performs the same
operation on two
different objects, those two different objects must both be subclasses of a
common parent.
This is class inheritance. C++ is going to force you to cast these objects
to a specific type
(the parent) before you can call the common method.  So now you have two
different objects of
the same type, but they are forced to share implementation since you had to
use class inheritance
to do it.  I'll avoid the technicalities of pure virtual/abstract classes
here, the point is the
same.
   Example again of the common widgets..in c++ the classes would be forced
to inherit from a 
common parent.  Class Widget would define Draw() operation.  Class Button
and Class Text would
be subclasses of Widget and would override Draw().  If you want to place
both of these objects into
a collection you'd have to specifically cast them to a Wdiget.

   With Smalltalk and Python things are much different.  Not only do you
have dynamic typing, but
classes are themselves objects that you can manipulate and call operations
on.  The dynamic typing
brings us back to the answers from your previous question.  Due to the
dynamic typing, two objects
only have to have the same operations defined to share types, REGARDLESS of
their implementations or
class relationships.  You can TRY and call any operation on ANY object, if
its there great all is well,
if not you'll get a runtime exception.
   Example again in Python of file-like object.  Assume your client code has
a collection of objects 
that it performs reads and writes to.  In python those objects can be
sockets, files, or stringIO's since
those object share the same type, ie all have read and write methods.
However none of those objects share any sort of implementation, meaning they
are not subclasses of a common ancestor.

  So class inheritance is standard classing and subclassing, whereas
interface inheritance is multiple
objects having the same type.  With interface inheritance you can swap in
and out different objects into
your client code regardless of the implementation/class that they are, and
all will work well.

For C++ interface inheritance REQUIRES class inheritance.  For
Python/Smalltalk, it does not.

-----Original Message-----
From: Eric Brunel [mailto:eric.brunel at pragmadev.com]
Sent: Friday, June 07, 2002 9:31 AM
To: python-list at python.org
Subject: Re: Program to an interface, not an implementation


Egbert Bouwman wrote:
> All your comments are crystal clear. Thanks.
> 
> My next problem is on nearly the same page (17) of the GoF book:
> class versus interface (== type) inheritance, but in the python context.
> 
> "In C++ and Eiffel inheritance means both interface and implementation
> inheritance" and
> "in Smalltalk inheritance means just implementation inheritance".
> I don't know any of these languages.
> 
> Now in my simplistic worldview, if you inherit from a class,
> you inherit everything, implementation and interface,
> because the implementation defines the interface as well.
> Even if all methods of the superclass only contain 'pass'.
> But I may be wrong.

I know C++ and I looked at SmallTalk a bit, and it seems the line between 
"implementation + interface" and "implementation only" inheritance is very 
thin... Apparently, it's quite directly related to strong typing and 
declarations. "Interface", in this context, seems to describe what's 
declared. So, as you noted in the beginning of your post, an "interface" is 
considered as a declared type. Since SmallTalk claims to have no types at 
all (which is also the case with Python...), they claim not to have such 
thing as an "interface", so obviously they cannot inherit it...

In fact, I don't see the point in this: as you noted, when a class inherits 
from another, it inherits everything, implementation and interface. So in 
practice, it's not really significant. As I understand it, the only 
practical consequence of this distinction *may* be in the following case:

class A:
  def m(self, i):
    pass
class B(A):
  def m(self, i, j):
    pass

In Python or SmallTalk, the preceding code is completely valid, so one 
could say that the "interface" of class A was not inherited by class B, 
since its method m has a different signature than A's method m.

Trying to do the same thing in interface-inheriting languages would not 
have the same effect at all: it would either be forbidden, or even have 
stranger consequences. In C++ or Java, it would declare a *second* method, 
also called m, that would be distinguished from the inherited one by its 
number of parameters or their type:

public class A {
  public int m(int i) {
    ...
  }
}
public class B inherits A {
  public int m(int i, int j) {
    ...
  }
}
...
o = new B();
o.m(1)      // Calls A.m
o.m(1, 2)   // Calls B.m

So, as you see, you probably can live without knowing anything about all 
this and still be able to do great stuff with whatever language you like 
:-).

However, if anybody knows a bit more than me about implementation/interface 
inheritance, I'd be glad to learn (I happen to teach OO sometimes and I'd 
like to be able to answer this question precisely if it ever gets answered 
to me ;-).
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com
-- 
http://mail.python.org/mailman/listinfo/python-list




From eppstein at ics.uci.edu  Tue Jun 25 12:49:20 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Tue, 25 Jun 2002 09:49:20 -0700
Subject: Suggestions for good programming practices?
References: <mailman.1024971992.26429.python-list@python.org> <roy-2CF124.07480625062002@reader2.panix.com> <mailman.1025012263.2719.python-list@python.org> <3D18868A.4050002@mxm.dk> <afa5sc$ptl$1@panix1.panix.com>
Message-ID: <eppstein-FD64AD.09492025062002@news.service.uci.edu>

In article <afa5sc$ptl$1 at panix1.panix.com>, aahz at pythoncraft.com (Aahz) 
wrote:

> At the very least, I think you ought to rewrite your example as
> 
> d = MyData(
>     firstName = 'Max M',
>     lastName = 'Rasmussen
>     )

I think rewriting it without the syntax error would be even better...

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From mgilfix at eecs.tufts.edu  Fri Jun  7 16:44:50 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Fri, 7 Jun 2002 16:44:50 -0400
Subject: Simple pychecker question
Message-ID: <20020607164450.G24428@eecs.tufts.edu>

  Pychecker doesn't seem to like named arguments (such as def
f(f='blah'), or sometimes, it's handy in __init__).  Why is that? Are
they considered bad style in some respects?  How about use in __init__
functions?

                    -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From deathtospam43423 at altavista.com  Sun Jun  9 03:15:05 2002
From: deathtospam43423 at altavista.com (netvegetable)
Date: 9 Jun 2002 07:15:05 GMT
Subject: Newby: Dictionaries as file objects?
Message-ID: <Xns9228AB206D06CAusYourStandingInIt@130.133.1.4>

Hi.

I want to be able to store a list on a file object, and I want each list 
item to be a dictionary, complete with its own set of keys and values. 

I know I can store a list in a file object, simply by writing each item to 
the file as a string , and making sure there's a "\n" character at the end 
of it. When I read the file object again, I merely need to open it as 
"open(object).readlines()" and hey presto, my list is back in order.

So is there a way to conveniently store a dictionary in each item as well?

Ta.

-- 
netvegetable at excite.com


From pyth at devel.trillke.net  Sat Jun  1 13:34:02 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 1 Jun 2002 19:34:02 +0200
Subject: Generating unique numbers?
In-Reply-To: <adav4g$o2b$1@panix1.panix.com>; from aahz@pythoncraft.com on Sat, Jun 01, 2002 at 01:09:36PM -0400
References: <3CF7EBC8.4080802@verio.net> <3CF8FC1C.7060903@lindbergs.org> <adav4g$o2b$1@panix1.panix.com>
Message-ID: <20020601193402.A30380@prim.han.de>

Aahz wrote:
> In article <3CF8FC1C.7060903 at lindbergs.org>, VanL  <news at lindbergs.org> wrote:
> >
> >The reason why I am wanting keys that are more likely to be globally
> >unique is because I will be building in hooks to make this outliner
> >network transparent (using ZEO), making it possible for multiple
> >outliners to be used together as a real-time collaborative tool.  I
> >want to be able to be able to have people merge stuctures that people
> >have made in the standalone instance into a shared instance without
> >tromping on another nodes data.
> >
> >Based on what people have said, I was thinking of using some
> >combination of DateTime.gmticks and the id of the obj, making it (I
> >think) extremely unlikely that there would be a collision.
> 
> id() is very, very bad for any kind of persistent ID; I still think it's
> at least very bad even if you add in any kind of time.
> 
> No, if you want to trade IDs across a network, you need GUIDs (DCE
> UUIDs).  There's just no way around that.  It's not the most perfect
> system in the world, but it's better than any other generally available
> scheme.

do you know of a good implementation of DCE UUIDs in python and/or C?

    holger




From maxm at mxm.dk  Sun Jun 23 05:36:22 2002
From: maxm at mxm.dk (Max M)
Date: Sun, 23 Jun 2002 11:36:22 +0200
Subject: Web templating/db tool with best designer/coder separation?
References: <af2gh6$2c2$0@216.39.172.122> <slrnah9mj3.rdk.jon+usenet@snowy.squish.net>
Message-ID: <3D159696.9010209@mxm.dk>

Jon Ribbens wrote:
> In article <af2gh6$2c2$0 at 216.39.172.122>, Bengt Richter wrote:
> 
> Take a look at http://jonpy.sourceforge.net/ . It almost completely
> separates code and HTML so that the designer and the coder do not
> interfere with each other.

I do understand that argument, but I almost certainly don't agree that 
it is of much use.

In smaller one-of systems it makes sense, but in an app-server like Zope 
where the idea is that different components can be joined together to 
create a bigger application, there has to be stronger rules for layout 
and design of the components so that they will automatically fit into 
the bigger whole.

Page template languages is the wrong approach in this regards.

The development of the Cmf and Plone are good examples of this wrongfull 
approach. They make it easy to create one single type of application, 
but makes it very hard to re-use their components to build other types 
of applications.

And I believe that the reason is that they are using pagetemplate 
languages to bind it together, which leads to this kind of design.

I my experience, tools which separates logic from view in the 
"pagetemplate" approach is doing it wrong.

The layout may be separated, but reuse of code/componentes greatly 
suffers from this approach.

regards Max M



From SBrunning at trisystems.co.uk  Tue Jun 25 09:31:48 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 25 Jun 2002 14:31:48 +0100
Subject: Suggestions for good programming practices?
Message-ID: <31575A892FF6D1118F5800600846864DCBD432@intrepid>

> From:	Emile van Sebille [SMTP:emile at fenx.com]
> Aahz
> > In Python 2.1.2 or higher, fire up the interactive interpreter and
> type
> > "import this".
> 
> Except ActiveState's
> 
> F:\Python22as>python
> Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
> Type "help", "copyright", "credits" or "license" for more information.
> >>> import this
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> ImportError: No module named this
 
It isn't in the 2.2 series until 2.2.1, IIRC.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.




From BPettersen at NAREX.com  Thu Jun 20 19:13:10 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 20 Jun 2002 17:13:10 -0600
Subject: Are generators in nested functions possible?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E201922151A9@admin56.narex.com>

I've got the following code to generate the indicies of all nearest
neighbours (span levels) in a multi dimensional qube:

def generateIndexes(length=5, span=1):
    spanvalues = range(-span, span+1)
    
    def genInd(res, length):
        if length == 1:
            for val in spanvalues:
                yield res + [val]
        else:
            for i in range(length):
                for val in spanvalues:
                    genInd(res + [val], length-1)
    
    return genInd([], length)

However, when I try to use it it doesn't seem to return anything:

  >>> for i in generateIndexes(3):
  ...   print i
  ...
  >>>

I tried getting the iterator (or is it generator?) directly and calling
next() on it, and it seems like it raises a StopIteration exception
immediately:

  >>> g = generateIndexes(3)
  >>> g.next()
  Traceback (most recent call last):
    File "<stdin>", line 1, in ?
  StopIteration
  >>>

Can anyone help me understand what is going on here?

-- bjorn




From tdelaney at avaya.com  Wed Jun 12 00:51:07 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 12 Jun 2002 14:51:07 +1000
Subject: Sorting list (maybe stupid)
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A37F@natasha.auslabs.avaya.com>

> From: Tim Peters [mailto:tim.one at comcast.net]
> 
> [Delaney, Timothy]
> > ...
> > Personally, I think the implementation should be documented to
> > use a stable sort - presumably mergesort (stable, minimal number of
> > comparisons which is very important in Python).
> 
> Write a mergesort as fast and memory-efficient as the sort 
> we've got and we'll consider it.  I wasn't able to ...
> you'll find that mergesort doesn't have enough of an
> advantage there to make up for the speed it loses due to greater data
> movement.

I'm not willing in the slightest to dispute this ...

My preference for a stable sort as default is that it is the "least
surprising" result. But I don't see it as a wart that the implementation is
not stable.

Tim Delaney




From skip at pobox.com  Thu Jun 13 03:31:49 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 13 Jun 2002 02:31:49 -0500
Subject: Extension objects in Python 1.5 and 2.2
In-Reply-To: <20020613071813.GA16764@strakt.com>
References: <20020613071813.GA16764@strakt.com>
Message-ID: <15624.19045.700125.215173@12-248-41-177.client.attbi.com>

    Martin> How do I write extension objects so that they work in both
    Martin> Python 1.5 and the 2.x series? 

Michael Hudson has been working on a new include file named pymemcompat.h.
Since it's only used for backward compatibility it's in the Misc directory.
You can grab it from CVS:

    http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/dist/src/Misc/pymemcompat.h

The idea is that you should be able to include it and program to the 2.3
memory api in code that runs under 1.5.2 or newer versions of the
interpreter.

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html






From usenet-jun02 at puzzling.org  Sun Jun  9 23:37:16 2002
From: usenet-jun02 at puzzling.org (Mary Gardiner)
Date: Mon, 10 Jun 2002 03:37:16 GMT
Subject: Sydney Australia: Python Interest Group
Message-ID: <M9VM8.3341$06.443812@nasal.pacific.net.au>

Hi,

A Python Interest Group has just started in Sydney under the banner of
the Sydney Linux Users Group (SLUG).

It's not restricted to Python under Linux and it is not restricted to
SLUG members or people involved in SLUG.

It meets on the third Monday of every month. Our second meeting is on
June 17th at 7:00pm at the University of Technology, Sydney, room
2.3.16.

The announcement is at
http://lists.slug.org.au/archives/announce/2002/06/msg00003.html

We don't have a webpage or mailing list independent of SLUG at present,
if you're interested in further announcements you should watch
http://www.slug.org.au/ or subscribe to SLUG's low volume announce list
at http://lists.slug.org.au/listinfo/announce

-Mary.

-- 
Mary Gardiner
<usenet at puzzling.org>


From garrett at bgb.cc  Mon Jun 10 14:51:43 2002
From: garrett at bgb.cc (Don Garrett)
Date: Mon, 10 Jun 2002 18:51:43 GMT
Subject: Newbie question communicating between 2 progs
References: <3D046BF0.DC36542C@raplounge.de> <3D04A938.627EE705@engcorp.com> <3D04E8ED.C0708C2B@raplounge.de>
Message-ID: <3D04F4EF.9040800@bgb.cc>

Ben Thursfield wrote:
> 
> Peter Hansen schrieb:
> 
> 
>>The question is ambiguous.  When you say "frames", the only thing
>>that comes to mind is that you are running Python on client-side
>>Internet Explorer under Windows, as scripts in two different parts
>>of a window (the frames) simultaneously.  I think you must have
>>meant something different however.
>>
>>Could you please describe more about your environment?  The
>>term "frame" can mean a number of things.
> 
> 
> That's right. I am running 2 cgi-scripts in two different frames of the
> browser, the one should only print an array into the frame and the other one
> communicates with a server. I decided to split the programm up because i want
> to refresh the printing frame every second. The main frame changes the array
> the second frame prints it out. My host has Python installed, so the
> #usr/bin/python method works fine.
> So the question is how can the programm responsible for printing out the array
> access the array defined in the other programm.

   This is a suprisingly hard problem.

   The two programs are only running while the frame they handle is being 
generated. They shutdown and start up on each refresh request. I'm assuming 
you aren't using some form of server push, which has it's own problems. They 
don't save any state at all between refreshes of the same page. This means you 
have to solve the problem of having a program communicate it's state to 
itself! The good news is that when the program can communicate to itself, you 
can use the same mechanism to communicate with the other frame.

   However, the best answer to your question is to use cookies. The contents 
should be visible to all pages on the local site (you can control this) so 
they will allow you scripts to communicate with themselves and each other. You 
can store a fair amount in one (I'm not sure what the size limit is), and it 
can be binary.

   A few warnings.

   Some people disable cookies. This can totally break your site.

   All cookies are always uploaded as part of every page request everywhere 
they are visible. This means that you shouldn't make them too large or you 
will slow your site. If they are big, try to restrict cookie visibility to 
just the pages that really need them. You can specify where the cookies are 
visible when you create them. You can even store the state info in a database 
and use the cookie as a key to retreive it.

   When it comes to security, you cannot trust the contents of the cookie. The 
contents need to be validated the same way the values from a form would be. A 
tricky black hat can force the cookie contents to be anything they want.

   Two last bits of advice from an old web developer (well as old as a web 
developer can be).

   Try to avoid frames. They cause a lot of problems for both you and for your 
users. For example, they often break cross-browser, they can break bookmarks, 
and they can break printing.

   Try to avoid refreshes, especially frequent ones. Someone with a poor link 
will spend more time looking at an empty page that hasn't loaded than they 
will looking at your content.

> I hope i described the problem better now.
> 
> Ben
> 

-- 
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc



From robin at jessikat.fsnet.co.uk  Sun Jun 30 18:38:15 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 30 Jun 2002 23:38:15 +0100
Subject: yield equivalent in C/JavaScript?
In-Reply-To: <20020630203856.GA59010@hishome.net>
References: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk>
 <20020630180139.GA38821@hishome.net> <AS+BzLARS2H9EwUy@jessikat.fsnet.co.uk>
 <20020630203856.GA59010@hishome.net>
Message-ID: <08$soXAXh4H9EwQ2@jessikat.fsnet.co.uk>

In message <20020630203856.GA59010 at hishome.net>, Oren Tirosh <oren-py-
l at hishome.net> writes
>On Sun, Jun 30, 2002 at 09:05:37PM +0100, Robin Becker wrote:
>> In message <20020630180139.GA38821 at hishome.net>, Oren Tirosh <oren-py-
>> l at hishome.net> writes
>> 
>> very smart! That Duff did it Fortran as I recall.
>
>I don't think there is switch with case fallthrough in fortran.  Duff's 
>device is most definitely a unique abuse of the C language.
>
>http://www.lysator.liu.se/c/duffs-device.html
>
>       Oren
>
I stand corrected, I thought this was Ian Duff who did marvellous things
with sparse matrices in Fortran.
-- 
Robin Becker




From boud at valdyas.org  Sat Jun  1 18:36:47 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Sun, 02 Jun 2002 00:36:47 +0200
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org> <3CF8E382.3A44F0DA@engcorp.com> <83ptzavk14.fsf@panacea.canonical.org> <3CF9408E.6015D2A@engcorp.com>
Message-ID: <3cf94ca7$0$31235$e4fe514c@dreader1.news.xs4all.nl>

Peter Hansen wrote:
> 
> Other options might include making sure modules are not loaded
> until they are needed, or loading them on a separate thread just
> after the main program starts.
> 

Convincing your users that they need to preload the libs might
work too -- for instance by creating a 'server' component that
should be started when the machine is booted that loads the lib.

> None of which seems to help the OP.  I'm curious why the Linux
> implementation is so much slower than the Win32.  My own machine
> is _not_ fast!  (It's only a 233 MHz Pentium MMX.)
> 

wxPython has to load gtk, probably, and the bindings. The underlying
gui library is already present in memory on a windows machine, leaving
only the bindings to be loaded.

I've also noticed that Python loads its own bytecode files rather
more slowly than other files. So any pixmaps shouldn't be put in the
code, but loaded separately. That can make quite a difference.

-- 
Boudewijn Rempt | http://www.valdyas.org


From hesterloli at hotmail.com  Wed Jun 19 04:17:13 2002
From: hesterloli at hotmail.com (George Hester)
Date: Wed, 19 Jun 2002 08:17:13 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <D7NP8.46301$GY.14487785@twister.nyroc.rr.com> <slrnagvdp4.1sg.gerhard@lilith.my-fqdn.de>
Message-ID: <d6XP8.47924$GY.15257912@twister.nyroc.rr.com>

You don't even have the balls to include a valid email address when you write to
me.

--
George Hester
_________________________________
"Gerhard H?ring" <gerhard at bigfoot.de> wrote in message
news:slrnagvdp4.1sg.gerhard at lilith.my-fqdn.de...
> George Hester wrote in comp.lang.python:
> > I just cannot believe what I ran into in this newsgroup. [...]
>
>
>                                                 .:\:/:.
>              +-------------------+             .:\:\:/:/:.
>              |   PLEASE DO NOT   |            :.:\:\:/:/:.:
>              |  FEED THE TROLLS  |           :=.' -   - '.=:
>              |                   |           '=(\ 9   9 /)='
>              |   Thank you,      |              (  (_)  )
>              |       Management  |              /`-vvv-'\
>              +-------------------+             /         \
>                      |  |        @@@          / /|,,,,,|\ \
>                      |  |        @@@         /_//  /^\  \\_\
>        @x@@x@        |  |         |/         WW(  (   )  )WW
>        \||||/        |  |        \|           __\,,\ /,,/__
>         \||/         |  |         |      jgs (______Y______)
>     /\/\/\/\/\/\/\/\//\/\\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
> --
> mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
> web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
> public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
> reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From cbrown at metservice.com  Tue Jun  4 21:47:08 2002
From: cbrown at metservice.com (Colin Brown)
Date: Wed, 5 Jun 2002 13:47:08 +1200
Subject: manipulating child processes under Windows
References: <lcptz7zw0m.fsf@gaffa.mit.edu>
Message-ID: <3cfd6d6d$1@news.nz.asiaonline.net>

"Douglas Alan" <nessus at mit.edu> wrote in message
news:lcptz7zw0m.fsf at gaffa.mit.edu...
>
> (2) What's the best way for me to solve the problem.  I.e., what's the
>     best way to get my Python script to wait for the real program to
>     terminate?
>
> |>oug

To locate a grandchild process and get its pid and handle on Win2000 you
could examine the output of the freeware program "pslist -t"
(www.sysinternals.com). After that use win32 routines as usual.

Colin Brown
PyNZ




From chris_mk at hotmail.com  Tue Jun  4 20:31:36 2002
From: chris_mk at hotmail.com (Christopher)
Date: 4 Jun 2002 17:31:36 -0700
Subject: Problem with urllib.urlopen()
References: <cc14131d.0206031445.5154de31@posting.google.com> <cc14131d.0206032128.5b2efcf6@posting.google.com>
Message-ID: <cc14131d.0206041631.157f11d1@posting.google.com>

Okay okay, here are links that 1) are short, 2) should work, and 3)
show my problem.

link 1:
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=PureSearch&db=PubMed&details_term=%28%22glucagon%22%5BMeSH%20Terms%5D%20OR%20glucagon%5BText%20Word%5D%29

link2:

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Display&db=PubMed&details_term=%28%22glucagon%22%5BMeSH%20Terms%5D%20OR%20glucagon%5BText%20Word%5D%29&dopt=XML&query_key=1

Look at link2 in your browser, then look at what urllib.urlopen or
.urlretrieve get.  Thanks.

Chris


From vvainio at tp.spt.fi  Thu Jun 13 08:38:32 2002
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 13 Jun 2002 05:38:32 -0700
Subject: Wanted:  Potential Python Authors
References: <ae7gke$mhb$1@ccis12.baylor.edu> 	<db3fgu40sbnt3pno5v1i2h9lnmpcoti2bk@4ax.com> <mailman.1023922824.32368.python-list@python.org>
Message-ID: <ad496f8.0206130438.2d9e77e8@posting.google.com>

Ian Bicking <ianb at colorstudy.com> wrote in message news:<mailman.1023922824.32368.python-list at python.org>...

> want to learn to program).  An expert programmer has already created an
> effective internal model and metaphor for programming, which is not
> shared by technical writers who are not experienced programmers (or
> experienced in the appropriate domain).

This reminds me of many of the C++ books, where the author 

1) Learned C
2) Looked up the class syntax and iostreams
3) Wrote a C++ book


From loewis at informatik.hu-berlin.de  Thu Jun  6 04:35:21 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 06 Jun 2002 10:35:21 +0200
Subject: Detecting OS and Window Manager
References: <mailman.1023309760.22815.python-list@python.org> <uft6liipd6m60e@corp.supernews.com>
Message-ID: <j4y9ds6bzq.fsf@informatik.hu-berlin.de>

"Andrew Wilkinson" <ajw126 at york.ac.uk> writes:

> Basically I want to have a program that hides itself in the taskbar
> of whatever the user is running. Unfortunately there isn't (or at
> least I don't know of) a portable way of doing this. I guess I'll
> just have to mess around with looking for the environment variables
> and then code it up for KDE/Gnome/whatever.

So you really don't care that much if the user is running KDE or
Gnome, but whether the gnome-panel is running or <put in the name of
the KDE taskbar application>. This is different from "running Gnome",
since you can combine the components of either package with each
other, or, say, run gnome without running a panel.

Regards,
Martin



From tim.one at comcast.net  Tue Jun 11 21:22:57 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 11 Jun 2002 21:22:57 -0400
Subject: lists and sequences
In-Reply-To: <200206120026.g5C0QiD24559@localhost.localdomain>
Message-ID: <LNBBLJKPBEHFEDALKOLCCEECPMAA.tim.one@comcast.net>

[Collin Monahan]
>>> E.g. when does n^2 time happen using it?

[Tim]
>> alist.insert(0, newthing) physically moves len(alist) elements.  Ditto

[Dave Reed]
> Just to be clear, these are order n, not order n^2.

"in a loop of appropriate duration" was left to the reader's imagination,
assuming they had one <wink>.  The obvious intent of Collin's question was
to discover which elementary operations required worse than O(1) time.  His
"e.g." was for the benefit of those who couldn't decipher his immediately
preceding "What type of situations should be avoided when using it?".
Knowing that an individual insertion or deletion "at the wrong end" can be
expensive is responsive to what he really wanted to know.

modern-bots-don't-take-human-words-at-face-value-ly y'rs  - tim





From sketerpot at chase3000.com  Tue Jun 11 11:54:55 2002
From: sketerpot at chase3000.com (Peter Scott)
Date: 11 Jun 2002 08:54:55 -0700
Subject: Cross platform spawning
Message-ID: <659f82ff.0206110754.775490d9@posting.google.com>

I am trying to make a script spawn a seperate process, like you would
do like this on unix:

os.system('python myscript.py myargument &')

But the & at the end (to make a separate process) doesn't work on
Windows. I want the script to run on both Linux (*BSD, whatever) and
Windows. I didn't want to have to find out the path to the python
interpreter, so os.spawnlp (with no blocking) sounded perfect until I
saw that it wasn't available on Windows.

Does anyone out there know what I can do in this situation?


From metaliu at yahoo.com  Fri Jun  7 20:08:14 2002
From: metaliu at yahoo.com (Bill)
Date: Fri, 7 Jun 2002 17:08:14 -0700
Subject: Stretching Images
Message-ID: <adrhtf$op$1@nntp1.jpl.nasa.gov>

I'm currently using Tkinter for GUI development.  I have a Tkinter label
which is showing an image.  Is there a way for me to stretch the image to a
specific size?

I tried creating a Frame and specifying its width and height.  The I tried
putting the label in the Frame and packing it using expand=YES and
fill=BOTH, but the Frame gets resized regardless.  I also ran the method
.grid_propagate(0) on the Frame to make the Frame keep its size attributes,
but it keeps resizing regardless.  Also, when I resize the Frame manually
with the mouse afterwards, the label with the image just centers itself in
the Frame.  It doesn't stretch the image to fill the Frame like I want.  Any
suggestions on how I can get the proper effect I'm looking for?

Thanks ahead of time,
Bill




From BPettersen at NAREX.com  Thu Jun 20 19:51:14 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 20 Jun 2002 17:51:14 -0600
Subject: Are generators in nested functions possible?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158EEE@admin56.narex.com>

> From: David Eppstein [mailto:eppstein at ics.uci.edu] 
> 
> In article <mailman.1024614876.11270.python-list at python.org>,
>  "Bjorn Pettersen" <BPettersen at NAREX.com> wrote:
> 
> > def generateIndexes(length=5, span=1):
> >     spanvalues = range(-span, span+1)
> >     
> >     def genInd(res, length):
> >         if length == 1:
> >             for val in spanvalues:
> >                 yield res + [val]
> >         else:
> >             for i in range(length):
> >                 for val in spanvalues:
> >                     genInd(res + [val], length-1)
>                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> This is a no-op.  Did you mean
>                       for x in genInd(res+[val],length-1): return x ?

Thanks! I knew it had to be something simple (got your other post about
using yield instead of return).

-- bjorn




From Chris.Barker at noaa.gov  Wed Jun 26 13:08:39 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Wed, 26 Jun 2002 10:08:39 -0700
Subject: Reference Count Confusion
References: <mailman.1025037754.18087.python-list@python.org>
Message-ID: <3D19F517.290A9CDE@noaa.gov>


Jeff Smith wrote:

> import Numeric
> import foo
> 
> a = Numeric.zeros(1000000).astype('f')
> b = Numeric.ones(1000000).astype('f')

A little nit unrelated to your question:

use: a = Numeric.zeros(1000000,'f')

instead. The result is the same, but it creates the Float array
directly, rather than creating an integer array, and then making a float
array from that.

>   return Py_BuildValue("O", c);

You might want to try:

return Py_BuildValue("N", c);

I'm a little fuzzy on al this too, but that works for me in a similar
situation. From Extending and Embedding the Python Interpreter:

"N" (object) [PyObject *] 
      Same as "O", except it doesn't increment the reference count on
the object. Useful when the object is created by a call to an object
constructor in the argument list. 

Good luck,

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From tatebll at aol.com  Fri Jun  7 09:14:34 2002
From: tatebll at aol.com (Bill Tate)
Date: 7 Jun 2002 06:14:34 -0700
Subject: Creating Dynamic Web Pages (ZOPE)
References: <46ca81a0.0206062016.789d3af6@posting.google.com>
Message-ID: <cb4ba455.0206070514.335e2cb2@posting.google.com>

hwcowan at hotmail.com (Hugh Cowan) wrote in message news:<46ca81a0.0206062016.789d3af6 at posting.google.com>...

[snip]
> My scenario is typical where I need to put together a company Intranet
> with a few front end Web-Pages where users can select information,
> various links, and enter search criteria.  The information is then
> retrieved from the MS-SQL Server and the results are sent back to the
> user.  It's nothing fancy, no shopping carts, or remembering sessions
> states, etc.. just really a simple front-end interface for
> non-technical people to easily retrieve information.
>  
> I don't know if ZOPE would fit my situation or not.  I know that it is
> / can be used to create complex E-commerce sites, but I am wondering
> if it is overkill for what I need done.  Would I be better off trying
> to find some sort of RAD Web Tool instead (anyone have any good
> suggestions) ?

I think its safe to say that there are any number of tools that could
fit the bill here including Zope.  There are aspects of zope that can
make your life much easier (e.g., zsql methods) but it does take some
time to get used to it and everything it offers.  Several recent books
on the subject I think would be very helpful to you in this regard. 
If you go to zope.org, there is some information on using WYSIWYG
tools such as Dreamweaver with zope that may also be of interest to
you as well.

I think zope works bests for you over the long haul where it is
important to keep things well organized and well structured.  The
ability to collaborate amongst individuals is also a great strength. 
If these issues are important to you, then I would say that Zope would
be an excellent choice.

Good luck!!!


From h.baumgartl at chello.NOSPAM.nl  Thu Jun 20 19:50:49 2002
From: h.baumgartl at chello.NOSPAM.nl (Henry Baumgartl)
Date: Thu, 20 Jun 2002 23:50:49 GMT
Subject: application development - separating content from function
References: <yroQ8.260949$i6.17986369@amsnews02.chello.com> <Xns9233ED737D952cliechtigmxnet@62.2.16.82>
Message-ID: <tTtQ8.266196$i6.18319309@amsnews02.chello.com>

"Chris Liechti" <cliechti at gmx.net> schreef in bericht
news:Xns9233ED737D952cliechtigmxnet at 62.2.16.82...

> tkinter or wxWindows come to mind, they should run right out of he box
> without special efforts of your side on those platforms, they're portable.

tkinter is not quite enough, tkinter + Tix would be an option, but I'm not
too sure about Tix on a Mac
wxWindows - same problem as far as wxPython on a Mac is concerned. Although
I understand that a port is in the making.
The question is then; Go for certain on tkinter and invest quite a lot of
effort in the future to deal with the application complexity? Go for
wxPython,which is very much 'under construction', accepting other
dependencies (i.e. mxDateTime) that are not really needed, and more or less
join the (fascinating) wx* effort?

The same goes, more or less, for design aspects of manipulating generated
data, and conversion of this data to information and knowledge for users.
Pickling flat files is flexible and fun, but in terms of quantitive
capacity, the ceiling is soon reached.
XML is still far from mature and, the way things look, may well become an
unwielding giant, generating far too much overhead to become really useful
for complex applications.
SQL - how much longer do we have to suffer this non-OO solution (if only it
didn't work so well ;-))

> the standard module ConfigParser provides .ini style configuaration. this
> is easy to save window positions, user settings in general and much more.

Will certainly be very useful and used to its limit. Can't do it all
though...

> if you need real plugin with dynamic loaded functions and classes:
> import+reload, module imp or execfile() can be used to load plugins (.py
> files) at runtime.
>
> in that case you migh define a simple interface like making it mandatory
> for an object to inherit from a special base class and or the file must
> provide an init() function to register itself in the main app.

Maybe it's unnecessary, but I have set myself the goal to minimize the use
of variables. Considering the enormous flexibility Python allows each
variable, and the extended functionality associated with the same, one can
only assume that too many variables do not exactly equal optimised use of
resources.
Even considered writing scripts that will generate the personalised modules
and use these in the actual user application. Does that make sense, or am I
going a little over the top there?

> > 3. And, last but not least ;-) the application needs to be
> > multi-lingual.

Following the advice of earlier respondents to my post (reduce complexity to
manageable proportions), and your suggestions, I feel that this issue may
wait until a later stage.

Regards, Henry Baumgartl
Dynamic Knowledge Network

*Please remove .NOSPAM from my e-mail address when abswering me directly.*




From twister at raplounge.de  Mon Jun 10 05:05:52 2002
From: twister at raplounge.de (Ben Thursfield)
Date: Mon, 10 Jun 2002 11:05:52 +0200
Subject: Newbie question communicating between 2 progs
Message-ID: <3D046BF0.DC36542C@raplounge.de>

Hi!

I am running two python scripts in 2 different frames simultanously, and
one of the programm needs to access a list defined in the other
programm. How do i manage to access that list?

I am thankful for any help.

Ben



From phr-n2002b at NOSPAMnightsong.com  Sun Jun 30 06:18:03 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 30 Jun 2002 03:18:03 -0700
Subject: yield equivalent in C/JavaScript?
References: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk> <7xy9cxhlvx.fsf@ruckus.brouhaha.com> <w3xdKUAuljH9EwgH@jessikat.fsnet.co.uk> <7xit41i7i9.fsf@ruckus.brouhaha.com> <14ms5VAygsH9Eww2@jessikat.fsnet.co.uk>
Message-ID: <7xelepnk84.fsf@ruckus.brouhaha.com>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:
> However the content.gz data has to be exactly compatible with gzip. My
> experiments with Python-2.1/zlib on freeBSD give me a compressed string
> that's slightly too long for some reason. As an example gzip --> 14099
> bytes while zlib.compress-->15008 bytes.
> 
> Anyone know what I need to do to make zlib work properly in this
> context? Or should I really be using the gzip module?

You're probably better off using mod_gzip (an apache module that
compresses the output automatically) rather than trying to compress
from Python.  Note that gzip encoding doesn't work so well in some IE
versions.


From emile at fenx.com  Tue Jun 11 23:17:21 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 12 Jun 2002 03:17:21 GMT
Subject: Filling a space...
References: <ae4v9h$3qg34$9@ID-108351.news.dfncis.de>
Message-ID: <53zN8.164672$cQ3.4444@sccrnsc01>

"Thor" <thor__00 at yahoo.com> wrote in message
news:ae4v9h$3qg34$9 at ID-108351.news.dfncis.de...
> Having this tupe: (2,3,4), which one would be the msot efficient way
to get
> this (parentheses ommited):
>
> 0,0,0  0,0,1  0,0,2  0,0,3
> 0,1,0  0,1,1  0,1,2  0,1,3
> 0,2,0  0,2,1  0,2,2  0,2,3
> 1,0,0  1,0,1  1,0,2  1,0,3....
>

>>> t = (2,3,4)
>>> [(a,b,c) for a in range(t[0]) for b in range(t[1]) for c in
range(t[2])]

I think I needed this once for a variable length tuple and parameterized
it further...

--

Emile van Sebille
emile at fenx.com

---------



From pyth at devel.trillke.net  Mon Jun 10 21:41:34 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 11 Jun 2002 03:41:34 +0200
Subject: Why does Python mix OO concepts and non OO concepts for operation s  on basic types?
In-Reply-To: <3D054A98.3090708@bgb.cc>; from garrett@bgb.cc on Tue, Jun 11, 2002 at 12:57:11AM +0000
References: <mailman.1022097199.9970.python-list@python.org> <3CEC1506.B0DA1787@earthlink.net> <3D054A98.3090708@bgb.cc>
Message-ID: <20020611034134.O6609@prim.han.de>

Don Garrett wrote:
> Hans Nowak wrote:
> > Besides, while Python is object-oriented, that isn't the only
> > paradigm in the language. People coming from a functional
> > background may find the len() function more natural than a
> > method.
> > 
> > YMMV,
> > 
> 
>    That's one thing I don't get. Why wasn't len() added as a member to the 
> various types during the recent rework that added so many other members to the 
> native types?

maybe duplicating '__len__' as 'len' is too much, well, duplication?

    holger




From pereira at cis.upenn.edu  Mon Jun  3 21:28:58 2002
From: pereira at cis.upenn.edu (Fernando Pereira)
Date: Mon, 03 Jun 2002 21:28:58 -0400
Subject: Graph distances, how can I speed it up?
References: <33803989.0206030235.4134d33e@posting.google.com> <d571ae53.0206030815.2e52aac@posting.google.com> <B92114F6.B79E%pereira@cis.upenn.edu> <eppstein-347A24.10122403062002@news.service.uci.edu>
Message-ID: <B921901A.B874%pereira@cis.upenn.edu>

On 6/3/02 1:12 PM, in article
eppstein-347A24.10122403062002 at news.service.uci.edu, "David Eppstein"
<eppstein at ics.uci.edu> wrote:

> In article <B92114F6.B79E%pereira at cis.upenn.edu>,
> Fernando Pereira <pereira at cis.upenn.edu> wrote:
> 
>> Dijkstra's algorithm is not a good choice here, because distances from all
>> sources appear to be needed, not from a single source.
> 
> Dijkstra is not a bad choice for all pairs shortest paths -- just repeat
> it for each source in the graph.
Think about the redundant traversal of some sub-graphs compared with
Floyd-Warshall. Worst-case bounds are not so interesting here, as we are
dealing with very particular graphs (WordNet). There's a huge literature on
all-pairs shortest-distance algorithms that may be worth looking at, for
instance <http://robotics.stanford.edu/~koller/papers/path.ps>.

-- F



From chris_mk at hotmail.com  Fri Jun 14 17:00:26 2002
From: chris_mk at hotmail.com (Christopher)
Date: 14 Jun 2002 14:00:26 -0700
Subject: PythonWin Problem!
References: <1a52c7f1.0206131455.6df50cf8@posting.google.com> <3D0973C5.60407@skippinet.com.au>
Message-ID: <cc14131d.0206141300.693aa7ae@posting.google.com>

Ethan,

>From the sound of it, I think Mark's analysis may be correct.  I ran
into the EXACT same problem a while back (which I was able to fix
using 'Move').  Incidently, I'm not sure what exactly happened to
cause my problem, all I can say is that the program was maximized when
I ran a script with a typo (it used a try..except: loop when I meant
try...except IndexError: and it locked up big-time).  When I started
it again, same as you, appeared in taskbar and tray but not on screen.
 I, however, was able to right-click, Maximize and that's how I
figured it was off the screen.  Good luck.

Chris


Mark Hammond <mhammond at skippinet.com.au> wrote in message news:<3D0973C5.60407 at skippinet.com.au>...
> Ethan wrote:
> > I have been using ActiveState's port of Python for a while. PythonWin
> > has been very helpful. Since a couple of days PythonWin developed a
> > problem. When I start it, it appears in the taskbar and tray but does
> > not appear on the screen.
> > 
> > My Win98 did recently develop a problem which prevents me from getting
> > the right click menu on the taskbar entries. This was secondary to
> > some registry cleaning I did. I am not sure if this is related to
> > Pythonwin problem but all others apps are working fine. I do get the
> > menu for the tray icon and the "Close Program" window does not say
> > "not responding"
> > 
> > I uninstalled Python - deleted the Python21 folder and reinstalled. It
> > did not help. Any suggestions? I really prefer PythonWin over IDLE.
> 
> Pythonwin has probably somehow remembered it's screen location as off 
> the screen.
> 
> Try deleting the "HKLM\Software\Python 2.1\Python for Win32" key from 
> the registry, and it should come back.
> 
> Alternatively, from the task-bar select "Move", then use the arrow keys 
> to try and bring it back on screen (presumably with the "Left" and "Up" 
> keys)
> 
> Mark.


From leonw at gmx.net  Mon Jun 17 12:54:50 2002
From: leonw at gmx.net (Leon Widdershoven)
Date: Mon, 17 Jun 2002 18:54:50 +0200
Subject: Numeric and PIL: 1D array from 2D-tuple
References: <3D0479FA.CEFFFAC2@ipm.fhg.de>
Message-ID: <3D0E145A.1D21E730@gmx.net>

Hi,

you probably wish to use something like:
text = Image.tostring()
arr  = Numeric.fromstring( text )

with a possible Numeric.reshape() if you wish to operate
on something else than a 1D array.

Regards,
Leon

Markus von Ehr wrote:
> 
> Hi,
> 
> I have a 2D tuple (an Image) and want to display it using ImageTk
> in PIL.
> Numeric.choose throws me an error:
> 
> Traceback (innermost last):
>   File "<pyshell#14>", line 1, in ?
>     res = Numeric.choose(MyImage, mask)
> ValueError: invalid entry in choice array
> 
> Anybody knows how to get a 1D-array to construct the
> ImageTk.PhotoImage?
> 
> Thanks for any comments,
> 
> Markus
> 
> code:
> 
> MyImage = self.cam.Image # MyImage is a 2D-tuple
> 
> mask = [1 for i in MyImage]
> 
> MyImage = Numeric.choose(MyImage, mask)
> 
> new_image = MyImage.tostring()
> self.im.fromstring(new_image, "raw", "RGBX", 0, -1)
> 
> self.photoimage = ImageTk.PhotoImage(self.im)


From rkr1410 at poczta.onet.pl  Mon Jun 24 04:05:29 2002
From: rkr1410 at poczta.onet.pl (hellboy)
Date: Mon, 24 Jun 2002 10:05:29 +0200
Subject: Newbie question (anyone programming with curses?)
References: <af4ovd$1bt3$1@pingwin.acn.pl>
Message-ID: <af6jrc$gm1$1@pingwin.acn.pl>

Thanks everyone, you helped me much.

cheers




From boud at valdyas.org  Tue Jun  4 06:26:54 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Tue, 04 Jun 2002 12:26:54 +0200
Subject: Looking for a list subclassing example...
References: <adi4dv$118tko$1@ID-146704.news.dfncis.de>
Message-ID: <3cfc994e$0$3851$e4fe514c@dreader4.news.xs4all.nl>

Shagshag13 wrote:

> Hello,
> 
> I'm looking for a list subclassing example, and i can't find it...
> 
> Do you have one to show ?
> 

I subclass lists in drift: http://www.valdyas.org/python/drift.html.

-- 
Boudewijn Rempt | http://www.valdyas.org


From martin at v.loewis.de  Fri Jun 14 14:51:18 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 14 Jun 2002 20:51:18 +0200
Subject: a possibly foolish question about slices
References: <yu99sn3prbqm.fsf@europa.research.att.com> <aedd8k$brf$1@scotsman.ed.ac.uk>
Message-ID: <m3znxxu21l.fsf@mira.informatik.hu-berlin.de>

eddie at holyrood.ed.ac.uk (Eddie Corns) writes:

> It frequently bugs me that you can't supply any number for b in <string>[a:b]
> to indicate spanning to the end of the string.
[...]
> So if the designers could fix this for version 2.3 I think we'll all be happy
> :)

Let me introduce Guido's time machine: you can always use sys.maxint
as the value of b.

Regards,
Martin


From giannivolontosky at TOGLIyahoo.it  Fri Jun 14 16:13:40 2002
From: giannivolontosky at TOGLIyahoo.it (LoGan)
Date: Fri, 14 Jun 2002 22:13:40 +0200
Subject: socket raw
Message-ID: <aedj3v$dk2$1@newsreader.mailgate.org>

hi, id like to make a python script that use socket raw, i found the btk
module but it cant handle received packets, it's really a good module but
useful only to send, anyone could help me ?
bye


From cmathur at cis.poly.edu  Mon Jun 17 14:34:08 2002
From: cmathur at cis.poly.edu (chandan mathur)
Date: Mon, 17 Jun 2002 14:34:08 -0400
Subject: Cookie based authentication
Message-ID: <3d0e2b38@news.poly.edu>

I am still stuck on how to use client cookie
(http://wwwsearch.sourceforge.net/ClientCookie/). I am trying to write a
script that automatically logs in to http://www.fool.com/login.asp (SSL
encryption) and downloads pages that i need. This is what i have tried so
far: I logged in using internet explorer and opened the cookie file from
windows explorer. Using set_cookie method in Cookies, i tried to set the
cookie as best as i could.. (i probably did something wrong here).

import ClientCookie
import urllib2
import urllib
request = urllib2.Request("http://www.fool.com/Login.asp")
result = urllib2.urlopen(request)
c = ClientCookie.Cookies()
c.extract_cookies(result, request)


# let's say this next request requires a cookie that was set in result
request2 = urllib2.Request("http://my.fool.com/index.htm")
c.add_cookie_header(request2)
c.set_cookie(None,"Wookie=Ref=","FoolV=5&Uid=123146592&LastVisit=6%2F17%2F20
02+1%3A24%3A20+PM&FirstVisit=6%2F17%2F2002+12%3A27%3A22+PM&Username=westlab"
,"/","fool.com/",None,None,None,None,None,None)
result2 = urllib2.urlopen(request2)
print result2.geturl()
print result2.info()  # headers
print result2.read()

it gets the wrong page


Thanx for your time

Chandan





From deathtospam43423 at altavista.com  Tue Jun 11 13:47:37 2002
From: deathtospam43423 at altavista.com (netvegetable)
Date: 11 Jun 2002 17:47:37 GMT
Subject: Newby: Dictionaries as file objects?
References: <Xns9228AB206D06CAusYourStandingInIt@130.133.1.4> <7396d2b2.0206101224.61005b7f@posting.google.com>
Message-ID: <Xns922B224A18783AusYourStandingInIt@130.133.1.4>

d_blade8 at hotmail.com (Lemniscate) wrote in
news:7396d2b2.0206101224.61005b7f at posting.google.com: 

> I agree with David.  However, depending on exactly you want to be
> doing (and be able to do), you may want to look into the dbm modules
> (such as dbm, dbhash, etc.).  These are, surprise, database modules
> but they allow for the storage of dictionary items.....

Damn! dbm doesn't appear to be available for windows. And dhash, doesn't 
seem to allow me to store a dictionary as a value.

Any other suggestions? 

> ......Now, pickle,
> shelve, and the like are perhaps the best things for storing Python
> objects (let's say a user-created object that contains mixed types)
> but the dbm modules have their uses too, based on EXACTLY what you
> want to do.  Anyway, have fun.
> 
> Lem
> 

-- 
netvegetable at excite.com


From cliechti at gmx.net  Tue Jun  4 14:44:12 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 4 Jun 2002 20:44:12 +0200
Subject: directory
References: <3cfcfd4a_1@news.iprimus.com.au>
Message-ID: <Xns9223D3529DC59cliechtigmxnet@62.2.16.82>

Gold Fish <occeanlinux at linuxmail.org> wrote in
news:3cfcfd4a_1 at news.iprimus.com.au: 

> How to read the subdirectories in the directory
> Support i have the directory /home/ which have some subdirectories
> /home/sub1
> /home/sub2
> /home/sub3
> I am using os.listdir('/home/') to read in the list so it will be 
> ['sub1','sub2','sub3'] now i want to read inside the sub directory
> sub2 once again i using os.listdir(os.listdir('/home/')[1]) but it
> came out with an errors: not a file or directory. Can anyone tell me
> what did i wrong. Anh how to fix it.

(probably because os.listdir returns only file-/diectorynames and not 
paths)

look at os.path.walk(), it's there to solve your problem ;-)

chris

-- 
Chris <cliechti at gmx.net>



From syver-en+usenet at online.no  Fri Jun 21 12:02:14 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Fri, 21 Jun 2002 16:02:14 GMT
Subject: python mode for emacs
Message-ID: <u660cfwo3.fsf@online.no>

I've downloaded the latest version from the python cvs tree, and it
works fine. But... I am not too happy with the default face for the
py-pseudo-keywords face (None, self....), and tried to change it by
customizing the face, which worked fine until I restarted emacs and
the face got it's ugly default. I looked at the font-lock hook and it
seems that the face is overwritten each time python-mode is loaded
with the keyword face. 

Is there a straight forward way of deciding how
the py-pseudo-keyword-face should look?
-- 

Vennlig hilsen 

Syver Enstad


From peter at engcorp.com  Sat Jun 29 08:30:39 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 29 Jun 2002 08:30:39 -0400
Subject: I'm an idiot
References: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16> <Xns923BD424CF53Cdavidgriswold1netsca@204.127.68.17> <afjfdg$mh2$1@slb7.atl.mindspring.net> <afjfti$e27$1@slb0.atl.mindspring.net>
Message-ID: <3D1DA86F.17869A93@engcorp.com>

Matt Gerrans wrote:
> 
> Oops, silly me, it was posted, but I think there was a typo in it:
> 
> > for line in open('c:/temp/temp.txt'):
> 
> is missing the ".readlines()" part.

Under newer Pythons (2.2?) it's valid because the file object
will be treated as an iterator and do the equivalent of 
xreadlines() (I think) automatically.  Not the most readable, 
perhaps, as written, but valid.

-Peter


From jj at void.si  Fri Jun  7 18:35:56 2002
From: jj at void.si (jj)
Date: Sat, 8 Jun 2002 00:35:56 +0200
Subject: win32com map attributes to methods
References: <mailman.1023403809.14634.python-list@python.org>
Message-ID: <3d0135be$1@news.perftech.si>

> Have you tried using property() in py2.2?
Yes, and its working, but i have to support older versions (at least py2.1).

My second attempt is done by the Mark's book  using _dynamic_ and
DynamicPolicy,  and its working for attributes. Ofcouse there is a BUT,
but its not working with methods.

Example
o.Flush  method is CALLED

# correct syntax causes proper method call too + exception:
o.Flush()
TypeError: 'NoneType' object is not callable, metod call invokes correct
method and because


# ----------- code -----------
mport sys, string
import pythoncom
import win32com.server.util
import win32com.client
#import win32com.server.policy


class ResponseIStream:
    '''Istream for response'''
    _public_methods_ = ['Write', 'Flush']
    _public_attrs_= ['ContentType', 'Expires']

    _com_interfaces_ = [pythoncom.IID_IStream, pythoncom.IID_IDispatch]

    def __init__(self):
        pass
    def set_Expires(self, value):
        print 'set_expires', value
        self.x = value
    def get_Expires(self):
        print 'get_expires'
        return self.x
    def write(self, x):
        print 'write:', x
    def Write(self, s):
        self.write(s)
        return len(s)
    def Flush(self):
        print 'Iresponse.Flush'


    def _dynamic_(self, name, lcid, wFlags, args):
        if wFlags & pythoncom.DISPATCH_METHOD:
            print 'method call', name
            item= getattr(self, name)
            return apply(item, args)

        elif wFlags & pythoncom.DISPATCH_PROPERTYGET:
            print 'propget'

            mname = 'get_'+name
            if hasattr(self, mname):
                return apply(getattr(self, mname), args)
            item = getattr(self, name)
            if callable(item):
                return
                #return apply(item, args)
            return item
        elif wFlags & pythoncom.DISPATCH_PROPERTYPUT:
            print 'propset'
            mname = 'set_'+name
            if hasattr(self, mname):
                return apply(getattr(self, mname), args)
            item = getattr(self, name)
            if callable(item):
                apply(item, args)
                return
            setattr(self, name, args)
if __name__=="__main__":

    o = win32com.server.util.wrap(ResponseIStream(),  usePolicy =
win32com.server.policy.DynamicPolicy)
    o = win32com.client.Dispatch(o)

    o.Flush
    print 'second flush !!!! is not working'

    o.Flush()
    o.Write('test')

    print 'set'
    o.Expires = 'test value'
    print 'get'
    print o.Expires


    xml = win32com.client.Dispatch("Msxml2.DOMDocument")
    xml.loadXML('<x>streaming</x>')
    xml.save(o)
    r = ResponseIStream()
    r.Expires = 9
    print r.Expires




From fperez528 at yahoo.com  Thu Jun 27 14:44:23 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 27 Jun 2002 12:44:23 -0600
Subject: connecting embedded Python audio samples with NumPy?
References: <3D1B2603.26EA9A9C@fxtech.com>
Message-ID: <affme7$82q$1@peabody.colorado.edu>

Paul Miller wrote:

> I have a C app that reads audio data, and I'd like to pass this audio
> data in array format to Python functions so I can use it with NumPy.
> 
> I have this working right now - generating a float array from the sample
> stream. But I want to put this array into a form usable by the FFT
> functions. I've read over this paper:
> 
> http://www.onlamp.com/pub/a/python/2001/01/31/numerically.html
> 
> Which indicates I need to put my 1D sample stream into a 2D array. Can
> anyone elaborate more on this? I'm wondering if I should provide the
> array pre-made for FFTs from my C code, or I should build the 2D array
> from the 1D array I have in Python.

I don't see why you need to convert the data stream to a 2d array. In fact, 
since you know your data is real, you can use FFT.real_fft() and 
FFT.inverse_real_fft() for the conversions, which guarantees that after 
processing you get real data back (without coefficients with small imaginary 
parts). That is, of course, as long as in your processing you don't break the 
symmetry assumptions on the fft data for a real signal :)

FFT comes with Numeric, so with a properly installed numeric it imports 
straight out. Here's a quick example:

In [1]: x = frange(0,2*pi,npts=8)

In [2]: y = sin(x) + sin(4*x)

In [3]: import FFT

In [4]: yhat = FFT.real_fft(y)

In [5]: yhat
Out[5]:
array([ -6.66133815e-16+0.j        ,   1.32978452e+00-3.21038382j,
             -8.67767478e-01+0.86776748j,  -2.41187287e+00+0.99903045j,
              3.89971165e+00+0.j        ])

In [6]: y2 = FFT.inverse_real_fft(yhat)

In [7]: y2
Out[7]:
array([ -5.55111512e-17,   3.47947743e-01,   1.75675939e+00,  -5.41044173e-01,
              5.41044173e-01,  -1.75675939e+00,  -3.47947743e-01,  
-1.27675648e-15])

In [8]: y2 - y
Out[8]:
array([ -5.55111512e-17,  -5.55111512e-17,   2.22044605e-16,   0.00000000e+00,
              0.00000000e+00,   0.00000000e+00,  -5.55111512e-17,  
-5.21501245e-17])

So within the accuracy of a C double, it all works as expected.

Cheers,

f



From bh at intevation.de  Sat Jun 22 13:05:21 2002
From: bh at intevation.de (Bernhard Herzog)
Date: 22 Jun 2002 19:05:21 +0200
Subject: shallow copy vs. deep copy
References: <MPG.177ec42fc19fcbf598968f@news.rz.uni-karlsruhe.de> <Xns9235BEC865DCDcliechtigmxnet@62.2.16.82>
Message-ID: <6qofe35jn2.fsf@abnoba.intevation.de>

Chris Liechti <cliechti at gmx.net> writes:
 
> > on the other side i could also:
> > 
> > new.__dict__ = old.__dict__.copy()
> 
> that's what i would call a swallow copy.

A european or an african swallow?

SCNR ;-)


   Bernhard 

-- 
Intevation GmbH                                 http://intevation.de/
Sketch                                 http://sketch.sourceforge.net/
MapIt!                                           http://www.mapit.de/


From gusraan at terra.com.br  Sun Jun 16 17:38:00 2002
From: gusraan at terra.com.br (Raphael Ribeiro)
Date: 16 Jun 2002 14:38:00 -0700
Subject: Visual Python
Message-ID: <337619fa.0206161337.546a958f@posting.google.com>

When i go to tools/debug and the ms-dos prompt start appears ,  an
error message appears , anyone knows why?


From peter at engcorp.com  Sat Jun  1 17:45:50 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 01 Jun 2002 17:45:50 -0400
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org> <3CF8E382.3A44F0DA@engcorp.com> <83ptzavk14.fsf@panacea.canonical.org>
Message-ID: <3CF9408E.6015D2A@engcorp.com>

Kragen Sitaker wrote:
> 
> Peter Hansen <peter at engcorp.com> writes:
> > How fast do you need it to be?  (That's a serious question.
> > Are you planning to open and close the application so frequently
> > that six seconds is unacceptable?)
> 
> From a human interface perspective, having to wait six seconds for a
> response to a command (including "start program") is annoying; this is
> the reason for "splash screens".  You don't want people to be annoyed
> every time they start your program.

I realize that.  That's why I was directing the discussion towards
alternatives, such as splash screens... :-)

Other options might include making sure modules are not loaded
until they are needed, or loading them on a separate thread just
after the main program starts.

None of which seems to help the OP.  I'm curious why the Linux
implementation is so much slower than the Win32.  My own machine
is _not_ fast!  (It's only a 233 MHz Pentium MMX.)

-Peter


From mhammond at skippinet.com.au  Sun Jun 16 18:57:14 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 16 Jun 2002 22:57:14 GMT
Subject: MS Word -- finding text
References: <mailman.1024247060.12127.python-list@python.org>
Message-ID: <3D0D1862.5050708@skippinet.com.au>

Mike Brenner wrote:
 > The COM objects (like Project, Word, Excel, etc.) sometimes return
 > stuff in Unicode format. When they do, the python str() function dies
 > when converting non-ASCII unicode characters.
 >
 > To avoid this problem, I use the following conversion routine. After
 > making the necessary check for None, it attempts a quick conversion
 > str() first. When necessary, it slowly goes through each character,
 > handling the exceptions that are raised.
 >
 > The default is a prime because that is the most common character that
 > hits me in Word and Excel documents. Instead of coding it as an ASCII
 > single-quote characters, these applications code it as a more
 > "beautiful" character, so it kills the python str() function.

Note that you should be able to convert a Unicode object obtained from a 
COM object simply by saying:

s = u_ob.encode("mbcs")

And should never (OK OK - rarely :) fail.  If it _does_ fail, then it 
means that the string can not be described in the mbcs code page, and 
you will need to determine the appropriate code page youself.

Mark.



From mcfletch at rogers.com  Mon Jun  3 12:38:41 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 03 Jun 2002 12:38:41 -0400
Subject: List
References: <20020602205610.31575.qmail@linuxmail.org> <009b01c20a7e$12a5ad80$0101010a@local> <mailman.1023062343.28782.python-list@python.org> <3cfb07cf_1@news.iprimus.com.au>
Message-ID: <3CFB9B91.2010002@rogers.com>

Okay, this is pretty basic stuff, I'll just go over it quickly and refer 
you to http://www.python.org/doc/Newbies.html which has introductory 
documentation for Python.  You'll particularly want to work on 
understanding the list as a data type and the manipulations you can use 
when working with it.

Imagine you have a list of elements returned by the function "into" 
(from the previous message):

pageData = [['p1','p2','p3'],['p4','p5','p6'],['p7','p8','p9'],['p10']]

You generate this list when the user requests the page, and look at the 
query value in the CGI environment to determine which page of results 
the user wants to retrieve:

	requestedPage = '2'

Now, you want requestedPage to show up as an integer index into your 
list of pageData, that is, it tells you what element in the compound 
list pageData should be used for generating the page.

	requestedPage = int(requestedPage)

(You'd want to do some error-checking there, making sure it really is an 
integer, catching a ValueError will work).

Now, to get the values for the particular page, you index into the 
pageData list using your new index:

	pageData[ requestedPage ]

Again, you want to do error checking catching an IndexError works there.

If you want to know the total number of pages, it's:

	len( pageData )

You then go about generating your page, being sure to enable/disable the 
forward (requestedPage-1)/backward (requestedPage+1) links depending on 
whether the user is on the first/last page (BTW: you can offer links to 
"first" and "last" on every page just by using 0 (first) and -1 (last) 
as the link indices).

Note: if there are no results at all, even 0 won't work as an index, 
you'll need to catch that condition and generate a "no results" page.

Hope this helps,
Mike

Gold Fish wrote:
...
> Sorry for the very stupid question but i still don't get my goal, that's is 
> using Mike C. Fletcher method i will generate the list of elements again 
> and don't know what size it is. For example :
> [['p1','p2','p3'],['p4','p5','p6'],['p7','p8','p9'],['p10']]
> I want to seperate this list again in order to using them later .It look 
> like you go to the realestate agent website and find the list of property 
> then you just want to display 3 properties per pages. And you can go to 
> other pages which also contain 3 properties as well until the last pages 
> which has only 1 items. For example:
> Page 1: ['p1','p2','p3']
> Page 2: ['p4','p5','p6']
> Page 3: ['p7','p8','p9']
> Page 4: ['p10']
> I didn't sleep well due to this problem Can any help me.
> 
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From peter at engcorp.com  Mon Jun 24 21:08:58 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 24 Jun 2002 21:08:58 -0400
Subject: Suggestions for good programming practices?
References: <WEuR8.2074$5l4.75657@ozemail.com.au>
Message-ID: <3D17C2AA.F694F74@engcorp.com>

Dianne van Dulken wrote:
> 
> if anyone has a list of things that they consider as good programming
> practice in Python (I thought this might be an interesting topic for
> discussion, anyway)

I'll amend my previous claim to point to testing as a very useful
part of Python development.  I don't mean running your code manually
after the fact, either.  I mean proper unit testing, which exercises
as much of the code in small pieces as you possibly can.

Unlike with statically typed languages, Python cannot check that
you are not, for example, trying to pass an integer into a routine
that can only accept a string until you actually execute the code.
You don't get the error on compilation as with C for example.
It can become frustrating if you have to run the code manually
each time you make a change and possibly go through a large
sequence of steps before you get to the part you want to test.

Therefore, learn to make your code testable.  Write it in small
modular pieces and have good tests for each.  When you make
a change, run the tests again and they will tell you whether
you've screwed up anything.  If something slips through and 
you find a bug, go back and write another test that fails
on the same problem, then fix the code (making the test 
pass).

No, not many people actually do this.  I've found it highly
effective, however, and it more than makes up for any 
limitations that might exist because of the dynamic nature
of Python compared to statically typed languages.  This
all comes from Extreme Programming, of course, but you can
easily benefit from proper unit testing even without other
changes to your development process.

And most of this advice actually applies to any language, 
but I think Python can take advantage of it even more than most.

-Peter


From phd at phd.pp.ru  Wed Jun  5 10:09:40 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 5 Jun 2002 18:09:40 +0400
Subject: new SmallEiffel elj-win32 release (-0.74)
In-Reply-To: <8cd6cc47.0206050602.1d69a237@posting.google.com>; from geoff@elj.com on Wed, Jun 05, 2002 at 07:03:00AM -0700
References: <8cd6cc47.0206050602.1d69a237@posting.google.com>
Message-ID: <20020605180940.C26674@phd.pp.ru>

On Wed, Jun 05, 2002 at 07:03:00AM -0700, Geoff Eldridge wrote:
> A new version of elj-win32 has been prepared and is now available from ..

   What does it have with Python?

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From bit_bucket5 at hotmail.com  Thu Jun  6 15:06:34 2002
From: bit_bucket5 at hotmail.com (Chris)
Date: 6 Jun 2002 12:06:34 -0700
Subject: Using AT on XP to run python script - MARK HAMMOND PLEASE READ
References: <mailman.1023320948.13063.python-list@python.org>
Message-ID: <fa7108b0.0206061106.339d547b@posting.google.com>

Interesting findings.  I know mine didn't work bcs the script I want
AT to execute sends me an email.  The script sends an email, then I
want it to run another AT command to set itself up to run again at a
later time.  But I can't get it to work.  If I run the script
manually, it runs, sends email and executes the proper AT command. 
But when AT fires off the task the script set up, nothing happens and
no more AT commands are set up.

"David LeBlanc" <whisper at oz.net> wrote in message news:<mailman.1023320948.13063.python-list at python.org>...
> This is most curious, but I believe the commands _are_ getting executed!
> 
> I tried the following (assume time is 16:30) "at 16:32 pythonw
> j:\python22\tools\idle\idle.py" (without the qoutes of course). At the
> appointed time, there was disk activity, but no idle window popped up (idle
> was chosen just to get something to pop up!). However, there _was_ a pythonw
> process running in the task manager! I think there might be a way to attach
> a console or get access to the GUI, but i'm not sure - maybe Mark Hammond
> might have a clue (he being the Python Win guru :) ).
> 
> I tried the following:
> #at_test.py
> import sys
> import time
> 
> tst = open("testfile.txt", 'a+')
> 
> tst.write(time.asctime() + "\n")
> 
> tst.close()
> 
> put into an 'at' job as: at xx:xx python j:\python22\at_test.py
> 
> I expected the output from this to be in j:\python22, but to my surprise, it
> was in c:\winnt\system32 where cmd.exe resides! testfile.txt contained the
> exact time the 'at' job was set to run at too!
> 
> I suspect your last example that produced no visible output did in fact work
> :)
> 
> In any case, it works - now if one could just get visible output!
> 
> David LeBlanc
> Seattle, WA USA
> 
> > -----Original Message-----
> > From: python-list-admin at python.org
> > [mailto:python-list-admin at python.org]On Behalf Of Chris Stromberger
> > Sent: Tuesday, June 04, 2002 17:35
> > To: python-list at python.org
> > Subject: Using AT on XP to run python script
> >
> >
> > I'm having trouble getting AT on XP to run a python script for me.
> > Have tried:
> >
> > at xx:xx "c:\python21\python.exe c:\scripts\script.py"
> > at xx:xx "start c:\scripts\script.py"
> > at xx:xx "c:\scripts\script.py"
> >
> > and get "the system cannot find the file specified" for the first two.
> > The last version does nothing apparently--no error indication or
> > anything.
> >
> > I imagine it is related to user accounts and such.  Anyone have any
> > pointers?
> > --
> > http://mail.python.org/mailman/listinfo/python-list


From nico at tekNico.net  Sat Jun  8 03:19:56 2002
From: nico at tekNico.net (Nicola Larosa)
Date: Sat, 08 Jun 2002 09:19:56 +0200
Subject: Medium to Large Scale Python Deployments
References: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com> <mailman.1023409393.3361.python-list@python.org> <3D003EC7.C72D27BE@engcorp.com> <3D006320.5040805@web.de> <3D00A209.EC24F671@engcorp.com> <adr2ao$bei$0@216.39.172.122> <3D0140BF.A239E39E@engcorp.com>
Message-ID: <3D01B01C.3000303@tekNico.net>

> (we rarely use doc strings, having avoided that aspect of
> Python when writing on an embedded system where all the .pyc 
> files were stored in our limited 1MB of flash along with the 
> interpreter...).

python -OO ?


-- 
"Too much cleverness in the parser can turn against you."
   Guido Van Rossum

Nicola Larosa - nico at tekNico.net






From gerhard at bigfoot.de  Tue Jun 18 01:35:31 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 18 Jun 2002 05:35:31 GMT
Subject: unicode strings and strings mix
References: <mailman.1024375941.29946.python-list@python.org>
Message-ID: <slrnagthhn.5tb.gerhard@lilith.my-fqdn.de>

Roman Suzi wrote in comp.lang.python:
> Is there any way to define "default" encoding

Yes.

> so the following error will not happen?
> 
>>>> unichr(0x3345) + "?????"
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> UnicodeError: ASCII decoding error: ordinal not in range(128)

'x' and 'A' are in the ASCII range, so this shouldn't produce an
exception. I also cannot reproduce it with sys.getdefaultencoding() ==
"ascii".

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From rhymes at myself.com  Sat Jun 29 14:54:43 2002
From: rhymes at myself.com (Rhymes)
Date: Sat, 29 Jun 2002 20:54:43 +0200
Subject: private
References: <pan.2002.06.29.17.38.51.789527.440@myself.com> <7x8z4y9eex.fsf@ruckus.brouhaha.com>
Message-ID: <pan.2002.06.29.20.54.42.370970.440@myself.com>

Thanks for the reply, you all

-- 
Rhymes
rhymes at myself.com
" ride or die "


From spokra at home.com  Mon Jun 10 22:20:44 2002
From: spokra at home.com (Steven Pokrandt)
Date: 10 Jun 2002 19:20:44 -0700
Subject: python and shared libaray
Message-ID: <a263f04a.0206101820.5461339f@posting.google.com>

I have a vendor supplied .so file, header file and docs.  I would like
to use this
.so file in python but don't have the src.  Is it possible to link
this library ie:

import filename.so

when I try the python complains about no init function..  can I add
this to the .so  without having the src?

can someone point me to a howto?

please send message to the following address..  spokra (AT) attbi
(dot) com

you get the meaning of the email address, and hopefully the spammers
don't


From j.bodemann at gmx.de  Sat Jun  1 11:17:07 2002
From: j.bodemann at gmx.de (Joern Bodemann)
Date: Sat, 1 Jun 2002 15:17:07 +0000 (UTC)
Subject: Jython: Getopt does not throw exception - 1 attachments
Message-ID: <Xns9220AFD3CD394jbodemanngmxde@62.72.90.23>


Hello all,

does anybody know why the following code does not throw an excetion if the 
script is started without parameter?

jython question.py




begin 644 question.py
M:6UP;W)T(&=E=&]P=`T*:6UP;W)T('-Y<PT*#0ID968@=7-A9V4H*3H-"@EP
M<FEN="`B4V]M92!A<F=U;65N=',@;6ES<VEN9R(-"@ES>7,N97AI="@P*0T*
M"0T*=')Y. at T*"6]P=',L(&%R9W,@/2!G971O<'0N9V5T;W!T*'-Y<RYA<F=V
M6S$Z72P@(G,Z93HB+"!;(G-T87)T/2(L(")E;F1E/2)=*0T*97AC97!T(&=E
M=&]P="Y'971O<'1%<G)O<CH-"@EU<V%G92 at I#0IO=71P=70@/2!.;VYE"F9O
M<B!O+"!A(&EN(&]P=',Z#0H):68@;R!I;B`H(BUS(BP@(BTM<W1A<G0](BDZ
M#0H)"7-T87)T(#T at 80T*"6EF(&\@:6X@*"(M92(L("(M+65N9&4](BDZ#0H)
2"65N9&4@/2!A"0T*"0D-"@D)
`
end



From wurmy at earthlink.net  Fri Jun 28 11:22:16 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Fri, 28 Jun 2002 15:22:16 GMT
Subject: metaclasses vs. inheritance
References: <ghTS8.58749$mh.1826422@news1.telusplanet.net> <mailman.1025259037.9256.python-list@python.org>
Message-ID: <3D1C7EF1.7060002@earthlink.net>

Aldo Cortesi wrote:

>>PS - what's the antonym of meta?
> 
> Hmmm... the meta- in "metaclass" or "metaphysics" really has
> no relation to the original Greek, which means "with", or
> "among" (as in "metatarsus" - the bones in the foot which
> lie next to the tarsus). So whatever antonym you come up
> with can't have anything to do with the original etymology
> of the prefix...  Erm... How about mundane? Or actual?
> Intrinsic? Essential? Or even "per se"....?

In ancient Greek, it also meant "over", like Latin "trans"... e.g. "over the 
river". I think the "meta" in "metaclass", "metadata" etc is related to this 
meaning.

Cheers,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/



From boud at valdyas.org  Sat Jun 15 05:18:09 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Sat, 15 Jun 2002 11:18:09 +0200
Subject: IDE recommendations??
References: <b9veea.l85.ln@majestix.gallien.de>
Message-ID: <3d0b06bd$0$52043$e4fe514c@dreader3.news.xs4all.nl>

Detlev Offenbach wrote:

> Hi all,
> 
> I have googled this list for IDE recommendations, but have found none
> that satisfied me. I am looking for an IDE that allows development of
> Qt/KDE applications using Python. So far I have found BlackAdder.
> However, it looks as if development has stalled. I could use individual
> applications (e.g. qt-designer for gui, scite as an editor, pyuic, ...),
> but this lacks some comfort. I would appreciate any recommendation for a
> suitable IDE
> 

I've tried to use most IDE's out there -- from Pythonwin to Pythonworks. 
None was really comfortable, except for Wing IDE, which is really powerful. 
It doesn't include integration of a GUI designer, though, which makes it 
little more than a glorified editor/debugger (but _very_ glorified). At one
point there was a version of KDevelop that supported Python development, but
that has disappeared, it seems.

I use bash & XEmacs & Designer nowadays, and that gives me a fair 
productivity.

-- 
Boudewijn Rempt | http://www.valdyas.org


From gerhard at bigfoot.de  Sun Jun  9 08:47:11 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 9 Jun 2002 12:47:11 GMT
Subject: Logic programing with python?
References: <3D0343EA.4070100@gmx.de>
Message-ID: <slrnag6jii.ql.gerhard@lilith.my-fqdn.de>

Ingo Linkweiler wrote in comp.lang.python:
> Hello,
> 
> is it possible to use Python as "logic lanuage" ?
> I have found the project "formula", a logical language coded with 
> python, and Chris Meyers "Prolog interpreter".
> I have heard from "holmes", but can not find any information about it.
> 
> Do you know other sources?

http://www.logilab.org/python-logic/

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From pyth at devel.trillke.net  Mon Jun 17 11:37:31 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 17 Jun 2002 17:37:31 +0200
Subject: what *is* a class?
In-Reply-To: <20020617143525.GB8814@unpythonic.net>; from jepler@unpythonic.net on Mon, Jun 17, 2002 at 09:35:26AM -0500
References: <MPG.17774343cc5a589098968a@news.rz.uni-karlsruhe.de> <Gxtt3F.upA@beaver.cs.washington.edu> <3D0DBC87.B1BDEB2D@ifib.uni-karlsruhe.de> <20020617143525.GB8814@unpythonic.net>
Message-ID: <20020617173731.Q15079@prim.han.de>

Jeff Epler wrote:
> This is not the Pythonic style.  Even when it would sometimes be
> convenient, methods do not tend to return "self".  That's why these
> don't work:
> 
>     # Iterate the sorted list
>     for item in l.sort(): ...
> 
>     # Create and pack a widget
>     x = Label(text="Wibble").pack()
> 
>     # Dump two objects to a pickle
>     p = pickle.Pickler()
>     p.dump(obj1).dump(obj2)
> 
> One motivation for this is that if list.sort returns None, nobody
> will write
>     for item in l.sort():
> thinking that a copy of l is sorted.  It'll immediately blow up.

I agree that this is a good thing. But it wouldn't hurt to have

    reverse, sort, extend

as standalone functions somewhere in the stdlib.  They would return
a copy, so that you can write:

    # iterate over a sorted copy of a list
    for item in sort(l):
        ...

reverse could probabally be made an iterator so that it gets really
cheap to write:

    for item in reverse(l):
        ...

And i think these functions should work with tuples and strings also.

cheers,

    holger




From jepler at unpythonic.net  Tue Jun  4 12:31:47 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 4 Jun 2002 11:31:47 -0500
Subject: self
In-Reply-To: <BI5L8.78$Vv1.4155449@newnews.cc.stevens-tech.edu>
References: <BI5L8.78$Vv1.4155449@newnews.cc.stevens-tech.edu>
Message-ID: <20020604113147.C24006@unpythonic.net>

On Tue, Jun 04, 2002 at 12:15:22PM -0400, Vojin Jovanovic wrote:
> Would it not be more logical
> to just use the name of the variable like it is for example in C++?

Nope. (Hint:
http://www.python.org/cgi-bin/faqw.py?req=show&file=faq06.007.htp)

Jeff
PS we love it this way.  You will too.




From peter at engcorp.com  Tue Jun 25 23:38:16 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 25 Jun 2002 23:38:16 -0400
Subject: Reminder of a python Project.
References: <2218.941T781T12224051threeseas@earthlink.net>
Message-ID: <3D193728.663CA774@engcorp.com>

Timothy Rue wrote:
> 
> First off, it's ok that I got such a high level of resistance to my
> initial posting here about this project. 

Obviously the level was not high enough.  Sorry, we'll try harder.

> $81 Billion = 31% of software development gets cancelled before complete
> $59 Billion = 53% of software development has cost over-runs of 189%
> 16% success - project success and failure ratio
[...]
> My 1997 estimate of software failures all around (total of industry and
> consumer level development and use):  Low estimate of 300 Billion.

Your information is out of date.  Agile software development processes 
developed since 1997 have improved the success rate of software development 
projects to the point where this is no longer an issue.  It is now possible
to deliver software with a 96% success rate, for less cost than previous
methods, with greater sustainability for the development teams, and at
a much higher quality level than appeared possible in the past.

I'm afraid there is no longer any need for an autocoder.

-Peter


From a.schmolck at gmx.net  Mon Jun 17 17:41:31 2002
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 17 Jun 2002 22:41:31 +0100
Subject: "Fatal Python error: GC object already in linked list"
Message-ID: <yfsfzzleg6s.fsf@black132.ex.ac.uk>

Just got the following in an interactive session:

    >>> reload(awmstools)
    Exception exceptions.RuntimeError: 'maximum recursion depth exceeded' in  ignored
    Fatal Python error: GC object already in linked list
    
    Process Python aborted
    
Any ideas as to what might be the cause?

alex


From graham at effectif.com  Wed Jun  5 19:44:32 2002
From: graham at effectif.com (Graham Ashton)
Date: Thu, 06 Jun 2002 00:44:32 +0100
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet>
Message-ID: <qkxL8.57201$wd3.9390318@news6-win.server.ntlworld.com>

On Wed, 05 Jun 2002 11:31:19 +0100, TuxTrax wrote:

> My question is this; I realize that Python compiles to bytecode
> automatically as part of the interpreters tokenization run, but is there
> a way to permanently compile python code outputting to an executable
> binary?

Not exactly, in general. Have a look for py2exe, psyco, and pyrex for info
on some of the things that you can currently do.

GvR alludes to compilation being of interest in the future in a recent
interview:

   http://www.onlamp.com/pub/a/python/2002/06/04/guido.html

-- 
Graham Ashton


From debl2Nonospawhammy at bellatlantic.net  Sat Jun 15 22:27:25 2002
From: debl2Nonospawhammy at bellatlantic.net (David Lees)
Date: Sun, 16 Jun 2002 02:27:25 GMT
Subject: Minor install annoyance on Activestate Python 2.2.1 build 222
Message-ID: <3D0BF7B4.21096694@bellatlantic.net>

I just downloaded and installed the latest ActiveState build of Python
2.2.1 off their site and could not get it to install on my win98SE box
even though I had previously run the windows installer.  The problem
turns on to be that the name of the file you download is
ActivePython-2.2.1-222.msi.exe
Simply remove the '.exe' extension so the name is:
ActivePython-2.2.1-222.msi and the installation goes just fine.

David Lees


From skip at pobox.com  Mon Jun 10 11:28:54 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 10 Jun 2002 10:28:54 -0500
Subject: Source formatting for long expressions.
In-Reply-To: <87hekdxypd.fsf@twistedmatrix.com>
References: <87hekdxypd.fsf@twistedmatrix.com>
Message-ID: <15620.50614.922005.965796@12-248-41-177.client.attbi.com>

    Chris> Does anyone know of a tool that will beak up a very, very long
    Chris> expression like this into multiple lines with sane indentation?

How about the pprint module?

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html





From peter at engcorp.com  Tue Jun 11 01:34:00 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 11 Jun 2002 01:34:00 -0400
Subject: Correct code/unit testing (was Re: Efficient python programming...)
References: <mailman.1023770019.8729.python-list@python.org>
Message-ID: <3D058BC8.1C4C3DFA@engcorp.com>

"Delaney, Timothy" wrote:
> 
> > From: Peter Hansen [mailto:peter at engcorp.com]
> > I have to confess I have never seen a successful code review, nor
> > an environment in which they were used regularly and which had a
> > high level of code quality.  Clearly as a result I'm fairly biased
> > against them -- I should keep a more open mind on the issue perhaps.
> 
> I presume you mean a full, formal review by a group for a large piece of
> code?

In that paragraph, yes, I meant that.

> I'm certain that with your experience with XP, you have seen and taken part
> in *many* successful reviews (where success means the code is improved as a
> result).

I think I tried to say that later in the message.  If pair programming
is considered the same as "review", then I agree.  Of course, pair 
programming is _much_ more than review-as-you-code, but that's another
topic.

-Peter


From uwe at rocksport.de  Mon Jun  3 12:10:40 2002
From: uwe at rocksport.de (Uwe Schmitt)
Date: 3 Jun 2002 16:10:40 GMT
Subject: different gui-toolkits pros/cons...
References: <adfsoo$dmik4$1@hades.rz.uni-sb.de> <_rLK8.12$fd2.979@news.ecrc.de>
Message-ID: <adg4e0$dq6mb$3@hades.rz.uni-sb.de>

Markus Jais <mjais at web.de> wrote:
| with pyQT there might be problems with the license
| if you want to make your programm freely available
| (but I am not a lawyer, so I might be wrong)

That's a very important point: what do I have to pay if I used pyQt
for development ??? I did not find any explicit prices at the
trolltec-page...

Greetings, Uwe.


-- 
Dr. rer. nat. Uwe Schmitt      ICQ# 159647634   Uwe.Schmitt at num.uni-sb.de
Universitaet des Saarlandes                     Angewandte Mathematik
Building 36.1 Room 4.17         PO-Box 151150   D-66041 Saarbruecken
Mobile:0177/6806587    Fax:+49(0)681/302-4435   Office:+49(0)681/302-2468


From jimmy at retzlaff.com  Tue Jun 25 18:12:47 2002
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Tue, 25 Jun 2002 15:12:47 -0700
Subject: SOLVED (Re: Python hits the spot)
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F56C5@MAYOR.retzlaff.com>

Bengt Richter wrote:
>>> Well, lucky guess on the heat. That _probably_ accounts for laptop
>>> going to 87% and then shutting down, but how about the other huge
>>> slowdowns in your big loop of 8? Did you get that diagnosed too?
>>> Just curious ;-)
>>
>>On my stationary machine definitely not. The 8 steps remain constant.
I 
>
>Still curious ;-)

I believe that newer mobile CPUs will throttle back their clock speed in
response to heat. This could explain the slow down. Apparently the
lowest clock speed wasn't low enough to keep the CPU happy...

Jimmy




From dfackrell at DELETETHIS.linuxmail.org  Thu Jun 20 14:09:12 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Thu, 20 Jun 2002 12:09:12 -0600
Subject: Proper number alignment
References: <aet5hu$9leln$1@ID-98166.news.dfncis.de>
Message-ID: <3d121a47$1_3@hpb10302.boi.hp.com>

I would avoid tabs altogether and figure out how many spaces I needed to get
there (tab stops can be set differently for every terminal).  Basically,
subtract the total length of the two strings from the total width you want
to fill, and use (' '*numberOfSpaces) as your filler.

--
Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.

"mdk" <mdk at mdk.com> wrote in message
news:aet5hu$9leln$1 at ID-98166.news.dfncis.de...
> I am using \t\t (tabs) to align these numbers (disregard the math):
>
> Total RAM                2,141,056
> Available RAM            12,934,032
> Used RAM                 1,024
>
> But they are left justified and I need them right justified.
>
> Any ideas?
>
> Thanks.




From anoop79 at myrealbox.com  Thu Jun 13 04:06:05 2002
From: anoop79 at myrealbox.com (Anoop P B)
Date: Thu, 13 Jun 2002 13:36:05 +0530
Subject: Threading tutorial wanted.
References: <20020613073828.GD1792@matijek.plusseven.com>
Message-ID: <3D08526D.9050203@myrealbox.com>

check this:
http://starship.python.net/crew/aahz/OSCON2000/TOC.HTM
it may help

also, wxpython has threading support. but i dont know how good it is. 
see the wxpython demo for more info.

i too am in need of answers to ur other questions on when to use, how 
safe, memory consumption, how to get rich ;), etc.

~anoop

Alex Polite wrote:
> Okidoki. My little webspider has reached the point where further
> optimization won't pay off. But my DSL pipe is not saturated by
> far. So I guess it's time for me to start looking into threads.
> 
> I few minutes of googling on python + threads doesn't yield a
> lot. Maybe someone here can point me to a good primer. Maybe someone
> here can write a good primer.
> 
> When is threading good, when is it bad? How much memory do they
> consume? Is it safe to let multiple threads use the same db
> connection? How safe is safe? How can I get rich?
> 
> Stuff like that.
> 
> 






From a.schmolck at gmx.net  Mon Jun 10 09:34:18 2002
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 10 Jun 2002 14:34:18 +0100
Subject: Could someone submit this to the sf-bug tracker, please?(Was: Re: __imul__ broken for 'objects' in 2.2.1)
References: <yfsd6v3qa86.fsf@black132.ex.ac.uk> <lk660tlnon.fsf@pc150.maths.bris.ac.uk>
Message-ID: <yfs660r8dgl.fsf_-_@black132.ex.ac.uk>

Michael Hudson <mwh at python.net> writes:

> Hmm.  You should file a bug on sf so this doesn't get forgotten
> (unless it's already been reported, but I don't think so).

Well yes, but then I *can't* which is why I've posted the previous
message. Could someone else please submit this bug to the sourceforge
bug-tracker? I tried again today, but logging in from the submission form has
no effect, I'm back to the same form that tells me to "please log-in" *after*
I have completed the log-in step (if I submit the wrong password, however I do
get an error message). I've tried 3 different browsers (mozilla .94, konqueror
and netscape 4.7) -- to no avail. Now helpfuly, just giving my email address
instead of logging in doesn't work either, because anonymous bug reports
aren't permitted.

If my problems don't reflect the weirdness of the proxy I'm forced to use, I'd
really suggest that either the decision not to allow anonymous bug reports
should be reviewed again.

alex


> Alexander Schmolck <a.schmolck at gmx.net> writes:
> 
> > OK, after unsuccessfully trying for about half an hour to submit this to the
> > sf bug tracker, I gave up:
> > 
> > class Breaky(object):
> >     def __imul__(self, other):
> >         print "imuling"
> >         return self
> > sq = Breaky()
> > sq *=1.
> > 
> > gives:
> > 
> > Traceback (most recent call last):[...] line 10, in ?
> >     sq *=1.
> > TypeError: can't multiply sequence to non-int
> > 
> > Unless I'm overlooking something, this is a fairly serious bug and I
> > can't see a way to work around it (getattribute would presumably
> > work, but slow everything down unacceptably, so the only 'solution'
> > seems to be to have no inplace-multiplication).
> > 
> > I had the same behavior on two different machines (running Mandrake 8.2 and
> > Suse 7.3).
> 
> Hmm.  You should file a bug on sf so this doesn't get forgotten
> (unless it's already been reported, but I don't think so).
> 
> Cheers,
> M.

-- 
Alexander Schmolck     Postgraduate Research Student
                       Department of Computer Science
                       University of Exeter
A.Schmolck at gmx.net     http://www.dcs.ex.ac.uk/people/aschmolc/


From walter at livinglogic.de  Wed Jun 19 07:03:48 2002
From: walter at livinglogic.de (=?ISO-8859-15?Q?Walter_D=F6rwald?=)
Date: Wed, 19 Jun 2002 13:03:48 +0200
Subject: PEP 293, Codec Error Handling Callbacks 
Message-ID: <3D106514.5030203@livinglogic.de>

There is a new PEP available (online at
http://www.python.org/peps/pep-0293.html)

Please comment!

Bye,
    Walter D?rwald

---------------------------------------------------------------------
PEP: 293
Title: Codec Error Handling Callbacks
Version: $Revision: 1.1 $
Last-Modified: $Date: 2002/06/19 03:22:11 $
Author: Walter D?rwald
Status: Draft
Type: Standards Track
Created: 18-Jun-2002
Python-Version: 2.3
Post-History:


Abstract

     This PEP aims at extending Python's fixed codec error handling
     schemes with a more flexible callback based approach.

     Python currently uses a fixed error handling for codec error
     handlers.  This PEP describes a mechanism which allows Python to
     use function callbacks as error handlers.  With these more
     flexible error handlers it is possible to add new functionality to
     existing codecs by e.g. providing fallback solutions or different
     encodings for cases where the standard codec mapping does not
     apply.


Specification

     Currently the set of codec error handling algorithms is fixed to
     either "strict", "replace" or "ignore" and the semantics of these
     algorithms is implemented separately for each codec.

     The proposed patch will make the set of error handling algorithms
     extensible through a codec error handler registry which maps
     handler names to handler functions.  This registry consists of the
     following two C functions:

         int PyCodec_RegisterError(const char *name, PyObject *error)

         PyObject *PyCodec_LookupError(const char *name)

     and their Python counterparts

         codecs.register_error(name, error)

         codecs.lookup_error(name)

     PyCodec_LookupError raises a LookupError if no callback function
     has been registered under this name.

     Similar to the encoding name registry there is no way of
     unregistering callback functions or iterating through the
     available functions.

     The callback functions will be used in the following way by the
     codecs: when the codec encounters an encoding/decoding error, the
     callback function is looked up by name, the information about the
     error is stored in an exception object and the callback is called
     with this object.  The callback returns information about how to
     proceed (or raises an exception).

     For encoding, the exception object will look like this:

        class UnicodeEncodeError(UnicodeError):
            def __init__(self, encoding, object, start, end, reason):
                UnicodeError.__init__(self,
                    "encoding '%s' can't encode characters " +
                    "in positions %d-%d: %s" % (encoding,
                        start, end-1, reason))
                self.encoding = encoding
                self.object = object
                self.start = start
                self.end = end
                self.reason = reason

     This type will be implemented in C with the appropriate setter and
     getter methods for the attributes, which have the following
     meaning:

       * encoding: The name of the encoding;
       * object: The original unicode object for which encode() has
         been called;
       * start: The position of the first unencodable character;
       * end: (The position of the last unencodable character)+1 (or
         the length of object, if all characters from start to the end
         of object are unencodable);
       * reason: The reason why object[start:end] couldn't be encoded.

     If object has consecutive unencodable characters, the encoder
     should collect those characters for one call to the callback if
     those characters can't be encoded for the same reason.  The
     encoder is not required to implement this behaviour but may call
     the callback for every single character, but it is strongly
     suggested that the collecting method is implemented.

     The callback must not modify the exception object.  If the
     callback does not raise an exception (either the one passed in, or
     a different one), it must return a tuple:

         (replacement, newpos)

     replacement is a unicode object that the encoder will encode and
     emit instead of the unencodable object[start:end] part, newpos
     specifies a new position within object, where (after encoding the
     replacement) the encoder will continue encoding.

     If the replacement string itself contains an unencodable character
     the encoder raises the exception object (but may set a different
     reason string before raising).

     Should further encoding errors occur, the encoder is allowed to
     reuse the exception object for the next call to the callback.
     Furthermore the encoder is allowed to cache the result of
     codecs.lookup_error.

     If the callback does not know how to handle the exception, it must
     raise a TypeError.

     Decoding works similar to encoding with the following differences:
     The exception class is named UnicodeDecodeError and the attribute
     object is the original 8bit string that the decoder is currently
     decoding.

     The decoder will call the callback with those bytes that
     constitute one undecodable sequence, even if there is more than
     one undecodable sequence that is undecodable for the same reason
     directly after the first one.  E.g. for the "unicode-escape"
     encoding, when decoding the illegal string "\\u00\\u01x", the
     callback will be called twice (once for "\\u00" and once for
     "\\u01").  This is done to be able to generate the correct number
     of replacement characters.

     The replacement returned from the callback is a unicode object
     that will be emitted by the decoder as-is without further
     processing instead of the undecodable object[start:end] part.

     There is a third API that uses the old strict/ignore/replace error
     handling scheme:

         PyUnicode_TranslateCharmap/unicode.translate

     The proposed patch will enhance PyUnicode_TranslateCharmap, so
     that it also supports the callback registry.  This has the
     additional side effect that PyUnicode_TranslateCharmap will
     support multi-character replacement strings (see SF feature
     request #403100 [1]).

     For PyUnicode_TranslateCharmap the exception class will be named
     UnicodeTranslateError.  PyUnicode_TranslateCharmap will collect
     all consecutive untranslatable characters (i.e. those that map to
     None) and call the callback with them.  The replacement returned
     from the callback is a unicode object that will be put in the
     translated result as-is, without further processing.

     All encoders and decoders are allowed to implement the callback
     functionality themselves, if they recognize the callback name
     (i.e. if it is a system callback like "strict", "replace" and
     "ignore").  The proposed patch will add two additional system
     callback names: "backslashreplace" and "xmlcharrefreplace", which
     can be used for encoding and translating and which will also be
     implemented in-place for all encoders and
     PyUnicode_TranslateCharmap.

     The Python equivalent of these five callbacks will look like this:

         def strict(exc):
             raise exc

         def ignore(exc):
             if isinstance(exc, UnicodeError):
                 return (u"", exc.end)
             else:
                 raise TypeError("can't handle %s" % exc.__name__)

        def replace(exc):
             if isinstance(exc, UnicodeEncodeError):
                 return ((exc.end-exc.start)*u"?", exc.end)
             elif isinstance(exc, UnicodeDecodeError):
                 return (u"\\ufffd", exc.end)
             elif isinstance(exc, UnicodeTranslateError):
                 return ((exc.end-exc.start)*u"\\ufffd", exc.end)
             else:
                 raise TypeError("can't handle %s" % exc.__name__)

        def backslashreplace(exc):
             if isinstance(exc,
                 (UnicodeEncodeError, UnicodeTranslateError)):
                 s = u""
                 for c in exc.object[exc.start:exc.end]:
                    if ord(c)<=0xff:
                        s += u"\\x%02x" % ord(c)
                    elif ord(c)<=0xffff:
                        s += u"\\u%04x" % ord(c)
                    else:
                        s += u"\\U%08x" % ord(c)
                 return (s, exc.end)
             else:
                 raise TypeError("can't handle %s" % exc.__name__)

        def xmlcharrefreplace(exc):
             if isinstance(exc,
                 (UnicodeEncodeError, UnicodeTranslateError)):
                 s = u""
                 for c in exc.object[exc.start:exc.end]:
                    s += u"&#%d;" % ord(c)
                 return (s, exc.end)
             else:
                 raise TypeError("can't handle %s" % exc.__name__)

     These five callback handlers will also be accessible to Python as
     codecs.strict_error, codecs.ignore_error, codecs.replace_error,
     codecs.backslashreplace_error and codecs.xmlcharrefreplace_error.


Rationale

     Most legacy encoding do not support the full range of Unicode
     characters.  For these cases many high level protocols support a
     way of escaping a Unicode character (e.g. Python itself supports
     the \x, \u and \U convention, XML supports character references
     via &#xxx; etc.).

     When implementing such an encoding algorithm, a problem with the
     current implementation of the encode method of Unicode objects
     becomes apparent: For determining which characters are unencodable
     by a certain encoding, every single character has to be tried,
     because encode does not provide any information about the location
     of the error(s), so

         # (1)
         us = u"xxx"
         s = us.encode(encoding)

     has to be replaced by

         # (2)
         us = u"xxx"
         v = []
         for c in us:
             try:
                 v.append(c.encode(encoding))
             except UnicodeError:
                 v.append("&#%d;" % ord(c))
         s = "".join(v)

     This slows down encoding dramatically as now the loop through the
     string is done in Python code and no longer in C code.

     Furthermore this solution poses problems with stateful encodings.
     For example UTF-16 uses a Byte Order Mark at the start of the
     encoded byte string to specify the byte order.  Using (2) with
     UTF-16, results in an 8 bit string with a BOM between every
     character.

     To work around this problem, a stream writer - which keeps state
     between calls to the encoding function - has to be used:

         # (3)
         us = u"xxx"
         import codecs, cStringIO as StringIO
         writer = codecs.getwriter(encoding)

         v = StringIO.StringIO()
         uv = writer(v)
         for c in us:
             try:
                 uv.write(c)
             except UnicodeError:
                 uv.write(u"&#%d;" % ord(c))
         s = v.getvalue()

     To compare the speed of (1) and (3) the following test script has
     been used:

         # (4)
         import time
         us = u"?a"*1000000
         encoding = "ascii"
         import codecs, cStringIO as StringIO

         t1 = time.time()

         s1 = us.encode(encoding, "replace")

         t2 = time.time()

         writer = codecs.getwriter(encoding)

         v = StringIO.StringIO()
         uv = writer(v)
         for c in us:
             try:
                 uv.write(c)
             except UnicodeError:
                 uv.write(u"?")
         s2 = v.getvalue()

         t3 = time.time()

         assert(s1==s2)
         print "1:", t2-t1
         print "2:", t3-t2
         print "factor:", (t3-t2)/(t2-t1)

     On Linux this gives the following output (with Python 2.3a0):

         1: 0.274321913719
         2: 51.1284689903
         factor: 186.381278466

     i.e. (3) is 180 times slower than (1).

     Codecs must be stateless, because as soon as a callback is
     registered it is available globally and can be called by multiple
     encode() calls.  To be able to use stateful callbacks, the errors
     parameter for encode/decode/translate would have to be changed
     from char * to PyObject *, so that the callback could be used
     directly, without the need to register the callback globally.  As
     this requires changes to lots of C prototypes, this approach was
     rejected.

     Currently all encoding/decoding functions have arguments

         const Py_UNICODE *p, int size

     or

         const char *p, int size

     to specify the unicode characters/8bit characters to be
     encoded/decoded.  So in case of an error the codec has to create a
     new unicode or str object from these parameters and store it in
     the exception object.  The callers of these encoding/decoding
     functions extract these parameters from str/unicode objects
     themselves most of the time, so it could speed up error handling
     if these object were passed directly.  As this again requires
     changes to many C functions, this approach has been rejected.


Implementation Notes

     A sample implementation is available as SourceForge patch #432401
     [2].  The current version of this patch differs from the
     specification in the following way:

       * The error information is passed from the codec to the callback
         not as an exception object, but as a tuple, which has an
         additional entry state, which can be used for additional
         information the codec might want to pass to the callback.
       * There are two separate registries (one for
         encoding/translating and one for decoding)

     The class codecs.StreamReaderWriter uses the errors parameter for
     both reading and writing.  To be more flexible this should
     probably be changed to two separate parameters for reading and
     writing.

     The errors parameter of PyUnicode_TranslateCharmap is not
     availably to Python, which makes testing of the new functionality
     of PyUnicode_TranslateCharmap impossible with Python scripts.  The
     patch should add an optional argument errors to unicode.translate
     to expose the functionality and make testing possible.

     Codecs that do something different than encoding/decoding from/to
     unicode and want to use the new machinery can define their own
     exception classes and the strict handlers will automatically work
     with it. The other predefined error handlers are unicode specific
     and expect to get a Unicode(Encode|Decode|Translate)Error
     exception object so they won't work.


Backwards Compatibility

     The semantics of unicode.encode with errors="replace" has changed:
     The old version always stored a ? character in the output string
     even if no character was mapped to ?  in the mapping.  With the
     proposed patch, the replacement string from the callback callback
     will again be looked up in the mapping dictionary.  But as all
     supported encodings are ASCII based, and thus map ? to ?, this
     should not be a problem in practice.


References

     [1] SF feature request #403100
         "Multicharacter replacements in PyUnicode_TranslateCharmap"
         http://www.python.org/sf/403100

     [2] SF patch #432401 "unicode encoding error callbacks"
         http://www.python.org/sf/432401


Copyright

     This document has been placed in the public domain.




Local Variables:
mode: indented-text
indent-tabs-mode: nil
sentence-end-double-space: t
fill-column: 70
End:





From UG at LY  Fri Jun 14 12:35:01 2002
From: UG at LY (Unconditional Grace)
Date: Fri, 14 Jun 2002 12:35:01 -0400
Subject: NEWbie asks "is Python a good tool for automated file/folder generation?"
Message-ID: <6k6kgukhuk8pfgos97a38h0d4r5jr0u3sr@4ax.com>

Hi all,

I'm so new that "newbie" doesn't even do me justice, and I'm writing
humbly for advice.

I have a series of scientific image files numbering, say, 1000, which
are really 100 groups, each group with its own identical string, but
appended with a different number from _01 to _10, or possibly some
other string.  The end enumeration is ALWAYS differentiated from the
constant string by a final underscore, but the strongs themselves
contain underscores too.

i.e.,
stringa_01.jpg
...
stringa_10.jpg
stringb_01.jpg
...
stringb_10.jpg
...


What I would like to do is largely string-based file and folder
manipulation:

1.  Create a folder for each group of 10, where the folder name is the
shared string for that group (no numbers).  This list of string names
could come directly from the directory, or I could perhaps port the
directory listing to a text file and parse that.

2.  Move each image into its appropriate folder.

3.  Within each folder, create one HTML page consisting of a table
with a link to each image, preferably with a thumbnail (can be
original image resized by HTML since this is all local = no bandwidth
issues).

4.  Create a top level HTML page that links to all 100 image HTML
pages created in step 3.  Can be text only.

5.  Optional:  in step 3, make each image part of an HTML page with
"previous image" and "next image" links, rather than just being called
by the browser as a jpg image.

---

I've never used Python, and I'm wondering if you, collectively, think
that Python is a good language to do this.  While it seems like Python
is an incredibly useful tool, I'm not sure I have time right now to
invest in learning it if it can't do this.  It seems I frequently want
to do repetitive Windows file/folder chores like this, so I want to
get smart-like.

If so, could someone please give me some pointers?  If not, other tool
recommendations?  PHP?

I'm not asking anyone to write such code for me, heh heh, but I'd love
it if someone could give me some advice (here or URLs) on folder and
file name operators, "for every item in list" type syntax, and/or
general string operators  (to truncate the numbers from the file
names, e.g.).

Anyhoo, in addition to any news group response, please cc: me on email
(H_J_Fong at hotmail.com) since my server is spotty and I'd hate to miss
your response(s).

Thanks for reading this far, it is a lot to ask, 
Homer


From cpgray at library.uwaterloo.ca  Mon Jun 24 14:24:18 2002
From: cpgray at library.uwaterloo.ca (Chris Gray)
Date: Mon, 24 Jun 2002 14:24:18 -0400
Subject: Handling HTTP Request Bodies (CGIHTTPServer)
Message-ID: <Pine.SOL.4.21.0206241417080.2626-100000@library.uwaterloo.ca>

The http server modules in the standard python library have one severe
limitation:  They don't handle request bodies.

CGIHTTPServer handles POST requests but doesn't process form values
submitted by the POST, only values submitted in the query string.

Where would I find examples (preferably built on SocketServer or
BaseHTTPServer) of what's necessary to handle request bodies properly?

Chris



From stefan1werner at hotmail.com  Tue Jun 11 16:52:51 2002
From: stefan1werner at hotmail.com (Stefan Werner)
Date: Tue, 11 Jun 2002 22:52:51 +0200
Subject: Embedding Python 2.2 in non-reentrant application
References: <ae5n1f$qtm$07$1@news.t-online.com> <m3k7p5wnt1.fsf@mira.informatik.hu-berlin.de>
Message-ID: <ae5o37$37q$03$1@news.t-online.com>

Hi,

"Martin v. Loewis" <martin at v.loewis.de> schrieb:

> No, you don't have to do anything. Inside a C function, you hold the
> GIL, and any other Python thread will block while your C function is
> running.

So, I can write Python modules in C just as I did for 1.5? What a
relief...keeping an eye on ownership and reference counting is enough work
already :)
Excuse my ignorance, but what does GIL mean?

Thanks so far,
  Stefan




From jones at jones.jones  Wed Jun 19 00:19:06 2002
From: jones at jones.jones (Jones)
Date: Wed, 19 Jun 2002 04:19:06 GMT
Subject: Writing a simple linux daemon type thing...
References: <K9eO8.393$LN6.261@newsread1.prod.itd.earthlink.net> <aeja1t$2ebi$1@news.idiom.com>
Message-ID: <_CTP8.4345$Fv1.455767@newsread2.prod.itd.earthlink.net>

I would have to guess Hungarian.... but I could guess by looking at the
words.. just the email addresses :-)

Thank you all for you suggestions.  It looks to me like the Python community
is very strong and helpful!  I'm glad I stumbled across it ;)


>
>  Searching google on "python daemon os.setsid" if find the following:
>
> http://mlf.linux.rulez.org/Archivum/l-code-l-199911/msg00060.html
>
>  Which contains some text in a language that I don't recognize
>  (Swedish?, Slavic?, who knows) and some code in Python:
>




From bokr at oz.net  Thu Jun 27 14:35:35 2002
From: bokr at oz.net (Bengt Richter)
Date: 27 Jun 2002 18:35:35 GMT
Subject: Web templating/db tool with best designer/coder separation?
References: <af2gh6$2c2$0@216.39.172.122>  <af8dtp$5qp$0@216.39.172.122> <mailman.1024976562.27993.python-list@python.org>
Message-ID: <affltn$fm6$0@216.39.172.122>

On 24 Jun 2002 22:41:51 -0500, Ian Bicking <ianb at colorstudy.com> wrote:

>On Mon, 2002-06-24 at 19:36, Bengt Richter wrote:
>> On 22 Jun 2002 18:43:50 GMT, bokr at oz.net (Bengt Richter) wrote:
>> and from the description, this would seem to have a real advantage over all of
>> the above, in that it uses attributes of html tags as opposed to special tags
>> that browsers or design tools won't generally understand and render usefully.
>
[...]
>You might be interested to look at XMLC (http://xmlc.enhydra.org/) --
>similar to ZPT, but with greater WYSIWYG support, and with considerably
>less transparency (no actual code is contained in the template).  It's
>kind of the extreme of that direction -- like accessing a DOM where the
>markup is annotated and the API is much more friendly.  It's not in
>Python, though maybe someone has created a clone (which shouldn't be
>hard).
I'd seen it in a list somewhere, but skipped it as I wasn't interested in a Java tool,
but I do like the minimal intrusion into the designer-mockup HTML world, using
"compilation" etc.

Thanks for the pointer.

Regards,
Bengt Richter


From gerhard at bigfoot.de  Tue Jun 18 21:49:11 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 19 Jun 2002 03:49:11 +0200
Subject: calling wxPython from the shell
In-Reply-To: <3D0FD764.4937BDB4@mitre.org>
References: <3D0FD764.4937BDB4@mitre.org>
Message-ID: <20020619014911.GA2739@lilith.my-fqdn.de>

* Angel Asencio <asencio at mitre.org> [2002-06-18 20:59 -0400]:
> I have been requested to write a function that will take a list, open
> a selection window, and then when the window is closed the function
> would return a list with the selected items.

> Looks like wxPython is quite favored, and I decided to use the Boa
> Constructor, finished the tutorial. Everything fine and dandy, until I
> wanted to create the function that you can call from the shell.
> 
> The scenario that I have tried is for the function to call the MainApp
> (that has the MainLoop) with the Frame that will do the selection. I
> understand about the events, but what ever I do with the event,
> 'myModule.myFunction( [ "a","collection","to","select"])' return
> nothing on the combinations I have tried.

Post source, or else nobody can help you with this particular problem.
Preferrable on the wxPython mailing  list, as that's where the
knowledgeable folks hang around (I'm just a beginner).

> Bottom line is that I want a function call that I can call in the
> shell, like as part of a  print, that I do the selection, and when I

If you want to spoil the fun of finding it out yourself, here's a quick
solution, hacked together with wxMultipleChoiceDialog and wxPySimpleApp.
Replace exec with print to see the source code. Or just execute it to
try it out ;-)

import zlib,base64
exec(zlib.decompress(base64.decodestring("""
eNptkE9LxDAQxe/5FEO9JLIWvS5UWFwQcVcXil5DbKftYP7RZmn99iapHlycQyDvZX4vM93oDMzL
6SsMzpbzAmS8GwNcs+6Po+mjbElp10+/T+bleNaBvMaHwVGD+2wz1mIHRpHlYssglvIeqkyqY6fG
nfdcsGxdwQ40TQFcB02GTFnXUaoixPMpjBsYle2R390Kkd18tLrP1P/+wF/eDocNFCdqPmFyBouL
S+SvKOqAR1JZD24+ulZpLqBK2Ke9fH3+GSCVH8kGKGrU2ARydhspqfERw7vSZ+QiRsDNPVzodYiN
fZo3RklplUEpU0QhZVqSlMUasm6MsW8XN3sP
""")))

Btw. I've learnt everything I know from studying the various parts of
the demo application, there are some very nice goodies hidden there.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 22.2 ?C      Wind: 0.9 m/s




From michael at damaru.com  Mon Jun 10 23:19:10 2002
From: michael at damaru.com (Michael Davis)
Date: Mon, 10 Jun 2002 23:19:10 -0400
Subject: newbie question - python blocking on int()?
References: <4LaN8.2488$Vr2.586019@news20.bellglobal.com> <Xns922A149F19723cliechtigmxnet@62.2.16.82> <pan.2002.06.11.03.00.52.273636.25803@fuckspam.hetdigitalegat.nl>
Message-ID: <FYdN8.3267$Ju2.867520@news20.bellglobal.com>

Wladimir wrote:

> On Tue, 11 Jun 2002 01:59:33 +0200, Chris Liechti wrote:
> 
>> also note that split() might not be safe for every case. e.g when the
>> user name contains a space (if possible) and surely when the filename
>> has a space in it.
> 
> Parsing the output of ftp.dir in any way is never comphrehensively safe
> because it it platform dependent according to the FTP RFC. Spaces in
> usernames are not possible though AFAIK, and most servers use a unix ls
> like semantic for dirs so it's *pretty safe*. (although I've seen ugly DOS
> like listings as well)

I know! I just want to make it work with two servers I use. I want to be 
able to detect when a file has changed, by a combination of modification 
time and file size. Unfortunately the ftp size() function is buggy, it 
always adds 1 whenever there is a \n in the file, (assumes, I guess, that 
you want the file size as though it were a unix text file that was going to 
be converted to dos)...

Thanks for replying!

-- 
Michael Davis
Damaru
Custom Programming - Web Development - Database Design
http://www.damaru.com
416-540-1284


From aahz at pythoncraft.com  Sat Jun 15 09:47:01 2002
From: aahz at pythoncraft.com (Aahz)
Date: 15 Jun 2002 09:47:01 -0400
Subject: Pronouncing '__init__'
References: <20020615132519.A14452@hishome.net> <mailman.1024141520.26347.python-list@python.org>
Message-ID: <aefggl$qtc$1@panix1.panix.com>

In article <mailman.1024141520.26347.python-list at python.org>,
holger krekel  <pyth at devel.trillke.net> wrote:
>Oren Tirosh wrote:
>>
>> There has been a thread about this issue in the past. Below are some of the
>> proposals people made.  I'd like to have an informal vote - how do *you*
>> say Python's magic names when talking with other Pythoneers?
>
>for oral language:
>
>    init constructor
>
>and if the context allows
>
>    init
>
>for written languange __init__
>
>Isn't it obvious? :-)

Yup.  Except that you're wrong: __init__ isn't a constructor, it's an
initializer.  __new__ is a constructor.  ;-)  So I just refer to the
"init method" so nobody gets the wrong idea.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From gb at cs.unc.edu  Sat Jun  1 07:32:35 2002
From: gb at cs.unc.edu (gb at cs.unc.edu)
Date: 01 Jun 2002 07:32:35 -0400
Subject: Code Folding with Emacs
References: <uelftqprw.fsf@cs.unc.edu> <m34rgo90d7.fsf@fkogtsp1.bmc.uu.se>
Message-ID: <un0uf6xpo.fsf@cs.unc.edu>

Thomas Sicheritz-Ponten <thomas at cbs.dtu.dk> writes:

> 
> Have you looked at folding.el?
> (fold-add-to-marks-list 'python-mode         "# {{{ "  "# }}}" nil t)
> (load "folding.el" 'nomessage 'noerror)

Yes, I looked at it but I didn't want to have to mark up my
files. Outline.el works really nicely with Python though I think
outline has a minor bug with respect to handling empty lines....

gb


From itamarst at yahoo.com  Mon Jun 10 18:45:40 2002
From: itamarst at yahoo.com (Itamar Shtull-Trauring)
Date: 10 Jun 2002 15:45:40 -0700
Subject: Python MTA
References: <mailman.1023489912.30571.python-list@python.org>
Message-ID: <637b94bc.0206101445.4d5b8315@posting.google.com>

"David LeBlanc" <whisper at oz.net> wrote in message news:<mailman.1023489912.30571.python-list at python.org>...

> Has anyone developed a simple/stupid Mail Transfer Agent using Python that
> does't rely on any external mail agents (and thus would be platform
> independent)?
> 
> What I have in mind is an app that sits between my mail client (outlook
> (dev/null=flames)) and, at most, a few mail hosts. It would act like a
> client to those mail hosts and as a host to outlook. The idea is to be able
> to insert a spam filter like ASK (or something better?).

Twisted comes with a SMTP server that can act as a smarthost -
forwarding requests on to other mailhosts. It can also act as an
incoming server, delivering mail to mailboxes which you can then read
with POP3 (also in Twisted), but this functionality is less developed.

Here's a quick guide to getting started - download and install
Twisted:
http://www.twistedmatrix.com

Debian unstable users:
apt-get install python2.1-twisted

Then:
$ mktap mail --help
$ mktap --xml mail ....   # with appropriate options

Now you have a mail.tax file you can run:
$ twistd -n -x mail.tax

mail.tax is a XML pickle, so you can edit it with a text file to
change the configuration.


From derek at wedgetail.com  Thu Jun 20 05:24:16 2002
From: derek at wedgetail.com (Derek Thomson)
Date: Thu, 20 Jun 2002 09:24:16 GMT
Subject: 'else' clause in 'for' and 'while'
References: <ad496f8.0206200107.291c5a40@posting.google.com>
Message-ID: <3D119F3F.5070600@wedgetail.com>

Ville Vainio wrote:
> How do you people use the else clause in for and while constructs? Are
> they used widely for some specific applications? Do they offer
> elegant, idiomatic solutions to some problems?
> 
> I haven't found any use for them yet myself, but knowing that they
> exist and still not using them troubles me, core python being quite
> 'small' after all.

It's good for searching, and saves mucking about with flags to achieve 
the same effect. For example:


for item in list:
     if item == match:
         # Item found, take some action
         item_found(item)
         break
else:
     # Item not found, take some other action
     item_not_found()


Regards,
Derek.



From ethanw at rediffmail.com  Thu Jun  6 18:07:24 2002
From: ethanw at rediffmail.com (Ethan)
Date: 6 Jun 2002 15:07:24 -0700
Subject: GUI Toolkit!
References: <1a52c7f1.0206011720.7c3ac7ea@posting.google.com> <kJsK8.4908$Oj4.3590@nwrddc03.gnilink.net>
Message-ID: <1a52c7f1.0206061407.4d5d4225@posting.google.com>

Thanks. I am downloading it now.

> If you need to make something very quickly, try pythoncard at
> pythoncard.sourceforge.net. It is a layer above wxWindows and comes with
> tons of samples - including an interface builder written in itself.


From peter at engcorp.com  Fri Jun  7 01:04:07 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 07 Jun 2002 01:04:07 -0400
Subject: Medium to Large Scale Python Deployments
References: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com> <mailman.1023409393.3361.python-list@python.org>
Message-ID: <3D003EC7.C72D27BE@engcorp.com>

Christopher Armstrong wrote:
> 
> > I'm also curious about large amounts of Python, period. Who knows of ~50K
> > Python lines-of-code (LOC) systems? 100K LOC? More?
> 
> chris at radii:~/Projects/Twisted$ find -name '*.py' | xargs wc -l
> ...
>   56051 total

How big is that when run through pycount?

(I.e. comments and blank lines don't count!)

-Peter


From observer at NOSPAM.space.pl  Wed Jun 19 07:42:08 2002
From: observer at NOSPAM.space.pl (Johann)
Date: Wed, 19 Jun 2002 13:42:08 +0200
Subject: PythonWorks Pro problem!
References: <q3nrgu019nlnc2bc19jaouhfo0833fknrj@4ax.com> <3lwP8.112157$pw3.6716@sccrnsc03> <8nrtgusniqrs9r3fs7l4nt8qka5fm7imi3@4ax.com> <afe3b146.0206182139.a01f19@posting.google.com>
Message-ID: <rar0huo3hp42lp1jcbg631lj2gpn3omphb@4ax.com>

On 18 Jun 2002 22:39:43 -0700, brian_l at yahoo.com (Brian Lenihan)
wrote:

>Open the "project settings" in the SYSTEM project.  You still have to
>right click on project and "reload actions" before the edit prefs
>change.

Yes. I is working now. I thought it may help me with another, greater
problem. PWpro has bugs dealing with non latin1 characters. I got a
response that it will be fixed in future. There is no way for writing
non-English characters  for now. :(

--
Johann


From jan_hermanns at gmx.de  Wed Jun  5 15:20:05 2002
From: jan_hermanns at gmx.de (Jan Hermanns)
Date: 5 Jun 2002 12:20:05 -0700
Subject: problems compiling Python2.2.1 on OpenBSD3.0
Message-ID: <cde057cb.0206051120.1b3c0894@posting.google.com>

I tried to compile Python 2.2.1 on OpenBSD3.0 (MacPPC Version),
but it does not work. After running the configure script. I tried
make. But make seems to run in an infinte loop!

Can anybody help me?

thank you

Jan Hermanns


From eric.brunel at pragmadev.com  Tue Jun  4 14:38:22 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 4 Jun 2002 18:38:22 +0000
Subject: self
References: <BI5L8.78$Vv1.4155449@newnews.cc.stevens-tech.edu>
Message-ID: <adiq31$e8f$1@wanadoo.fr>

Vojin Jovanovic wrote:
> I have been working on a project trying to implement some advanced
> functionality.  Now, one thing that has been a constraint for me is the
> concept of self.  Why is it that within a class one has to use self. (or
> "any word". ) to refer to instance variables? Would it not be more logical
> to just use the name of the variable like it is for example in C++?

This is *not* a variable, this is an attribute. Hence the "self.". I always 
found the C++/Java convention particularly disturbing. With this 
convention, there's no obvious means to tell a local variable from an 
object's attribute:

public class C {
  public int a;

  public void m() {
    int b;

    a = 0;  // What's the difference
    b = 0;  // between these 2 lines?
  }
}

The notation is exactly the same for two quite different things:
- "a = 0;" means "set the value of attribute a of method's receiver to 0", 
setting that will still exist after the method's end
- "b = 0" means "set the value of the variable b, local to method m, to the 
value 0", setting that will be discarded after the method's end

Not using any prefix in front of attribute names in methods is IMHO a 
dangerous shortcut, and a confusing one... I really prefer Python's 
notation, which is indeed a bit "heavier", but far less confusing. You'll 
certainly find your code far more readable in a few months than it would be 
in C++ or Java.
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From bokr at oz.net  Sun Jun 23 13:17:41 2002
From: bokr at oz.net (Bengt Richter)
Date: 23 Jun 2002 17:17:41 GMT
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at>
Message-ID: <af4vrl$ffe$0@216.39.172.122>

On Sun, 23 Jun 2002 11:05:02 +0200, Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote:

>Bengt Richter wrote:
>
>> Unless you are doing a lot of i/o from the program, sounds like maybe that
>> 13% (100-87) may be waiting for swapping to get more virtual space? Or it could be
>> coincidental background stuff? (Do you trigger any other processes at some point?)
>
>To cut it here: I tested the Mie code again and it works! The images 
>what I get are exactly the same as the one in a specific peer reviewed 
>paper. After that calculation I wait 20 minutes in order to watch
        ^^^^^
Not sure what you mean: Do you mean that the program completes, and then
you wait a further 20 minutes after that, expecting a possible "freeze?"
Is that something you have observed before? I.e., "freezing" _after_
successfully completing calculations, with something else working before
the "freeze"? BTW, what will Suse with your configuration do if you press
Alt-F2 etc? Will it give you a console login screen? Can you boot without
X and keep two logged-in sessions, with maybe one as root to see if you
can Alt-Fn to that when things "freeze" running your app from the other?
BTW2: you're not running your app as root, right? Does your app need X
for display or could you run it from a plain terminal console? The latter
would have a lot less baggage when trying to switch between sessions.

BTW. I heard there was one problematic version of SuSe in the past. You don't
happen to have that version I hope. Would someone set that FUD straight?
  
>whether the system will freeze or not: it freezes not. THE calculation 
>does not use any swap! I have got 256MB RAM but the calculation consumes 
>about 20 MB on Linux; there is no need for swapping (side note: my swap 
>is 756MB).
Hm, ok. 20MB and not growing. I missed that point. BTW, are you getting
your 87% number from top? If you toggle with "i" (no-idle) what else is left?
Anything besides top?

I wonder how/whether the percentage calculated is affected by power management
that slows down the CPU clock to keep it cool? Wonder if that could be happening.
I've never looked into power management stuff. I guess there is a temperature
where it just shuts down too, to protect itself?
>
>In turn I tried the following combinations:
>
>1. The complete calculations without the Mie code (note: the Mie code 
>works) and dummy values instead.
>Result: The first 10 minutes the system consumes 99.8% of the processor 
>(memory consumption is up to 5%) but after 10 minutes it sucks down to 
>87% and after 15 minutes the system freezes (I mean there is only the 
>power off option then!).
How is your system configured to respond to Ctl-Alt-Del? Does it respond
to that? If so, it's not _absolutely_ "frozen." Likewise, if you run without
a GUI with plain virtual consoles, you can often switch to another even if
if a current one appears "frozen." If you are on a local network and can
log into another machine, you could also try to telnet in to the "frozen"
machine, depending on how it's configured. You could enable telnet just for
a local test. Plain old Ctrl-C or Ctrl-Break don't do anything when it's
"frozen," right?

>
>2. I did the same without the Mie and Fortran code. Same situation as in 
>point 1.
>
>3. I did the same without Fortran/C/Satellite binary reading -> same 
>situation as in point 1 or 2.
Well, seems like it must be something about your Python code?

>
>4. I turned off every power management for my laptop, but that won't 
>cure the situation.
I don't know how that works, but I would think a system shouldn't shut
itself down if it has work to do, unless it's an emergency like running
out of battery power or overheating. So I don't understand this shutdown.
>
>
>Summarize: The system frezzes after 15 minutes and sucks down to 87% 
                       ^^^^^^^                  ^^^
In that order? How can you get a display of "87%" on a "frozen" system??

>system performance, without any notice why (I cannot see why under "top" 
>; sometimes I notice that many "kdeinits" appear right after the python 
>process)? Exactly the same situation appears on Windows: system goes 
>down from 99% to 87% after some time, but on Windows I do not get a 
>frozen system.
I guess I need a clearer definition of "frozen" ;-)
>
>I then wrote a small Python script in order to test the swap. The Python 
>script does constantly claim memory and my swap memory consumption goes 
>up to 500MB (system perfromance is only some 5%). After 30 minutes the 
>calculation is completed and I can work on without any problems.
>
Ok. Sounds normal.
>
>That means, if I have got a calculation of the form: processor 
>consumption 99% and memory consumption 5% the system fails; if I have
What exactly do you see or try that makes you conclude the system has "failed?"
 
>got a calculation in the reverse order: 90% memory and 5% processor the 
>system remains stable.
Does your program accumulate a lot of state to keep referring to as it
progresses? E.g., if information were saved in lists that grew and had
to be traversed a lot, you'd expect one kind of behavior. If you were
simulating some population of things with behavior that was either easy
or hard to calculate, and the progress of simulation forced changes to
a more complex mode of calculation (e.g., adaptive adjustments to the
resolution of a grid etc.), then you could expect increasingly bogged
down behavior, which might be proper. But if there's a bug so that
new iterations decide to shift gears using leftover parameters, you could
get degraded performance (e.g., because of unneeded high resolution calculations),
yet finally come out with valid numbers (and a false impression that Python sucked ;-)

I have no idea what your app is doing, so I am just speculating, in the hope
that something useful will trigger a helpful thought.
>
>
>Maybe this all can be explained that the ActivePython version has a bug 
>or something like that. Or that the combination Linux/Python/laptop is 
>not optimal. But I am not willing to install the normal Python version 
>again (I did it until I noticed that idle was not installed; I  invested 
>2 days in order to install idle but was not successfull; ActiveState was 
>my isle in the following).
Why not run from a plain text mode virtual terminal?

>
>The offer remains valid: if someone is interested in my source code he 
>can get it.
Ok.
>
>I am not any longer willing to try the calculation on my Linux box and 
>Python 2.2 (20 Linux power off  shut downs are enough in a day; should I 
Do you know why it's shutting down? What are the manufacturer's criteria
for shutting down? Or can some user software be calling for shutdown for
some reason? Is a screen saver involved? It must be maddening ;-)

>destroy my laptop?). In the afternoon I will decide whether I will wait 
>6 hours for the calculation on Windows or not. And I will also decide
Are there parameters you can input to the problem to control how accurately
you want things to be calculated? Can you make a low accuracy/resolution/whatever
run in a short time? It might let you do debug experiments more quickly.
Or maybe there's a bug in how that is controlled? Especially if it is adaptive.
Maybe config parameters are getting modified so all 8 (IIRC) major loops
don't see the same values?

Have you any profiling information at all to know where all the time is being
spent, and how the proportions change as things bog down?

>whether it will be a good idea to rewrite it in Fortran 90. Python can 
>then be used in combination with DISLIN for the plots.
>
Still no real diagnosis, though ;-/

Regards,
Bengt Richter


From ianb at colorstudy.com  Mon Jun  3 02:32:35 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 03 Jun 2002 01:32:35 -0500
Subject: Extracting data from HTML
In-Reply-To: <837klhrpv3.fsf@panacea.canonical.org>
References: <82096df4.0205311152.5891a17b@posting.google.com>
	<1022883953.6150.5802.camel@localhost>
	<mailman.1022908262.30070.python-list@python.org> 
	<837klhrpv3.fsf@panacea.canonical.org>
Message-ID: <1023085956.8048.87.camel@localhost>

On Sun, 2002-06-02 at 22:41, Kragen Sitaker wrote:
> Geoff Gerrietts <geoff at gerrietts.net> writes:
> > Both techniques are worth knowing -- but better than either would be
> > finding a way to get the information you're after via XML-RPC or some
> > other protocol that's designed to carry data rather than rendering
> > instructions.
> 
> You seem to imply that XML-RPC is better suited to carrying data
> rather than rendering instructions than HTTP is.  I disagree with this
> implication, and I adduce the following evidence:
> - the thousands of RSS feeds (see www.syndic8.com) using HTTP
> - people downloading Python via HTTP
> - the fact that XML-RPC runs over HTTP

I think you are misinterpreting Geoff's response, and you seem to have a
chip on your shoulder about it.  He did not compare XML-RPC to HTTP, but
to HTML (at least, that's clearly implicit because this thread was
talking about HTML parsing).  HTML is clearly a poor way to exchange
machine-readable information, there are too many layout-related tags
that are usually only appreciated by humans.

An alternative interface meant for programmatic parsing is clearly
easier and more robust to deal with fetching data, be that with XML-RPC,
or plain HTTP with normal GET/POST variables and XML response -- even a
CSV response, or newline-delimited list, or what have you would be
easier than nearly any HTML you'll find out there.

  Ian






From buzzard at urubu.freeserve.co.uk  Sun Jun 23 20:19:11 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Mon, 24 Jun 2002 01:19:11 +0100
Subject: ? and %s placeholders, help?
Message-ID: <af5oq7$sj1$1@news8.svr.pol.co.uk>

Can anyone please explain the following.  Basically I need to know the
difference between the ? placeholder and the %s placeholder.  I have only
been able to find examples of the use of '?' in situations where it clearly
works (integers?).  But I have no good idea why I cannot use it as below.
Or am I stuck with having the statement parsed each time I execute it?

Currently my INSERTs seem to be slow.  In fact it's 15 times quicker to
write my data to a text file and then dump it to a MySQL table, so I'm
guessing I should be able to speed this up somehow?  Cheers.  TIA.

Duncan

>>> import MySQLdb
>>> conn = MySQLdb.connect(db='mytests')
>>> curs = conn.cursor()
>>> query = "INSERT INTO %s (%s) VALUES %s"
>>> tbl = 'AA_temp'
>>> vars = 'var1, var2, var3, var4, var5, var6, var7'
>>> vals = '("level1", "level1", "level1", "level1", "level1", "level1",
"level1")'
>>> curs.execute(query % (tbl, vars, vals))
1L
>>> query = "INSERT INTO ? (?) VALUES ?"
>>> curs.execute(query, (tbl, vars, vals))
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
  File "C:\Python22\Lib\site-packages\MySQLdb\cursors.py", line 70, in
execute
    raise ProgrammingError, m.args[0]
ProgrammingError: not all arguments converted
>>>




From jdhunter at nitace.bsd.uchicago.edu  Fri Jun 28 12:08:24 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Fri, 28 Jun 2002 11:08:24 -0500
Subject: Python objects
References: <mailman.1025272177.1237.python-list@python.org>
Message-ID: <m21yar9yiv.fsf@mother.paradise.lost>

>>>>> "Teja" == Teja Sastry <kurugant at denshi.ece.utk.edu> writes:

    Teja> Could anybody share their experience on how to call the C++
    Teja> variable to python, I'm not quite familiar with python . I
    Teja> see there is a way to declare that as a python object in C++
    Teja> program and export it to python. I would really appreciate
    Teja> if anybody can clarify me on how do this prgramming.

This is called 'Extending Python' and there is a helpful manual
written by  Guido van Rossum and Fred Drake:
http://www.python.org/doc/current/ext/ext.html.

There are tools to automate this process.  The boost python library is
probably the best; it's template based C++ library which has extensive
support for wrapping C++ classes in python
http://www.boost.org/libs/python/doc/.  You may also want to look at
SWIG http://www.swig.org, which will generate python extensions for
C/C++ code but has limited support for some C++ features.

There are more ....

Have fun,
John Hunter


From Bart.Bartels at vu.edu.au  Thu Jun 20 02:04:00 2002
From: Bart.Bartels at vu.edu.au (Bart Bartels)
Date: Thu, 20 Jun 2002 16:04:00 +1000
Subject: Win32com Python script/MS-Access Please help!!
References: <aeo6h1$itf$1@netnews.upenn.edu> <WfdQ8.36351$I82.3834202285@newssvr12.news.prodigy.com>
Message-ID: <3D117050.E39766A1@vu.edu.au>

Sprung !!!!!

However, I would like to know as well.  I live in Aust and find it
hard (and expensive) going to school/uni in other parts of the world
:-)

Bart

Robert Oelschlaeger wrote:
> 
> Dave:
> 
> You've done a good job of re-posting (my) code from
> http://aspn.activestate.com/ASPN/Mail/Message/1225515 as
> your own work. Yes, you changed two "int"s to "float"s and
> added a "c:/" to the "open" call, and removed the "def main():",
> but that's about it.
> 
> Good luck with your homework. Let's see, that would be "OPIM
> 101: Introduction to Computers as Analysis Tools" at the Wharton
> School, University of Pennsylvania. See also
> http://opim-web.wharton.upenn.edu/home/coursedes.html#OPIM101
> 
> Next time, you might just ask for help at opim-help at wharton.upenn.edu
> 
> One of the stated aims of that course is "Build student skill and
> comfort using the computer to solve problems"; I don't
> think that they mean for you to use the computer to find
> others to do your work for you.
> 
> Bob
> 
> "Dave" <dave5774 at hotmail.com> wrote in message
> news:aeo6h1$itf$1 at netnews.upenn.edu...
> > Hey all,
> > I'm having trouble writing a Python script relating to Microsoft
> > Access.  If
> > anyone could help me I would greatly appreciate it.  Here's the
> > problem :
> >
> > I built an MS-Access table with 5 columns(fields) : OrderID, OrderDate,
> > PlantID, ProductID, OrderQty, with OrderID as the primary key.  All
> > fields
> > are numeric, except for OrderDate(which is a date field).
> >
> > Now I want to populate this table with data from a web page
> > (http://opim.wharton.upenn.edu/~opim101/spring02/Plant1.html).  please
> > check
> > the site.
> >
> > My object is to read the the data and insert it record by record into my
> > Access table.
> >
> > Here's what I have so far :
> >
> > import urllib, string
> > Plant1 =
> > urllib.urlopen
> > ('http://opim.wharton.upenn.edu/~opim101/spring02/Plant1.html'
> > )
> > lines1 = Plant1.readlines()
> > import win32com.client
> > engine = win32com.client.Dispatch("DAO.DBEngine.35")
> > db=engine.OpenDatabase(r'd:\Plants.mdb')
> >
> > I have several problems :
> > Firstly, when I read the lines, I notice that not all of them are
> > relevant.
> > I need to ignore <html>, <body>, etc. at the beginning of the file, as
> > well
> > as the column headers, so that the action really begins from line 4 of
> > the
> > file and ends with </body>
> > Also, I need to eliminate <p> from every line in the file in order to
> > get to
> > the five comma-delimited fields that follow, and then split the rest of
> > the
> > string to get to each field.  Here's what I think part of the code
> > should
> > look like :
> >
> > for line in lines1.readlines():
> >     line=line.rstrip()
> >     inOrderID,inOrderDate,inRegionID,inProductID,inOrderQty =
> > line.split(",")
> >
> > Something like that, I'm not sure...
> > Lastly, now that I've read all the fields, I need to insert them each
> > as a
> > single record into my table of my access database using Python DAO
> > using the
> > Insert statement.  Also, the OrderDate field that I read in, looks like
> > a
> > text type to Access, so I need to pad it with the # sign on either side
> > before inserting it into the Access database, so that Access interprets
> > it
> > as a Date/Time type.
> >
> > Finally, I need to get the same information from 2 other plants and tag
> > them
> > onto the same table...from the web page
> > http://opim.wharton.upenn.edu/~opim101/spring02/PlantPortal.html
> > where plant1's data shows up first, plant2's data shows up below it, and
> > plant3's data shows up last...so all plants' data in the same table.
> >
> > I would like
> > to use more standard commands.  I can use SQL, and I know Python -
> > Access
> > commands like OpenRecordSet(), TableDefs.count(), Execute(), while not
> > rs.EOF, Fields().....
> > Is there any other way to write the program without SGML?  To give you
> > more
> > of an idea of the way I'd like the program, I wrote another program for
> > doing a similar thing in Excel, not Access, for a file named sales.txt
> > and
> > the same 5 columns, but put from a file into 5 columns on excel, and
> > certain
> > parameters to enter at the beginning... :
> >
> > ProductID = float(raw_input("Please enter the ProductID: "))
> > OrderQty = float(raw_input("Please enter the cutoff order quantity: "))
> > CSV = open('c:/sales.txt','r')
> > import win32com.client
> > x1=win32com.client.Dispatch('Excel.Application')
> > x1.visible=1
> > x1.Workbooks.Add()
> > x1.Range("A1").Select()
> > x1.Range("A1:B1").Value = ("ProductID:",ProductID)
> > x1.Range("A2:B2").Value = ("OrderQty:",OrderQty)
> > x1.ActiveCell.Offset(4).Activate()
> > x1.ActiveCell.Range("A1:E1").Value=
> > ("OrderID","OrderDate","RegionID","Produc
> > tID","OrderQty")
> > x1.ActiveCell.Offset(2).Activate()
> > for line in CSV.readlines():
> >     line=line.rstrip()
> > inOrderID,inOrderDate,inRegionID,inProductID,inOrderQty = line.split
> > (",")
> > if float(inProductID) == ProductID:
> >        if float(inOrderQty) >= OrderQty:
> >            xl.ActiveCell.Range("A1:E1").Value =
> > (inOrderID,inOrderDate,inRegionID,inProductID,inOrderQty)
> > xl.ActiveCell.Offset(2).Activate()
> >
> > I just don't know how to adjust this type of thing for Access...
> >
> > If anyone could help me by writing the script I would greatly appreciate
> > it...
> > Thank you so much.
> > Sincerely,
> > Dave
> >
> >
> >
> >
> >
> >


From pyth at devel.trillke.net  Tue Jun 11 18:04:25 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 12 Jun 2002 00:04:25 +0200
Subject: RfD: Automagically making a class instance 'synchronized' (draft impl.)
In-Reply-To: <20020611205423.GA28113@lilith.my-fqdn.de>; from gh_pythonlist@gmx.de on Tue, Jun 11, 2002 at 10:54:23PM +0200
References: <20020611205423.GA28113@lilith.my-fqdn.de>
Message-ID: <20020612000425.U6609@prim.han.de>

Gerhard H?ring wrote:
> For one of my projects, I'll need something not quite unlike the
> following code. It's not tested, yet, but it looks ok and complete to
> me. I thought that it could be useful to other developers who need to
> use multithreading, so I'm posting it here. Maybe something like this
> would even be a useful addition to the Standard Library?

note that Ian Kjos some weeks ago send a nice module 'morethreading.py'
to the dev-list. IIRC, it still waits for beeing reviewed and extended
with other ideas (like synchronized methods). 

> from threading import RLock
> ...
>     def __call__(self, *args, **kwargs):
>         self.proxy.__dict__["rlock"].acquire(1)
> ...

I think for inclusion into the standard lib it *might* be
worth to reduce overhead (doubly-indirect function calls + lookups).
A quick google search showed there has already been some discussion
on this issue and there seems to be a 'synchronizing' metaclass.

anyway, another comment:

>         try:
>             self.obj(*args, **kwargs)
>         except:
>             self.proxy.__dict__["rlock"].release()
>             raise
>             return
>         self.proxy.__dict__["rlock"].release()

this can be safely substituted and then even returns the correct value<wink> with
          try:
              return self.obj(*args, **kwargs)
          finally:
              self.proxy.__dict__["rlock"].release()

the finally statement is executed in all cases and doesn't hinder
an exception to be raised eventually. I think it was designed
for use cases exactly like this :-)

regards,

    holger




From NineOfSix at gmx.de  Wed Jun 19 02:55:46 2002
From: NineOfSix at gmx.de (Axel Grune)
Date: Wed, 19 Jun 2002 08:55:46 +0200
Subject: python cgi + IIS
References: <3D0F1813.3010304@gmx.de> <x0NP8.165989$Fp1.91529@atlpnn01.usenetserver.com>
Message-ID: <3D102AF2.5010204@gmx.de>

Steve Holden schrieb:

>     http://www.e-coli.net/pyiis.html

Thats what i read and did ;-(

Thanks,
Axel



From peter at engcorp.com  Tue Jun 25 09:47:09 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 25 Jun 2002 09:47:09 -0400
Subject: Thread Memory Leak
References: <pan.2002.06.25.12.37.12.535288.17223@pobox.com> <mailman.1025010754.9180.python-list@python.org>
Message-ID: <3D18745D.554A7C44@engcorp.com>

Skip Montanaro wrote:
> 
>     Marco> while 1:
>     Marco>     listThread = []
> 
>     Marco>     for i in range(50):
>     Marco>         listThread.append( threading.Thread( None, runthread ) )
>     Marco>     for i in listThread:
>     Marco>         i.start()
>     Marco>     for i in listThread:
>     Marco>         i.join()
> 
>     Marco>     pid = os.getpid()
>     Marco>     print "==================================="
>     Marco>     os.system( "ps -o vsize,rss %d" % pid )
> 
> You never remove any thread objects from your listThread list.  It fills up
> with completed threads.

Actually, the list is rebound to a new empty list each pass through the 
while loop...

I can't duplicate the symptoms on my machine (RedHat 7.1, Python 2.0).

-Peter


From stagnoNOSPAM at prosa.it  Mon Jun 24 16:38:21 2002
From: stagnoNOSPAM at prosa.it (Marco Stagno)
Date: Mon, 24 Jun 2002 20:38:21 GMT
Subject: [ANN] my_gui: a gui module for MySQL . First public release (0.24)
Message-ID: <eo0fhu8rt6v9678ncbnm5k5de9438t9j8u@4ax.com>

This is the first public release for "my_gui". 

my_gui.py is a simple Grafic User Interface for MySql 

It can be used as Python module to create your own MySQL interface. 

Sorry, no documentation yet

GPL licence

download it at http://www.prosa.it/projects/projects

                                  MAS!



From whisper at oz.net  Wed Jun 19 20:46:05 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 19 Jun 2002 17:46:05 -0700
Subject: Flame fests (was Re: python version?)
In-Reply-To: <aer53s$7is$1@panix1.panix.com>
Message-ID: <GCEDKONBLEFPPADDJCOEKEFCDKAA.whisper@oz.net>

Darn Aahz - I was hoping to see a report that you had passed flame testing
;-)

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Aahz
> Sent: Wednesday, June 19, 2002 16:46
> To: python-list at python.org
> Subject: Flame fests (was Re: python version?)
>
>
> Everybody have fun playing "kick the newbie"?  Maybe it's time to quit?
>
> (Yes, yes, I've read the thread -- I still think it's time to revert
> c.l.py to its usual habits.)
> --
> Aahz (aahz at pythoncraft.com)           <*>
> http://www.pythoncraft.com/
>
> Project Vote Smart: http://www.vote-smart.org/
> --
> http://mail.python.org/mailman/listinfo/python-list





From spam at fisher.forestry.uga.edu  Sat Jun  1 10:20:32 2002
From: spam at fisher.forestry.uga.edu (Chris Fonnesbeck)
Date: 1 Jun 2002 07:20:32 -0700
Subject: passing large arrays from C to Python as arguments
Message-ID: <c3835e5f.0206010620.5a277bb2@posting.google.com>

I am looking for a way of passing relatively large arrays to python as
an argument to a method call, without hard-coding the length of the
array using Py_BuildValue. Is there an obvious solution to this, like
passing it in a dictionary, or something? At any rate, it is very
awkward to have to pass each element of each array into the argument
individually.

Thanks,
cjf


From jbublitzNO at SPAMnwinternet.com  Sat Jun 29 04:42:56 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Sat, 29 Jun 2002 08:42:56 GMT
Subject: I'm an idiot
References: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16> <afj27n023m7@enews2.newsguy.com> <3D1D5C5E.3A5EB81D@earthlink.net>
Message-ID: <3D1D72E1.8060000@SPAMnwinternet.com>

Joseph A Knapka wrote:
> "Beej J?rgensen" wrote:

>>Close--but I don't believe readline() throws an exception on EOF;
>>instead it just returns "".

> Now, personally, I'd like to have any read after EOF throw
> an exception. That's a nice way to break out of a loop, and
> it removes the confusion that newbies and occasionally
> oldbies (cough) sometimes experience: "If it returns '' at
> EOF, what the heck does it return for empty lines???".

> But I'm sure there are Very Good Reasons not to do this,
> and that some wise person will presently educate me :-)

It's the readline() method, meaning it reads/returns
lines, which are everything up to and including a
newline. In a text file (where this is meant to be
used), an 'empty' line contains at least a newline,
which is what readline() will return (not ''). The
only place an 'empty line' can exist is after the
final newline or after the last character after the
final newline for an unterminated last line. By
definition, that's EOF.

That's not a good explanation of why it doesn't throw
an exception - just an explanation of how readline()
works and why the exception isn't necessary. Generally,
reading past EOF isn't considered an error in C/C++
either. Look at read() or readline() in C for example.
If you want readline() to throw an exception, you can
define your own comparable function that does that.


Jim



From kosh at aesaeion.com  Fri Jun 21 03:40:15 2002
From: kosh at aesaeion.com (kosh at aesaeion.com)
Date: Fri, 21 Jun 2002 01:40:15 -0600 (MDT)
Subject: compile python witout tkinter
In-Reply-To: <20020621074746.GA765@lilith.my-fqdn.de>
Message-ID: <Pine.LNX.4.44.0206210139380.585-100000@aesaeion.com>

On Fri, 21 Jun 2002, Gerhard [iso-8859-15] H?ring wrote:

> * kosh at aesaeion.com <kosh at aesaeion.com> [2002-06-21 01:26 -0600]:
> > On a machine I need to build python on I need to build it without tkinter.
>
> Why? Even if Tkinter gets compiled, it doesn't change anything for the
> rest. You could even delete the tkinter modules afterwards, if you're
> short on disk (quota).

Because the libraries it depends on are not installed on that machine and
there is no point installing them since they are not needed.





From david.abrahams at rcn.com  Thu Jun  6 18:38:13 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Thu, 6 Jun 2002 18:38:13 -0400
Subject: numeric limits?
References: <adofuv$6v$1@bob.news.rcn.net> <200206061327.46814.gherron@islandtraining.com>
Message-ID: <1e8501c20dab$7e717a50$6601a8c0@boostconsulting.com>

Thanks, that's a big help.

-Dave
----- Original Message -----
From: "Gary Herron" <gherron at islandtraining.com>
To: "David Abrahams" <david.abrahams at rcn.com>; <python-list at python.org>
Sent: Thursday, June 06, 2002 4:27 PM
Subject: Re: numeric limits?


> On Thursday 06 June 2002 01:15 pm, David Abrahams wrote:
> > Is there some straightforward way to find out the range of a Python Int
(in
> > code)?
>
> Use sys.maxint :
>
> >>> import sys
> >>> sys.maxint
> 2147483647
>
> Gary Herron
>
>





From mike.christie at athensgroup.com  Fri Jun 21 14:20:39 2002
From: mike.christie at athensgroup.com (Mike Christie)
Date: Fri, 21 Jun 2002 18:20:39 GMT
Subject: [newbie] Overwriting class definitions
References: <3D13563D.AB80D324@athensgroup.com> <3d135729$1_2@hpb10302.boi.hp.com>
Message-ID: <3D136E77.2F6197E4@athensgroup.com>

Daniel Fackrell wrote:
> You'll need to reload(circle), as import will simply reuse a version that
> has already been loaded.

Thanks!  Much appreciated.

Mike


From ypoi at spray.se  Tue Jun 11 04:52:56 2002
From: ypoi at spray.se (DR)
Date: 11 Jun 2002 01:52:56 -0700
Subject: Getting Tkinter to accept cyrillic text
Message-ID: <83034362.0206110052.351cb526@posting.google.com>

I need Tkinter entry and label widgets to support cyrillic letters
(koi8-r). I read in some post that you should change
sys.setdefaultencoding(), but that method doesn't exist in my Python
(I use version 2.2.1) -Has this feature been changed or is my
installation bad? Also there is this locale.setlocale() thing. What
effect does changing it have?

I really would appreciate some kind of checklist, like: Do this, then
change that, then evereything works! Or a patch or something (Win95)

Is there any comprahensive recource for this type of problem? I've
only found some scattered posts so far.

Please help me out, i'm stuck!


regards, DR


From david at dataovation.com  Thu Jun 27 23:35:09 2002
From: david at dataovation.com (David McInnis)
Date: Thu, 27 Jun 2002 20:35:09 -0700
Subject: Dialing out on MODEM
Message-ID: <000601c21e54$cdd6ae80$221d7041@bigdaddy>

I need to write a small script that opens com3 and writes an ATDT to
dial out, answer the line and send a numeric message.  Basically, I
paging utility. 

I am using pyserial and cannot seem to get my modem to dial with 

Modem.write("ATDT5555555")

David McInnis





From FooWeissBarMike at hotmail.com  Mon Jun 17 10:37:46 2002
From: FooWeissBarMike at hotmail.com (Michael Weiss)
Date: Mon, 17 Jun 2002 10:37:46 -0400
Subject: System Configuration
References: <Pine.LNX.4.44.0206170010400.15076-100000@calypso.cs.brandeis.edu>
Message-ID: <3d0df505$1@news.mhogaming.com>

I'd suggest getting the win32all extensions for Python and using the Windows
API do to all of this.

http://msdn.microsoft.com/library/default.asp provides an extensive docs for
everything MS development related.
http://starship.python.net/crew/skippy/ is the site for all win32-python
goodness.


"Rajat Chopra" <rchopra at calypso.cs.brandeis.edu> wrote in message
news:Pine.LNX.4.44.0206170010400.15076-100000 at calypso.cs.brandeis.edu...
> I am looking to create a utility program that will query various aspects
> of my system, such as the hardware configuration and the software programs
> installed.
>
> I have looked into the Python module list but could not find much to do
> with system analysis. At the moment, I would like my program to function
> only on Windows 2000 systems. What modules can I use to determine such
> things as:
>
> -How much memory is installed in the system?
> -How much disk space is available on the system?
> -What is the processor speed of the system?
> -What is the bus speed?
> -Is Micrsoft Office installed on the system?
> -What version of Internet Explorer is on the system?
> -Certain registry actions, such as inserting a key, querying a key, etc.
>
> Thank you in advance for your help.
>
> Rajat
>




From david.abrahams at rcn.com  Tue Jun  4 08:01:56 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Tue, 4 Jun 2002 08:01:56 -0400
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <06da01c208f4$d69003c0$6601a8c0@boostconsulting.com> <200206010121.g511LxX19223@pcp742651pcs.reston01.va.comcast.net> <073501c20960$23f367e0$6601a8c0@boostconsulting.com> <200206011334.g51DY1D21669@pcp742651pcs.reston01.va.comcast.net> <07d801c20970$cdebe050$6601a8c0@boostconsulting.com> <mailman.1022940122.3421.python-list@python.org> <Id5K8.171932$Po6.340972@rwcrnsc52.ops.asp.att.net> <oWIK8.179611$%u2.31494@atlpnn01.usenetserver.com> <cHJK8.181202$%u2.93890@atlpnn01.usenetserver.com> <ctKK8.185481$Po6.352104@rwcrnsc52.ops.asp.att.net> <mailman.1023128707.5413.python-list@python.org> <3CFC50BD.CB9CB70C@replyto.address.invalid>
Message-ID: <adiaha$h3e$1@bob.news.rcn.net>

Funny that this topic should come up over here, since:

http://aspn.activestate.com/ASPN/Mail/Message/python-dev/1222411

minds-think-alike-ly y'rs,
dave

"Greg Ewing" <look at replyto.address.invalid> wrote in message
news:3CFC50BD.CB9CB70C at replyto.address.invalid...
> Jeff Epler wrote:
> >
> > Maybe the bytecode for inplace operations like
> >     f[i] += v
> > should implement the
> > pseudocode
> >     _temp = f[i]
> >     _new = inplace_add(_temp, v)
> >     if _new is not _temp:
> >         f[i] = _new
> >     del _temp, _new
>
> That's a nice idea...
>
> > However, this would probably double the bytecode size of the sequence,
> > and hurt performance similarly....
>
> It needn't, if:
>
> (1) the INPLACE_xxx bytecodes leave the first operand
>     on the stack
>
> (2) there are new STORE_xxx_IF_UNCHANGED bytecodes that
>     compare the value to be stored with the previous one.
>
> The total number of bytecodes should then be the same,
> and all the new work would be done in C code, which
> ought to be pretty fast (a pointer comparison and
> a few extra inc/decrefs).
>
> --
> Greg Ewing, Computer Science Dept,
> University of Canterbury,
> Christchurch, New Zealand
> http://www.cosc.canterbury.ac.nz/~greg




From vjovanov at stevens-tech.edu  Tue Jun  4 15:49:06 2002
From: vjovanov at stevens-tech.edu (Vojin Jovanovic)
Date: Tue, 4 Jun 2002 15:49:06 -0400
Subject: self
References: <mailman.1023208220.19083.python-list@python.org> <TL7L8.79$4p7.4100181@newnews.cc.stevens-tech.edu>
Message-ID: <ZQ8L8.80$1e2.4189336@newnews.cc.stevens-tech.edu>

I don't see how that is going to work.  In fact I tried things like that
before without success.  Just sticking the dictionary into eval/exec is
missing the point.  Every time parsing is done flow needs to go through
__getattr__ special method to evaluate nested equations.  Without self. you
can't achieve that effect.  Try the approach you suggested and you'll get
'#U'.


Vin


"Vojin Jovanovic" <vjovanov at stevens-tech.edu> wrote in message
news:TL7L8.79$4p7.4100181 at newnews.cc.stevens-tech.edu...
> > Which of i, j and k are instance variables?  Which are locals?  Which
are
> > globals?  For all its virtues, Python isn't telepathic <wink>.
>
> Well, now that you pointed it out, I think that Python should be
telepathic.
> Those i,j and k variables should have been found in class C, otherwise
they
> should have been searched outside of the class.  I read the FAQ page that
> somebody sent me and I really think that the main reason for self is ...
> "When classes were added to Python, this was (again) the simplest way of
> implementing methods without too many changes to the interpreter. " which
is
> the main and really lousy argument.  All, other arguments are debatable.
>
> Now let me give you the problem with self which is making my life
> complicated. Consider this class.
>
> class Foo:
>  def __init__(self):
>   self.__dict__['Equations']={
>     'a':str(5),
>     'b':'self.a+5',
>     'c':'self.b*self.a'}
>
>  def __getattr__(self,arg):
>   if self.Equations.has_key(arg) == 0:
>    raise AttributeError,  "Foo instance has no attribute '" + arg + "'"
>   try:
>    try:  #first check if this is a simple expression
>     return eval(self.Equations[arg])
>    except:  #if not try to execute it
>     exec self.Equations[arg]
>   except:   #if even that didn't work then just return undefined
>    return '#U'
>
>  def __setattr__(self,arg,val):
>   self.__dict__['Equations'][arg]=str(val)
>
>  def __delattr__(self,arg):
>   del self.__dict__['Equations'][arg]
>
> This class implements a kind of symbolic behavior with the ability to add
> new equations on the fly.  For example...
>
>
> >>> import foo
> >>> A=foo.Foo()
> >>> A.Equations
> {'b': 'self.a+5', 'c': 'self.b*self.a', 'a': '5'}
> >>> A.f='self.b*self.c'
> >>> A.f
> 500
> >>> A.Equations
> {'f': 'self.b*self.c', 'b': 'self.a+5', 'c': 'self.b*self.a', 'a': '5'}
>
>
> As you can see, equations get added on the fly and the parsing gets done
as
> the attribute is being called.  Now, notice that ugly self. thing in the
> equations.  Instead of just typing >>>A.f='a*c' I have to add self in
front
> a and c to achieve the desired effect.  I could not find any other way to
> implement the class in order to avoid typing self. which of course it
> doesn't mean that there isn't any. If anybody has an idea how to do this
> please let me know.  Otherwise, here is the situation where the self
> concept is bad and having such a dynamic language like Python I think
there
> has to be a way out of this problem.
>
> Why is this all relevant? Well, I am working on embedding an interpreted
> language in a large engineering application and Python is  one of few
> candidates.  So I have been looking for its pros and cons.  So far it
passed
> all the tests except the above.  Because, imagine if we required from the
> user of our system that he needs to type self. or even s. in front of
every
> new variable that he has in his equation, it would be a big turn off.
>
> Vin
>
>
>




From peter at engcorp.com  Thu Jun 20 21:04:09 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 20 Jun 2002 21:04:09 -0400
Subject: RE strings (was: Variable Interpolation - status of PEP 215)
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <4PfQ8.44245$n4.10307683@newsc.telia.net> <KhiQ8.44269$n4.10312432@newsc.telia.net> <29e28c51.0206201156.7e61a7f5@posting.google.com>
Message-ID: <3D127B89.EE1C17F4@engcorp.com>

Cimarron Taylor wrote:
> 
> I like it!
> 
> To me
> 
>   m = re'^EMP:([^,]*),([^,]*),([^,]*),([^,]*)$'.match(line)
>   if m:
>         ...
> 
> is much clearer and more maintainable than
> 
>   import re
>   emp_re = re.compile(r'^EMP:([^,]*),([^,]*),([^,]*),([^,]*)$')
>   m = re.match(emp_re, line)
>   if m:
>       ...

But is it really much clearer and more maintainable than

import re
m = re.match(r'^EMP:([^,]*),([^,]*),([^,]*),([^,]*)$', line)
if m:
    ...

Given that the only real difference is you skip an import, and
at the cost of modifying the language definition in an unusual
manner, is there any point to this at all?

-Peter


From milos.prudek at tiscali.cz  Sat Jun  8 08:03:17 2002
From: milos.prudek at tiscali.cz (Milos Prudek)
Date: Sat, 08 Jun 2002 14:03:17 +0200
Subject: htmllib and parsing data sourrounded by HTML tags
Message-ID: <3D01F285.5010103@tiscali.cz>

I studied htmllib hard and I was able to subclass it and do some
processing of HTML tags. I managed to extend it to process <TD>, for
instance.

But what about the text that is surrounded by HTML tags? Can sgmllib,
htmllib and formatter be leveraged to find data sourrounded by HTML tags
and process them?

An example HTML:
<TABLE>
<TR>
<TD>First Name:</TD><TD>Peter</TD>
<TD>Last Name:</TD><TD>Smith</TD>
</TR>
</TABLE>

The task is to locate second row in this table, and pull string from
second column of this row, i.e. "Smith".

Naturally this can be done without htmllib, but for the sake of
systematic approach I'm interested whether it can be done with htmllib.
It seems that htmllib walks on <HTML> tags but leaves everything that is
not a HTML tag unprocessed and writes it out "as is", and there are not
hooks to attach to. Also it seems that there would have to be a "row
counter" and "column counter" attributes in sgmllib, but I do not feel
smart enough to understand sgmllib enough to subclass it...

Should I forget about htmllib because it is unsuitable for this?

-- 
Milos Prudek






From dryose at juno.com  Mon Jun  3 16:02:39 2002
From: dryose at juno.com (yose)
Date: 3 Jun 2002 13:02:39 -0700
Subject: mySQL on Solaris .8
References: <79d1869.0206030646.512af52c@posting.google.com>
Message-ID: <79d1869.0206031202.228c4387@posting.google.com>

dryose at juno.com (yose) wrote in message news:<79d1869.0206030646.512af52c at posting.google.com>...
> I am trying to install a mySQL database in solaris 2.8. I have a copy
> of the distro for mySQL-python-0.9.1. When I try to run setup.py, I am
> told that I am missing distutils.core
> 
> Where do I get this? What package is it part of?
> 
> Thanks!


Let me follow-up to my original message:

I need to get a mySQL database working in Solaris, and I 
need to access it via the Python API.

I would appreciate any help that you can offer!!

Joe


From derek at wedgetail.com  Mon Jun 24 22:30:15 2002
From: derek at wedgetail.com (Derek Thomson)
Date: Tue, 25 Jun 2002 02:30:15 GMT
Subject: I'd give up Perl tomorrow if only...
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au> <3D12F6FC.3000608@wedgetail.com> <slrnahekqq.12v.msoulier@pmerd071.ca.nortel.com>
Message-ID: <3D17D5B6.1090203@wedgetail.com>

Michael P. Soulier wrote:
> On Fri, 21 Jun 2002 09:50:54 GMT, Derek Thomson <derek at wedgetail.com> wrote:
> 
>>Among other things, it's one of the reasons I'm tinkering with a Perl ORB.
> 
> 
>     There's also XMLRPC, and SOAP, something already supported by both
> languages. 

These are no good for reasons I've described elsewhere on this thread. 
To summarize, the major problem is this - "How do you represent object 
references?". It's the difference between a marshalling scheme (which is 
what XMLRPC, SOAP, and CORBA-IIOP are) and a scalable architecture for 
distributed systems, which is what CORBA is (distributed as opposed to 
merely "client-server").

The simple answer for those who have large bandwidth, and like to waste 
it, and therefore want to use SOAP and/or XMLRPC as an on-the-wire 
protocol, is to add a CORBA-like infrastructure on top of, say XMLRPC, 
which I have in fact done with Fnorb in the xmlrpc branch on SourceForge.

Regards,
Derek.



From tebeka at cs.bgu.ac.il  Mon Jun 17 08:54:06 2002
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 17 Jun 2002 05:54:06 -0700
Subject: Help: win32gui and threads, process won't shut down
Message-ID: <33803989.0206170454.4609ebab@posting.google.com>

Hello All,

Sorry if this is the 2'nd time I'm sending this. My browser hang up
sending the 1'st one.

I've wrote a simple logger daemon which has UI through Window's
taskbar.
(See code below)

However when I close it from the taskbar the windows shuts down but
the pythonw process still lives and I have to kill it explicitly.

Any ideas how to make is shutdown properly?

TIA
Miki

--------------
#!/usr/bin/env python
'''Logger daemon

Opens socket server and write messages to log file
'''

__author__ = 'Miki Tebeka <miki_tebeka at amat.com>'
__version__ = '$Revision: 1.3 $'

# $Source: /home/miki/.cvsroot/logging/logd.py,v $

import SocketServer
import socket
import Queue
import threading
import time
import re

from win32api import *
from win32gui import *
import win32con
import sys
import os, os.path

MESSAGE_QUEUE = Queue.Queue(0) # Thread safe message queue with no
limit

# These two are not protected by mutex since only the main thread
write there and
# the worker thread reads them
WRITE = 1 # Write flag
NEW_LOG = 0 # Create new file

def error(msg):
    '''Error function'''
    MessageBox(win32con.MB_OK, msg)

class SocketListener(SocketServer.BaseRequestHandler):
    '''Socket listener. Reads line from socket and place it in message
queue'''
    
    def __init__(self, request, client_address, server):
        '''Constructor'''
        SocketServer.BaseRequestHandler.__init__(self, request,
client_address, server)

    def handle(self):
        '''Handler function. Reads line from socket and writes to
queue'''
        msg = ''
        while 1:
            try:
                c = self.request.recv(1)
                if not c:
                    return
                if (c == '\n'):
                    MESSAGE_QUEUE.put(msg)
                    msg = ''
                else:
                    msg += c
            except socket.error, e:
                error('Error: %s' % e)
                return

class LoggerDaemon(SocketServer.ThreadingTCPServer, threading.Thread):
    '''Logger Daemon. Listens on given port and threads listeners upon
requests'''
    def __init__(self, port):
        self.port = port
        self.addr = ('', self.port)
        threading.Thread.__init__(self)
        SocketServer.ThreadingTCPServer.__init__(self, self.addr,
SocketListener)

    def run(self):
        self.serve_forever()

class LogWriter(threading.Thread):
    '''Log writer. Get messages from queue and write to file'''
    def __init__(self, log_dir):
        self.log_dir = log_dir
        if not os.path.isdir(self.log_dir):
            try:
                os.makedirs(self.log_dir)
            except OSError, ose:
                error("Can't create log directory %s (error was: %s)"
% (self.log_dir, ose))
                sys.exit(1)
        self.new_log()
        threading.Thread.__init__(self)

    def run(self):
        '''Thread function'''
        global NEW_LOG, WRITE
        while 1:
            msg = MESSAGE_QUEUE.get()
            if NEW_LOG:
                self.new_log()
                NEW_LOG = 0 # This is a sync NO-NO but we can live
with that
            if WRITE:
                print >> self.fp, msg.strip()
                self.fp.flush()

    def new_log(self):
        '''New log file'''
        logname = '%s/%s' % (self.log_dir, re.sub('(\s+)|:', '_',
time.ctime()))
        self.log_file = logname
        self.fp = open(self.log_file, 'a')

class MainWindow:
    '''Main window for handling taskbar notifications'''
    def __init__(self):
        '''Constructor'''
        self.ID_TBAR = win32con.WM_USER + 20
        self.icon_file = './logd.ico'
        self.title = 'BeeLogger Daemon'
        self.IDM_TOGGLE, self.IDM_NEWLOG, self.IDM_QUIT = 1023, 1024,
1025
        self.resume_pause_msg = ('Resume', 'Pause')
        message_map = {
            win32con.WM_DESTROY: self.OnDestroy,
            win32con.WM_COMMAND: self.OnCommand,
            self.ID_TBAR : self.OnTaskbarNotify,
        }
        
        # Register the Window class.
        wc = WNDCLASS()
        hinst = wc.hInstance = GetModuleHandle(None)
        wc.lpszClassName = "BeeLoggerDaemon"
        wc.style = win32con.CS_VREDRAW | win32con.CS_HREDRAW;
        wc.hCursor = LoadCursor( 0, win32con.IDC_ARROW )
        wc.hbrBackground = win32con.COLOR_WINDOW
        wc.lpfnWndProc = message_map # could also specify a wndproc.
        classAtom = RegisterClass(wc)
        # Create the Window.
        style = win32con.WS_OVERLAPPED | win32con.WS_SYSMENU
        self.hwnd = CreateWindow( classAtom, self.title, style, \
                    0, 0, win32con.CW_USEDEFAULT,
win32con.CW_USEDEFAULT, \
                    0, 0, hinst, None)
        UpdateWindow(self.hwnd)

        # Try and find a custom icon
        if not os.path.isfile(self.icon_file):
            icon_flags = win32con.LR_LOADFROMFILE |
win32con.LR_DEFAULTSIZE
            self.hicon = LoadIcon(0, win32con.IDI_APPLICATION)
        else:
            self.hicon = LoadImage(hinst, self.icon_file,
win32con.IMAGE_ICON, 0, 0, icon_flags)

        flags = NIF_ICON | NIF_MESSAGE | NIF_TIP
        nid = (self.hwnd, 0, flags, self.ID_TBAR, self.hicon,
self.title)
        Shell_NotifyIcon(NIM_ADD, nid)

    def OnDestroy(self, hwnd, msg, wparam, lparam):
        nid = (self.hwnd, 0)
        Shell_NotifyIcon(NIM_DELETE, nid)
        PostQuitMessage(0) # Terminate the app.

    def OnTaskbarNotify(self, hwnd, msg, wparam, lparam):
        global WRITE
        if lparam==win32con.WM_LBUTTONUP or
lparam==win32con.WM_LBUTTONDBLCLK or lparam==win32con.WM_RBUTTONUP:
            menu = CreatePopupMenu()
            AppendMenu(menu, win32con.MF_STRING, self.IDM_TOGGLE,
self.resume_pause_msg[WRITE])
            AppendMenu(menu, win32con.MF_STRING, self.IDM_NEWLOG, 'New
Log')
            AppendMenu(menu, win32con.MF_STRING, self.IDM_QUIT,
'Quit')
            pos = GetCursorPos()
            TrackPopupMenu(menu, win32con.TPM_LEFTALIGN, pos[0],
pos[1], 0, self.hwnd, None)
        return 1

    def OnCommand(self, hwnd, msg, wparam, lparam):
        global WRITE, NEW_LOG
        id = LOWORD(wparam)
        if id == self.IDM_TOGGLE:
            WRITE = 1 - WRITE
        elif id == self.IDM_NEWLOG:
            NEW_LOG = 1
        elif id == self.IDM_QUIT:
            DestroyWindow(self.hwnd)
        else:
            raise ValueError('Unknown command: %s' % id)


# MAIN
if __name__ == '__main__':
    if len(sys.argv) != 3:
        print >> sys.stderr, 'usage: logd LOG_DIR PORT'
        sys.exit(1)
    
    # Start writer
    writer = LogWriter(sys.argv[1])
    writer.start()

    # Start TCP server
    logd = LoggerDaemon(int(sys.argv[2]))
    logd.start()
        
    # Start UI
    w=MainWindow()
    PumpMessages()

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


From martin at strakt.com  Tue Jun  4 03:24:37 2002
From: martin at strakt.com (Martin =?iso-8859-1?Q?Sj=F6gren?=)
Date: Tue, 4 Jun 2002 09:24:37 +0200
Subject: 'for every' and 'for any'
In-Reply-To: <7934d084.0206031636.5c29b31f@posting.google.com>
References: <20020526135944.A32690@hishome.net> <20020526091742.A987@unpythonic.net> <mailman.1022567685.7803.python-list@python.org> <slrnaf7af3.8kn.quinn@schilling.ugcs.caltech.edu> <mailman.1022658469.27441.python-list@python.org> <7934d084.0206031636.5c29b31f@posting.google.com>
Message-ID: <20020604072437.GA21777@strakt.com>

On Mon, Jun 03, 2002 at 05:36:46PM -0700, Andrae Muys wrote:
> 
> Now I just don't understand how this position can possibly make sense?
>  C dosn't have any 32-bit int types standardised, so what do _you_ use
> when you need one?  Please don't tell me you don't use int, or I'll
> have to hunt you down and put you out of your misery ;).  I personally
> have absolutely no problem with any of the 3 examples you gave above:

Doesn't C99 specify <stdint.h>? With int8_t, int16_t, ..., uint8_t,
uint16_t, ...


Martin

-- 
Martin Sj?gren
  martin at strakt.com              ICQ : 41245059
  Phone: +46 (0)31 7710870       Cell: +46 (0)739 169191
  GPG key: http://www.strakt.com/~martin/gpg.html




From dougz at cs.washington.edu  Mon Jun 24 18:57:06 2002
From: dougz at cs.washington.edu (Douglas Zongker)
Date: Mon, 24 Jun 2002 22:57:06 GMT
Subject: adding attributes to builtin functions
Message-ID: <Gy8Fr6.J7D@beaver.cs.washington.edu>

I know that when you def a function in Python you can add arbitrary
attributes to it:

   def foo():
      return 4
   foo.special = 6

Is it possible to do this with a function that comes from a C
extension module?

   import mymodule
   print mymodule.myfunction
   mymodule.myfunction.special = 6

gives me:

   <built-in function myfunction>
   Traceback (most recent call last):
     File "<stdin>", line 1, in ?
   TypeError: 'builtin_function_or_method' object has only read-only
     attributes (assign to .special)

Is there a way to define a builtin function whose attributes aren't
read-only?

thanks,
dz

   



From bobx at linuxmail.org  Thu Jun 13 17:35:03 2002
From: bobx at linuxmail.org (Bob X)
Date: Thu, 13 Jun 2002 21:35:03 GMT
Subject: Python 2.2.1
References: <II7O8.622$e27.275217555@twister1.starband.net>
Message-ID: <3D09100A.4040201@linuxmail.org>

If you go to www.activestate.com and get the ActivePython setup it will 
do it all for you.  : )

Dave wrote:
>     Hi, I am 12 years old and would like to begin learning Python over 
> the summer. I have Windows 98 SE. I have python 2.2.1 and all of its 
> utilities on my computer. I am taking a tutorial 
> from  http://softwaredev.earthweb.com/sdopen/article/0,,12382_626311,00.html  
> If you go to it and scroll down to the paragraph entitled *Getting 
> Started* there is a paragraph that says, *"You will need to cause the 
> directory containing the file named python.exe to be listed in your 
> system environment variable named path. If you already know how to set 
> the path, go ahead and do it.  If you don't already know how, you may 
> need to get some help." *
> ** 
>  
>     How do I set the path and what should the name of the path be? All 
> help and information would be appreciated. Thanks.:)



From altis at semi-retired.com  Fri Jun 14 12:11:13 2002
From: altis at semi-retired.com (Kevin Altis)
Date: Fri, 14 Jun 2002 09:11:13 -0700
Subject: ANN: PythonCard 0.6.7
Message-ID: <gyoO8.45$YM3.33091@news.uswest.net>

PythonCard is a GUI construction kit for building cross-platform desktop
applications on Windows, Mac OS X, and Linux.

Release 0.6.7 includes over 30 sample applications, new additions include
chat, webserver, pictureViewer, slideshow, and webgrabber. Four of the
samples have been promoted to tools status: codeEditor, findfiles,
resourceEditor, and textEditor. This release supports the new wxPython 2.3.3
preview for Mac OS X.

All the information you need about PythonCard can be found on the project
web page at:
http://pythoncard.sourceforge.net/

The installation instructions and walkthroughs are available on the main
documentation page:
http://pythoncard.sourceforge.net/documentation.html

You can download the latest release at:
http://sourceforge.net/project/showfiles.php?group_id=19015

For a list of some of the samples that have been built with PythonCard and
screenshots of them in action go to:
http://pythoncard.sourceforge.net/samples.html
A description of each sample is included in the readme.txt file in each
sample directory.

The kind people at SourceForge host the project:
http://sourceforge.net/projects/pythoncard/

If you want to get involved the main contact point is the Mailing list:
http://lists.sourceforge.net/lists/listinfo/pythoncard-users

PythonCard requires Python 2.1.x or later and wxPython 2.3.2.1 or later.
wxPython can be downloaded at http://www.wxpython.org/

ka
---
Kevin Altis
altis at semi-retired.com





From jiri at baum.com.au  Sun Jun  2 04:17:38 2002
From: jiri at baum.com.au (Jiri Baum)
Date: Sun, 2 Jun 2002 18:17:38 +1000
Subject: How to call functions inside a cgi script from the html interface?
References: <adc4sv$vjrdc$2@ID-49758.news.dfncis.de>
Message-ID: <2bkcda.c25.ln@10.0.0.1>

Ken:
> How to call functions inside a cgi script from the html interface?

> Eg. If there is a link on the html display call "Search", how do I get it
> to call the function call "def search( )" inside the cgi script?

You need to get yourself a CGI programming reference, or even a good HTML
book, and refer to them before you post questions on the (wrong) newsgroup.

Also read the docs for the python cgi module (eg at www.python.org).

HTTP doesn't have any facilities for doing this directly. Every time a user
selects a link, a request is made to the server for that page. This can
either be a completely different cgi script, or it can be the same script
with different parameters. Either way, it's a separate connection, so don't
expect variables to stay.


Jiri
-- 
Jiri Baum <jiri at baum.com.au>           http://www.csse.monash.edu.au/~jirib
  MAT LinuxPLC project --- http://mat.sf.net --- Machine Automation Tools


From pyth at devel.trillke.net  Tue Jun 18 13:07:09 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 18 Jun 2002 19:07:09 +0200
Subject: why len(list) instead of list.len()  ?
In-Reply-To: <c5HP8.243362$cQ3.10419@sccrnsc01>; from emile@fenx.com on Tue, Jun 18, 2002 at 02:03:52PM +0000
References: <EhBP8.1$Qn1.1124@news.ecrc.de> <c5HP8.243362$cQ3.10419@sccrnsc01>
Message-ID: <20020618190709.D15079@prim.han.de>

Emile van Sebille wrote:
> "Markus Jais" <mjais at web.de> wrote in message
> news:EhBP8.1$Qn1.1124 at news.ecrc.de...
> > hello
> >
> > I way wondering why one has to write
> >
> > list = [2, 3, 5]
> > print len(list)
> >
> > instead of list.len()
> >
> > in Ruby I can write
> >
> > array = [2, 3, 4]
> > puts array.length()
> >
> >
> 
> Because a two-line wrapper goes a long way?
> 
> Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
> Type "help", "copyright", "credits" or "license" for more information.
> >>> class myList(list):
> ...     len = list.__len__
> ...
> >>> l = myList([1,2,3,4])
> >>> l
> [1, 2, 3, 4]
> >>> l.len()
> 4
> >>> l.extend([5,6,7])
> >>> l.len()
> 7

this 'hands-on fix' has come up before, too. It bears the problem that 
you would have to wrap every list that you work with. Otherwise you get nasty
inconsistencies. 

Actually i think that there might a solution for the 'why does len not
work as expected'- problem.  Once the builtin types cannot not only be
extended by inheritance but also by directly binding names in the types
attribute space.  Something like:

    list.len=list.__len__

could then really fix this problem.  Of course there would still be major
combats over the question whether it should be included by default :-) 

    holger




From dfackrell at DELETETHIS.linuxmail.org  Mon Jun 24 10:47:14 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Mon, 24 Jun 2002 08:47:14 -0600
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com> <lkznxqjqx2.fsf@pc150.maths.bris.ac.uk> <3D11A21D.9010809@mxm.dk> <uh3e1usajle42c@news.supernews.com> <3D11DA9B.163A7BD0@info.unicaen.fr> <uh446elev7af68@news.supernews.com> <3D12EEAA.9C3A3D29@info.unicaen.fr> <6v30fa.sm1.ln@grey.aerojockey.localdomain>
Message-ID: <3d1730f2$1_1@hpb10302.boi.hp.com>

After skimming the linked page to get a feel for the CAME FROM statement, is
it conceptually different from short procedures?  Does the only difference
lie in the fact that for compiled languages one of the instances will be
inline?  I do understand that it would have some effect on variable scoping,
but this can be easily handled in other ways.

So what's the benefit that might outweigh the opacity of the code created?

--
Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.


"Carl Banks" <imbosol at vt.edu> wrote in message
news:6v30fa.sm1.ln at grey.aerojockey.localdomain...
> Jerzy Karczmarczuk wrote:
> > (Oh, I adored that wonderful piece of programming madness, an "article"
> > where somebody proposed a new super-Fortran where GO TO was replaced by
> > CAME FROM statement, and the language had an inspiring non-deterministic
> > conditional:   If <<condition>> THEN WHAT?.
> > Any references, anybody?)
>
>
> http://www.fortran.com/fortran/come_from.html
>
>
> --
> CARL BANKS                                http://www.aerojockey.com
> "Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."




From observer at NOSPAM.space.pl  Thu Jun 13 04:57:21 2002
From: observer at NOSPAM.space.pl (Johann)
Date: Thu, 13 Jun 2002 10:57:21 +0200
Subject: How to check version of Python under CGI
References: <rbmggus7ln1uiutktf5libf4bbngu29ji7@4ax.com> <873cvrin1q.fsf@tux.ntw23.fr>
Message-ID: <cgnggu40ethjhpqcmltnfb06d5qvuf078k@4ax.com>

On 13 Jun 2002 10:46:09 +0200, Sylvain Thenault <thenault at nerim.net>
wrote:

>> I would like to write a simple CGI script to get as much info about
>> installed Python as possible. But I even don't know how to check
>> version of Python. :( From shell command  it is easy, but how to do it
>> from CGI script? Eg.
>
>print "Python version: %s" % sys.version

Thanx. Do you know how to check what modules are installed? E.g. I
would like to check if PYANA or other xml modules are installed.

--
Johann


From aurelienviolet at noos.fr  Sat Jun  1 22:30:41 2002
From: aurelienviolet at noos.fr (aurel-le-gnou)
Date: Sun, 02 Jun 2002 04:30:41 +0200
Subject: Help for Tkinter and python's programming
Message-ID: <3CF98351.95763821@noos.fr>

Hi,


I'm french and just beginning in Python's programming.
I'm doing a really simple GUI with the module "Tkinter" for the
implementation of a meta search engine based on Google and altavista.

This Gui is quite finished but i've a problem. I'm sure it's really
stupid but I really not understand the reasons of this pb.
If someone can Help me, it would be really fine.


Here is my problem:

I create a GUI with some buttons and entry, textbox, radiobuttons...
Then there is a mainloop wainting for events.
When the button 1 is pressed on a button, i want to get each content of
these entries... and the state of the radiobuttons...
Then launch a program with the differents values and contents found just
before.
I don't success in associating the event and the function to execute
with the different arguments


All my tkinter widgets are created in a simple main by calling some
functions (no object implementation, this interface is really simple).
Particularly these two functions define naturally before the main:


def make_button(frame, text, color, side, padx, pady):

    button = Button(frame, text=text, fg=color, bg='white')
    button.pack(side=side,padx=padx, pady=pady)
    return button


def make_call(frame_left, frame_right, list_one, list_two,
button_number, button_search, button_language, button_rank,
button_engine):

    ....
    ....
    
    return button_language_content, button_number_content,
button_search_content, list_one, list_two, button_rank_content,
button_engine_content



The goal of the function make_call is:
- to get the differents arguments in the diferents Entry and the state
of different radiobuttons
- then call an other program program with these arguments.



In my main:

...

button_go = make_button(bottom,'GO','black',LEFT, '5m',0)

...
...


button_go.bind("<Button-1>", make_call(frame_left, frame_right,
list_one, list_two, button_number, button_search, button_language,
button_rank, button_engine))  
# the arguments of make_call() are visible and defined in the main when
I do this "bind".

....

root.mainloop()


So, I just want to execute the fonction make_call with some arguments
(which corresponds to the different entries and radiobuttons to check
their content) when the button1 is pressed on my button_go. (this
arguments are visible in my main) 
But the problem is that even if the user doesn't click on the button the
associate function is executed.
I've also problem in giving arguments to make_call() 
(the number of arguments is sometimes wrong ... Is there a default
argument "event" given to the function with some informations ?)


Thanks a lot for your help and your criticm about what i've done wrong.



Aur?lien Violet

aurelienviolet at noos.fr




From mgilfix at eecs.tufts.edu  Mon Jun 10 08:49:17 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Mon, 10 Jun 2002 08:49:17 -0400
Subject: Does Python need a '>>>' operator?
In-Reply-To: <Pine.GSO.4.33.0206100850550.12342-100000@techst02.technion.ac.il>; from cben@tx.technion.ac.il on Mon, Jun 10, 2002 at 08:57:15AM +0300
References: <mailman.1018919132.7451.python-list@python.org> <a9gnvs$42p$0@216.39.172.122> <Pine.GSO.4.33.0206100850550.12342-100000@techst02.technion.ac.il>
Message-ID: <20020610084917.C23641@eecs.tufts.edu>

On Mon, Jun 10 @ 08:57, Beni Cherniavksy wrote:
> I just got another idea: use 0x1234 for 0-filled numbers and 1xABCD for
> 1-filled ones.  That way you impose no restrictions on what follows the
> prefix and keep backward compatibility.  0xFFFFFFFF stays a 2^n-1
> _positive_ number, as it should be.  The look of 1x is weird at first but
> it is very logical...

  Just a vote of confidence. I like this solution a lot. Seems the
best of both worlds, since it leaves the current scheme alone and it
doesn't mean fiddling with expected sizes.

                        -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From robin at jessikat.fsnet.co.uk  Sun Jun 30 07:53:27 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sun, 30 Jun 2002 12:53:27 +0100
Subject: yield equivalent in C/JavaScript?
References: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk>
 <7xy9cxhlvx.fsf@ruckus.brouhaha.com> <w3xdKUAuljH9EwgH@jessikat.fsnet.co.uk>
 <7xit41i7i9.fsf@ruckus.brouhaha.com> <14ms5VAygsH9Eww2@jessikat.fsnet.co.uk>
 <7xelepnk84.fsf@ruckus.brouhaha.com>
Message-ID: <WrXUtJA3EvH9EwER@jessikat.fsnet.co.uk>

In article <7xelepnk84.fsf at ruckus.brouhaha.com>, Paul Rubin <phr-
n2002b at NOSPAMnightsong.com> writes
>Robin Becker <robin at jessikat.fsnet.co.uk> writes:
>> However the content.gz data has to be exactly compatible with gzip. My
>> experiments with Python-2.1/zlib on freeBSD give me a compressed string
>> that's slightly too long for some reason. As an example gzip --> 14099
>> bytes while zlib.compress-->15008 bytes.
>> 
>> Anyone know what I need to do to make zlib work properly in this
>> context? Or should I really be using the gzip module?
>
>You're probably better off using mod_gzip (an apache module that
>compresses the output automatically) rather than trying to compress
>from Python.  Note that gzip encoding doesn't work so well in some IE
>versions.
Unfortunately I don't think the server uses apache, but maybe it can be
persuaded to do it automatically.

The IE stuff seems to rely too much on the URL extension so provided it
thinks it's htm or html IE 5.x seems OK.
-- 
Robin Becker


From morningstar at yubc.net  Sun Jun 23 13:35:35 2002
From: morningstar at yubc.net (Ivan J.)
Date: 23 Jun 2002 10:35:35 -0700
Subject: printing on the browser
Message-ID: <d9531d76.0206230935.6e7a7cab@posting.google.com>

Hi!
This is actually cgi quesion, but ...
Is there a way to print both HTML and IMAGE parts using:
>>>print "Conent-type: ...",
statement, if not is there any other way to create web page with text
and images without savingimages onto disk?
Thank you!


From "aglyportat\" at n-o.spa__mmm.yahoo dott com>  Sat Jun 15 22:40:43 2002
From: "aglyportat\" at n-o.spa__mmm.yahoo dott com> (Anton Graph)
Date: Sat, 15 Jun 2002 19:40:43 -0700
Subject: RH7.2 2.4.X off_t: long or long long?
References: <ugnrs8ke4fl17d@corp.supernews.com>
Message-ID: <ugnth4s84tdoe1@corp.supernews.com>

Anton Graph wrote:
> I have an odd problem trying to link my app:
> 
> glog.o(.text+0x100a): undefined reference to `loadfile(char const *, 
> long long *)'
> collect2: ld returned 1 exit status
> 
> 
> in loadfile.C:
> 
> char *
> loadfile(const char *path, off_t *length)
> {
>     ...
> 
> and the declaration in the header is:
> 
> char *loadfile(const char *path, off_t *size=0);
> 
> nm -C libgen.a |grep loadfil
>          U loadfile(char const *, long *)
> loadfile.o:
> 00000000 T loadfile(char const *, long *)
>          U loadfile(char const *, long long *)
> 
> Why are there two references differing in the last arg? Is off_t 
> supposed to be long or long long?
> 
> Thank you!
> 


Nevermind, this had something to do with Python.h that I've included.
Removed the inclusion of Python.h and everything works fine.

For some odd reason

/* The number of bytes in an off_t. */
#define SIZEOF_OFF_T 8

is defined in pyconfig.h which is used by Python.h but oddly there is no 
redefinition of off_t anywhere in Python files. Kinda puzzling.



From ehab_teima at hotmail.com  Wed Jun  5 19:14:14 2002
From: ehab_teima at hotmail.com (Ehab Teima)
Date: 5 Jun 2002 16:14:14 -0700
Subject: minidom and encoding problem
Message-ID: <17aafe08.0206051514.4af054d2@posting.google.com>

Hi,

I'm using Python 2.1. I wrote classes to create xml document from
scratch. The code worked fine until I hit an encoding problem. The
classes can read text and insert it as is to xml document using
creatTextNode. This text had characters > 127, and I got this error.

    self._doc=xml.dom.minidom.parse(self._xml_filename)
  File "D:\Python21\lib\xml\dom\minidom.py", line 910, in parse
    return _doparse(pulldom.parse, args, kwargs)
  File "D:\Python21\lib\xml\dom\minidom.py", line 902, in _doparse
    toktype, rootNode = events.getEvent()
  File "D:\Python21\lib\xml\dom\pulldom.py", line 234, in getEvent
    self.parser.feed(buf)
  File "D:\Python21\lib\xml\sax\expatreader.py", line 92, in feed
    self._err_handler.fatalError(exc)
  File "D:\Python21\lib\xml\sax\handler.py", line 38, in fatalError
    raise exception
xml.sax._exceptions.SAXParseException: <unknown>:75:1: not well-formed

I know it's not possible to add an enconding attribute using writexml,
so the generated document only has <?xml version="1.0"?>. Is there any
way to get around this problem. I'd like to be able at least to parse
the document while reading using the proper encoding, such as,
encoding="ISO-8859-1". I'm only using minidom, any ideas how?

Another question:
Does any body know how to get the rootnode of a document? If I know
the root node, I can add the proper header and then write the root
node using writexml.

Thanks,
Ehab


From wurmy at earthlink.net  Fri Jun 28 11:33:01 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Fri, 28 Jun 2002 15:33:01 GMT
Subject: Be the first on your block to register your computer.
References: <mailman.1025211444.21440.python-list@python.org> <afg1ks$e6lj5$1@ID-71831.news.dfncis.de> <3D1B6E69.1090204@earthlink.net> <IXWS8.368185$o66.975748@news-server.bigpond.net.au>
Message-ID: <3D1C8169.2040800@earthlink.net>

Neil Hodgson wrote:

>    I suspect that a Python interpreter could be built out of a set of
> Outlook message processing rules although I wouldn't be the one to volunteer
> to do it whereas I will volunteer to do the Excel port if it should ever be
> needed.

Cool! OutlookPython! I can see the ads now...

"Virus developers! Are you tired of having to use a shoddy language like VB for 
your products? Then upgrade now to OutlookPython! Combining all the prized 
features of Outlook(TM) with a real programming language, including:
- improved standard library makes reading mailboxes a breeze!
- attach any file with just one function call!
- reuse code to build your own virus library!
- excellent introspection features allow for self-modifying code!
- as an additional bonus, network features even allow for DoS attacks!

Register now!"

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/



From logiplexsoftware at earthlink.net  Tue Jun 18 17:14:30 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 18 Jun 2002 14:14:30 -0700
Subject: python version?
In-Reply-To: <i0NP8.46300$GY.14481287@twister.nyroc.rr.com>
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com>
	<aegfck$6tkre$1@ID-114079.news.dfncis.de>
	<wEPO8.16066$uk2.8447171@twister.nyroc.rr.com>
	<3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com>
	<gYgP8.42356$GY.12361719@twister.nyroc.rr.com>
	<aekcug$7js8n$1@ID-114079.news.dfncis.de>
	<9YBP8.44418$GY.13865843@twister.nyroc.rr.com>
	<mailman.1024423525.24359.python-list@python.org>
	<i0NP8.46300$GY.14481287@twister.nyroc.rr.com>
Message-ID: <20020618141430.76aa68ec.logiplexsoftware@earthlink.net>

On Tue, 18 Jun 2002 20:48:14 GMT
George Hester wrote:

> Bah humbug.

Ah, the Christmas spirit, all year round ;)

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From chris.lyon at tgm.co.uk  Tue Jun 25 06:46:28 2002
From: chris.lyon at tgm.co.uk (Chris Lyon)
Date: 25 Jun 2002 03:46:28 -0700
Subject: My .py CGI script works in IE but not Netscape
References: <79b54d28.0205311837.731002aa@posting.google.com> <mailman.1024943033.9314.python-list@python.org>
Message-ID: <7652d5e7.0206250246.4b6e72ec@posting.google.com>

Netscape is much less forgiving than IE,
Check table tags and complete html,head and body tags alway's help.
Content type will make a difference,
and do include a 'proper DOCtype'

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
above the head. It all helps


From gerhard.haering at gmx.de  Fri Jun 21 14:46:43 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 21 Jun 2002 20:46:43 +0200
Subject: ssl on windows
In-Reply-To: <uh6qucjo5omi5e@corp.supernews.com>
References: <uh6qucjo5omi5e@corp.supernews.com>
Message-ID: <20020621184643.GB13964@lilith.my-fqdn.de>

* Scott Hathaway <slhath at charter.net> [2002-06-21 13:09 -0500]:
> I would like to use ssl on Windows with Python 2.2.  Is this available yet?
> If so, how can I go about getting it installed?  When I try to get an https
> address with urllib, it tells me that https is an unknown protocol.

I've once built the socket module for Python 2.2 with SSL support. I
don't think I've updated it for Python 2.2.1, yet. But you can still use
it for 2.2.1, as these versions are binary compatible. I'll update it to
2.2.1 during the next day(s).

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From a.schmolck at gmx.net  Fri Jun  7 13:17:45 2002
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 07 Jun 2002 18:17:45 +0100
Subject: __imul__ broken for 'objects' in 2.2.1
Message-ID: <yfsd6v3qa86.fsf@black132.ex.ac.uk>

OK, after unsuccessfully trying for about half an hour to submit this to the
sf bug tracker, I gave up:

class Breaky(object):
    def __imul__(self, other):
        print "imuling"
        return self
sq = Breaky()
sq *=1.

gives:

Traceback (most recent call last):[...] line 10, in ?
    sq *=1.
TypeError: can't multiply sequence to non-int

Unless I'm overlooking something, this is a fairly serious bug and I can't see
a way to work around it (getattribute would presumably work, but slow
everything down unacceptably, so the only 'solution' seems to be to have no
inplace-multiplication).

I had the same behavior on two different machines (running Mandrake 8.2 and
Suse 7.3).

alex


From calves at coelce.com.br  Thu Jun 20 08:19:36 2002
From: calves at coelce.com.br (Alves, Carlos Alberto - Coelce)
Date: Thu, 20 Jun 2002 09:19:36 -0300
Subject: RES: module string.py
Message-ID: <29A97D00F387D411AC7900902770E148058F93C4@lcoeexc01.coelce.net>

See answer below:

> -----Mensagem original-----
> De: r_proietti at yahoo.com [mailto:r_proietti at yahoo.com]
> Enviada em: quinta-feira, 20 de junho de 2002 08:39
> Para: python-list at python.org
> Assunto: module string.py
> 
> 
> Hi, I'm a Python newbye so I'm sorry if my questions will be a little
> stupid...
> 
> I have two trouble:
> 
> 1)  looking at the module "string.py" I have seen that are described
> some functions like "lower", "rjust", etc. Ther problem is that it
> seems not to describe exactly how this functions are made, but there
> is only a comment about their properties. So in which file are these
> functions described?

Try to see the documentation that comes with python distribuition at
Doc/lib/module-string.html subdirectory of python directory. Or see Python
Lybrary Reference that comes with python too.

> 
> 2) I tried to cancel the file "string.py". After that, the Python
> Interpreter doesn't work anymore (it can't even start working). Why?
> 
> Thank you very much, Nop.
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020620/c1c741b2/attachment.html>

From m2 at plusseven.com  Thu Jun 13 03:38:28 2002
From: m2 at plusseven.com (Alex Polite)
Date: Thu, 13 Jun 2002 09:38:28 +0200
Subject: Threading tutorial wanted.
Message-ID: <20020613073828.GD1792@matijek.plusseven.com>

Okidoki. My little webspider has reached the point where further
optimization won't pay off. But my DSL pipe is not saturated by
far. So I guess it's time for me to start looking into threads.

I few minutes of googling on python + threads doesn't yield a
lot. Maybe someone here can point me to a good primer. Maybe someone
here can write a good primer.

When is threading good, when is it bad? How much memory do they
consume? Is it safe to let multiple threads use the same db
connection? How safe is safe? How can I get rich?

Stuff like that.


-- 
Alex Polite
http://plusseven.com/gpg/




From siegfried.gonzi at kfunigraz.ac.at  Wed Jun 26 03:08:44 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Wed, 26 Jun 2002 09:08:44 +0200
Subject: SOLVED (Re: Python hits the spot)
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at> <af4vrl$ffe$0@216.39.172.122> <3D16DD5C.1050708@kfunigraz.ac.at> <uhdume2n1mhbba@corp.supernews.com> <3D171291.2060905@kfunigraz.ac.at> <af8ib7$k51$0@216.39.172.122> <3D181BA2.3090009@kfunigraz.ac.at> <afai04$97n$0@216.39.172.122>
Message-ID: <3D19687C.8040708@kfunigraz.ac.at>

Bengt Richter wrote:

> That can be read two ways ;-) 
> 1) Definitely no diagnosis, and 8 steps remain different as they were
> 2) Definitely no slowdowns, and 8 steps now execute with ~constant timing.
> 
> Still curious ;-)

Hi,

point 2) is appropiate.


> I think the laptop manufacturer should buy it for you (unless you were
> running it outside specs for ambient temperature and ventilation).
> Good luck.

They made me the offer to collect my laptop and look inside. Their 
stance actually ist that the cooler itself does wobbling and that is the 
reson for the strange beahvior.

If my paper is over I will see to give them my laptop. I will then 
report any changes.


Regards,
S. Gonzi



From peter at engcorp.com  Sun Jun 23 21:45:35 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 23 Jun 2002 21:45:35 -0400
Subject: HTTP authentication
References: <TWtR8.86654$Am4.23833590@twister.columbus.rr.com> <3D166ED4.FBE9E2C8@alcyone.com>
Message-ID: <3D1679BF.7F2D3A1@engcorp.com>

Erik Max Francis wrote:
> 
> Preston Tamkin wrote:
> 
> > hello all. I've just recently jumped on the python bandwagon, and I
> > need
> > some help!
> 
> Derive from the urllib.FancyURLopener class, and override the
> proper_use_password method to return the (user, password) 2-tuple you
> want:
> 
>         http://www.python.org/doc/current/lib/urlopener-objs.html

I think another solution is just to specify the username and 
password in the URL like so:

 ref = urllib.urlopen("http://username:password at host.domain/path/file")

-Peter


From mcfletch at rogers.com  Thu Jun 20 23:24:19 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Thu, 20 Jun 2002 23:24:19 -0400
Subject: newbie string question
References: <aeu40s$9jg2s$1@ID-145503.news.dfncis.de>
Message-ID: <3D129C63.40109@rogers.com>

Think of the indices as pointing between the characters. 0 is before the 
first character, len(sequence) is just after the last character, like so.

[ a,  b,  c,  d,  e,  f,  g  ]
  0   1   2   3   4   5   6   7
-7  -6  -5  -4  -3  -2  -1

Asking for a single index gives you the index after the pointer, as if 
you were saying "move to pointer position 5, then read a single item 
(moving forward)".  Asking for a negative index just sets that index = 
len(sequence)+index and does the same thing as normal.

Asking for everything between two pointers/indices (slicing) starts from 
one pointer, and reads all the items up to the end pointer (which is 
sitting just before the character you'd get by asking for pystr[end]). 
If your end-pointer is before or equal to the beginning pointer, it's 
not going to read anything (there's nothing between the pointers moving 
in the forward direction).

So, in the example above,
     0:2  -> a,b
     0:0  ->
     2:3  -> c
     4:-1 -> e,f
     4:-3 ->
     4:-4 ->

HTH,
Mike

PS: Apologies to Tim Peters for mangling his explanation, it's been a 
long time since I've read it.

Don Low wrote:
> There's probably a very simple explanation, but I can't see it for now, so
> bear with me.
> 
> 
>>>>pystr = 'Python'
>>>>pystr = [5]
>>>
> 'n'
> 
>>>>pystr [2:5]
>>>
> 'tho'
> 
> The question is, why isn't pystr [2:5] 'thon' instead of 'tho'. Could
> somebody explain?
> 


-- 
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From merkosh at hadiko.de  Wed Jun 12 13:45:24 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Wed, 12 Jun 2002 19:45:24 +0200
Subject: socket.inet_aton - bug?
Message-ID: <MPG.1771a71d361193fc989683@news.rz.uni-karlsruhe.de>

Hi,

socket.inet_ntoa() accepts the string '\xff\xff\xff\xff' which is 
excactly the same i'd expect socket.inet_aton('255.255.255.255') to 
produce if it accepted the string '255.255.255.255' as an IP adress.
socket.inet_aton() dies with socket.error "illegal IP adress".

i know this is no valid ip adress, but it is a valid netmask. the 
purpose of converting the netmask to a binary format was that i wanted 
to isolate the subnet of an ip adress.
i wanted to use test_ip & netmask == ip_range to allow subnets to access 
a tcp server. 
is there another way of controlling this or why does socket.inet_aton() 
behave like that?

Ciao
Uwe


From mbell at cs.pitt.edu  Thu Jun 13 11:51:31 2002
From: mbell at cs.pitt.edu (Matthew Bell)
Date: 13 Jun 2002 08:51:31 -0700
Subject: email forwarding program
Message-ID: <e21498ef.0206130751.2ef784a7@posting.google.com>

Hi,

I have a question regarding the email handling
libraries in Python.  I'm trying to write a
program that, based on some information it
extracts from an email, forwards it to another
person.  I've about figured out imaplib, but
don't see how I can use it to forward on emails.

One possibility I've considered is to write out
the email to a temp file, then make a call to
the operating system to use the UNIX program
mail -- but would like to avoid doing that if
at all possible.

Any advice?

Thanks,
Matt


From imbosol at vt.edu  Tue Jun 25 18:13:41 2002
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 25 Jun 2002 18:13:41 -0400
Subject: Suggestions for good programming practices?
References: <mailman.1025039833.18335.python-list@python.org>
Message-ID: <lupafa.691.ln@grey.aerojockey.localdomain>

brueckd at tbye.com wrote:
> On Tue, 25 Jun 2002, Carl Banks wrote:
>> You seem to have this idea, based on this last paragraph, that a good
>> way to keep newbies from misusing eval and exec is to keep them in
>> ignorance.  I think that's absurd.  I don't know if you're just
>> grasping for arguments, because you previously seemed to think it's
>> best to explain WHY, as opposed to simply saying, "Don't do it."
> 
> Um, what's absurd is how you reached that conclusion from the paragraph I
> wrote. <0.4 wink> I said that people WON'T get hurt if they are NOT
> ignorant - I'm advocating an explanation of risk in place of the
> categorical "stay away!".

What's really absurd is how you reached the conclusion that I reached
a conclusion when I only said, "you seem to." <wink>  So why did you
ask what kind of teacher calls a newbie's attention to eval?  You
seemed (not to be confused with "I have concluded") to have been
advocating ignorance.

Are you merely saying it's better to say nothing than to say, "don't
do this?"  Because I would disagree with that, too.


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From matt at mondoinfo.com  Thu Jun 27 22:28:55 2002
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Fri, 28 Jun 2002 02:28:55 GMT
Subject: Is Python growing?
References: <3D1B92FB.59B590A0@pop.ntlworld.com>
Message-ID: <slrnahnif7.1c6.matt@happy-hour.mondoinfo.com>

On Thu, 27 Jun 2002 23:34:36 +0100, a.clarke11
<a.clarke11 at pop.ntlworld.com> wrote:

> My question is, I use 2.1 and by the time I need to update, probably
> the version number will have reached Python 3.xx. Will that version
> have the inherent simplicity that makes 2.1 so usable? Or will it
> become more complex, and lose some of its original ease of use? By
> more complex, I mean will it have a steeper learning curve for
> newbies, will it be slower, will the download take longer, and will
> the featurelist be significantly longer?  Other things develop until
> they lose their function, eg flowers, saplings, clouds: can Python
> avoid increasing complexity, is that the present trend in
> development?

Dear Tony,
Entire flamewars have been written on the subject <wink> but here's my
opinion: I suspect that you'll be OK. Indeed, I'm placing my own bets
that way.

Here's why I think that: It's not just complexity that's good to
avoid, it's certain kinds of complexity. Python will surely get new
features, but Python doesn't have to be much harder to learn and can
be easier to use if (only) the right features are included.

Python is wonderfully easy to get started with. But someone who's new
to programming doesn't need to learn everything about Python in order
for it to be useful and fun. (That sounds obvious but it's less true
in many other languages.) So if Guido adds a feature here and there,
that doesn't necessarily make the learning curve much worse. I've been
using Python for some years now and there are corners of the standard
library that I haven't explored because I haven't needed to. Indeed,
there are aspects of the core language that I haven't found out much
about because I haven't needed them.

On the other hand, some features make Python much easier to use and
read. A while ago, I replaced:

def lengthOfLongestStr(list):
  l=len(list[0])
  for count in range(1,len(list)):
    if len(list[count])>l:
      l=len(list[count])
  return l

with

def lengthOfLongestStr(myList):
  return max([len(s) for s in myList])

To my eye, the second one is much easier to read. Similarly, writing
something like:

countWidgetsFound=countWidgetsFound+1

as

countWidgetsFound+=1

makes the intention clearer and avoids the chance of making a typo.

I'm happy to trust Guido to add the right features.

Regards,
Matt


From debl2noohnospammyyy at bellatlantic.net  Sun Jun  9 02:37:06 2002
From: debl2noohnospammyyy at bellatlantic.net (David Lees)
Date: Sun, 09 Jun 2002 06:37:06 GMT
Subject: Determination of screen resolution?
References: <3D02F4C8.C2DF2EE0@bellatlantic.net>
Message-ID: <3D02F7A4.89CE4537@bellatlantic.net>

OK, I should have done a little more digging.  I just found the answer
to my question in the documentation which is simply:

v = canvas.winfo_vrootheight()
h = canvas.winfo_vrootwidth()

David Lees wrote:
> 
> I would like to use tkinter and have the canvas object automatically
> scaled fit the screen.  Is it possible to read the screen resolution
> (800x600, 1024,768...) from somewhere.
> 
> David Lees


From peter at engcorp.com  Mon Jun 24 21:03:04 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 24 Jun 2002 21:03:04 -0400
Subject: Suggestions for good programming practices?
References: <mailman.1024957832.5633.python-list@python.org>
Message-ID: <3D17C148.6CC8B3D2@engcorp.com>

David LeBlanc wrote:
> 
> The one that always gets me is "lat,lon,timezone = getLocation("seattle")".
> I.E. the ability to return multiple distinct values as opposed to returning
> multiple values in the form of a struct as in C or C++.

You _are_ returning them "in a struct".  It's spelled "tuple"
but other than that what's the difference?

-Peter


From gmcm at hypernet.com  Thu Jun  6 08:37:07 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 06 Jun 2002 12:37:07 GMT
Subject: Installer vs py2exe: Windows DLLs
References: <u1ooxxn8.fsf@morpheus.demon.co.uk> <Xns9220730326E0Fgmcmhypernetcom@199.171.54.215> <u1ogoiqu.fsf@morpheus.demon.co.uk>
Message-ID: <Xns922557D7E23EFgmcmhypernetcom@199.171.54.213>

Paul Moore wrote:

[Installer's dependency on win32api / pywintypes]

> OK, I analyzed this a bit further. This is all for Python 2.2 on
> Windows 2000.
> 
> The win32api dependency (and hence the PyWinTypes22 dependency) comes
> from iu.py, in the RegistryImportDirector. Two questions come to mind
> - first is whether it is appropriate for Installer-built
> applications to look in the registry for Python entries in any case, as
> the whole point is to run independently of any installed copy of
> Python. I imagine that something later in the process disables picking
> up registry entries, but that leaves the RegistryImportDirector code as
> dead code (which isn't worth worrying about, except for the spurious
> dependency).

Well iu.py is a drop-in import replacement. I guess I'll have to
think about refactoring it so Installer can use it without drawing
in the registry code.

> The second question is whether RegistryImportDirector she
> translation is pretty much automatic - I attach a patch. This gives a
> dependency on _winreg.pyd (pity - I'd assumed _winreg would be built
> in on Windows :-() but that is *much* smaller than win32api +
> PyWinTypes22. Minimal testing shows no problems, but a better test
> would be worth doing...

_winreg.pyd didn't exist as of Py 1.5, but your patch looks
good.
 
> The dependency on _sre comes from posixpath (in expandvars()), which
> is imported conditionally by os. And os is imported by archive.py, to
> get at os.path.basename, and os.path.splitext. Would it not be
> possible here to directly use {dos,mac,nt,posix}path?

Sure. In fact, iu is already doing that, so archive could probably
get at them that way (although it currently doesn't know about iu).
But archive only imports os when *building* an archive, so if that's
the only import of os, you're free to exclude it.
 
> I know that in the grand scheme of things, removing dependencies like
> this is fairly pointless for any "real" application, but it appeals to
> my sense of minimalism. 

You can get a cross-reference of dependencies from mf.py for this
exact reason (deciding what can be safely excluded).

-- Gordon
http://www.mcmillan-inc.com/



From missive at frontiernet.net  Fri Jun 21 19:15:05 2002
From: missive at frontiernet.net (Lee Harr)
Date: Fri, 21 Jun 2002 23:15:05 -0000
Subject: Newbie-Question: Oberserving an directory
References: <3d1352a7@news.piro.net>
Message-ID: <uh7crpjvqc51a0@corp.supernews.com>

In article <3d1352a7 at news.piro.net>, Marcus Klein wrote:
> Hi there,
> 
> I want to observe an directory to check if a new file has been put in, if 
> that happens I want to start an event.
> 
> Does anyone have a good idea how to start, I've tried it with signal, but 
> either I am too stupid or it is the complete wrong way :-/
> 

You do not say which OS you are using. If FreeBSD, you might look
at PyKQueue.

The first hit from google with pykqueue will take you there.



From tim.one at comcast.net  Mon Jun 10 08:34:10 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 10 Jun 2002 08:34:10 -0400
Subject: Where is Berkeley DB on your system? Can you live without DB 1.85?
In-Reply-To: <GCEDKONBLEFPPADDJCOECEMFDHAA.whisper@oz.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCAELHPLAA.tim.one@comcast.net>

[David LeBlanc]
> ...
> Hmmm... I have Python 2.2.1 binary installed from Sourceforge -
> what is that bsddb.pyd file dated 4/9/2002 in the dlls directory?

It's compiled from Sam Rushing's Win32 port of Berkeley's DB library,
version 1.85, from

    http://www.nightmare.com/software.html

under the "bsd db" link (*not* the "bsddb" link).

It's a bug pit, through no fault of Sam's.  The only reason we ship it is
that we've always shipped it, and nobody has volunteered the work to package
something better.

> ...
> <gasp> could Guido be so heartless as to imperil people's data by
> allowing the inclusion bsddb in the standard Python distro?!?

Yes <wink>.  It wasn't known to be badly damaged at the time it was first
shipped, and now it's supplied for backward compatibility.  A better
alternative would be most welcome, but PythonLabs has no bandwidth to work
on it.





From peter at engcorp.com  Tue Jun 18 20:51:30 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 18 Jun 2002 20:51:30 -0400
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <yXBP8.44417$GY.13865423@twister.nyroc.rr.com>
Message-ID: <3D0FD592.93BFBCDB@engcorp.com>

George Hester wrote:
> 
> Listen you aren't my daddy.  If you do not want to help constructively then
> please get off my boat.

I thought I was being constructive in two distinct ways, although
perhaps I was more blunt than you were able to handle.  If so,
my apologies for appearing to patronize you.  That was not my
intention.

I did point out two helpful things (I thought).  One was to
show you by explicit example how you can't access sys.version
until you've imported the sys module.  You didn't seem to 
understand that was required from your response in another post,
and I believe you might have missed the point again in 
reading my response.  If you read the post again, you'll see
that I got the same error you did until I did "import sys"
and then the problem went away.  As that would have solved
your initial problem, I felt it would be fairly helpful.

Secondly, and I believe I'm still correct here, I pointed out
that you might have a lot of trouble trying to juggle two 
things at the same time: learning Python, and using it in
an unforgiving environment as an ASP scripting language.
I have been in your position many times in the past, and 
every time I have been burned by ignoring my own instincts.
Never attempt to take on two difficult tasks at once when
you can partition the problem and approach one task at a
time.  Learn Python first (without ASP involved), _then_ 
try to use it for ASP.

You may choose to ignore my advice.  It was, after all,
given freely and, as always, is worth what it costs you...

Cheers,
-Peter


From insyte-clp at emt-p.org  Thu Jun 20 16:02:21 2002
From: insyte-clp at emt-p.org (Ben Beuchler)
Date: 20 Jun 2002 20:02:21 GMT
Subject: HTTPSHandler ??? missing?
References: <3d121907$1@news.poly.edu>
Message-ID: <slrnah4d4m.r77.insyte-clp@petra.bitstream.net>

In article <3d121907$1 at news.poly.edu>, chandan mathur wrote:
> I am trying to login to a ssl secure server using urllib2 and clientcookie
> but can't find HTTPSHandler() in urllib2.py as mentioned in the manual!

I believe that's only built if your 'socket' module supports SSL.  And
your socket module will only link against the SSL libraries if you already
have them (OpenSSL) installed on your system.

-Ben

-- 
Ben Beuchler                                           There is no spoon.
insyte at emt-p.org                                            -- The Matrix


From occeanlinux at linuxmail.org  Tue Jun  4 13:49:32 2002
From: occeanlinux at linuxmail.org (Gold Fish)
Date: Tue, 04 Jun 2002 17:49:32 GMT
Subject: directory 
Message-ID: <3cfcfd4a_1@news.iprimus.com.au>

How to read the subdirectories in the directory
Support i have the directory /home/ which have some subdirectories
/home/sub1
/home/sub2
/home/sub3
I am using os.listdir('/home/') to read in the list so it will be 
['sub1','sub2','sub3'] now i want to read inside the sub directory sub2
once again i using os.listdir(os.listdir('/home/')[1]) but it came out with 
an errors: not a file or directory. Can anyone tell me what did i wrong. 
Anh how to fix it.


From amckay at merlintechnologies.com  Mon Jun 10 14:11:50 2002
From: amckay at merlintechnologies.com (Andy McKay)
Date: Mon, 10 Jun 2002 11:11:50 -0700
Subject: activating a webpage's click command from python
In-Reply-To: <3D04EA96.B3518A37@engcorp.com>
References: <mailman.1023727993.12908.python-list@python.org> <3D04EA96.B3518A37@engcorp.com>
Message-ID: <20020610181150.GB2547@mckay.merlintechnologies.com>

> does the job for Nortel on the TSE (disclaimer: anyone even 
> thinking of buying Nortel should probably just keep thinking :-).

Oh come on its a bargain, a fraction of what it was, good buying
opportunity ;)

not-putting-my-money-there'ly yours
-- 
  Andy McKay
  Merlin Technologies




From igor.stroh at wohnheim.uni-ulm.de  Wed Jun 12 15:15:06 2002
From: igor.stroh at wohnheim.uni-ulm.de (Igor Stroh)
Date: Wed, 12 Jun 2002 21:15:06 +0200
Subject: methods to parse big foreign ascii-file?
References: <mailman.1023899956.13534.python-list@python.org>
Message-ID: <3d079dbc$1@sol.wh-hms.uni-ulm.de>

On Wed, 12 Jun 2002 18:05:51 +0200, "holger krekel"
<pyth at devel.trillke.net> wrote:

> - come up with a grammar and let a parser do
>   the hard work (which one?)

if the file is structured in a certain form, that is you can describe the
data structure with a EBNF, then have a look at 'simpleparse' (I've lost
the link, but google will certainly have some for you :) )...

HTH,
Igor

-- 
The idea behind optimistic concurrency control is surprisingly simple:
just go ahead and do whatever you want to, without paying attention
to what anybody else is doing. If there is a problem, worry about it
later. Many politicians use this algorithm, too. (c) Andrew S. Tanenbaum


From jadestar at idiom.com  Sun Jun 16 18:37:06 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 16 Jun 2002 22:37:06 GMT
Subject: newbie system() question
References: <3d035180$0$7017$bc7fd3c@news.sonofon.dk>
Message-ID: <aej3ui$1df4$1@news.idiom.com>

Dan Larsen <dl at bestpriceeu.com> wrote:

> Hi all :o)
...
> ... Isn't there a function, where all output
> from the command, is parsed to a variable?

 I think you're thinking of popen()
 There are several versions of popen() depending on whether you 
 want the just one pipe (stdin or stdout) or a pair of pipes (one
 for input and another for output) a triplet of pipes (stdin,stdout,
 stderr) or a pair with stdout and stderr combined into one stream,
 etc.

 There is also a popen2 module which allows you to subclass your
 own popen class and then use the poll() or wait() methods to retrieve
 the error/return code from the terminated subprocess.

> Thanx in advance :o)
> Dan Larsen




From dougz at cs.washington.edu  Tue Jun 25 18:08:34 2002
From: dougz at cs.washington.edu (Douglas Zongker)
Date: Tue, 25 Jun 2002 22:08:34 GMT
Subject: Reference Count Confusion
References: <mailman.1025037754.18087.python-list@python.org>
Message-ID: <GyA86B.16A5@beaver.cs.washington.edu>

Jeff Smith <smithj at kryos.colorado.edu> wrote:
: I'm a little puzzled about reference counts from Python objects created
: in a C extension. [...]
:
: static PyObject *clean(PyObject *self, PyObject *args)
: {
:   PyArrayObject *c;
:
:   PyArg_ParseTuple(args, "O", &c);
:
:   printf("clean: %d\n", c->ob_refcnt);
:   if (c->data != NULL) free(c->data);
:   if (c->dimensions != NULL) free(c->dimensions);
:   if (c->strides == NULL) free(c->strides);
:   Py_DECREF(c);
:   printf("clean: %d\n", c->ob_refcnt);
:
:   return Py_BuildValue("");
: }

You shouldn't decref c here.  When you use the "O" specifier with
PyArg_ParseTuple(), it gives you a borrowed reference to the object.
You're losing references each time you call this function, and that's
probably messing things up.

dz






From ftobin at neverending.org  Sun Jun 30 16:02:12 2002
From: ftobin at neverending.org (Frank Tobin)
Date: Sun, 30 Jun 2002 16:02:12 -0400
Subject: I'd give up Perl tomorrow if only...
In-Reply-To: <mailman.1024681648.27687.python-list@python.org>
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au>
 <mailman.1024681648.27687.python-list@python.org>
Message-ID: <Pine.LNX.4.44.0206301550120.7807-100000@palanthas.neverending.org>

Christopher A. Craig, on 2002-06-21, wrote:

> Given the hardware and bandwidth, it shouldn't be all that hard to
> create a CPAN for Python.  I suspect one of the main reasons it hasn't
> been done is that Google works so well.

I don't think you have a good grasp of what CPAN really is.  It's much 
more than simply a big FTP site.

The key thing about CPAN is not only the breadth of modules it has, but 
the strong testing system they employ, namely cpan-testers.  I am the 
author of a module on CPAN, and it's great to be able to upload module, 
and have it tested on a diverse set of architecures.

http://testers.cpan.org/

Note that in order to have something be as successful as cpan-testers,
modules needs to have:

1) a standard extraction mechanism (Perl and Python do have)
2) a standard testing mechanism (Perl has ('make test'), Python doesn't)
3) a standard installation mechanism (Perl and Python do have)
4) a standard README system (Perl has, Python doesn't)

What is most missing for Python is a standard testing suite system.  And
until it gets this, it's pretty much not possible to get anything close to
the quality of CPAN with cpan-testers.

Another great benefit of CPAN is that the modules on it are somewhat
'sanctioned'.  Sure, it doesn't take much to get onto CPAN, but the
difference between the 'something' it does take, and 'nothing', is
enormous in speaking for the quality of the modules you'll find in
contrast to searching Google and coming across Joe Schmoe's Python module.
A similar scenario is comparing random RPM's you might find on Google to
FreeBSD/Gentoo's ports trees; the centralization somewhat sanctions the
software, and this is critical, IMO.

Furthermore, I find the de-facto multi-tier naming system that Perl
employs for modules to be a good thing; it encourages users dividing up
their modules in the global namespace.  This is critical for finding what
you need.  For example, I can go into the DBD branch on CPAN and find
every database driver it has.  Python doesn't have an equivalent (packages
just aren't the same thing; they are owned by 'one' thing, most of the
time).

-- 
Frank Tobin			http://www.neverending.org/~ftobin/



From whisper at oz.net  Tue Jun 18 01:37:04 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 17 Jun 2002 22:37:04 -0700
Subject: Listing all available (not just loaded) modules
In-Reply-To: <aem9om$888fb$1@ID-140151.news.dfncis.de>
Message-ID: <GCEDKONBLEFPPADDJCOEAENBDJAA.whisper@oz.net>

Idle, PythonWin and Boa (at least) do this, with PythonWin being the most
detailed - it shows classes and methods within the modules. The version of
Pythonwin I have does not show the contents of the site-packages tree, but
that would be easily fixable.

HTH,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Michael Steuer
> Sent: Monday, June 17, 2002 20:35
> To: python-list at python.org
> Subject: Listing all available (not just loaded) modules
>
>
> I've been searching the newsgroup for information on this but so
> far have been
> unsuccessful:
>
> Is there a way to dynamically list all available modules (not
> just those that
> are already loaded) in python? By available I mean those modules,
> which are
> installed on the machine executing python (in the lib path(s)).
>
> Cheers,
>
> Mik.e
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From chris.gonnerman at newcenturycomputers.net  Tue Jun 18 08:08:03 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Tue, 18 Jun 2002 07:08:03 -0500
Subject: Newbie - Process Management
References: <1024387201.371093@ernani.logica.co.uk>
Message-ID: <002d01c216c0$d7a2d560$0101010a@local>

----- Original Message -----
From: "Dan Scholey" <scholeyd at logica.com>


> Hi,
>
> I am hoping you can help me, I am fairly new to python (and Unix) but am
> trying to solve the following problem...
>
> I want to be able to check if a parent process has died.
>
> At the moment I can check that a child process has died using...
>
>         status = (os.waitpid(child, os.WNOHANG))[0]
>         if status!=0:
>
> I've looked at several manuals and can't find a similar function for
parent
> processes.
> Obviously I know both the parent and child process id ( get_pid() /
> get_ppid() ).

Here is a partial screen-capture from my Linux system.  If you are using any
standard Unixoid OS this should work:

<< partial ps listing >>

15321 pts/1    S      0:00 -bash
15377 ?        S      0:00 /usr/sbin/httpd -f /etc/httpd/httpd.conf
15521 pts/1    R      0:00 ps ax

office:~ $ python
Python 2.1.2 (#1, Jan 16 2002, 17:04:23)
[GCC egcs-2.91.66 19990314/Linux (egcs-1.1.2 release)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> import os
>>> os.kill(15321,0)
>>> os.kill(15322,0)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
OSError: [Errno 3] No such process
>>>

In other words, PID 15321 exists (I knew it did, it's my shell) but 15322
doesn't.  So:

    def parentcheck():
        try:
            os.kill(os.get_ppid(), 0)
            return 1
        except OSError:
            return 0


> I have a workaround of....
>
>         ps_line=string.split(commands.getoutput("ps -lp
%s"%my_pid),"\012")
>         if string.atoi(string.split(ps_line[1])[4])!=daddy:
>
> ... but I would rather not use ps.
>
> Any ideas?

Hope this helps.

> Cheers
> Dan

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net






From whisper at oz.net  Fri Jun 14 16:25:28 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 14 Jun 2002 13:25:28 -0700
Subject: Zip makes sense! was: Re: why not "'in' in 'in'"?
In-Reply-To: <lk8z5imc9r.fsf@pc150.maths.bris.ac.uk>
Message-ID: <GCEDKONBLEFPPADDJCOEAEDPDJAA.whisper@oz.net>

Aaaah, a logger. Obviously the first on his block to be a logger. The joys
of bravely loggin where no one has logged before! ;->

Perhaps if he was an english major, or english was his first language, he
could tell that the common meaning of yield isn't produce. I have no idea
what zip means beyond being a name for a compression utility or a verb
meaning to exercise a zipper.

If-you-want-to-make-a-beginner's-language-don't-fsck-it-up-with-obscure-mean
ingsly-yours,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Michael Hudson
> Sent: Friday, June 14, 2002 2:38
> To: python-list at python.org
> Subject: Re: Zip makes sense! was: Re: why not "'in' in 'in'"?
>
>
> "David LeBlanc" <whisper at oz.net> writes:
>
> > Haskell - that's a mountain range in New York state? :p)
>
> No, it's the first name of a pioneering logician.
>
> Cheers,
> M.
>
> --
>   I really hope there's a catastrophic bug insome future e-mail
>   program where if you try and send an attachment it cancels your
>   ISP account, deletes your harddrive, and pisses in your coffee
>                                                          -- Adam Rixey
> --
> http://mail.python.org/mailman/listinfo/python-list





From markus.vonehr at ipm.fhg.de  Fri Jun 14 07:24:10 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Fri, 14 Jun 2002 13:24:10 +0200
Subject: PIL and miscolored images
Message-ID: <3D09D25A.4C04E03@ipm.fhg.de>

Hi everybody,

I'd like to show a greylevel image as miscolored image,
anyone has an idea how i can do it and maybe adjust
the color scale?

Thanks for any answer,

Markus

-------------- next part --------------
A non-text attachment was scrubbed...
Name: markus.vonehr.vcf
Type: text/x-vcard
Size: 245 bytes
Desc: Visitenkarte f?r Markus von Ehr
URL: <http://mail.python.org/pipermail/python-list/attachments/20020614/6b6d089d/attachment.vcf>

From krissepu at vip.fi  Mon Jun  3 05:53:11 2002
From: krissepu at vip.fi (Pekka Niiranen)
Date: Mon, 03 Jun 2002 12:53:11 +0300
Subject: Iterating through two list + moaning
References: <3cee0ae6_1@news3.newsgroups.com> <mailman.1022234742.24014.python-list@python.org> <3cee2338_2@news3.newsgroups.com> <mailman.1022241402.21254.python-list@python.org> <1022297508.675568@yasure> <831ybrwl1v.fsf@panacea.canonical.org> <87sn44lusn.fsf@tleepslib.sk.tsukuba.ac.jp>
Message-ID: <3CFB3C87.1F0E9EEB@vip.fi>

Right,

1)   now if somebody would:
    a)    change lisp syntax to as clear as Python's is and
    b)    wrap GUI library that supports grid with unicode (wxwindows) to it

    I would change to it immediately.

2) There has also been discussions to remove reduce(), map() and filter()
    from python, but as long as for -loops and list -comprehensions are
    as slow as presently, I am worried of it.

3) What I would like to see is:
    a) a bold move to fix python (to stackless for example) without
        any consideration of backwards compability
    b) Make distrubution smaller and consistent (no len() but a.len() ), since its
standard
        library is resembling a bloatware allready.
    c) Development decisions made only on engineering bases.
        To me GvR and the rest of the developers are not owing anything to us
"freeloaders".
        If they are into money/career, shit happens.

    With current development direction, I see no other result than Microsoft
-syndrome, where you
    cannot fix code because of the wide userbase.

 -pekka-

"Stephen J. Turnbull" wrote:

> >>>>> "Kragen" == Kragen Sitaker <kragen at pobox.com> writes:
>
>     Kragen> "Donn Cave" <donn at drizzle.com> writes:
>
>     >> The function that carries data is cool enough.  Lisp does it
>     >> easier than Python, and in fact does "functional programming"
>     >> better than Python by a long shot.
>
>     Kragen> That's debatable.
>
> If you're going to take the Python side of that, I think you've found
> yourself a real challenge.  After all, Lisp was invented by writing an
> automatic interpreter for a purely functional logic notation.
>
> While you can do functional programming in Python, the syntax and the
> ubiquity of side effects makes it hard to emphasize that style.
>
> At least for me.
>
> Hm.  Maybe I should take a look at M. Pinard's topy tool, and see if
> maybe I shouldn't be writing my Python programs in Lisp (Scheme, to be
> precise)<wink>.
>
> --
> Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
> University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
>  My nostalgia for Icon makes me forget about any of the bad things.  I don't
> have much nostalgia for Perl, so its faults I remember.  Scott Gilbert c.l.py



From daniel.dittmar at sap.com  Mon Jun 10 08:45:35 2002
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Mon, 10 Jun 2002 14:45:35 +0200
Subject: Program to an interface, not an implementation
References: <D54BF26DB5A7D511B36B00508B9596A402F8EC63@sdgcaexch2.peregrine.com> <mailman.1023704594.18063.python-list@python.org>
Message-ID: <ae271r$aiq$1@news1.wdf.sap-ag.de>

Egbert Bouwman wrote:
> This I read as:
> For Python interface_inheritance does not require class_inheritance.
>
> So you can inherit an interface without inheriting the class,
> but interface_inheritance by way of class_inheritance is possible as
> well. That means at least two things:
> - python has interface inheritance (which some people deny),
> - interface_inheritance without class_inheritance is a kind of
> cultural   (not genetical) inheritance: give it the same name and all
> is well.   But for me that is only polymorphism !

Perhaps it would be better to skip the term interface_inheritance
altogether. In Javaspeak, you IMPLEMENT an interface, you promise that you
conform to a specification. In Java, this can be checked at compile time. In
Python, you could probably create interface objects which can check if a
given object or class implements all the methods that are required.

It is only polymorphism in the way the language runtime handles this. But it
is a difference in the way you document your software.

> I do all this hair splitting because the GoF say (still page 17)
> that many patterns depend on the distinction between interface- and
> implementation inheritance. How do you define in python the
> interface inheritance the GoF talk about ?

An example from the module pickle:
  dump(object, file[, bin])
  ...
  file must have a write() method that accepts a single string argument. It
can thus be a file object opened for writing, a StringIO object, or any
other custom object that meets this interface.


You specify the methods that will be called. Having a well known object with
the required behaviour is always helpful.

The important part about 'programming to an interface' is this: If you're
implementing the 'dump' method, don't call arbitrary methods of a file
object just because you happen to test it with a file object.

Daniel




From d_blade8 at hotmail.com  Mon Jun 10 16:24:24 2002
From: d_blade8 at hotmail.com (Lemniscate)
Date: 10 Jun 2002 13:24:24 -0700
Subject: Newby: Dictionaries as file objects?
References: <Xns9228AB206D06CAusYourStandingInIt@130.133.1.4>
Message-ID: <7396d2b2.0206101224.61005b7f@posting.google.com>

I agree with David.  However, depending on exactly you want to be
doing (and be able to do), you may want to look into the dbm modules
(such as dbm, dbhash, etc.).  These are, surprise, database modules
but they allow for the storage of dictionary items.  Now, pickle,
shelve, and the like are perhaps the best things for storing Python
objects (let's say a user-created object that contains mixed types)
but the dbm modules have their uses too, based on EXACTLY what you
want to do.  Anyway, have fun.

Lem


netvegetable <deathtospam43423 at altavista.com> wrote in message news:<Xns9228AB206D06CAusYourStandingInIt at 130.133.1.4>...
> Hi.
> 
> I want to be able to store a list on a file object, and I want each list 
> item to be a dictionary, complete with its own set of keys and values. 
> 
> I know I can store a list in a file object, simply by writing each item to 
> the file as a string , and making sure there's a "\n" character at the end 
> of it. When I read the file object again, I merely need to open it as 
> "open(object).readlines()" and hey presto, my list is back in order.
> 
> So is there a way to conveniently store a dictionary in each item as well?
> 
> Ta.


From ffjhenon at club-internet.fr  Tue Jun 18 15:36:33 2002
From: ffjhenon at club-internet.fr (Fabien =?iso-8859-1?Q?H=E9non?=)
Date: Tue, 18 Jun 2002 21:36:33 +0200
Subject: For those who are interested
Message-ID: <3D0F8BC1.13E432FE@club-internet.fr>

I have written an editor for POV-RAY ( the raytracer) using Tkinter that

features syntax highlighting.
It fires/stops up raytraces, save,....

It can be found at :
http://pyvon.sourceforge.net

Fabien




From rowen at cesmail.net  Wed Jun 26 16:02:02 2002
From: rowen at cesmail.net (Russell E. Owen)
Date: Wed, 26 Jun 2002 13:02:02 -0700
Subject: Help with an algorithm wanted
Message-ID: <rowen-CE3C0F.13020226062002@nntp1.u.washington.edu>

I'd like some help solving a problem in Python.

The basic idea is I have an item of data that can be represented in any 
of four ways: A,B,C and D. I also have conversion functions between 
certain pairs of representations, such that there's always a path 
between any two data representations. In this case suppose I know A<->C, 
B<->C and C<->D, or graphically:

A <--+
     +--> C <---> D
B <--+

I'd like to create objects that have one of these items (A,B,C,D) as a 
given, then be able to query any such object for any data representation.

For instance, an object with A as the given would return D by solving 
A->C->D.

What I'm trying to figure out is a simple and reasonably elegant way to 
code such objects.

I am pretty sure recursion is the heart of the solution. An object with 
A as the given could have a method getD that returned D<-C and a method 
getC that returned C<-A.

The real issue is creating such objects without hand coding each one 
(since in the real world I have more representations and the number of 
casese proliferates badly).

I suspect I should use recursion to create the object, too, but it 
sounds like it'll need a lot of introspection. For instance to create an 
object with A as the given, one might:
- start by defining method getA to return A
- look through the set of conversion functions that output A; the only 
one is A<-C, so method getC returns C<-A.
- repeat with all methods that return C and don't have methods already 
defined; this gives method getB returns B<-C and method getD returns 
D<-C.

Alternatively, I can imagine making each object basically identical 
except for some kind of meta-data that tells each method which data item 
to query; the appropriate function can then be looked up as needed (e.g. 
via a dictionary that takes as a key the output and input 
representation). This would slow down the methods but avoid the 
introspection.

Any suggestions?

-- Russell


From claird at starbase.neosoft.com  Wed Jun  5 11:44:45 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 5 Jun 2002 10:44:45 -0500
Subject: writing a text in hyphenless justification
References: <mailman.1023270008.12374.python-list@python.org>
Message-ID: <C4C16D17B2F131A3.2CF2A390D6D00225.5C67BAD6A5120DA1@lp.airnews.net>

In article <mailman.1023270008.12374.python-list at python.org>,
David LeBlanc <whisper at oz.net> wrote:
>Justifying text, with or without hyphens, is SO not a trivial job to do
>right.
			.
			.
			.
?  There are a couple of senses of the original question 
which might make this easier than first appears.  In any
case, <URL: http://wiki.tcl.tk/justify > has a robust
implementation of full-blown fixed-width justification.
While it's written in Tcl, I don't think it'll present
any problems for Pythoneers.
-- 

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


From kkto at csis.hku.hk  Sun Jun  2 00:23:48 2002
From: kkto at csis.hku.hk (Isaac To)
Date: 02 Jun 2002 12:23:48 +0800
Subject: Why no open(f, "w").write()?
References: <mailman.1022702940.7544.python-list@python.org> <slrnafagek.64g.huaiyu@gauss.almadan.ibm.com> <slrnafcsor.12v.msoulier@pmerd071.ca.nortel.com> <20020531225504.248c92ce.larooy@xtar.co.nz> <m2lm9zh39j.fsf@mycroft.actrix.gen.nz> <20020601194350.6366b7be.larooy@xtar.co.nz>
Message-ID: <7iu1omtijv.fsf@enark.csis.hku.hk>

>>>>> "John" == John La Rooy <larooy at xtar.co.nz> writes:

    John> Good point. If I had infinite memory and was allowed an infinite
    John> number of open files, I'd never need to close or flush my file
    John> objects, because the optimisation between my writing to a file and
    John> the actual event taking place will have no effect on the behaviour
    John> of my code. But it DOES have an effect doesn't it? Buffering is an
    John> optimisation, isn't it?

Is it remotely possible to make buffering to be transparent to the
programmer?  I don't think so: at least, when the language needs to define
the concurrency semantics: when two programs open the same file, when is it
guaranteed that the second program will correctly read the things written by
the first program.  It is not something that the programmer can say "okay,
I'll let the system to figure that out"... or do you have some new ideas
about it?

Regards,
Isaac.


From mwh at python.net  Mon Jun 10 09:25:45 2002
From: mwh at python.net (Michael Hudson)
Date: Mon, 10 Jun 2002 13:25:45 GMT
Subject: PyCObject subtype PyObject
References: <3D022727.2147C50E@doc.ic.ac.uk> <lk1ybflaua.fsf@pc150.maths.bris.ac.uk> <3D0483AF.5ACB350@doc.ic.ac.uk>
Message-ID: <lkvg8rjmet.fsf@pc150.maths.bris.ac.uk>

Benjamin Tai <bt98 at doc.ic.ac.uk> writes:

> Michael Hudson wrote:
> 
> > > 2) For function "PyCObject_FromVoidPtr()", it takes in a "void*" as the
> > > first argument.
> > > For function "PyCObject_AsVoidPtr()", it returns a "void*".
> > > When implementing Python extension, what is the potential danger of
> > > casting pointers, of different size, to and from "void*"?
> >
> > Doesn't C guarantee that you can cast a pointer of type (T*) to
> > (void*) and back again for all T?  I was fairly sure it did.
> >
> 
> I am extending a C++ application with Python. Before attempting other
> possibilities, like Boost, I just want to document the work I have done so
> far.
> 
> Simply speaking I am hiding a C++ pointer, as a PyCObject*, inside a Python
> instance.
> 
> The scheme works OK. The C++ class hierarchy is reflected by Python classes.
> The pair of C++ and Python object is passed around between the two
> languages.

Oh, that makes sense.  Hmm, you're probably going to be OK but I doubt
you get a standard-backed guarantee any more.

> > > 4) Apart from exporting function between modules, are there any
> > > examples of using PyCObject?
> >
> 
> The example, in the Python/C API document, is talking about exporting
> functions between Python modules. So I am wondering if my method is valid?

I think so.

Cheers,
M.

-- 
  I'm a keen cyclist and I stop at red lights.  Those who don't need
  hitting with a great big slapping machine.
                                           -- Colin Davidson, cam.misc


From rodrigo.senra at ic.unicamp.br  Wed Jun 19 12:55:43 2002
From: rodrigo.senra at ic.unicamp.br (Rodrigo Senra)
Date: 19 Jun 2002 09:55:43 -0700
Subject: any users of ExtensionClass out there?
References: <mailman.1024440443.17628.python-list@python.org>
Message-ID: <8b04704.0206190855.5696145b@posting.google.com>

Jeremy Hylton <jeremy at zope.com> wrote in message news:<mailman.1024440443.17628.python-list at python.org>...

> There has been little perceived interest in ExtensionClass releases
> because Zope is moving away from ExtensionClass.  I wonder, though, if
> there are other users of ExtensionClass who would be interested in a
> new release.  Please let me know if you are.


 Even though I have no production code based in ExtensionClass, I'd used
 it sometimes to explore Python reflection capabilities over C extensions.
 Therefore, I'm interested in any changes you & your team may have come up 
 with.

 best regards,
 Senra


From erict at millfilm.co.uk  Thu Jun 20 04:43:42 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Thu, 20 Jun 2002 09:43:42 +0100
Subject: object object
Message-ID: <3D1195BE.9B10EE00@millfilm.co.uk>

I didn' t find a lot about class object.
and I would  like to know what I can
inherit from it and what is the difference
between, class foo(object): and class foo:

Thanks




From SteveC at nospam.innocent.com  Wed Jun 19 17:48:36 2002
From: SteveC at nospam.innocent.com (Steven M. Castellotti)
Date: Wed, 19 Jun 2002 21:48:36 GMT
Subject: Dispatching a Win32 COM in the background?
Message-ID: <U_6Q8.39803$ks6.8500@nwrddc02.gnilink.net>

Hey all-

	I'm running a Python application under windows, and I want to make use
of the Microsoft Speech API to sythesize Text-To-Speech output, without
halting the execution of my program.

I'm making my calls like this:

(intialization)

from win32com.client import constants
import win32com.client
self.speech = win32com.client.Dispatch("SAPI.SpVoice")



(during execution)

self.speech.Speak(text)


	I'm not very familiar with win32com calls, but I'm wondering if there's
an easy to to either:

1) Make the call such that the process occurs in the background, and
application execution continues to flow

2) Make the call inside a thread which produces the same effect.


	Can anyone please offer me some tips/sample code to achieve this?


Thanks in advance.


Cheers,

Steve Castellotti
SteveC (at) innocent.com
http://cogengine.sourceforge.net/


From gh_pythonlist at gmx.de  Tue Jun 18 18:20:11 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 19 Jun 2002 00:20:11 +0200
Subject: results of stat[ST_ATIME]???????
In-Reply-To: <3D0F9DAE0000034E@www.zipmail.com.br>
References: <3D0F9DAE0000034E@www.zipmail.com.br>
Message-ID: <20020618222010.GB1879@lilith.my-fqdn.de>

* jubafre at zipmail.com.br <jubafre at zipmail.com.br> [2002-06-18 18:58 -0300]:
> when i use st[ST_ATIME] a int number is returned, how i can transform
> this number to a "time number"???????

time.gmtime and time.localtime, depending on wether you want GMT or time
in the local timezone. Check out the docs for the time module for
details. If you want to have a readable string instead of the tuple
gmtime/localtime give you, you need to format using strftime.

Example:

time.strftime("%Y-%m-%d %H:%M:%S",
    time.localtime(os.stat("/tmp/a")[stat.ST_ATIME]))

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 25.0 ?C      Wind: 1.1 m/s




From Chris.Barker at noaa.gov  Tue Jun 11 12:55:49 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 11 Jun 2002 09:55:49 -0700
Subject: Sorting list (maybe stupid)
References: <slrnagbh7m.h03.jarvin24@lehtori.cc.tut.fi> <ae4gnp$l18$1@news.tudelft.nl> <ae4hn9$jm4$1@wanadoo.fr>
Message-ID: <3D062B7C.FE6A4963@noaa.gov>


Eric Brunel wrote:

> l = [(102, 'foo'), (2, 'bar'), (1, 'baz')]
> l.sort()
> print l
> 
> Apparently, the default sort function already sorts tuples according to
> their first element...

Note that this only works if the rest of the elements in the tuples are
also sortable. In this case, they are strings, so they are.

-Chris




-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From whisper at oz.net  Mon Jun 24 22:10:57 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 24 Jun 2002 19:10:57 -0700
Subject: Suggestions for good programming practices?
In-Reply-To: <3D17CC8E.ABB72D14@engcorp.com>
Message-ID: <GCEDKONBLEFPPADDJCOEEEOMDLAA.whisper@oz.net>

Somebody seems to be working overtime to miss the point here...

BTW, struct is a module name :->

David LeBlanc
Seattle, WA USA 

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Peter Hansen
> Sent: Monday, June 24, 2002 18:51
> To: python-list at python.org
> Subject: Re: Suggestions for good programming practices?
> 
> 
> Brian Quinlan wrote:
> > 
> > Peter Hansen wrote:
> > > You _are_ returning them "in a struct".  It's spelled "tuple"
> > > but other than that what's the difference?
> > 
> > struct members are access by name while tuple elements are accessed
> > positionally.
> 
> Okay.  So why is that a problem?
> 
> Or to ask it from the other side, what's stopping you from 
> returning items in an object?
> 
> class Struct:
>     pass
> 
> def returnStruct():
>     result = Struct()
> 
>     result.anInt = 5
>     result.aString = 'foo bar baz'
>     result.spam = ('eggs', 'brian')
>  
>     return result
> 
> -Peter
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From dalke at dalkescientific.com  Sun Jun  2 01:53:26 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sat, 1 Jun 2002 23:53:26 -0600
Subject: What does sys.exit(1) means?
References: <adcb22$100q9d$1@ID-49758.news.dfncis.de>
Message-ID: <adcbq8$4hc$1@nntp9.atl.mindspring.net>

Ken:
>What does sys.exit(1) means?

It's the same as

  raise SystemExit(1)

That is, it raises an exception.  See the documentation for more
information about what that means.  If nothing else catches the
exception (which may happen accidentally with a blanket 'excect:')
then the top-level interpreter catches that exception, gets the
exit value, quits the main loop, and returns that value (in this
case, the integer '1') as the exit code for the program.

                    Andrew
                    dalke at dalkescientific.com





From timr at probo.com  Mon Jun 10 01:56:22 2002
From: timr at probo.com (Tim Roberts)
Date: Sun, 09 Jun 2002 22:56:22 -0700
Subject: Removing ^M
References: <mailman.1023549912.12643.python-list@python.org> <adteeb$250i1$4@ID-100555.news.dfncis.de> <adteif$250i1$5@ID-100555.news.dfncis.de>
Message-ID: <ppf8gu4llpreg1gmt8hft1k4rjhcvvs8a3@4ax.com>

Michael Kelly <mkelly2002 at adelphia.net> wrote:

>Michael Kelly wrote:
>
>> If you're on a unix flavor this works pretty well
>> 
>> tr -d '\015' <infile >outfile
>
>Whoops!  Should just be
>tr -d '015' <infile >outfile
>
>No need for the backslash.

There is if you want to remove carriage returns!  Without the backslash,
your command removes all occurrances of the digits '0', '1', and '5' from
the file.  Not entirely useless, but certainly much less useful than your
FIRST example.

In other words, you were right the first time...
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From mis6 at pitt.edu  Wed Jun 19 08:15:21 2002
From: mis6 at pitt.edu (Michele Simionato)
Date: 19 Jun 2002 05:15:21 -0700
Subject: Tkinter and the Menu Widget
Message-ID: <2259b0e2.0206190415.1f232447@posting.google.com>

Hello everybody !

I just started playing with Tkinter. I have a problem with the Menu
Widget. I want to create a menu like that

    File   Help
    -----  -----
    File1  Help1
    File2  Help2

Actually this is a simplified version of what I need since for my application
the number of items N has to be dynamic; here for sake of semplicity I took 
N=2. Notice that I want to store the labels for the various items 
in a dictionary like that:

    menuD={'File':['File1','File2'],'Help':['Help1','Help2']}

This dictionary is created dynamically by another program and can have
much more complicated items.
When I invoke a specific menu a procedure has to be called which
do something on a string of the form "invoked menu=invoked item" (for instance 
"Help=Help1" if I invoked the item Help1 of the menu Help). Suppose for 
sake of simplicity that dosomething(s) simply print the string s, even if
in the real case dosomething() calls an external program which requires
the string s as input. My attempt has been the following code:

    import Tkinter
    def dosomething(s):
        print s
    root=Tkinter.Tk()
    mn=Tkinter.Menu(root) #attach the menu mn to root
    menuD={'File':['File1','File2'],'Help':['Help1','Help2']}
    smn={} #dictionary keys->submenu-objects defined below
    for key in menuD.keys():
        smn[key]=Tkinter.Menu(mn) #define a submenu of mn
        mn.add_cascade(label=key,menu=smn[key]) #attach smn[key] to mn
        for submenu in menuD[key]:
            smn[key].add_command(label=submenu,command=lambda :
                                 dosomething(key+'='+submenu))
    root.config(menu=mn) #display the menu mn
    Tkinter.mainloop()

I used the lambda statement since <command=> requires a callable object
as argument; notice that this code only runs in Python2.2 (in Python2.1
with the statement from __future__ import nested_scopes).

This code correctly creates the menu; however since at the end of the
loop the value of key is 'Help' and the value of submenu is 'Help2',
when I invoke any item of the menu with the mouse, dosomething always 
print the string 'Help=Help2'. Instead, I would like to print the string 
corresponding to the actually called menu ! 
The Tkinter documentation I have is not helpful, I need a way to pass
to dosomething some identification of the called menu. Any suggestion ? 
 
--
Michele Simionato - Dept. of Physics and Astronomy
210 Allen Hall Pittsburgh PA 15260 U.S.A.
Phone: 001-412-624-9041 Fax: 001-412-624-9163
Home-page: http://www.phyast.pitt.edu/~micheles/


From jh at web.de  Fri Jun  7 03:39:12 2002
From: jh at web.de (=?ISO-8859-1?Q?J=FCrgen_Hermann?=)
Date: Fri, 07 Jun 2002 09:39:12 +0200
Subject: Medium to Large Scale Python Deployments
References: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com> <mailman.1023409393.3361.python-list@python.org> <3D003EC7.C72D27BE@engcorp.com>
Message-ID: <3D006320.5040805@web.de>

Peter Hansen wrote:
> Christopher Armstrong wrote:
> 
>>>I'm also curious about large amounts of Python, period. Who knows of ~50K
>>>Python lines-of-code (LOC) systems? 100K LOC? More?
>>
>>chris at radii:~/Projects/Twisted$ find -name '*.py' | xargs wc -l
>>...
>>  56051 total
> 
> 
> How big is that when run through pycount?
> 
> (I.e. comments and blank lines don't count!)
> 
> -Peter

Summary on "twisted"
====================
files                  256
lines                51654
bytes              1728211
comments              5993
empty lines          10029
non-commentary lines 29910



From romberg at smaug.fsl.noaa.gov  Wed Jun 12 12:31:45 2002
From: romberg at smaug.fsl.noaa.gov (Mike Romberg)
Date: 12 Jun 2002 10:31:45 -0600
Subject: setup.py breaks Modules/Setup.local
Message-ID: <ul3bsagsbke.fsf@smaug.fsl.noaa.gov>

  I am attempting to get Modules/Setup.local working to build python
2.2.1.  But there seems to be a bad interaction with setup.py.  I have
a Modules/Setup.local which contains a definition for the _tkinter
module.  I would like to use a Tcl/Tk release from a location that
setup.py does not find.  It looks like this:

# Edit this file for local setup changes
*shared*

_tkinter _tkinter.c tkappinit.c -DWITH_APPINIT \
        -L/usr/local/tcltk-8.3.2/lib \
        -I/usr/local/tcltk-8.3.2/include \
        -I/usr/X11R6/include \
        -ltk8.3 -ltcl8.3 \
        -L/usr/X11R6/lib \
        -lX11

# We don't want ssl sockets 'cause this opens
# up a whole can of DLL heck (often missing on stock redhat setups).
_socket socketmodule.c

*static*

  Now when I build python I see this:

  1.  _tkinter.o is built with the -I switches I want (this is good)

  2.  _tkinter.so is linked with the -L switches I want (also good)

  3. setup.py runs and relinks _tkinter.so with the incorrect -L
     switches.  (not good)

gcc -fPIC -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include -DHAVE_CONFIG_H  -DWITH_APPINIT -I/usr/local/tcltk-8.3.2/include -I/usr/X11R6/include -c ./Modules/_tkinter.c -o Modules/_tkinter.o
gcc -shared  Modules/_tkinter.o Modules/tkappinit.o  -L/usr/local/tcltk-8.3.2/lib -ltk8.3 -ltcl8.3 -L/usr/X11R6/lib -lX11  -o Modules/_tkinter.so
case $MAKEFLAGS in \
*-s*) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -g -O3 -Wall -Wstrict-prototypes' ./python -E ./setup.py -q build;; \
*) CC='gcc' LDSHARED='gcc -shared' OPT='-DNDEBUG -g -O3 -Wall -Wstrict-prototypes' ./python -E ./setup.py build;; \
esac
running build
running build_ext

[snip]

building '_tkinter' extension
skipping /scratch/Python-2.2.1/Modules/_tkinter.c (build/temp.linux-i686-2.2/_tkinter.o up-to-date)
skipping /scratch/Python-2.2.1/Modules/tkappinit.c (build/temp.linux-i686-2.2/tkappinit.o up-to-date)
gcc -shared build/temp.linux-i686-2.2/_tkinter.o build/temp.linux-i686-2.2/tkappinit.o -L/usr/X11R6/lib -L/usr/local/lib -ltk8.3 -ltcl8.3 -lX11 -o build/lib.linux-i686-2.2/_tkinter.so


  Is there some way to prevent this?  Or is there something that I'm
not doing correctly?

Thanks,

Mike Romberg (romberg at fsl.noaa.gov)

  


From whisper at oz.net  Mon Jun 17 16:29:17 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 17 Jun 2002 13:29:17 -0700
Subject: Long pathes on windows 
In-Reply-To: <7647A9F4B298E74DB6793865DA67285004AD86@exchange.adrembi.com>
Message-ID: <GCEDKONBLEFPPADDJCOEGELDDJAA.whisper@oz.net>

I'm not sure how you convert long paths containing spaces into the short
mangled names that Windows 98 etc. used. I know they are on Windows 2000,
since i've had a long filename "fold" itself for no reason I could figure
out, but I don't know how to _make_ it happen.

You actually don't have to get rid of the spaces (although it's inconvenient
having them imo): all you need to do is quote the whole path as in: f =
open("c:\\Documents and Settings\\Administrator\\Local
Settings\\Temp\\conf.h", 'rw').

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Roman Yakovenko
> Sent: Monday, June 17, 2002 11:18
> To: python-list at python.org
> Subject: Long pathes on windows
>
>
> Good evening. I have big problem Windows 2000 and small one.
> The small one is long pathes on windwos. How can I convert long
> path with ' '
> into short path?
>
> Thanks for the help. Roman.
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From beej at piratehaven.org  Fri Jun 28 21:24:07 2002
From: beej at piratehaven.org (Beej Jørgensen)
Date: 29 Jun 2002 01:24:07 GMT
Subject: I'm an idiot
References: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16>
Message-ID: <afj27n023m7@enews2.newsguy.com>

In article <Xns923BB72B1BAF7davidgriswold1netsca at 204.127.202.16>,
David  <david_griswold1 at yahoo.com> wrote:
>Any help would be appreciated.

Ok, I'm no pro and don't have the manuals in front of me, so I hope I
don't lead you too astray here:

>f=open('c:\\temp\\temp.txt', 'r')
>g=open('c:\\temp\\temp1.txt', 'w')
>while 1:
>    try:
>        s=f.readline
             ^^^^^^^^

This is a method, so I think it should be "f.readline()"

>        g.write(s.split())

Er, did you mean strip()?  And maybe you have to use String.strip(s),
but I can't remember.

>    except IOError:
>        break

Close--but I don't believe readline() throws an exception on EOF;
instead it just returns "".

>g.close
>f.close

So something like this:

    while 1:
        s = f.readline()
        if s == "": break
        g.write(String.strip(s))  # or s.strip(), if that doesn't work

Sorry if I hosed something here, but maybe it'll be enough to get you
going even if the info isn't entirely accurate.

-Beej



From mfranklin1 at gatwick.westerngeco.slb.com  Wed Jun  5 09:37:00 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Wed, 5 Jun 2002 13:37:00 +0000
Subject: Fastest way to read / procsess / write text?
In-Reply-To: <200206051234.g55CYsB06651@helios.gatwick.geco-prakla.slb.com>
References: <3CFD5229.2060803@excite.com> <200206051234.g55CYsB06651@helios.gatwick.geco-prakla.slb.com>
Message-ID: <200206051240.g55CeDB10732@helios.gatwick.geco-prakla.slb.com>

On Wednesday 05 Jun 2002 1:31 pm, you wrote:
> On Tuesday 04 Jun 2002 11:47 pm, you wrote:
> > I'm somewhat new to python and programing in general and I trying to
> > write a simple script that I can use to format large text files so that
> > they are suitable for importing into MySQL. So far, I have came up with
> > this:
> >
> > #!/usr/bin/python2
> >
> > import sys, re
> >
> > infile = sys.stdin
> > data = infile.read()
> > infile.close()
>
> This was the standard idiom (untill xreadlines came along I guess)
>
>
> while 1:
>     data=infile.readlines(100000) # hint to readlines for 100000 bytes....
>     if not data:
>         break ## EOF reached
>     for line in data:
>         ## process data
>
> > data = re.sub('[.](?=\d\d\d\d\d\d)|[.](?=\w+[ ][<|>])|[.](?=\w+[:])|[
> > ](?!0x)', '\t', data)
> >
> > outfile = open("/mnt/storage/output.txt", "w")
> > outfile.write(data)
> > outfile.close()



Havine re-read you original version this is how to do it;-)

while 1:
    data=infile.read(100000)
    if not data:
        break ## EOF reached
    ## process data.....

















> >
> >
> > This works beautifully most of the time (it's super fast), except when I
> > pipe large files (>50megs) to it and then it usually dies half way
> > through complaining of memory errors because it ran out of ram.
> >
> > Short of buying more ram, is there a way I can make this more effecient
> > without taking a big peformance hit? I tried the above where I used a
> > "for line in data:" loop so it would only process one line at a time but
> > this seemed to take forever! I'm imagining there's a way to process the
> > data in something like 1 meg chunks, but I'm not too sure how to do that.
> >
> > I'm using python 2.2.1 on a RH linux 7.2 system with 256 ram and a
> > Athlon XP1700 if it makes any diffrence.
> >
> > Any suggestions?
> >
> > TIA,
> > Brent




From chris.gonnerman at newcenturycomputers.net  Tue Jun 18 08:24:13 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Tue, 18 Jun 2002 07:24:13 -0500
Subject: Newbie - Process Management
References: <1024387201.371093@ernani.logica.co.uk>
Message-ID: <006a01c216c3$0f8b0cc0$0101010a@local>

As an addendum to my previous answer, I'd recommend you spend a little time
studying the online manuals and info files.  (Personally I HATE the GNU info
system but hey, we're stuck with it.  Perhaps if I was an EMACS user the
insane key bindings would make sense to me.)  Python's os module is, after
all, an interface to the operating system, so more subtle or detailed
information must come from the operating system documentation.

To read the man page for the kill() system call:

    man 2 kill

(at the shell prompt of course, not the Python prompt.)  Access the info
system with

    info

but read the help right away because the key bindings are nasty.  Current
documentation for GNU command-line tools will be in the info system, but
system calls are still in the man system.

Note that the first parameter to man can be omitted if you are sure that
the word you are looking for will be found in the right place (yeah, right).
On my system,

    man kill

does get you the right page, but some other keywords may turn up first in
the TCL documentation for instance; that bites me regularly.  The manual
sections are based on the numbering of the old AT&T Unix manuals.  Typing

    man man

will tell you a bit about using the system, including this list of standard
manual sections:

       1   Executable programs or shell commands
       2   System calls (functions provided by the kernel)
       3   Library calls (functions within system libraries)
       4   Special files (usually found in /dev)
       5   File formats and conventions eg /etc/passwd
       6   Games
       7   Macro packages and conventions eg man(7), groff(7).
       8   System administration commands (usually only for root)
       9   Kernel routines [Non standard]

There may be additional sections which are non-standard on your system.

So, say you don't know where information on the kill() system call is:

    apropos kill

gets this on my system:

SDL_KillThread (3)   - Gracelessly terminates the thread.
pthread_kill (3thr)  - handling of signals in threads
pthread_kill_other_threads_np (3thr) - terminate all threads in program
except calling thr
ead
kill (1)             - terminate a process
kill (2)             - send signal to a process
killall (1)          - kill processes by name
killall5 (8)         - send a signal to all processes.
killchar (3x) [curs_termattrs] - curses environment query routines
killpg (2)           - send signal to a process group
killpg (3)           - send signal to all members of a process group.
killproc (8)         - Send signals to processes by full path name
rotatelogs (8)       - rotate Apache logs without having to kill the server
skill (1)            - report process status
yes (1)              - output a string repeatedly until killed
XKillClient (3x)     - control clients
xkill (1x)           - kill a client by its X resource

and it turns out that the kill in section 2 is the one we are looking for.

<<Sigh>> Everyone has to have their own help system; KDE and GNOME each have
built-in hypertext help which you will have to learn about to separately.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net





From gmcm at hypernet.com  Sat Jun  1 11:39:29 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 01 Jun 2002 15:39:29 GMT
Subject: Gordon McMillan's Installer - Data files in the archive
References: <y9e0xy2x.fsf@morpheus.demon.co.uk> <uofewjgv4.fsf@ctwd0143.fitlinxx.com>
Message-ID: <Xns922076BD73935gmcmhypernetcom@199.171.54.215>

David Bolen wrote:

> Paul Moore <gustav at morpheus.demon.co.uk> writes:

[a bunch of Installer questions that David answers 
 as well as I could...]

>> On a (semi-) related note, is it possible to make Installer *not*
>> track dependencies back into the installed copy of Python, and so
>> produce executables which assume the existence of an installed copy of
>> Python on the target system? 

[snip]

> I've also wanted to do this but haven't had a chance to run tests, so
> this is just my thinking on how I would do it when I get time.

[snip]

I'll add another hint. TOC's support set operations (intersection, 
union and difference in particular). So if you had a TOC describing
the std lib, you could do
 apppure = a.pure - stdlib
 appbinaries = a.binaries - stdbinaries

Another hint: look at the .toc files in the buildxxx directory
and note how easy it is to cut-and-paste.

In some ways, wxPython is easier, since you can exclude it from
the Analysis (in the latest Installer release).

I believe David's right that you'll have to modify launch.c.
On Windows, commenting out the sys.path manipulations should
do it. On Linux / Unix you'd have to modify the make to use
Python's getpath.c, not my hacked version.

-- Gordon
http://www.mcmillan-inc.com/


From brueckd at tbye.com  Sun Jun 16 20:52:52 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Sun, 16 Jun 2002 17:52:52 -0700 (PDT)
Subject: What If.....    Strong Types
In-Reply-To: <3D0C0F0D.4030900@bgb.cc>
Message-ID: <Pine.LNX.4.44.0206161735370.6273-100000@ns0.tbye.com>

On Sun, 16 Jun 2002, Don Garrett wrote:

>    Primarily, what if classes always had rigidly defined interfaces. I
> mean that public members had to be declared to exist, and that methods
> couldn't be modified on an instance. Private members could exist but
> would be really private.

:( IMO this would add little value and we would lose a lot in the process. 
I wouldn't be bothered if some sort of optional interface support was 
built into the language itself, but I don't have the problem with the lack 
of it today (same with private members - it just doesn't cause me any 
problems the way it is). On the downside, some of the most Pythonic 
solutions to problems that'd be messy to solve in other languages are 
quite elegant in Python because of its "dynamicness". For example, it's 
sooo convenient to attach/detach members to objects as they pass through a 
program's workflow. Having to declare them all up front or whatever would 
be messy.

> All types would be classes (including ints and such).

I never understood all the hooplah over ints not being classes. I've only 
ever seen a handful of cases where it might be useful, and lots of cases 
where it'd get abused, but some people really seem to want it.

>    My belief is that almost all the convenience of Python would be
> maintained, but that compilation, optimization, and automated code
> analysis would be simpler.

I'm _sort of_ in favor of some optional type annotation for optimization 
purposes, and sort of against it, the idea being that after your code is 
done you could go back and annotate (or even have a tool suggest 
annotation) variables with type information such that you are saying that 
a particular variable name will always refer to integers, for example. My 
fear, though, is that it would be used too often, i.e. people would 
provide type information in places where it wouldn't help at all in terms 
of execution speed. One of the beauties of Python is the clean syntax, the 
readable code. A good portion of that is the code is not encumbered with 
all sorts of variable and type declarations. I like the approach of Pyrex 
because it's close to Python but still separate enough to prevent people 
from using it too much - if you make it too easy to optimize the code then 
people will optimize it too often. Bleh.

>    I'm just wondering if the change would be good or bad. Would it really 
> break anything important in Python today?

You might not consider it important, but it'd break a lot of _my_ code, so 
it'd be important to me! :) Python doesn't force you to use all its 
dynamic abilities, but I regard that as one of the tools of the Python 
toolbox, so yes, I'd be hurt by its removal.

> Would it prevent any common errors? 

It'd be most useful for people who skimp on testing, but the added 
complexity might actually introduce errors, so who knows what the net 
effect would be - perhaps even buggier code than without it.

> Would it help with automated code analysis (compilation and optimization 
> included) as much as I think it would?

Dunno - how much do you think it would? ;-) A better question, is how
often do we really need compilation and optmization at all anyway? If it's
less than 10% (which I believe it is), then the payoff generally wouldn't
be worth language-wide changes. A plain extension module (especially with
the help of Boost, SWIG, PyInline, etc.) isn't too painful, and projects
like Pyrex might have a permanent niche too, but in all cases they are one 
step removed from standard Python - which is exactly where they belong.

-Dave





From fperez528 at yahoo.com  Wed Jun 19 16:21:42 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 19 Jun 2002 14:21:42 -0600
Subject: OS X debugger?
References: <aeqotj$dm$1@panix1.panix.com>
Message-ID: <aeqp4m$c09$2@peabody.colorado.edu>

Aahz wrote:

> Anyone know a good visual debugger for OS X???I'd?originally?been
> planning to just use shell windows for the demos of my OSCON tutorial,
> but it occurs to me that a single-stepping debugger with watch variables
> would be even better

You can probably build ddd for OS X. I've used it for python in the past and 
it's quite nice (but I've only used it in Linux, so I don't know how tricky 
the osX build would be. It's Motif based).

cheers,


f


From marklists at mceahern.com  Thu Jun 27 09:00:12 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 27 Jun 2002 08:00:12 -0500
Subject: sys.argv[0] - Returns path and filename.
In-Reply-To: <c07d0edc.0206270431.5f2709b4@posting.google.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEMJLGAA.marklists@mceahern.com>

> How do I get the current path of the script I am running, on the mac
> osx and the linux box ? (Any work around would be great, I don't want
> to have to hardcode the path into the script, or even give the path as
> an argument,I want the script to know where it is.)

import os
print os.getcwd()

// m
-





From derek at wedgetail.com  Fri Jun 21 07:02:09 2002
From: derek at wedgetail.com (Derek Thomson)
Date: Fri, 21 Jun 2002 11:02:09 GMT
Subject: converting an array of chars to a string
References: <3d12dda2_6@news.newsgroups.com> <3D12E228.6030501@wedgetail.com> <mailman.1024648440.1200.python-list@python.org>
Message-ID: <3D1307AF.3070308@wedgetail.com>

Gerhard H?ring wrote:
> 
> 
> Unless Python 1.5.2 compatibilty is a requirement, just use string
> methods:
> 
>     lst = ['a', 'b', 'c', 'd']
>     s = ''.join(lst)

And they say you can't write obfuscated code in Python :)

I would have expected "join" to be a method on the sequence, if 
anything. As in "s = lst.join('')". In fact, I looked for it, as I 
*knew* I'd seen it used before as a method, I just didn't expect it to 
be in the string class.

I think I'll just stick with string.join for the time being. As it turns 
out, there's more than one way to do it, even in Python ;)

< It's also not a good idea to replace the builtin list function, thus
 > I'm using 'lst' here.

You're right, of course.

Regards,
Derek.





From cliechti at gmx.net  Fri Jun 28 17:04:17 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 28 Jun 2002 23:04:17 +0200
Subject: joining 2 lists together - an elegant way?
References: <pan.2002.06.28.15.03.29.911402.1787@presidency.com> <pan.2002.06.28.15.06.32.33132.1787@presidency.com>
Message-ID: <Xns923BEB2573E3Ccliechtigmxnet@62.2.16.82>

Rajarshi Guha <rajarshi at presidency.com> wrote in 
news:pan.2002.06.28.15.06.32.33132.1787 at presidency.com:

> On Fri, 28 Jun 2002 15:03:29 -0400, Rajarshi Guha wrote:
> 
>> Hi,
>>   is there a more elegant way to join to lists together?
>> 
>> Right now I do:
>> 
>> newlist = []
>> for i in l1:
>>   newlist.append(i)
>> for i in l2:
>>   newlist.append(i)
>> 
>> It does the job - but a one liner (if possible) would be nice :) TIA,
> 
> Sorry :( Found it!

Fine! now if you would post it also you would 1) get feedback, 2) help 
others/newbees when they search the news archive...

just for the record:
>>> l1, l2 = [1,2,3], [4,5,6,7]
>>> l1 + l2    	    	#oneliner
[1, 2, 3, 4, 5, 6, 7]
>>> n = l1[:]    	#copy and
>>> n.extend(l2)    	#extend
[1, 2, 3, 4, 5, 6, 7]

chris

-- 
Chris <cliechti at gmx.net>



From emile at fenx.com  Wed Jun 19 06:58:50 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 19 Jun 2002 10:58:50 GMT
Subject: CPU Usage Windows
References: <ad381f5b.0206190233.4fa1bf58@posting.google.com>
Message-ID: <KtZP8.8424$EP.267@sccrnsc03>

Rad
> I've noticed that when I run Python script on
> Windows 2k Pro
> Dual Intel Xeon (2.2GHz)
> CPU Usage for pythonw.exe is only 50%?!
>
> Nothing else is running on the machine (apart form the OS). Some of
> the python programs are running for hours (large data sets) and even
> though python shell freeze's (doesn't refresh) CPU usage is still only
> 50%.
>
> System Idle Process is about 40%-50%
>
> Any ideas why is this and can I do anything about it?
>
> Is it something to do with python not being able to use the power of
> both processors?
>

IIRC, when I noticed something similar I found out both that task
manager has some issues with dual processor systems and that the 50%
meant 100% of one processor.

--

Emile van Sebille
emile at fenx.com

---------



From dougz at cs.washington.edu  Tue Jun 18 18:40:42 2002
From: dougz at cs.washington.edu (Douglas Zongker)
Date: Tue, 18 Jun 2002 22:40:42 GMT
Subject: results of stat[ST_ATIME]???????
References: <mailman.1024437565.26109.python-list@python.org>
Message-ID: <GxxAzv.10Dt@beaver.cs.washington.edu>

jubafre at zipmail.com.br wrote:
: when i use st[ST_ATIME] a int number is returned, how i can transform
: this number to a "time number"???????

What do you mean by "time number"?  The file times given by stat() are
just like the values returned by time.time(), except that they're ints
instead of floats.  You can do all the same things with them:

    >>> import os, time
    >>> access = os.stat('/tmp/foo').st_atime
    >>> access
    1024439265
    >>> time.ctime(access)
    'Tue Jun 18 15:27:45 2002'
    >>> time.localtime(access)
    (2002, 6, 18, 15, 27, 45, 1, 169, 1)

dz


From skip at pobox.com  Mon Jun 24 21:54:06 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 24 Jun 2002 20:54:06 -0500
Subject: How to stop a SocketServer?
In-Reply-To: <3D17BF72.2E1CCF40@engcorp.com>
References: <7xhejv9af2.fsf_-_@ruckus.brouhaha.com>
        <3d1744d4$1_1@hpb10302.boi.hp.com>
        <3D1757DF.776A428F@engcorp.com>
        <3d176a0c$1_3@hpb10302.boi.hp.com>
        <3D17BF72.2E1CCF40@engcorp.com>
Message-ID: <15639.52542.299814.852145@12-248-8-148.client.attbi.com>

    Peter> By the way, look into "timeoutsocket.py" which is a plug-in
    Peter> replacement for the standard socket.py and which might do the
    Peter> trick for you.

Note socket objects grow a settimeout() method in 2.3.  If you get Python
from CVS you can have it today:

    settimeout(value)

    Set a timeout on blocking socket operations.  The value argument can be
    a nonnegative float expressing seconds, or None.  If a float is given,
    subsequent socket operations will raise an error exception if the
    timeout period value has elapsed before the operation has completed.
    Setting a timeout of None disables timeouts on socket operations.
    s.settimeout(0.0) is equivalent to s.blocking(0); s.settimeout(None) is
    equivalent to s.setblocking(1).  New in version 2.3.

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html




From cbbrowne at acm.org  Sat Jun  8 12:42:36 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: 8 Jun 2002 16:42:36 GMT
Subject: Where is Berkeley DB on your system? Can you live without DB 1.85?
References: <15617.3834.292130.446065@12-248-41-177.client.attbi.com> <GCEDKONBLEFPPADDJCOEAEGJDHAA.whisper@oz.net> <mailman.1023535391.14446.python-list@python.org>
Message-ID: <adtc5s$232o1$2@ID-125932.news.dfncis.de>

The world rejoiced as Oleg Broytmann <phd at phd.pp.ru> wrote:
> On Fri, Jun 07, 2002 at 01:14:59PM -0700, David LeBlanc wrote:
>> If Sleepycat will allow it, i'm all in favor of updating to the latest
>> version! However, they want money for commercial use of versions > 1.85
>> AFAIK.
>
>    They don't. You can freely incorporate BSDDB into an opensource project.
>    http://www.sleepycat.com/faq.html#A22

Ah, but that presumably _wouldn't_ mean that incorporating it into
Python would allow it to be freely incorporated into a
"non-OS-project" that uses Python.
-- 
(reverse (concatenate 'string "gro.mca@" "enworbbc"))
http://www3.sympatico.ca/cbbrowne/wp.html
Mental health is overrated!! 


From anoop79 at myrealbox.com  Tue Jun 18 15:41:51 2002
From: anoop79 at myrealbox.com (Anoop P B)
Date: Wed, 19 Jun 2002 01:11:51 +0530
Subject: tray icon using tkinter
References: <mailman.1023950718.20009.python-list@python.org> <e08b1d4.0206171530.563e67fd@posting.google.com> <4a7fde99.0206180244.11eea205@posting.google.com>
Message-ID: <3D0F8CFF.1000905@myrealbox.com>

thanks a lot for your help. appreciate it.

actually, i havent done much tkinter or mfc before. i have been playing 
with wxpython for a month. i didnt like the idea that i had so many 
widget-ID numbers floating about in wxpython that i had to keep track 
of. and i dont understand why the concept of widgetID numbers - why not 
just link the widget-name with the handler?

anyways, so i decided to check out tkinter and see if i can write my 
current project in tkinter instead. it appears as though i need to learn 
mfc to do what i want (the tray icon). so, i am thinking i'd rather go 
back to using wxpython.

the TrayIt thing suggested by John looks like a gr8 tool but wont serve 
my purpose here since i want to also learn how to integrate this feature 
into my python program. Besides, deployment on the client's machine will 
be a problem. secondly, tkinter3000 seems to be in pre-alpha and the 
tkinter3k docs that i could find on the net dont seem to have anything 
on howto do the tray icon thing. or have i missed something there?

~anoop

Ryan wrote:
> john at johnnypops.demon.co.uk (John Popplewell) wrote in message news:
> 
>>Anoop P B <anoop79 at myrealbox.com> wrote in message news:
>>
>>>	is there a way that i can provide a tray icon ie. next to the clock (on 
>>>win32) for my program (using tkinter) ? i know that there is a way of 
>>>doing this in wxpython but i need to know whether its possible in tkinter.
>>>
>>>thanks,
>>>anoop
>>
>>
>>Have you had a look at "Tray It!" ?
>>
>>http://www.teamcti.com/trayit/index.html
>>
>>cheers,
>>John.
> 
> 
> also look at the win32 extensions, I remember going through all the
> demo code that there is a way to do it (MFC I think but don't know for
> sure).  It is simple to do a try except block around the import and
> logic to use it since it will only be for windows.  It is used for the
> windows IDE also.
>   Also check out TKinter3000 at 
> http://www.pythonware.com/products/tkinter/tkinter3000.htm, they may
> have something (or you can ship the one from win32 extensions with
> your app)






From gcordova at hebmex.com  Mon Jun 17 13:49:06 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Mon, 17 Jun 2002 12:49:06 -0500
Subject: Divorce property objects from attribute access dependency?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233012DE3D4@mail.hebmex.com>

> 
>  >>> class C(object):
>  ...     def __init__(self,v):
>  ...         self._v = v
>  ...     def get_v(self): return self._v
>  ...     def set_v(self,v): self._v = v
>  ...     v=property(get_v,set_v)
>  ...     vlist = [v,v]
>  ...
>  >>> c=C('something')
>  >>> c.v
>  'something'
>  >>> c.v='other'
>  >>> c.v
>  'other'
>  >>> c.vlist
>  [<property object at 0x0082D4A0>, <property object at 0x0082D4A0>]
>  >>> c.vlist[0]
>  <property object at 0x0082D4A0>
> 
> Why not trigger get_v() in these accesses too, and have 
> single mechanism to suppress it instead of a single
> mechanism to effect it?
>

You mean that you don't recognize the difference between
evaluating "c.v" and "c.vlist"?

Well, for one:

"c.v" is evaluated at runtime, and it's recognized as
an instance's access to a class' property attribute,
so all the background magic happens (which is what
you expected). Just a note: get_v and set_v are
methods, which take "self", and all that. Ok.

On the other hand, "vlist" is created at compile
time, and it's a list, not a property attribute;
it's only a list with two objects. When you're
evaluating "c.vlist", the items in the list are not
in the correct context to be evaluated as property
attributes, so the background magic does not happen.
For one, there is no "self" to pass to the methods
get_v and set_v; another, they're "property attributes",
not "magical list items". :-)
 
> ISTM kind of a hack to rig access via instance attribute as a 
> mechanism for controlling referenced-object behavior (this is
> my impression, I haven't checked code ;-). IMO default behavior
> of an object in lhs or rhs context should not depend on how a
> reference to it was acquired and dereferenced.
> 
> (Obviously, if property objects always acted as they now do 
> through instance attribute access, "vlist=[v,v]" in this
> example would have to have some kind of access override like
> "vlist = [obj(v),obj(v)]", or else it would generate 
> ['something','something'] right there. Using obj(p) a property
> object could be passed around without evaluation and the
> potential side effects thereof).
> 

A property object *can* be passed around without side effects,
it's only when accesed as an attribute of an instance of a
class which defines them, that the expected behaviour
triggers.

For example, try this:

>>> class klass(object): pass
...
>>> ob = klass()
>>> ob._blah = 10
>>> def _get_blah(self): return self._blah
>>> def _set_blah(self, b): self._blah = b
>>> ob.blah = property(_get_blah, _set_blah)

What behaviour do you expect from this? Give it a try. :-)

> BTW, it would be nice to be able to define a property at 
> global module scope also, and perhaps this would tie in nicely.
> 
> Just musing ;-) Thoughts? Am I missing a big gotcha?
> One thing is that it will effectively make possible a ()-less 
> function call, which will upset some people ;-)

You can do that without properties:

>>> class Caller:
...    def __init__(self, func, *args, **kw):
...       self.func, self.args, self.kw = func, args, kw
...    def __str__(self): return self.func(*args, **kw)
...    __repr__ = __str__
...
>>> def PrintTen(): print 10
...
>>> ten = Caller(PrintTen)
>>> ten
10

Another thing, you *can* create "magical list items"
with this class; since __repr__ is also defined,
then when the list which contains them is rendered,
the __repr__ method of the items is called. Neat. :-)

I Know, it's a simple-minded example, but you know
what I mean. You can use this Caller class to execute
a function when you type the name of the object that's
related to the function.

Good luck :-)

-gustavo

> 
> Regards,
> Bengt Richter
> 




From shalehperry at attbi.com  Tue Jun  4 16:56:04 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 04 Jun 2002 13:56:04 -0700 (PDT)
Subject: self
In-Reply-To: <ZQ8L8.80$1e2.4189336@newnews.cc.stevens-tech.edu>
Message-ID: <XFMail.20020604135604.shalehperry@attbi.com>

On 04-Jun-2002 Vojin Jovanovic wrote:
> I don't see how that is going to work.  In fact I tried things like that
> before without success.  Just sticking the dictionary into eval/exec is
> missing the point.  Every time parsing is done flow needs to go through
> __getattr__ special method to evaluate nested equations.  Without self. you
> can't achieve that effect.  Try the approach you suggested and you'll get
> '#U'.
> 

calc.setEquation('f = a*c')
calc.valueFor('f')

sometimes you can just redesign.




From emile at fenx.com  Sun Jun 23 21:22:15 2002
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 24 Jun 2002 01:22:15 GMT
Subject: mail-problem over t-online
References: <mailman.1024866209.9154.python-list@python.org>
Message-ID: <bvuR8.302454$cQ3.16047@sccrnsc01>

"Klaus Reinhardt" <K.Rdt at TU-Berlin.DE> wrote in message
news:mailman.1024866209.9154.python-list at python.org...
> Am 23.06.02 22:57:06, schrieb Peter Hansen <peter at engcorp.com>:
>
> >Klaus Reinhardt wrote:
> >>         host=smtplib.SMTP('mailszrz.zrz.tu-berlin.de')
> >>         host.sendmail(zwi_eck(From),zwi_eck(Tp),al)
> >>         host.quit
> >> authentification required .. Providing my log and pass
> >> produces errors, too.
>
> >Try connecting manually with TELNET first, to ensure you
> >have a clear channel to that server and port.  Many ISPs
> >will block port 25 except to their own mail server, to
> >prevent you using them to send spam through other systems
> >and having them suffer the consequences...
> >
> >-Peter
> ---------------------------------------------------------------------
> Hi
> Thanks for the hint (port 25); the telnet is working
> and on anis.tu I have a python, so I can launch a
> script; but this is the long way. My goal is to fiddle all
> together to one script. For now I couldn't establish
> a working telnet a.s.o. - But these are my python-
> beginnings. -)
> K at Rdt
> ---------------------------------------------------------------------
>

The intent of the hint was to manually telnet into port 25 and send mail
to see what was going on -- not as an alternative method of
implementation.  Did you do that and see that you can send mail?


--

Emile van Sebille
emile at fenx.com

---------



From jepler at unpythonic.net  Tue Jun 25 09:34:14 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 25 Jun 2002 08:34:14 -0500
Subject: Suggestions for good programming practices?
In-Reply-To: <roy-2CF124.07480625062002@reader2.panix.com>
References: <mailman.1024971992.26429.python-list@python.org> <roy-2CF124.07480625062002@reader2.panix.com>
Message-ID: <20020625133409.GA4435@unpythonic.net>

On Tue, Jun 25, 2002 at 07:48:06AM -0400, Roy Smith wrote:
> "David LeBlanc" <whisper at oz.net> wrote:
> > Structs and tuples are different. They're both of the same general category
> > of aggregate data, but they are different as Brian notes.
> 
> I've come to really dislike using tuples for data structures, and 
> instead create small data classes.  Even if the class is no more 
> complicated than:
> 
> class myData:
>    def __init__ (self, foo, bar):
>       self.foo = foo
>       self.bar = bar
> 
> you've added quite a bit of functionality.  First, the code becomes a 
> little more self-documenting.  "power = ckt.volts * ckt.amps" is a lot 
> easier to read and understand than "power = ckt[0] * ckt[1]".

(the following applies most obviously to modern versions of python.  I
use subclassing of builtin objects, __slots__ and zip().  Even so, the
myStruct class is useful in any version of Python back to at least 1.5,
if you make myStruct a regular class, and include a definition
of zip())

You can do even better (lower memory overhead) with
    class myData(object):
	__slots__ = ['foo', 'bar']
        def __init__ (self, foo, bar):
            self.foo = foo
            self.bar = bar

you could go a little farther and make a superclass that has a __init__
that works for any subclass:
    class myStruct(object):
	def __init__(self, *args):
	    if len(args) != len(self.__slots__):
		raise TypeError, \
		    "__init__() takes exactly %d arguments (%d given)" \
		    % (len(self.__slots__)+1, len(args)+1)
	    for name, arg in zip(self.__slots__, args):
		setattr(self, name, arg)

    class myData(myStruct):
	__slots__ = ['foo', 'bar']

Jeff




From bokr at oz.net  Mon Jun 24 15:00:21 2002
From: bokr at oz.net (Bengt Richter)
Date: 24 Jun 2002 19:00:21 GMT
Subject: converting an array of chars to a string
References: <3d12dda2_6@news.newsgroups.com> <aevj41$fg2$0@216.39.172.122> <3d17038e_6@news.newsgroups.com> <af7jei$3kr$0@216.39.172.122> <3d1757c7_1@news.newsgroups.com>
Message-ID: <af7q85$lit$0@216.39.172.122>

On Mon, 24 Jun 2002 19:40:17 +0200, JB <jblazi at hotmail.com> wrote:

>Bengt Richter wrote:
>> No cause for desperation ;-)
>> 
>> Are you saying that you want modulo 256 arithmetic on
>> unsigned 8-bit data? No problem. But what does your data
>> represent? And what does adding a number to such a data
>> item mean in the real world?
>
>Of course you are right. I really needed only some 8-bit 
>shifting. The point is, that there is no need for the 
>overhead that is involved. If anybody uses the array module 
>which, you can argue, is a low level module, he should know 
>what he is doing.
>
It is hard to be very helpful without knowing the "big picture."
You originally said:

"How can I convert an array of chars to string?

The problem is this: I should like to perform 8 bit 
calculations in a buffer. After this has been done, the 
buffer has to ba written to a binary file (and file.write 
takes strings instead of lists)."

And I still have no idea why you want to do that, or what your
data represents, or how much of it there is, or where it comes
from, or how fast and/or often you have to process it.

Who knows, there may be a whole application out there that
does what you want, and more. But a narrow question about
8 bit calculations and buffers and binary files will not
generally evoke the association, unless someone happens to be
doing an implementation, and says, "Hey, that's just what I'm working
on in my blarglefoo framework. Are you by any chance working
on a similar app? It's for processing ..."

I'm not intending to criticize, I'm just trying to point out a typical
feature of newsgroup technical questions. They tend to say, "I need
a fast square root" and leave out that they don't have floating point
(ok, that'd be a probable inference), and they are checking distances
between things against a threshold. Whereas if they'd revealed what they
were trying to do, it could've been pointed out that they didn't need a
square root at all (just compare with the threshold distance squared).

Anyway, good luck with it ;-)

Regards,
Bengt Richter


From marklists at mceahern.com  Sat Jun 15 10:14:01 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sat, 15 Jun 2002 09:14:01 -0500
Subject: Pronouncing '__init__'
In-Reply-To: <20020615165915.A17174@hishome.net>
Message-ID: <NCBBLFCOHHDIKCAFGCFBGENFLEAA.marklists@mceahern.com>

> I used __init__ just as an example.  I meant how do you pronounce names
> with leading and trailing double underscore in general.  The magic name
> __init__ may be by far the one most commonly encountered in everyday
> programming but __there__ __are__ __many__ __others__.

As Aahz just kindly pointed out my (shared) mistake in thinking of __init__
as a constructor, I withdraw that vote.  ;-)

But I think saying, for example, the str method, in the context of a user
class, is clear enough to indicate __str__.  If you're talking about len(),
you probably just say "len."  If you're talking about __len__, you probably
say, "the len method of x" or just "the len method."  Or "the magic len
method."

Cheers,

// mark

-





From andersjm at dancontrol.dk  Mon Jun 17 07:14:06 2002
From: andersjm at dancontrol.dk (Anders J. Munch)
Date: Mon, 17 Jun 2002 13:14:06 +0200
Subject: Type subclassing: bug or feature
References: <aeau48$111$1@panix1.panix.com>
Message-ID: <3d0dc4dc$0$80428$edfadb0f@dspool01.news.tele.dk>

"Aahz" <aahz at pythoncraft.com> wrote:
> Consider the following code:
> 
> class MyStr(str):
>     def contains(self, value):
>         return self.find(value) >= 0
> 
> s = MyStr("hello, world!")
> s = s.capitalize()
> if s.contains('Hello'):
>     print "Found it!"
> 
> It fails with an AttributeError when it calls s.contains(), because
> s.capitalize() returned a str instead of a MyStr.  Anyone want to take a
> whack at defending this as the correct behavior?

Inheriting from a class/type with value semantics is usually bad
design.  Use containment or helper functions instead.

it-so-happens-that-Python-has-support-for-creating-thusly-flawed-designs-
but-that-doesn't-make-it-a-good-idea-ly y'rs, Anders




From occeanlinux at linuxmail.org  Sun Jun  2 15:56:58 2002
From: occeanlinux at linuxmail.org (Gold Fish)
Date: Sun, 02 Jun 2002 19:56:58 GMT
Subject: List
References: <3cfa6ed8_1@news.iprimus.com.au> <addrjh$imq$1@news.tudelft.nl>
Message-ID: <3cfa7619_1@news.iprimus.com.au>

Charl P. Botha wrote:

> In article <3cfa6ed8_1 at news.iprimus.com.au>, Gold Fish wrote:
>> Assume we got 2 list
>> list 1 = ['ABC','CDG','DAV','FE','FECA','FEAFA','FEA']
>> list 2 = []
>> how can we copy element from list 1 to list 2 say i just want to copy the
>> 3 element 'CDG','DAV','FE' to list2 only.
> 
> list1 = ['ABC','CDG','DAV','FE','FECA','FEAFA','FEA']
> list2 = list1[1:4]
> 
> Is that what you meant?
> 
No i mean if i got from list1 i wanna take 3 elements in any position to 
list2 but i don't know how many item in that list.


From SteveC at nospam.innocent.com  Fri Jun 21 15:06:49 2002
From: SteveC at nospam.innocent.com (Steven M. Castellotti)
Date: Fri, 21 Jun 2002 19:06:49 GMT
Subject: Dispatching a Win32 COM in the background?
References: <mailman.1024531044.10120.python-list@python.org> <9OpQ8.90552$ks6.8015@nwrddc02.gnilink.net> <3D127FE0.74D47795@engcorp.com>
Message-ID: <dPKQ8.1057$cE5.316@nwrddc02.gnilink.net>

On Thu, 20 Jun 2002 20:22:40 -0500, Peter Hansen wrote:

> "Steven M. Castellotti" wrote:
>> 
>> Thanks a lot guys, your suggestions have worked perfectly!
>> 
>>         Just for the record, here's what the full implementation looks
>>         like.
> 
> Beautiful... thanks for the followup Steven.  I especially like your use
> of deferred imports to speed up initial program loading.

	Thank you for the compliment. Speeding up program loading by deferring
imports is really only a side bonus though. The program is actually
cross-platfrom, meant to be run under both Linux and Windows. Since I'm
making calls to the Festival TTS program under linux, I have a seperate
class object, depending on what gets returned by os.name.


> [...]
>>       if not(self.text_queue.empty()):
>>          text = self.text_queue.get()
>>          speech.Speak(text)
>>          time.sleep(.001)
> 
> Note that the time.sleep() as shown above is doing nothing useful for
> you.  I suspect you had it unindented but your posting probably used
> tabs and spaces which makes it hard to see properly on some readers.
> 
> The reason it does nothing useful is that Python will automatically
> switch between different threads every so often anyway, so the concept
> of giving up the CPU for what amounts to a tiny fraction of a second is
> somewhat unnecessary.

	That makes sense to me, I think I'll remove the time.sleep() call. The
reason I had it in there in the first place (and it was unindented) was
because I have a similar call in a loop to Pygame (a wrapper to the SDL
graphics library), which is not multithreaded. I needed the calls in the
Pygame loop in order to prevent CPU usage from sitting at 100%, but if
multithreading will happen properly without it, I won't use it.


Cheers,

Steve Castellotti
SteveC (at) innocent.com
http://cogengine.sourceforge.net/


From david_griswold1 at yahoo.com  Fri Jun 28 23:51:16 2002
From: david_griswold1 at yahoo.com (David Griswold)
Date: Sat, 29 Jun 2002 03:51:16 GMT
Subject: I'm an idiot
References: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16>
Message-ID: <Xns923BD424CF53Cdavidgriswold1netsca@204.127.68.17>

Thank you, everyone, that took the time to help a newbie.

David



From fredegar at haftmann-online.de  Tue Jun  4 08:48:46 2002
From: fredegar at haftmann-online.de (Florian Fredegar Haftmann)
Date: Tue, 04 Jun 2002 13:48:46 +0100
Subject: Creating explorer link on Win32
Message-ID: <3CFCB72E.4010405@haftmann-online.de>

Hi!

I'm searching for a possibility to create an explorer link (*.lnk) to a 
file on Win32. Has anyne heard of a python module able to do that in a 
simple way?

Thanks,
FFH



From Uwe.Mayer at ifib.uni-karlsruhe.de  Mon Jun 17 06:40:07 2002
From: Uwe.Mayer at ifib.uni-karlsruhe.de (Uwe Mayer)
Date: Mon, 17 Jun 2002 12:40:07 +0200
Subject: what *is* a class?
References: <MPG.17774343cc5a589098968a@news.rz.uni-karlsruhe.de> <Gxtt3F.upA@beaver.cs.washington.edu>
Message-ID: <3D0DBC87.B1BDEB2D@ifib.uni-karlsruhe.de>

Douglas Zongker wrote:
> "Group()" creates a new object which is an instance of the class
> Group.  It calls the object's initializer ("__init__" method), then
> returns the newly created object.
> 
> : i want to write a method which returns the instance, so that f.e. the
> : following is possible:
> :   g = Group()
> :   g2 = g.foobar()
> :   g2 is g
> : so that g2 is identical to g.
> : any idea?
> 
>     class Group:
>         def __init__( self ):
>             print 'creating instance'
> 
>         def foo( self ):
>             print 'in the foo method'
>             return self
> 
>     >>> g = Group()
>     creating instance
>     >>> g2 = g.foo()
>     in the foo method
>     >>> g2 is g
>     1
>     >>>
> 
> Note that there's usually no reason for a method (such as "foo") to
> return the self object, because the caller should already have the
> object to call the method in the first place.  I could leave "return
> self" out of the definition of "foo", and just write:
>
>     g = Group()
>     g.foo()
>     g2 = g
> 
> The result would be the same.

I am very sorry. Of course I tried that, returning self, but perhaps I
called: g2 = g.foo   instead of g2 = g.foo()
:-(((

The reason why I wanted that was for lazyness. I am writing classes
which should provide random access to arbitrary binary file formats with
a special focus on RIFF.
I wanted to create an object model of the file (without actually reading
all of it - especially skipping the large data chunks).
I had to distinguish between creating the object model and actually
reading the file in. Thus I decided to use the constructor for creating
the object model and provide a parse() method. I didn't like writing:

f = file('dummy.avi','rb')
avi = RIFF(<specify format>)
avi.parse(f)

instead I wanted to write:

avi = RIFF(<specify format>).parse(f)

but then of course parse() must return an instance object.
On the one side all parse() methods must explicitly "return self" (which
is not nice). On the other side from the semantics I thought it was ok
for parse to return an instance object of RIFF.

What do you think?
Thanks for your help!
Ciao
Uwe


From sholden at holdenweb.com  Wed Jun 26 09:53:32 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 26 Jun 2002 09:53:32 -0400
Subject: enlightenment (Re: python cgi + IIS)
References: <3D0F1813.3010304@gmx.de> <x0NP8.165989$Fp1.91529@atlpnn01.usenetserver.com> <3D102AF2.5010204@gmx.de> <u4rfyc4gj.fsf@online.no> <3D13522E.6010703@gmx.de>
Message-ID: <UGjS8.43504$mF.5186@atlpnn01.usenetserver.com>

"Axel Grune" <NineOfSix at gmx.de> wrote in message
news:3D13522E.6010703 at gmx.de...
> > for, and so forth) give an account of all the steps you have performed
>
> This steps: http://www.e-coli.net/pyiis.html
>
> This doc tells you to change the 'App Mapping' in the properties of
> [IISAdmin] to enable the execution of python files for the entire server
> but for any reason it doesn't work.
> In the end our admin suggested to change the properties of 'Default Web
> Site' (the node above [IISAdmin]) and then it works.
>

In fact the document we are referring to says "Right click on the place you
want the associations to apply to", by which the author means "Right-click
on the server or one of its real or virtual subdirectory icons".

Anyway, glad you got to where you needed to be.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From timr at probo.com  Sun Jun  2 22:50:26 2002
From: timr at probo.com (Tim Roberts)
Date: Sun, 02 Jun 2002 19:50:26 -0700
Subject: HTML and python
References: <ad94md$viphf$1@ID-49758.news.dfncis.de> <gstK8.148299$%u2.116279@atlpnn01.usenetserver.com> <Pine.SOL.4.30.0206022022220.23311-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <t8mlfu4apvov308874l3ummgepuji7rjn1@4ax.com>

"John J. Lee" <jjl at pobox.com> wrote:

>On Sun, 2 Jun 2002, Steve Holden wrote:
>> "Ken" <ken at hotmail.com> wrote:
>>
>> > How do you call to a static HTML page from the script?
>> >
>> > <FORM METHOD="POST" ACTION="frame2a.html" target=_parent>
>> > This doesn't seem to be working..... Can someone help?
>>
>> It really makes no sense to use method POST to access a static page. Using
>> method GET will probably owrk, though you will see the for contents in the
>> URL.
>
>Why doesn't it make sense?
>
>(following from memory, may be wrong)
>
>from urllib2 import urlopen
>from urllib import urlencode
>result = urlopen(url, urlencode(data))
>
>Sorry if this is no help -- I didn't see the original message.

How could you not see it?  It was part of your reply!  See the section that
starts "How do you call..."?  That is the original message.

What you posted is not a static HTML page.  It is a Python CGI script.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From andrewm at object-craft.com.au  Thu Jun 20 00:40:38 2002
From: andrewm at object-craft.com.au (Andrew McNamara)
Date: Thu, 20 Jun 2002 14:40:38 +1000
Subject: Variable Interpolation - status of PEP 215 
In-Reply-To: Your message of "Thu, 20 Jun 2002 00:09:46 -0400."
             <3D11558A.14451441@engcorp.com> 
Message-ID: <20020620044038.9416438F5A@coffee.object-craft.com.au>

>> I am a relative newbie to Python. I have a reasonably good perl background
>> (don't worry I am not about to start a language flame war :-), and one of
>> the things I really miss is variable interpolation in strings. I searched
>> the net and found PEP 215 (http://www.python.org/peps/pep-0215.html). This
>> seems to be exactly what I need. The proposal seems to be quite old, so I'd
>> like to know if it's still in consideration or has it been dropped
>> completely?
[...]
>>>> foo = "trick"
>>>> bar = 95.5
>>>> print "This %(foo)s is loved by %(bar)s%% of programmers." % locals()
>This trick is loved by 95.5% of programmers.
>
>You can use any dictionary, but locals() or globals() can be handy
>in many cases.

It took me a while to come to love the % operator - it really is a
neat idea. It means that you don't have to worry about escaping the
interpolation characters unless you explicitly want interpolation. In
perl, I seemed to be forever forgetting to escape them, producing obscure
bugs as a result. And it doesn't require any special rules in the grammer - 
it's just another operator.

And in C, the problem is even worse (although it's not a fair comparison) - 
you have to allocate the memory first, worry about bounds checking,
call sprintf (really snprintf), then worry about releasing the memory.

-- 
Andrew McNamara, Senior Developer, Object Craft
http://www.object-craft.com.au/




From pixie888 at hotmail.com  Wed Jun 19 09:23:15 2002
From: pixie888 at hotmail.com (pixie888 at hotmail.com)
Date: Wed, 19 Jun 2002 13:23:15 GMT
Subject: SafeArray problem
References: <3d105710.83643750@news.skynet.be> <3d108014$1@news.mhogaming.com>
Message-ID: <3d108545.95472593@news.skynet.be>

On Wed, 19 Jun 2002 08:55:31 -0400, "Michael Weiss"
<FooWeissBarMike at hotmail.com> wrote:

>How is the safearray parameter defined in the IDL on this method? [in, out]
>[in] [out, retval], etc...?

I have not written an idl-file myself, I just used the classwizard in
Visual Studio to write my method on the automation object (using MFC).
What do you really want to know? (I looked for idl-files in my project
but did not find any)

><pixie888 at hotmail.com> wrote in message
>news:3d105710.83643750 at news.skynet.be...
>> Hi all,
>>
>> I am using an Automation object which has a method which takes a
>> SafeArray as an argument. In my Python module I call this method and
>> give a list of longs as an argument.
>>
>> While debugging (using the VC++6 IDE) I notice that the type of the
>> argument (upon entering my automation object) is VT_ERROR (argument
>> not found, actually). Now, what is really odd is the following: I
>> extended my method: it takes two arguments, the first being a
>> SafeArray and the second a long (which will not be used). When I call
>> this method again from my Python module (with a list of longs and a
>> dummy long as arguments) then everything works fine.
>>
>> Beats me, really.
>>
>> Any idea where the problem lies?
>>
>> Thanks,
>>
>> Henk
>
>



From prema at prema.co.nz  Sat Jun 15 02:49:32 2002
From: prema at prema.co.nz (Prema)
Date: 14 Jun 2002 23:49:32 -0700
Subject: MS Word -- finding text
Message-ID: <ae58a20e.0206142249.54dfa746@posting.google.com>

Hi All !
Just wondering if anyone has some sample code around how to open a
document, find a string and return the paragraph(s) containing the
string.

I have got a good part of the way but running up against Unicode
decoding

Can post some snippets if needed
Thanks very much in advance

Kind regards 
Mike


From missive at frontiernet.net  Sun Jun 23 05:49:03 2002
From: missive at frontiernet.net (Lee Harr)
Date: Sun, 23 Jun 2002 09:49:03 -0000
Subject: DerivVar instance has no attribute '__float__'
Message-ID: <uhb6cfbuvmmr45@corp.supernews.com>

I just installed the Scientific module and am trying it out:

>>> from Scientific.Functions.Derivatives import DerivVar
>>> from math import sin
>>> print sin(DerivVar(2))
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
AttributeError: DerivVar instance has no attribute '__float__'
>>> 


I googled for DerivVar float and found a message from one
person last year with the same problem, but no posted solution.

I am using:
Python 2.2
FreeBSD 4.5
py22-numeric-20.3
py22-scientific-2.2

Anyone have a hint?



From peter at engcorp.com  Sun Jun 23 16:53:36 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 23 Jun 2002 16:53:36 -0400
Subject: Redirecting StdOut
References: <1024856420.145085@seux119> <Xns9236D1FF016AAcliechtigmxnet@62.2.16.82> <1024860416.580309@seux119>
Message-ID: <3D163550.E525D840@engcorp.com>

"Ugo Garc?a" wrote:
> 
> perfect. This is what I was looking for. Now I'd make a C module to handle
> the messages written to stdout & stderr in order to show them in the screen.
> Thanks a lot

Why would you bother with C for a game that relies on simple print
statements for its interface?  Can't you implement the game entirely
in Python?

-Peter


From mcherm at destiny.com  Wed Jun 19 14:23:02 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Wed, 19 Jun 2002 14:23:02 -0400
Subject: Pop a list from beginning ? and memory saving...
Message-ID: <3D10CC06.3010700@destiny.com>

Holger Krekel wrote:
>> [Tim Peters]
>>> [James T. Dennis]
>>>> ...
>>>>  So, the critical question is:
>>>> 	Is the list's reverse() method done in place
>>> Yes.
>>>> and does it run in linear time
>>> Yes.
>>>> (or better?
>>> No.  It does len(list)/2 swaps.
> 	O(n/2) is linear, but still slightly better than 
> 	O(n)
> 
>> in another thread we had some discussion about a
>> reverse-iterator.  You could write
>>     for item in iters.reverse(list):
>>         # item iterates over list's items backwards
>>         # but the list itself stays unmodified. 
>> Do you agree that this would be a good idea?
>> please-share-your-brain-ly y'rs, holger
> 


James T Dennis wrote:
> It might be better to have a hack to the implementation of the
>  list primitives --- so they defer the shift of remaining elements
>  until GC.  The head of the list could point at the first none 
>  deleted item.  Then pop(0) could be implemented in O(1) time and
>  the garbage collector would eventually come along to do it's 
>  job.


I'm not sure that makes sense in Python. Most garbage is NOT collected 
by a separate GC thread (as in Jython and Java) or in bursts of GC 
activity. In cPython, nearly everything is garbage collected via 
reference counting... the object is GC'ed when the last reference to it 
is released.

As for the original question (is it a good idea), I'd say YES, but with 
one caveat. I think that it is important that the "iterable" utilities 
operate on ANY iterator ... and that isn't possible with the intelligent 
implementation of reverse. So I think that if iterable.reverse() were to 
check the type of it's argument and use the clever approach (counting 
backwards) for types list and tuple, but fall back on creating a temp 
list and reversing it for other types, then that'd be quite useful. 
Optimize the common case, but don't lose the universality.


-- Michael Chermside











From sheershion at mailexpire.com  Wed Jun 12 16:21:11 2002
From: sheershion at mailexpire.com (Robert Amesz)
Date: 12 Jun 2002 20:21:11 GMT
Subject: newbie question - python blocking on int()?
References: <4LaN8.2488$Vr2.586019@news20.bellglobal.com>
Message-ID: <Xns922BDDB6BCBC2rcamesz@amesz.demon.nl>

Michael Davis wrote:

> I'm writing a specialized ftp client. I'm parsing the output of
> ftp.dir, which gives me a string like this (call it str):

There already is a Python extension module for parsing that type of
data. I found it on:

http://freshmeat.net/redir/ftpparsemodule/20709/url_tgz/ftpparsemodule-0.93.tar.gz

I haven't used it yet, so I can't comment on it.


Robert Amesz


From FooWeissBarMike at hotmail.com  Wed Jun  5 08:48:12 2002
From: FooWeissBarMike at hotmail.com (Michael Weiss)
Date: Wed, 5 Jun 2002 08:48:12 -0400
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet>
Message-ID: <3cfe0a36@news.mhogaming.com>

While not exactly compiled to a native exe, there are other options:

Look at this: (Py2Exe)
http://starship.python.net/crew/theller/py2exe/

I understand there are other options similar to py2exe that I've never
tried... like "freeze" or the stuff over at
http://www.mcmillan-inc.com/install1.html



"TuxTrax" <TuxTrax at fortress.tuxnet.net> wrote in message
news:slrnafrqit.lbm.TuxTrax at fortress.tuxnet...
> Hello all
>
> I am new to this forum, and to Python, please excuse the newbie question.
>
> I have started reading the o'reilly book, "learning python" and find
python
> to be an amazing language so far. My first impression of it is a language
that
> is easier than basic to learn and use, and more powerful than many other
> high level languages such as pascal.
>
> My question is this; I realize that Python compiles to bytecode
automatically
> as part of the interpreters tokenization run, but is there a way to
permanently
> compile python code outputting to an executable binary?
>
> Cheers,
>
> Mathew
>
> --
> TuxTrax   (n.) An invincible, all terrain, Linux driven armored assault
> vehicle that can take as much fire as it gives ;-)
>
> <Carnivore activation string>
> ASSASINATION ANTHRAX PRESIDENT NUCLEAR TALIBAN AMMONIUM NITRATE
> </Carnivore activation string>
>
> Yes, I am a Penguin cult high priest. Fin readings upon request.
> ROT13 this email address to mail me: uvtuqrfregzna at lnubb.pbz
>




From maxm at mxm.dk  Thu Jun 20 05:36:29 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 20 Jun 2002 11:36:29 +0200
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com> <lkznxqjqx2.fsf@pc150.maths.bris.ac.uk>
Message-ID: <3D11A21D.9010809@mxm.dk>

Michael Hudson wrote:
 > "John Roth" <johnroth at ameritech.net> writes:
 >
 >
 >>There's no equivalent of a goto command in Python, for good reason,
 >>which I won't repeat here. Prof. Djikstra did it very well in 1974,
 >>in his letter "Goto Considered Harmful."
 >
 >
 > There are also technical difficulties implementing it (the
 > blockstack), or I'd have probably done goto functionality in
 > bytecodehacks :)


What do you mean???


Goto's are simple ;-)


def goto(lable):
     lable()


def start():

     print 'starting'
     goto(lable2)

def lable1():

     print 'at lable 1'
     goto(stop)

def lable2():

     print 'at lable 2'
     goto(lable1)

def stop():
     print 'stopping'

goto(start)

 >>> starting
 >>> at lable 2
 >>> at lable 1
 >>> stopping



From candiazoo at attbi.com  Thu Jun 27 13:34:26 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Thu, 27 Jun 2002 17:34:26 GMT
Subject: Here is the result of my profile...
References: <3d1a6993.248534625@netnews.attbi.com>
Message-ID: <3d1b4c89.306637140@netnews.attbi.com>

         Completed at Thu Jun 27 10:50:13 2002

         11623 function calls in 152.079 CPU seconds

   Ordered by: standard name

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
        1    0.001    0.001  151.962  151.962 <string>:1(?)
      429    0.140    0.000    0.279    0.001 catalog_model.py:109(__init__)
      429    0.148    0.000    0.148    0.000 catalog_model.py:114(set_base_fee)

      429    0.014    0.000    0.014    0.000 catalog_model.py:118(set_page_fee)

      429    0.015    0.000    0.015    0.000 catalog_model.py:121(to_string)
     1703    0.104    0.000    0.104    0.000 catalog_model.py:509(fix_date_rang
e)
      848    0.157    0.000    0.157    0.000 catalog_model.py:57(__init__)
        1  150.396  150.396  151.961  151.961 catalog_model.py:573(generate_cata
log)
      848    0.027    0.000    0.027    0.000 catalog_model.py:66(set_wrk_inst)
      848    0.020    0.000    0.020    0.000 catalog_model.py:69(set_title)
      848    0.017    0.000    0.017    0.000 catalog_model.py:72(set_pub_name)
      848    0.017    0.000    0.017    0.000 catalog_model.py:75(set_std_num)
      848    0.016    0.000    0.016    0.000 catalog_model.py:78(set_date_range
)
      848    0.137    0.000    0.137    0.000 catalog_model.py:81(set_source_id)

       33    0.000    0.000    0.000    0.000 catalog_model.py:820(test_refresh)

       36    0.353    0.010    0.353    0.010 catalog_model.py:823(test_echo)
        1    0.000    0.000    0.000    0.000 catalog_model.py:826(test_cancel)
      419    0.015    0.000    0.033    0.000 catalog_model.py:89(__init__)
      419    0.136    0.000    0.136    0.000 catalog_model.py:92(to_string)
        1    0.000    0.000    0.000    0.000 connections.py:85(cursor)
        1    0.011    0.011    0.189    0.189 cursors.py:108(__do_query)
        1    0.022    0.022    0.022    0.022 cursors.py:135(_fetch_row)
        1    0.000    0.000    0.000    0.000 cursors.py:145(_check_for_warnings
)
        1    0.178    0.178    0.178    0.178 cursors.py:160(_get_result)
        2    0.002    0.001    0.002    0.001 cursors.py:162(close)
        1    0.000    0.000    0.212    0.212 cursors.py:167(_query)
      671    0.027    0.000    0.036    0.000 cursors.py:174(fetchone)
        1    0.000    0.000    0.000    0.000 cursors.py:21(__init__)
        1    0.000    0.000    0.000    0.000 cursors.py:28(__del__)
        2    0.000    0.000    0.000    0.000 cursors.py:31(close)
      671    0.009    0.000    0.009    0.000 cursors.py:36(_check_executed)
        2    0.000    0.000    0.000    0.000 cursors.py:46(_get_db)
        1    0.000    0.000    0.212    0.212 cursors.py:51(execute)
        0    0.000             0.000          profile:0(profiler)
        1    0.117    0.117  152.079  152.079 profile:0(test.generate_catalog())


From mfranklin1 at gatwick.westerngeco.slb.com  Tue Jun 25 10:24:19 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Tue, 25 Jun 2002 14:24:19 +0000
Subject: Embedding in Tkinter (reprise)
In-Reply-To: <1907a5e4.0206250504.4b417e9a@posting.google.com>
References: <1907a5e4.0206240803.6ca0a4c@posting.google.com> <mailman.1024936355.12699.python-list@python.org> <1907a5e4.0206250504.4b417e9a@posting.google.com>
Message-ID: <200206251327.g5PDRwM28367@helios.gatwick.geco-prakla.slb.com>

On Tuesday 25 Jun 2002 1:04 pm, Pier Paolo Glave wrote:
> I tried to follow the hint by Jeff, but it seems that the TkSteal
> package hasn't been updated for a long time: the last available
> version is for Tk 4.x.
>
> Anyway, I made some further investigation on Tcl/Tk, and I found out
> that there exists an extension package to Tk 8.x, named BLT, which has
> a "container" object that can fit for my needs.
> Now, my problem is that I'm not an expert in Python/Tk integration,
> and I don't know how to wrap this into a Python class.
>
> Does anyone have any suggestions?

Pmw (Python Mega Widgets) has an interface to BLT but I havn't used it so I 
don't know for sure that you can use this.....

try:  http://pmw.sourceforge.net/

for more info.

Martin





From whisper at oz.net  Fri Jun 14 04:33:30 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 14 Jun 2002 01:33:30 -0700
Subject: Zip makes sense! was: Re: why not "'in' in 'in'"?
In-Reply-To: <332c7c11.0206132333.199ba913@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEGEBHDJAA.whisper@oz.net>

Haskell - that's a mountain range in New York state? :p)

David LeBlanc
Seattle, WA USA 

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Benno
> Sent: Friday, June 14, 2002 0:33
> To: python-list at python.org
> Subject: Zip makes sense! was: Re: why not "'in' in 'in'"?
> 
> 
> "David LeBlanc" <whisper at oz.net> wrote in message 
> news:<mailman.1024021164.12598.python-list at python.org>...
> > > -----Original Message-----
> > > From: python-list-admin at python.org
> > > [mailto:python-list-admin at python.org]On Behalf Of Tim Peters
> > > Sent: Thursday, June 13, 2002 18:57
> > > To: python-list at python.org
> > > Cc: Grant Griffin
> > > Subject: RE: why not "'in' in 'in'"?
> > >
> 
> <snip discusion about 'in'>
>  
> > Hmmm... How about calling it "contains" instead of "in"? "if 
> 'feefifofum'
> > contains 'fi'" seems nicer then "if 'fi' in 'feefifofum'" maybe?
> > 
> > Of course maybe this makes too much sense for a language where 
> zip has zip
> > to do with compression and yield nothing to do with threads.
> 
> 
> I'm guessing you are mentioning this fairly tounge in cheek, 
> however zip makes
> a lot of sense to anyone coming from Haskell. If I was after the 
> zip compression
> algorithm i would look in the hypothetical compression package.
> 
> it-made-perfect-sense-to-me-ing-ly yours Benno
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From emile at fenx.com  Sun Jun  2 10:27:49 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 02 Jun 2002 14:27:49 GMT
Subject: How to delete file?
References: <add9ai$107clv$1@ID-49758.news.dfncis.de>
Message-ID: <FXpK8.90000$352.4832@sccrnsc02>

"Ken" <ken at hotmail.com> wrote in message
news:add9ai$107clv$1 at ID-49758.news.dfncis.de...
> How do you delete file?
>

import os
os.unlink(path)
os.remove(path)

> If you write to file from beginning position of an existing file, do
you
> overwrite the existing data or insert from it?
>


--or-- to overwrite a file (or create if not already there) just open
it:

>>> print open.__doc__
open(filename[, mode[, buffering]]) -> file object

Open a file.  The mode can be 'r', 'w' or 'a' for reading (default),
writing or appending.  The file will be created if it doesn't exist
when opened for writing or appending; it will be truncated when
opened for writing.  Add a 'b' to the mode for binary files.
Add a '+' to the mode to allow simultaneous reading and writing.
If the buffering argument is given, 0 means unbuffered, 1 means line
buffered, and larger numbers specify the buffer size.



HTH,

--

Emile van Sebille
emile at fenx.com

---------



From paul at fxtech.com  Thu Jun 27 10:49:38 2002
From: paul at fxtech.com (Paul Miller)
Date: Thu, 27 Jun 2002 14:49:38 GMT
Subject: connecting embedded Python audio samples with NumPy?
Message-ID: <3D1B2603.26EA9A9C@fxtech.com>

I have a C app that reads audio data, and I'd like to pass this audio
data in array format to Python functions so I can use it with NumPy.

I have this working right now - generating a float array from the sample
stream. But I want to put this array into a form usable by the FFT
functions. I've read over this paper:

	http://www.onlamp.com/pub/a/python/2001/01/31/numerically.html

Which indicates I need to put my 1D sample stream into a 2D array. Can
anyone elaborate more on this? I'm wondering if I should provide the
array pre-made for FFTs from my C code, or I should build the 2D array
from the 1D array I have in Python.


From ponderor at lycos.com  Fri Jun 14 03:20:30 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 14 Jun 2002 00:20:30 -0700
Subject: Python 2.2.1
References: <II7O8.622$e27.275217555@twister1.starband.net>
Message-ID: <e81be8b2.0206132320.56e3ca56@posting.google.com>

>     How do I set the path and what should the name of the path be? All 
> help and information would be appreciated. Thanks.:)
> --

Windows 9X -> Millenium get increasingly harder to setup environment
details.

a. Autoexec.bat MAY work for you.

b. Typign MSCONFIG from your RUN menu (Start~Run, or Windows key+R)
will bring you to an environment configuratin screen.

c. These don't necesarily apply to the "DOS Mode" settings, which may
require you setting up another autoexec.bat.

d. The suggestions for using PythonWin are most appropriate.  

After rebuilding (for me this is re-installing OS+Formatting, not
simply reinstalling.) my PC and not installing it, I missed PythonWIN.
(But for a personal perspective training reason I want to use the
default tools, plus run PythonCard with fewer "environment" variables.
(Did you see the plug for PythonCard? If not here's a few more: 
Python-Tutor mailing list.  Useless Python web site. Google those.)) 
PythonWIN includes the doc's in Windows Style help. I find those a
handy in ways but prefer local HTML help files.  PyDoc is also handy(
http://www.onlamp.com/pub/a/python/2001/04/18/pydoc.html ).

How's that for vague cues?  I have Windows ME on my home laptop. Too
cheap to buy XP or take the time to install Linux and train myself and
wife.

Your question is most valid.  I searched the Windows help, google,
microsoft.com, and finally got the answer from a friend trying to find
why my autoexec.bat changes didn't take affect, and the diamond in the
rough: MSCONFIG.

-Dean


From peter at engcorp.com  Tue Jun 25 00:02:24 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 25 Jun 2002 00:02:24 -0400
Subject: Suggestions for good programming practices?
References: <mailman.1024971129.7460.python-list@python.org>
Message-ID: <3D17EB50.E5A6781D@engcorp.com>

David LeBlanc wrote:
> 
> Somebody seems to be working overtime to miss the point here...

Don't be rude please.  Your offline message pointed out that you
_like_ this behaviour, and so do I, but your original message
appeared to indicate you did not.  I was trying to get at
why you didn't like it, not to be pedantic or anything.

David LeBlanc originally wrote:
> The one that always gets me is "lat,lon,timezone = 
> getLocation("seattle")".  I.E. the ability to return multiple 
> distinct values as opposed to returning
> multiple values in the form of a struct as in C or C++.

Sorry for the confusion...

-Peter


From talon34 at hotmail.com  Wed Jun 26 23:36:08 2002
From: talon34 at hotmail.com (Talon)
Date: Thu, 27 Jun 2002 03:36:08 GMT
Subject: Installation problems DCOracle2 & W2000
References: <mailman.1025115815.15029.python-list@python.org>
Message-ID: <talon34-674734.20343426062002@netnews.attbi.com>

In article <mailman.1025115815.15029.python-list at python.org>,
 brueckd at tbye.com wrote:

> On Wed, 26 Jun 2002, Talon wrote:
> 
> > OK, here you go:
> > 
> > Path = C:\Python21\;  (I don't have a specific PYTHONPATH variable set. 
> > This is in the general PATH variable)
> > 
> > The error I get is
> > 
> > File "C:\Python21\myfiles\htsa_data_loader_v2.py", line 9, in ?
> >   import os,sys,re,string,DCOracle2
> > File "C:\Python21\DCOracle2\__init__.py", line 91, in ?
> >   import DA
> > File "C:\Python21\DCOracle2\DA.py", line 90, in ?
> >   from db import DB
> > File "C:\Python21\DCOracle2\db.py", line 89, in ?
> >   import DCOracle2, DateTime
> > ImportError: No module named DateTime
> > 
> > DateTime is located in C:\Program Files\ZopeDev\lib\python\DateTime
> 
> Aha! Here's what I think the problem is: the DCOracle2 bundle off the Zope 
> site comes all wrapped up in the Zope database adapter (ZOracleDA), so 
> your above script is actually trying to import the DA and not the 
> DCOracle2 package itself (I can see why this is confusing - both 
> directories are named DCOracle2 but one is a subdirectory of the other!).
> 
> So... if you look in c:\python21\DCOracle2 you should find a subdirectory 
> in there also called DCOracle2. This is the only one you should need for 
> doing standalone (none-Zope) work with Oracle. The easiest thing to do is 
> rename c:\python21\DCOracle2 to DCOracle2bak or something, and then move 
> its DCOracle2 subdirectory into c:\python21 and then try your script 
> again. 
> 
> Does this make sense? Please let me know if it doesn't or if it still 
> doesn't work after that.
> 
> -Dave
> 
> 
> 

Dave,
That was dead on.  I did as you said and now it works perfectly.  I 
thought I had done this originally, but I guess I had moved the outer 
folder instead.  Thank you very much.  Thanks to Fazal also for his 
input.  It is much appreciated.

Mark


From recmorg at charter.net  Mon Jun 10 19:44:14 2002
From: recmorg at charter.net (TuxedoKamen)
Date: Mon, 10 Jun 2002 23:44:14 -0000
Subject: Making an About Box...
Message-ID: <Xns9229BE9B416A2endymion84charternet@216.168.3.40>

Hello.  I'd like to make an About Box using Tkinter.  I'm using a grid 
inside a Toplevel widget, but I don't really like this solution.  I'd like 
a window that did not have the minimize and maximize buttons, like the 
tkMessageBox derived boxes showinfo, showwarning, etc.  Is it possible to 
make a window that has this type of title bar and put my own widgets in it?


From peter at engcorp.com  Wed Jun 19 22:03:14 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 19 Jun 2002 22:03:14 -0400
Subject: Unicode hexadecimal characters
References: <3eeda89d.0206191615.71b0629d@posting.google.com> <3D112C60.9C3D2F9F@engcorp.com> <njaQ8.101599$6m5.87417@rwcrnsc51.ops.asp.att.net>
Message-ID: <3D1137E2.7CB8F814@engcorp.com>

Thinkit wrote:
> 
> "Peter Hansen" <peter at engcorp.com> wrote in message
> news:3D112C60.9C3D2F9F at engcorp.com...
> > Thinkit wrote:
> > >
> > > It's a bit OT but does Unicode support the digits ten through fifteen
> > > (for hexadecimal)?
> >
> > Do you mean "does Unicode have unique glyphs representing the
> > hexadecimal digits ten through fifteen other than the standard
> > A, B, C... ?"
> 
> Yes, that's what I mean.  I looked at the spec and saw a lot of different
> numerals, including 1-20 for circled numbers.  Will have to examine it
> again.

Ah, those are just for typographical use, as in for glossy brochures
and such.  Not intended to be used to represent digits in any
mathematical or computer sense I believe.  It's unlikely there is any 
aspect of Unicode which covers "pretty" representations of hexadecimal
digits...

-Peter


From mwh at python.net  Mon Jun 10 05:52:41 2002
From: mwh at python.net (Michael Hudson)
Date: Mon, 10 Jun 2002 09:52:41 GMT
Subject: PyCObject subtype PyObject
References: <3D022727.2147C50E@doc.ic.ac.uk>
Message-ID: <lk1ybflaua.fsf@pc150.maths.bris.ac.uk>

Benjamin Tai <bt98 at doc.ic.ac.uk> writes:

> Hi,
> 
> 
> Any comments to clarify my confusion would be appreciated.
> 
> In Python/C API Reference Manual, section 7.5.9 CObjects,  the document
> mentions that:
> 
> "PyCObject This subtype of PyObject represents an opaque value, ..."
> 
> 
> 
> 1) What does it mean by "subtype"? As a C interface, is the document
> talking about two structs which share some common fields?

Yes.

> 2) For function "PyCObject_FromVoidPtr()", it takes in a "void*" as the
> first argument.
> For function "PyCObject_AsVoidPtr()", it returns a "void*".
> When implementing Python extension, what is the potential danger of
> casting pointers, of different size, to and from "void*"?

Doesn't C guarantee that you can cast a pointer of type (T*) to
(void*) and back again for all T?  I was fairly sure it did.

Python is implemented in C not C++, you know...

> 3) What if I can always determine the original type of the pointer?

Huh?

> 4) Apart from exporting function between modules, are there any
> examples of using PyCObject?

Not sure what you mean here either...

Cheers,
M.

-- 
  The PROPER way to handle HTML postings is to cancel the article,
  then hire a hitman to kill the poster, his wife and kids, and fuck
  his dog and smash his computer into little bits. Anything more is
  just extremism.                                 -- Paul Tomblin, asr


From donn at drizzle.com  Wed Jun  5 02:04:34 2002
From: donn at drizzle.com (Donn Cave)
Date: Wed, 05 Jun 2002 06:04:34 -0000
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <3CFCF3E7.AE891EBE@noaa.gov> <adiumj$22g2$1@nntp6.u.washington.edu> <adjduf$92g$0@216.39.172.122> <slrnafqp12.4ea.huaiyu@gauss.almadan.ibm.com>
Message-ID: <1023257072.657314@yasure>

Quoth huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu):
...
| That is one proposal.  My proposal is to disallow += for immutables.
| Another symbol would be used for rebinding operations.  (Or maybe swap the
| role of these symbols.)  Thus for mutables both operations would be
| available.

That suits me, except the mutable/immutable distinction.  That's
how we have been describing the distinction, but it's worth noting
that it isn't really what we mean.

For a silly example, suppose that in 2.3, None supports +=, kind of
like the UNIX /dev/null.  This would not make it mutable in any
reasonable sense of the word, but it's in the "in place" category.
A class instance that implements __iadd__ to return something other
than self, would not be, though it's certainly mutable.

	Donn Cave, donn at drizzle.com


From eric.brunel at pragmadev.com  Mon Jun  3 10:56:57 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 3 Jun 2002 14:56:57 +0000
Subject: Interrupting a continuous script
References: <_6JK8.1535$376.74360@newsfep1-win.server.ntli.net>
Message-ID: <adfonq$flb$1@wanadoo.fr>

Julian Gollop wrote:
> I have a few programs continually processing text files on a Linux
> machine. Occasionally I need to stop them, but I don't want to use a
> keyboard interrupt because it may cause the script to stop in the middle
> of doing something.
> Does anyone know a simple way to allow me to stop the process by pressing
> a key on the keyboard - without stopping inside something crucial?

I'd use an interrupt handler on the signal SIGINT, which is sent to your 
process when you press Ctrl-C. It may sound complicated, but is in fact 
quite straightforward (as often with Python :-):

---------------------------
import sys, os, signal, time

interrupted = 0

def ctrlCHandler(*whatever):
  global interrupted
  interrupted = 1
  print "Interrupt caught. Please wait..."

signal.signal(signal.SIGINT, ctrlCHandler)
print "Something really important start"
for i in range(50): time.sleep(0.1)
print "Something really important end"
if interrupted: sys.exit()
print "Something really important start"
for i in range(50): time.sleep(0.1)
print "Something really important end"
---------------------------

Try to run this script and press Ctrl-C while it runs. You'll see it stops 
only when the "important" things are over. You'll just have to put a few 
lines like:
if interrupted: sys.exit()
at strategic places in your code, i.e. between important things.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From egbert at bork.demon.nl  Sat Jun 22 13:57:15 2002
From: egbert at bork.demon.nl (Egbert Bouwman)
Date: Sat, 22 Jun 2002 19:57:15 +0200
Subject: Tkinter & Python: how to get the return of a call associated to an event on my gui
In-Reply-To: <3D12A2E4.6A566789@noos.fr>; from aurelienviolet@noos.fr on Fri, Jun 21, 2002 at 05:52:04AM +0200
References: <3D12A2E4.6A566789@noos.fr>
Message-ID: <20020622195714.A689@bork.demon.nl>

A callback may do something, but it does not return anything,
at least not something you may use. I suppose that you need 'global',
ie an export of a name from within a function (the local namespace)
to the surrounding global one: the namespace of the module in which 
the function is defined. I think the line 'global result' in your
callback will help.

As long as you stay within one class-instance, you can solve it in 
another way: let your callback produce a 'self.result', which is 
available in the whole class-instance.

Things become interesting when you have two classes, A and B,
and A needs the data that will be collected by a gui in B.
I call it interesting because I am still struggling with it.

A and B should agree on the layout of the data structure that goes
from B to A, but how B collects these data, from which entries or
canvases, is entirely his own business. In the same way: B knows 
nothing of the way A handles these data.

First the solution, and then some explanations.
    
    class B:
        def __init__(self,callback=None):
            # build a gui ...
            bouton = Button( ....., command=callback)
            bouton.grid( ...)
           
        def retour(self):
            # collect the gui-data in retour_list[],
            # and probably close this b-gui.
            return retour_list
        
    class A:
        def __init__(self):
            b = B( ..., callback=self.sign)
            
        def sign(self,event=None):    
            self.data_list = b.retour()
            process_data()
            
        def process_data(self)
            # do something usefull with self.data_list
            # exit (button) the application
            
Within the usual Tkinter loop you start this with:
    a = A()

A passes in an argument the name of a callback to B.
This callback ('sign') is a method of A.
The chain of activities is started by the button-push in B
that activates the callback in A.
The callback sends a message ('retour') to B to collect 
and return the data, and A processes them.

How B collects the data is hidden in 'retour', and what A
does with them is hidden in 'process_data'.

Now I have two questions:

am I making things much too complicated, or
does some standard pattern solve this problem in a better way ?

egbert
-- 
Egbert Bouwman - Keizersgracht 197 II - 1016 DS  Amsterdam - 020 6257991
========================================================================




From K.Rdt at TU-Berlin.DE  Sun Jun 23 15:08:14 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Sun, 23 Jun 2002 21:08:14 +0200
Subject: mail-problem over t-online 
Message-ID: <SOFDA6ROQNHBPKBAXVHFSQFEGC96YMH.3d161c9e@lap>

---------------------------------------------------------------------
Hi

Testing python and email I adapted this script:
	
	host=smtplib.SMTP('mailszrz.zrz.tu-berlin.de')
	host.sendmail(zwi_eck(From),zwi_eck(Tp),al)
	host.quit

This is working, when I'm directly connected to my
old ISP TU-Berlin via ISDN. But the same is failing,
when I'm connected at t-online via ADSL, with
authentification required .. Providing my log and pass
produces errors, too.

Pop3 is working in both cases.

Who can give me a hint?
Tia

		K at Rdt
---------------------------------------------------------------------







From cliechti at gmx.net  Sat Jun 29 16:44:26 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 29 Jun 2002 22:44:26 +0200
Subject: can't import generators
References: <m2hejlc060.fsf@mother.paradise.lost> <afl5je$4iu$1@panix1.panix.com>
Message-ID: <Xns923CE7C804E6Fcliechtigmxnet@62.2.16.82>

aahz at pythoncraft.com (Aahz) wrote in news:afl5je$4iu$1 at panix1.panix.com:

> In article <m2hejlc060.fsf at mother.paradise.lost>,
> John Hunter  <jdhunter at nitace.bsd.uchicago.edu> wrote:
>>
>>I am having trouble working with generators.  Anyone know what might
>>cause this behavior:
>>
>>Python 2.2.1 (#1, Apr 22 2002, 21:20:54) [GCC 3.0.4] on linux2
>>Type "help", "copyright", "credits" or "license" for more information.
>>>>> from __future__ import generators
>>Traceback (most recent call last):
>>  File "<stdin>", line 1, in ?
>>ImportError: cannot import name generators
> 
> Can't do this in interactive mode.  Use a script.

?!?!? since when?

$ python
Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> from __future__ import generators
>>> def g(x):
...     for p in x: yield p
...
>>> [x for x in g(range(3))]
[0, 1, 2]
>>>

definetly works....

chris
-- 
Chris <cliechti at gmx.net>



From sholden at holdenweb.com  Tue Jun  4 08:47:04 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 4 Jun 2002 08:47:04 -0400
Subject: Printing a text centered, it would be a newbie prob
References: <c7bfc769.0206040012.20aa0a10@posting.google.com>
Message-ID: <2J2L8.225207$%u2.29779@atlpnn01.usenetserver.com>

"Backflip" <Backflip at gmx.de> wrote in message
news:c7bfc769.0206040012.20aa0a10 at posting.google.com...
> Hello World (=
> I?ve got a little problem, i want to put out an import string centert. For
example:
> for word in raw_input("bitte hier ihren text eingeben:").split():
>    print word
>
> text=dies ist ein test
>
> now the output looks like this
>
> dies
> ist
> ein
> test
>
> but i want that it looks like this one
>
> dies ist
> ein test
>
> that both rows are simular to each other, could smb help me?


>From your description, are we to assume that you only want two rows of
output? Is the "similarity" referring to the number of words or the number
of characters? Since the former is easier, would this answer your problem:


>>> w = "this is an example".split()
>>> wl = len(w) // 2
>>> l1 = w[:wl]
>>> l2 = w[wl:]
>>> print "%s\n%s" % (" ".join(l1), " ".join(l2))
this is
an example
>>>

I suspect there will be more that you haven't yet told us, however: the
specification was not rigorous.

regards
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From neal at metaslash.com  Thu Jun  6 16:50:28 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Thu, 06 Jun 2002 16:50:28 -0400
Subject: ANN: PyChecker v0.8.11
Message-ID: <3CFFCB14.22BB35E6@metaslash.com>

A new version of PyChecker is available for your hacking pleasure.

        PyChecker is a tool for finding bugs in Python source code.
        It finds problems that are typically caught by a compiler for less
        dynamic languages, like C and C++.  It is similar to lint.

Comments, criticisms, new ideas, and other feedback is welcome.

Changes from 0.8.10 to 0.8.11:

  * Improve error message for syntax errors from user files
  * Fix pychecker.bat so it should work now
  * Add a warning for using __coerce__ in new-style classes
  * Add --deprecated option for using deprecated modules or functions
  * Add a warning for using functions with security problems (os.t[e]mpnam)
  * Add a warning for comparing against True/False or defining True/False
  * Add --badexcept option to warn when using string exceptions or
        classes not derived from Exception to raise/catch exceptions
  * Fix spurious warnings from using (test and 'true' or 'false)

Various other bugs were also fixed, including some problems 
using PyChecker in Python 2.2.


PyChecker is available on Source Forge:
    Web page:           http://pychecker.sourceforge.net/
    Project page:       http://sourceforge.net/projects/pychecker/
    Mailling List:      pychecker-list at lists.sourceforge.net

Neal
--
pychecker-list at lists.sourceforge.net




From emile at fenx.com  Sat Jun 15 10:22:09 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 15 Jun 2002 14:22:09 GMT
Subject: fileupload error!
References: <mailman.1023972676.28575.python-list@python.org>
Message-ID: <l4IO8.81651$pw3.1985@sccrnsc03>

penguin
> Hi I try to upload a picture with a python cgi script. Tried many
versions of the code, this is the current.
> Here's the code and error!
>
> The error:
>
> Traceback (most recent call last): File "C:\cgi-bin\upload.py", line
179,
> in ? xmldokument = skjema['dok'].value File "C:\Python22\lib\cgi.py",
line
> 550, in __getitem__ raise KeyError, key KeyError: dok
>

I didn't read through your code, but the error you're getting means that
'doc' is not a key in cgi.FieldStorage().

skjema = cgi.FieldStorage()
xmldokument = skjema['dok'].value


There are many discussions and examples of cgi module usage at:

http://groups.google.com/groups?as_q=cgi.FieldStorage&as_ugroup=comp.lan
g.python


HTH,


--

Emile van Sebille
emile at fenx.com

---------



From jadestar at idiom.com  Thu Jun 27 18:53:23 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 27 Jun 2002 22:53:23 GMT
Subject: How to get rid the new line
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org> <3D1AD534.A74EC9F4@ipm.fhg.de> <3D1B1A76.563E5394@engcorp.com> <pan.2002.06.28.00.14.48.789371.16627@linuxmail.org>
Message-ID: <afg513$2426$1@news.idiom.com>

SiverFish <occeanlinux at linuxmail.org> wrote:

> On Thu, 27 Jun 2002 14:00:22 +0000, Peter Hansen wrote:

>> Markus von Ehr wrote:
>>> 
>>>         f = open(filename, 'r')
>>>         lines = f.readlines()
>>>         line1 = lines[0] # exemplarily for first line line1 =
>>>         line1[0:len(line1)-1]
>>> 
>>> or:
>>>         f = open(filename, 'r')
>>>         lines = f.readlines()
>>>         line1 = lines[0][0:len(line1)-1] # exemplarily for first line
>> 
>> I believe this is unsafe.  The final line may not be terminated with \n.
>> 
>> -Peter
> Any method like chomp in perl
> I just one to cut read only the line that not empty into the list

 Could try this:

	def chomp(line):
		if line[-1]=='\n':
			line=line[:-1]
		return line

 (or variations thereon, to satisfy other semantic requirements).




From peter at engcorp.com  Sat Jun 29 00:23:16 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 29 Jun 2002 00:23:16 -0400
Subject: I'm an idiot
References: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16> <mailman.1025316799.2919.python-list@python.org> <Xns923BD3E551D3Ddavidgriswold1netsca@204.127.68.17>
Message-ID: <3D1D3634.C89FCF95@engcorp.com>

David Griswold wrote:
> 
> I had no idea it could be this compact.  Thank you for your help!
> 
> David
> 
> Carel Fellinger <carel.fellinger at iae.nl> wrote in
> news:mailman.1025316799.2919.python-list at python.org:
> 
>  and write it like this:
> >
> >    g = open('c:/temp/temp1.txt', 'w')
> >    for line in open('c:/temp/temp.txt'):
> >        g.write(line.strip())

It can't!  At least, it's pretty ugly and not very Pythonic
code (says yet another self-proclaimed expert :-).

Also, it's going to write out all lines in the file as one
big line because it removes the \n at the end of each
but doesn't add any.  

Try some of the other, slightly more verbose, suggestions if
you want your code to be readable.

-Peter


From calves at coelce.com.br  Mon Jun 17 12:05:17 2002
From: calves at coelce.com.br (Alves, Carlos Alberto - Coelce)
Date: Mon, 17 Jun 2002 13:05:17 -0300
Subject: tab size
Message-ID: <29A97D00F387D411AC7900902770E14805853F72@lcoeexc01.coelce.net>

Hi all,
Consider the following code:
>>> for i in range(10):
	print "%d\t"%i,

	
0	1	2	3	4	5	6	7	8	9

>>> 

How can I change the default size of tab in formating string \t?!

Carlos Alberto
COELCE/DPRON-Departamento de Projetos e Obras Norte
Fone: 677- 2228
e-mail: calves at coelce.com.br
\|||/
(o o)
--ooo0-(_)-0ooo--

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020617/0f151b87/attachment.html>

From baf at texas.net  Sun Jun 30 11:49:16 2002
From: baf at texas.net (Ben Fairbank)
Date: Sun, 30 Jun 2002 15:49:16 GMT
Subject: [Newbie] Is the text delimiter changeable?
Message-ID: <3d1f26a4.5915886@news.texas.net>

I have to process large amounts of text and the text includes numerous
apostrophes and quotation marks.  It would be very handy if I could
temporarily assign a substitute for Python's use of the caracters "
and ' to delimit text strings.  If I could use, for example, the @
character or the # character, neither of which appears in the text I
have to work with, and if I could disable the " and ' delimiters, then
my task would be greatly simplified.  I realize I can backslash escape
these characters in some circumstances, but I do not think that will
work when a user's input is a word such as "can't" or "don't."
Anyway, short of doing something truly drastic, such as recompiling
the system (which I have no intention of doing), is there a workaround
that will permit this?

Thank you,

BAF


From peter at engcorp.com  Wed Jun 19 21:18:56 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 19 Jun 2002 21:18:56 -0400
Subject: directory
References: <mailman.1023215768.17394.python-list@python.org> <aer7ev$2em8$1@news.idiom.com>
Message-ID: <3D112D80.830FAB33@engcorp.com>

"James T. Dennis" wrote:
> 
>  os.path.walk() uses an awkward interface.  I wrote and posted this a couple
>  of months ago:
> 
> #!/usr/bin/env python2.2
> ## Change the shebang line to match your needs!
> import os
> def dirwalk(startdir=None):
>         if not startdir:
>                 startdir="."
....


I guess this needs the following line before the "import os" ?

  from __future__ import generators

I'm using standard 2.2.1, I think, and although I vaguely thought
generators were already enabled by default I guess they are not.

Have you modified your site.py or something?

-Peter


From rajarshi at presidency.com  Fri Jun 28 15:03:29 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Fri, 28 Jun 2002 15:03:29 -0400
Subject: joining 2 lists together - an elegant way?
Message-ID: <pan.2002.06.28.15.03.29.911402.1787@presidency.com>

Hi,
  is there a more elegant way to join to lists together?

Right now I do:

newlist = []
for i in l1:
  newlist.append(i)
for i in l2:
  newlist.append(i)

It does the job - but a one liner (if possible) would be nice :)
TIA,


From rddlog at prodigy.net  Sat Jun 22 13:01:47 2002
From: rddlog at prodigy.net (Russ Dilley)
Date: Sat, 22 Jun 2002 17:01:47 GMT
Subject: Numerical Methods in Python
Message-ID: <%32R8.853$xv1.36390126@newssvr15.news.prodigy.com>

I am interested in the numerical capability of Python as well as it's
applicability to system administration  tasks. I am a thermal analyst and
deal rather extensive with numerical solutions to problems, such as
finite-difference methods, interpolation, and numerical integration. I
understand that there is a numerical package for Python called PyNum and was
wondering if anyone can give me any insight into the capacities of this
package.

My dilemma is that I also have responsibilities in UNIX system
administration and it seems to me that I might be able to "kill two birds
with one stone" by learning the Python language. I currently use Perl and
Shell scripts for system admin tasks and a mixture of FORTRAN, C, MathCAD,
Matlab, and Excel for numerical work. As you might guess, it tends to get
confusing constantly jumping from one to another. The syntax and application
tends to blur and it is somewhat inefficient to "relearn" each to some
extent whenever I have a new task to address.

I would like to focus more attention to become proficient with a single
language and, based on what I've seen, it seems that Python might be
appropriate for use in system administration AND numerical programming.

Does anyone have any insight to the applicability of Python to either, or
both, of these tasks?








From tjreedy at udel.edu  Wed Jun 12 16:34:53 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 12 Jun 2002 20:34:53 GMT
Subject: get a list of classes in this module
References: <mailman.1023907851.3426.python-list@python.org>
Message-ID: <NfON8.146292$4i.13334503@bin2.nnrp.aus1.giganews.com>

"Mark McEahern" <mark at mceahern.com> wrote in message
news:mailman.1023907851.3426.python-list at python.org...
> Suppose I have a list of classes:
>
> # validation.py
> class CreditCardValidation:
...
> and I want to make a helper method in the module to translate card
type
> strings into their appropriate class constructor.
...
> I know I can use inspect.getmembers() to do this, but what if I want
to put
> the helper method in the same file as the classes (i.e.,
validation.py)?
> How do I refer to *this* module?

I believe the name of a modulue is the same inside as out.  I do not
know inspect or whether it can work on a module under construction.
However, if the subclasses are too numerous to build the dict by hand,
or the set so dynamic that you do not trust yourself to update it
properly, I believe something like the following (untested) should
work:

def propersubclasstuple((classname,classobj)): return <boolean
expression>

  class_dict = dict(filter(propersubclasstuple, globals().items() ))

> Should I use globals()?  I mean, that would work--so I guess the
question is
> whether there's a more Pythonic approach?

Correct and simple (in that order) *is* Pythonic ;<)

Terry J. Reedy





From DLNXPEGFQVEB at spammotel.com  Mon Jun 10 08:47:00 2002
From: DLNXPEGFQVEB at spammotel.com (Christos Georgiou)
Date: Mon, 10 Jun 2002 15:47:00 +0300
Subject: __imul__ broken for 'objects' in 2.2.1
References: <yfsd6v3qa86.fsf@black132.ex.ac.uk>
Message-ID: <3n79guskvoaeqtr4srg11kaqse7jcqutt4@4ax.com>

On 07 Jun 2002 18:17:45 +0100, rumours say that Alexander Schmolck
<a.schmolck at gmx.net> might have written:

>class Breaky(object):
>    def __imul__(self, other):
>        print "imuling"
>        return self
>sq = Breaky()
>sq *=1.
>
>gives:
>
>Traceback (most recent call last):[...] line 10, in ?
>    sq *=1.
>TypeError: can't multiply sequence to non-int

Just in case: it is indeed your intention to multiply the object with a
float, right?

PS It seems that a Breaky instance is not (incorrectly) considered a
sequence in 2.3a0 (20020603).
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')


From merkosh at hadiko.de  Mon Jun 17 10:39:00 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Mon, 17 Jun 2002 16:39:00 +0200
Subject: what *is* a class?
References: <MPG.17774343cc5a589098968a@news.rz.uni-karlsruhe.de> <aejn8r$pnh$1@nntp9.atl.mindspring.net>
Message-ID: <MPG.177812f0f2706e8e98968b@news.rz.uni-karlsruhe.de>

[This followup was posted to comp.lang.python and a copy was sent to the 
cited author.]

In article <aejn8r$pnh$1 at nntp9.atl.mindspring.net>, 
mgerrans at mindspring.com says...
> It creates and returns an instance (or object, if you prefer) of the class
> Group.   g is referring to that instance, so you haven't lost it, unless you
> assign g to something else.   Since you have a reference to the object you
> created (g), you don't need to "get ahold of it again" -- you still have a
> hold of (on?) it.

well, yes. and no.
if you want to create an object structure, mapping f.e. a binary file 
you have to somehow 'create' the structure. for this you don't have to 
read the file in. the problem was how to code this. assuming i have 
classes for BYTE, WORD, DWORD, LONG, FLOAT, etc. i could pass the root 
node (here: f.e. AVI()) a list of classes describing the header 
information:

AVI(BYTE, DWORD, BYTE, BYTE, BYTE, ...)

If you then build compound classes they perhaps consist of a length 
DWORD followed by 'length' number of bytes. it would be natural to 
compose this new structure LIST of more basic types (at least for the 
length DWORD, certainly not for the chain of data):

class LIST(...):
    def __init__(self, f):
        self.file = f
        self.length = BYTE().parse(f)
        ...

the last line creates a byte structure and reads it from file. i *have* 
to use a separate method call here, because imagine the AVI example from 
above having a string of 17 characters as an argument:

AVI(BYTE, DWORD, STR(len=17), ...)

STR(len=17) already instanciates a STR object while BYTE and DWORD only 
refer to classes.
In the above LIST declaration I could write:

self.length = BYTE()
self.length.parse(f)

... which looks a little humble to me.

 
> But if you insist on having a method to do it, you could define it like so:
> 
> class Group:
>    # ... other code.
>    def getRef(self):
>       return self
> 
> Then, instead of "g2 = g", you could do this:
> 
> g2 = g.getRef()

I guess when I tried that I wrote:

g2 = Group().getRef

instead of 

g2 = Group().getRef()
 
(ouch!)
which of course returned an "unbound method object" (IIRC)

> I can't think of any reason why you would want to do this though, unless you
> just need the extra typing practice and want to confuse someone who might be
> reading the code...

lol. :)
does the above explanation make sense now or is there a more common 
approach to what i'm doing?
 
> Perhaps I misunderstand your question and what you want is a factory.   That
> would be a method that creates and returns objects.   Of course, the
> constructor does this, but you might want a factory if you will instantiate
> different types of objects based on some runtime consideration.   If this is
> what you are after, then reply to that effect for more elaboration.

Yes. That sounds good, too. Depending on how you can define and create 
them.
I will be parsing arbitrary binary files. Up till now I'm just writing 
some base classes for later use. 

If you got some spare time I'd greatly appreciate it if you could either 
elaborate on factories or perhaps name a link with some related 
information to it.

Thanks
Ciao
Uwe


From eppstein at ics.uci.edu  Thu Jun 20 19:38:22 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Thu, 20 Jun 2002 16:38:22 -0700
Subject: Are generators in nested functions possible?
References: <mailman.1024614876.11270.python-list@python.org>
Message-ID: <eppstein-9035CF.16382220062002@news.service.uci.edu>

In article <mailman.1024614876.11270.python-list at python.org>,
 "Bjorn Pettersen" <BPettersen at NAREX.com> wrote:

> def generateIndexes(length=5, span=1):
>     spanvalues = range(-span, span+1)
>     
>     def genInd(res, length):
>         if length == 1:
>             for val in spanvalues:
>                 yield res + [val]
>         else:
>             for i in range(length):
>                 for val in spanvalues:
>                     genInd(res + [val], length-1)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
This is a no-op.  Did you mean
                      for x in genInd(res+[val],length-1): return x
?

>     return genInd([], length)

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From marklists at mceahern.com  Sat Jun 15 09:49:54 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sat, 15 Jun 2002 08:49:54 -0500
Subject: Pronouncing '__init__'
In-Reply-To: <20020615132519.A14452@hishome.net>
Message-ID: <NCBBLFCOHHDIKCAFGCFBCENELEAA.marklists@mceahern.com>

My vote(s):

  the constructor

or

  init

when it makes sense it context.

// m
-





From stephen.boulet at motorola.com  Wed Jun  5 10:59:46 2002
From: stephen.boulet at motorola.com (Stephen Boulet)
Date: Wed, 05 Jun 2002 09:59:46 -0500
Subject: os.popen and place in file
Message-ID: <3CFE2762.519E4344@motorola.com>

I'm running windows NT.

The command: 

  f = os.popen('dir')

returns a file object 'f'. I can't however use the seek command on it:

for i in range(3):
   print f.readline()

f.seek(0) # Should put me back at the top of the "file"

for i in range(3):
   print f.readline()

This should (I think) print the same three lines again, but it doesn't.
Anyone know why?

-- Stephen

(By the way, f = os.popen('dir', 'rb') doesn't work either.)


From pyth at devel.trillke.net  Wed Jun 12 06:09:36 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 12 Jun 2002 12:09:36 +0200
Subject: JIT Complier for python?
In-Reply-To: <8d3f4438.0206111905.7934664b@posting.google.com>; from d2002xx@myrealbox.com on Tue, Jun 11, 2002 at 08:05:09PM -0700
References: <8d3f4438.0206111905.7934664b@posting.google.com>
Message-ID: <20020612120936.W6609@prim.han.de>

d2002xx wrote:
> Is there any JIT Compiler for Python?

you might want to take a look at

http://homepages.ulb.ac.be/~arigo/psyco/

--holger




From gh_pythonlist at gmx.de  Thu Jun 13 22:09:24 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 14 Jun 2002 04:09:24 +0200
Subject: Type subclassing: bug or feature
In-Reply-To: <aeau48$111$1@panix1.panix.com>
References: <aeau48$111$1@panix1.panix.com>
Message-ID: <20020614020924.GC590@lilith.my-fqdn.de>

* Aahz <aahz at pythoncraft.com> [2002-06-13 16:08 -0400]:
> Consider the following code:
> 
> class MyStr(str):
>     def contains(self, value):
>         return self.find(value) >= 0
> 
> s = MyStr("hello, world!")
> s = s.capitalize()
> if s.contains('Hello'):
>     print "Found it!"
> 
> It fails with an AttributeError when it calls s.contains(), because
> s.capitalize() returned a str instead of a MyStr.  Anyone want to take a
> whack at defending this as the correct behavior?

I once had the same problem with subclassing the list type/class: you'd
need to override every single method of the base class to make
subclassing of builtin Python types really useful. I ended up just
implementing the parts of the sequence protocol that I needed in the
particular case.

I'll look at the other answers if there is some clever soluiton using
__getattr__ or a metaclass. One day I'll have to figure out that
metaclass stuff, anyways. I still have no idea what it's all about.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 17.1 ?C      Wind: 1.2 m/s




From peter at engcorp.com  Sat Jun 22 18:52:37 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 22 Jun 2002 18:52:37 -0400
Subject: The best GUI library in python
References: <8d3f4438.0206220457.b820418@posting.google.com> <3D1481DE.C0331204@engcorp.com> <8d3f4438.0206221422.6f71287a@posting.google.com>
Message-ID: <3D14FFB5.90808865@engcorp.com>

d2002xx wrote:
> 
> Peter Hansen <peter at engcorp.com> wrote in message news:<3D1481DE.C0331204 at engcorp.com>...
> > How fast do you need your GUI to be?  Are you talking about
> > an unacceptably slow loading time, or response time within
> > the application?
> 
> Response time within the application, "idle" is. I think it's not due
> to python, because boa-constructor (using wxWindows) and some examples
> of PyQT work much faster than idle and others based on Tkinter.
> 
> Well, if you doubt what I said, try to use grail. Or write a listbox
> containing 10,000 items, then wait......

So your application needs to work in a speedy fashion with listboxes
containing 10,000 items?  How fast?  Is your test simply to scroll
through the list, or is it to time the display the first time?

Grail?  I didn't think that was supported.  I did try downloading it 
once, as I recall, a long time ago, but didn't get it to work.  It
was probably just me but I don't feel like trying again.

Anyway, how fast do you need the GUI to be?  The grass is always 
greener...

-Peter


From gerhard at bigfoot.de  Tue Jun 18 18:41:40 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 18 Jun 2002 22:41:40 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <D7NP8.46301$GY.14487785@twister.nyroc.rr.com>
Message-ID: <slrnagvdp4.1sg.gerhard@lilith.my-fqdn.de>

George Hester wrote in comp.lang.python:
> I just cannot believe what I ran into in this newsgroup. [...]


                                                .:\:/:.
             +-------------------+             .:\:\:/:/:.
             |   PLEASE DO NOT   |            :.:\:\:/:/:.:
             |  FEED THE TROLLS  |           :=.' -   - '.=:
             |                   |           '=(\ 9   9 /)='
             |   Thank you,      |              (  (_)  )
             |       Management  |              /`-vvv-'\
             +-------------------+             /         \
                     |  |        @@@          / /|,,,,,|\ \
                     |  |        @@@         /_//  /^\  \\_\
       @x@@x@        |  |         |/         WW(  (   )  )WW
       \||||/        |  |        \|           __\,,\ /,,/__
        \||/         |  |         |      jgs (______Y______)
    /\/\/\/\/\/\/\/\//\/\\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From eric.brunel at pragmadev.com  Thu Jun  6 05:40:06 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Thu, 6 Jun 2002 09:40:06 +0000
Subject: Freeze questions
References: <mailman.1023297729.24131.python-list@python.org>
Message-ID: <adn39m$mdv$1@wanadoo.fr>

Sung H. Kim wrote:
> Can the Freeze tool be used with C libraries? i.e. extensions?
> 
> Will it create a truly standalone app (on Windows)?
> 
> Does it support tk/tkinter apps yet?
> 
> Finally, is there a good how-to on the Windows platform?

Does it have to be freeze? I used it quite a lot and to be honest, it's 
quite difficult to get Tkinter applications working after freezing if you 
do not have a shared tcl/tk installation... There are other solutions that 
are far more straightforward for this business. I personally use Gordon 
McMillan's installer ( http://mcmillan-inc.com/install1.html ) which is 
very simple to configure and use. And it natively supports Tkinter 
applications.

There is another installer called py2exe (just google for it), but I never 
tried it, so I don't know if it suits your needs.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From paul at boddie.net  Tue Jun 11 07:49:20 2002
From: paul at boddie.net (Paul Boddie)
Date: 11 Jun 2002 04:49:20 -0700
Subject: Compiling python2.2.1 on redHat 6.1 - doesn't work
References: <bb06817.0205210533.d8d5ffb@posting.google.com> <23891c90.0205220343.4aa581c8@posting.google.com> <bb06817.0205230212.7b6a388c@posting.google.com> <23891c90.0205230743.7ea59255@posting.google.com> <23891c90.0206030901.26402dd0@posting.google.com> <bb06817.0206100526.55f36028@posting.google.com>
Message-ID: <23891c90.0206110349.15aa7861@posting.google.com>

matthew.russell at securetrading.com (Matt Russell) wrote in message news:<bb06817.0206100526.55f36028 at posting.google.com>...
> Thanks Paul, 
> am afraid being a "c" newbie, I have no idea on how to apply the
> patch, would be great if you could provide an example.

Well, the stated bug in question is supposed to be fixed, although the
thread on SourceForge hardly seems to indicate complete satisfaction,
and I'm not entirely convinced that fixing configure.in changes the
underlying issue. Anyway, I think it's enough for you to do the
following:

  1. Copy posixmodule.c somewhere safe. You shouldn't need to do this
     but it makes me feel better!

  2. In the Python-2.2.1 directory (the top level of the source tree)
     download the patch as Python-2.2.1-posixmodule.

  3. Try something like this in that directory:

     patch -p0 < Python-2.2.1-posixmodule

The program should work out which file is to be patched automatically.
You should be able to complete the build from there on. In other
words:

  make

Remember that I haven't tested any of functionality affected by this
patch, and I suppose that what I've done hasn't been rubber-stamped by
the people who really maintain this code, so beware!

Paul


From phr-n2002b at NOSPAMnightsong.com  Fri Jun 21 06:33:57 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 21 Jun 2002 03:33:57 -0700
Subject: I'd give up Perl tomorrow if only...
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au> <3D12F6FC.3000608@wedgetail.com>
Message-ID: <7xd6ukex9m.fsf@ruckus.brouhaha.com>

Derek Thomson <derek at wedgetail.com> writes:
> Among other things, it's one of the reasons I'm tinkering with a Perl ORB.

Wow, that sounds cool.

> Yes, it's stop-gap, but when you already have a non-trivial Perl
> module, either on CPAN or in-house, it can save a lot of
> rewriting. What we *really* need to do is dynamically create CORBA
> interfaces for Perl modules, but that might be impossible given that
> Perl is like Python in that module and class interfaces aren't
> pre-declared :(

Why not some more dynamic remote object protocol instead of messing
with Corba?


From whisper at oz.net  Fri Jun 14 17:28:58 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 14 Jun 2002 14:28:58 -0700
Subject: Associating a C++ class with a Python class
In-Reply-To: <2cc96dc7.0206141212.5b6ca5df@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEMEEBDJAA.whisper@oz.net>

Try the Boost Python library that does all this and more - you can write
python classes callable from C++ and vice versa. The author says the latest
and greatest (version 2) is best obtained from CVS if you're starting a new
project.

www.boost.org

HTH,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of John Dunn
> Sent: Friday, June 14, 2002 13:13
> To: python-list at python.org
> Subject: Associating a C++ class with a Python class
>
>
> I would like to add a new class type to my embedded python engine and
> I would like this class to wrap an exisiting C++ class instance. All
> of the examples I have found are more 'c implementation' than 'c
> wrappers', meaning that the python instance is not associated with an
> different c++ class. The methods of the python class will just be
> straight calls through to the C++ class.
>
> These classes will only be created from C, so I am basically looking
> for a way to stuff a pointer to my C++ class into the python object. I
> guess I could have a map of PyObject->C++ class, but I was hoping
> there was a cleaner way to accompish this task. Is there a standard
> way to do this? Am I missing something obvious?
>
> Thanks-
>
> John
> --
> John Dunn
> Peak Audio, a division of Cirrus Logic, Inc.
> http://www.peakaudio.com
> --
> http://mail.python.org/mailman/listinfo/python-list





From jubafre at zipmail.com.br  Tue Jun 18 21:01:07 2002
From: jubafre at zipmail.com.br (jubafre at zipmail.com.br)
Date: Tue, 18 Jun 2002 22:01:07 -0300
Subject: =?iso-8859-1?Q?I=20have=20=2Epy=2C=20I=20want=20=2Eexe=2C=20how=3F=3F?=
Message-ID: <3D0F6623000012D8@www.zipmail.com.br>

i have a file.py 
how i make e executable file .exe for run in windows???

Juliano Freitas
www.gebrasil.hpg.com.br



------------------------------------------
Use o melhor sistema de busca da Internet
Radar UOL - http://www.radaruol.com.br







From jonathan at onegoodidea.com  Wed Jun 26 14:39:54 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Wed, 26 Jun 2002 19:39:54 +0100
Subject: Python 2.2 __slots__ Bug?
References: <mailman.1025089474.17780.python-list@python.org> <afcrp8$c64$1@panix1.panix.com>
Message-ID: <B93FC90A.D01E%jonathan@onegoodidea.com>

On 26/6/2002 17:57, in article afcrp8$c64$1 at panix1.panix.com, "Aahz"
<aahz at pythoncraft.com> wrote:

> Interesting.  Using Python 2.2.1 on Mandrake 8.1, I get a segfault
> without printing "I'm fine".  Go ahead and file a bug.

I finally got Sourceforge to talk to me long enough to submit a bug report
for this:

    #574207 Chained __slots__ dealloc segfault

Feel free to followup with any information I've missed.

Jonathan



From ponderor at lycos.com  Sat Jun 15 20:45:42 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 15 Jun 2002 17:45:42 -0700
Subject: Newbie deluxe
References: <hdDO8.29984$Hn4.969427@news1.east.cox.net>
Message-ID: <e81be8b2.0206151645.9fa4b30@posting.google.com>

"Zach" <zsmith at ponymail.com.nospam> wrote in message news:<hdDO8.29984$Hn4.969427 at news1.east.cox.net>...

> I've just started taking a look at Python (experienced in Delphi) and so far
> it looks very clean and comprehensible.  My question is, what are the
> fundamental differences and similarities between it and Perl?  I hear so
> much about these two languages, especially in Linux circles and just
> wondered what sets them apart.  

This is an interesting article: http://www.prescod.net/python/why.html

> Can full fledged graphical applications be
> built with Python or does it lean more towards scripting?  

If you do like WXPython & Rapid GUI (You /did/ mention Delphi), also
check out the PythonCard project: http://pythoncard.sourceforge.net/

>> I'm aware that it
> has OOP capabilities so I assume it reaches beyond just scripting but like I
> said, I've no experience with it.  I would like to broaden my coding skills
> with the addition of another language so any general opinions with regards
> to its usefulness are appreciated.

Broadening my programming language perspective was one of the primary
reasons I dug into Python.   I realized how important this was
studying Java (from a C/C++ background), and realized how wonderful
this CAN BE studying Python.  The shell/procedural perspective gives
you a "try it while it's fresh in my head" opportunity, which quickly
and easily allows you to wrap your procedural research in OO
classes...OO purists require that you map all classes before
implementation, but I don't find this to be the best case, especially
when you're getting familier with the language/technology. I digress,
yet that is where I find Python very useful, and wish other languages
(C#) provided a shell epxerience.  BTW, Jython is on my To-Do list,
and where I'd consider starting if you're heading into a Java project.
 ( http://www.oreilly.com/catalog/jythoness/ ,
http://www.onjava.com/pub/a/onjava/2002/03/27/jython.html ... and
since I'm here, O'Reilly's Safari is a fairly cheap way to get your
hands on the book resources you're not sure you want to invest in. )

Python's "batteries included" and open source initiative takes you
beyond learning the language to learning and experimenting with modern
technologies, and if you're looking into deep comp. sci. studies I
don't think you'll be disappointed with opportunities to find and
experiment with classical algorithm implementation or specialized
projects.

I'm not sure what meant by "coding skills", but by shear speed of
coding & it's consistency with many coder's common thinking ("How I
Think") should support your endeavor.

Python's OOP will give you all you want. Unless you're a
purist...Python has multiple inheritance, and requires some extra
syntax for interfaces, and other mostly minor nuances, nuances are
familier to all "OO" languages....but if that's your style you'll
probably grumble about indention and : usage and will get over it
shortly after you've got a good grasp on the language. (Note the
/after/ part, you'll probalby tune into Python quickly enough to start
pondering fundamentals, instead of still memorizing syntactic
features/stumbling blocks of other languages.) (For the c.l.py record,
I haven't been following the recent "why doesn't ... OO" discussions,
and I'm not qualified nor want to comment on Python as a functional
language. But mention it in case that's what Zach may be looking for
in expanding his coding skills.)

When you get perplexed, the python mailing lists, especially the Tutor
list, are friendly environments for your questions/befuddlements.

My 2 cents flattened by a train of associations,

-Dean


From gerhard at bigfoot.de  Thu Jun 20 09:10:34 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 20 Jun 2002 13:10:34 GMT
Subject: RES: latex-tex-packages
References: <mailman.1024575481.1722.python-list@python.org>
Message-ID: <slrnah3kub.18g.gerhard@lilith.my-fqdn.de>

Alves, Carlos Alberto - Coelce wrote:
> This message is in MIME format. Since your mail reader does not understand
> this format, some or all of this message may not be legible.

Please don't do that and configure your mail-/newsreader to send text-only.

> Thanks,
> Pardon me, but I forgot to say it's to install in my linux box at home. So I
> appreciate packages for linux box based on debian distribuition (I use
> kernel series 2.4).

I'd let the Debian package management worry about the dependencies, and just

$ apt-get install lyx

which will install the tetex-bin, among others.

Btw. my preferred frontend to the Debian package management system is aptitude.
Or you can use the somewhat clumsy, but powerful dselect to select packages to
install.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From gh_pythonlist at gmx.de  Tue Jun 11 01:49:46 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Tue, 11 Jun 2002 07:49:46 +0200
Subject: wxPython performance
In-Reply-To: <3D04CF8D.4050407@bgb.cc>
References: <slrnafhmoa.389.ak@ak.silmarill.org> <ufk3bdnh675sc1@corp.supernews.com> <3D04CF8D.4050407@bgb.cc>
Message-ID: <20020611054946.GA10730@lilith.my-fqdn.de>

* Don Garrett <garrett at bgb.cc> [2002-06-10 16:12 +0000]:
> RPM1 wrote:
> >"Andrei Kulakov" <ak at silmarill.org> wrote
> >> "hello world" type program from wxPython tutorial takes 6
> >>seconds to start on my cel 366 128mb system (running Debian). Is
> >>this typical? Why is it so slow? I'm looking for a graphic
> >>toolkit that'd be fairly quick. I don't want to use Tk because of
> >>the looks.
> 
> This whole thread is somewhat strange.... one of the things I liked about 
> wxWindows was that it seemed like one of the snappier cross plateform 
> libraries I've worked with. [...]

I agree. Some people are mistaking "hello world" with typical
applications and startup time with performance ;-)

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 12.1 ?C      Wind: 2.7 m/s




From R.Barrett at ftel.co.uk  Mon Jun 10 12:45:30 2002
From: R.Barrett at ftel.co.uk (Richard Barrett)
Date: Mon, 10 Jun 2002 17:45:30 +0100
Subject: module name differences depending on how generated
Message-ID: <5.1.0.14.2.20020610171311.037b8c98@pop.ftel.co.uk>

Compiling Python 2.2.1 from source on Suse Linux 7.3.

It appears that if the nis (Sun Yellow Pages support) module is generated 
by $build/setup.py, the resulting shared library is called nis.so 
(presumably per the Extension constructor in $build/setup.py).

If, on the other hand, you enable the nis support module's production in 
the $build/Modules/Setup file the module gets called nismodule.so (after 
the source file name).

I'll assume there is a good reason for this difference and would appreciate 
it if anyone could tell me what it is.

It happens that I initially did make install (which installed nis.so). 
Later, while making other changes, I  uncommented the nismodule line 
$build/Modules/Setup and ended up with both nis and nismodule installed. As 
I subsequently figured out, nis.so will be loaded in preference to 
nismodule.so when in response to an import nis statement.

The problem I then had was that in patching nismodule.c to fix a documented 
problem in the nismdoule.c code running under Linux (see 
http://sourceforge.net/tracker/index.php?func=detail&aid=233084&group_id=5470&atid=105470) 
running make install to generate a revised nis support module produced and 
installed a new nismodule.so. But this had no effect as the originally 
generated nis.so was being picked up in preference. Of course nis.so was 
not being regenerated and installed (presumably because setup.py detected 
that nismodule.so was being produced) so that at first blush the patch 
appeared to be ineffective.

Having wasted some time with this module naming problem, I thought I'd ask 
if it was for a good purpose.





From claird at starbase.neosoft.com  Fri Jun 14 15:34:23 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 14 Jun 2002 14:34:23 -0500
Subject: NEWbie asks "is Python a good tool for automated file/folder generation?"
References: <6k6kgukhuk8pfgos97a38h0d4r5jr0u3sr@4ax.com>
Message-ID: <91B0759CECFC4C3A.944B7AF3C2197D61.C612F005CF9D6D07@lp.airnews.net>

In article <6k6kgukhuk8pfgos97a38h0d4r5jr0u3sr at 4ax.com>,
Unconditional Grace  <UG at LY> wrote:
			.
			.
			.
>I have a series of scientific image files numbering, say, 1000, which
>are really 100 groups, each group with its own identical string, but
>appended with a different number from _01 to _10, or possibly some
>other string.  The end enumeration is ALWAYS differentiated from the
			.
			.
			.
>I've never used Python, and I'm wondering if you, collectively, think
>that Python is a good language to do this.  While it seems like Python
Python's an absolutely fine language for this.
No broadly-popular language will do the job
much better, and it certainly would be consider-
ably harder in several.
			.
			.
			.
-- 

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


From chris.lyon at spritenote.co.uk  Sun Jun 23 14:30:49 2002
From: chris.lyon at spritenote.co.uk (chris lyon)
Date: Sun, 23 Jun 2002 19:30:49 +0100
Subject: Newbie to HTMLgen
Message-ID: <3d16132a$0$27325$afc38c87@news.easynet.co.uk>

I'm porting an app from Cold Fusion to Python, for various reasons.

And am looking at the possibility of using HTMLgen

I am trying to place the Submit button of a form into a table element using
Tablelite
since the Submit button is generated by the Form Class is there any way to
put it into the table header tag,
of a table that has rows of table data and inputs.

I can't seem to find a way to place the submit button into a table header
cell where the client likes it.



What I'm trying to produce is :-


<HTML>
<HEAD>
<TITLE>General Input Form</TITLE>
</HEAD>
<BODY BGCOLOR=#A0A0A0>&nbsp
<TABLE BORDER="3" CELLPADDING="5" CELLSPACING="0" bgcolor="#C0C0C0">
    <TR bgcolor=#A0A0A0>
        <TH><H3>Add A New System</H3></TH>
        <FORM ACTION="adm-gen-insert.cfm" METHOD="post"
ENCTYPE="multipart/form-data">
        <TH>
            <INPUT TYPE="submit" VALUE="Add System">
        </TH>

<TR><TD></TD><TD>This produces entries in the octavia database</td></tr>

 <TR><TD>DSN:</TD><TD><INPUT TYPE="text" NAME="gendsn" SIZE="10"
MAXLENGTH="80"></TD></TR>

<TR><TD>Webroot value</TD><TD>C:\INETPUB\WWWROOT</TD></TR>
<TR><TD>Title:</TD><TD><INPUT TYPE="text" NAME="gentitle" SIZE="20"
MAXLENGTH="80"></TD></TR>
<TR><TD>Webroot:</TD><TD><INPUT TYPE="text" NAME="genwebroot"
value="C:\INETPUB\WWWROOT" SIZE="40" MAXLENGTH="255"></TD></TR>
<TR><TD>Image (large):</TD><TD><INPUT TYPE="file" NAME="filegenimagebig"
SIZE="40" MAXLENGTH="80"></TD></TR>
<TR><TD>Image (small):</TD><TD><INPUT TYPE="file" NAME="filegenimagesmall"
SIZE="40" MAXLENGTH="80"></TD></TR>
<TR><TD>Logo:</TD><TD><INPUT TYPE="file" NAME="filegenlogo" SIZE="40"
MAXLENGTH="80"></TD></TR>
<TR><TD>Bkgd Colour:</TD><TD><INPUT TYPE="text" NAME="genbgcolor"
VALUE="#A0A0A0" SIZE="10" MAXLENGTH="80"></TD></TR>
<TR><TD>Text Colour:</TD><TD><INPUT TYPE="text" NAME="gentxcolor"
VALUE="#000000" SIZE="10" MAXLENGTH="80"></TD></TR>


    </FORM>
</TABLE>

</BODY>
</HTML>




From emile at fenx.com  Fri Jun  7 09:17:51 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 07 Jun 2002 13:17:51 GMT
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com> <Lz1M8.123774$352.6153@sccrnsc02> <3D00ABB6.6FAFD567@engcorp.com>
Message-ID: <3o2M8.10509$pw3.96@sccrnsc03>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3D00ABB6.6FAFD567 at engcorp.com...
> Emile van Sebille wrote:
> >
> > Peter Hansen
> > > You forgot the even more important first thing for a beginner:
> > > get it correct!  The only good way to do this is to write unit
> > > tests that prove it.  If you don't have tests, you don't have
> > > working code (or how can you prove it?).
> >
> > We always used absence of bug reports as a gauge... and always
> > questioned if that indicated usage or correctness.  ;-)
>
> For those who are serious when they say that, it's all well
> and good if you can get your customer to test the code for
> you.... ;-)
>

Even better, as most of the time they're not sure what they asked for in
the first place, blurring the line separating implementation error and
user specification error.  ;-))

(we really did think like that then -- for some set of we that includes
me anyway)

--

Emile van Sebille
emile at fenx.com

---------



From fwang2 at yahoo.com  Tue Jun 11 02:19:29 2002
From: fwang2 at yahoo.com (oliver)
Date: 10 Jun 2002 23:19:29 -0700
Subject: python, shell, environment variable
Message-ID: <a7c2c93c.0206102219.28e2d2f7@posting.google.com>

hi, folks,

I am trying to get a set of python and bash scripts to work, and I
constantly run into trouble with environment variables:

First, my python script need to modify environment variable, say,
PATH, and want to keep the modification even after the script is done.
os.environ["PATH"]= ... doesn't seem to work, any idea?

Second, it would be better if my python script can call bash shell and
still keep the environment variable modification done by bash. I tried
os.popen("source some_shell"), it doesn't work. ?

Thanks for help.

-oliver


From jblazi at hotmail.com  Mon Jun 24 07:40:51 2002
From: jblazi at hotmail.com (JB)
Date: Mon, 24 Jun 2002 13:40:51 +0200
Subject: converting an array of chars to a string
References: <3d12dda2_6@news.newsgroups.com> <aevj41$fg2$0@216.39.172.122>
Message-ID: <3d17038e_6@news.newsgroups.com>

Bengt Richter wrote:

> Have a look at the array module. It can construct an array
> of signed or unsigned character elements optionally
> initialized with values from a list or string. 
[...]

Thank you very much. In principle, this was exactly, for 
what I was looking. But most unforunately if I declare such 
an array a, then

a[0]=210
a[0] += 46

produces an error instead of the correct result 0. Now I am 
a bit desperate.

This is like many things in the Python library: they look 
good at first sight but when you take a closer look at 
them, you see, that...

Thank you anyway!

-- 
Janos Blazi


-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From plugin at supercable.es  Tue Jun 25 14:11:18 2002
From: plugin at supercable.es (Ugo García)
Date: Tue, 25 Jun 2002 20:11:18 +0200
Subject: Threads on embedded Python
References: <1024952585.111238@seux119> <Xns9237EECAA9CC5cliechtigmxnet@62.2.16.82> <1024957605.771843@seux119> <Xns9238944DA0FEcliechtigmxnet@62.2.16.82>
Message-ID: <1025028733.290520@seux119>

thanks Chris. The URL you give me is very interesting. Pufff, this is a big
mess... Don't know exactly how to do it. Perhaps could be a good idea to
have the main loop in Python...

But, one question: you say that I don't have to have a main loop in order to
have threads running using Py_BEGIN_ALLOW_THREADS, but don't known exactly
how to use it. I explain: from the main loop in my C application I launch
threads in python using PyRun_SimpleString. But these threads doesn't run
unless I call Python to do something. How I have to use
Py_BEGIN_ALLOW_THREADS in order not to call Python continuosly?

Thanks again,
--ugo

"Chris Liechti" <cliechti at gmx.net> escribi? en el mensaje
news:Xns9238944DA0FEcliechtigmxnet at 62.2.16.82...
> "Ugo Garc?a" <plugin at supercable.es> wrote in
> news:1024957605.771843 at seux119:
>
> > Let see if I get something clear... I don't have any main loop in
> > Python. I only have some functions that 'control' certains object in
> > my operation. To be a little more clear, the task of the functions are
> > for example, move an object in the world generated by a game with a
> > little intelligence. For example: (a simple example)
> >
> > FUNCTION 1:
> >     - Wait until main character is in correct position
> >     - Play a sound
> >     - Move an enemy from pos X1,Y1 to X2,Y2
> >     - Play an animation
> >     - ....
> >     - ....
> >
> > So, I don't have a main loop, only function like that (and other kind
> > of functions) that a run as a thread to have all them running
> > simultaneous. Is there other way to do that without having threads?
>
> sure, it's called polling ;-)
>
> and i think this one could be an iteresting read for you:
> http://www-106.ibm.com/developerworks/library/l-pythrd.html
> it's a good way to distribute CPU power to different 'microthreads' while
> having  them well separated. altough you will need polling as microthreads
> don't support blocking IO.
>
>
> why not hanging it up the other way round? i think if you have the main
> loop in python the thing will get easier. you can then call back to your C
> module to do the screen updates etc.
>
> > Do
> > I have to have a main loop in order to have the threads running?
>
> when you're in a C module the Global Interpreter Lock stopps all python
> activity util you enable it explicitly (Py_BEGIN_ALLOW_THREADS) or you
> return to python.
> the GIL makes writing extensions to python peanuts...
>
> > Perhaps running Python in a separate thread could be a good idea.
>
> yes someone has to spend the CPU cycles...
>
> > Thanks,
> > --ugo
> > "Chris Liechti" <cliechti at gmx.net> escribi? en el mensaje
> > news:Xns9237EECAA9CC5cliechtigmxnet at 62.2.16.82...
> >> "Ugo Garc?a" <plugin at supercable.es> wrote in
> >> news:1024952585.111238 at seux119:
> >>
> >> > Hi all. I have an applicattion who uses Python to control some
> >> > processes. I've made a little script library to work with threads
> >> > and each proccess is a Python thread. The main application make
> >> > calls to Python in order to create these proccesses. The problem is
> >> > that the proccess doesn't run unless I call Python to do something.
> >> > Now in the main loop I have a line like this:
> >> >     PyRun_SimpleString("dummy=1");
> >> >
> >> > This make Python the threads to run. The problem is that this is a
> >> > little slow 'cause if I call Python from tha main application too
> >> > many times the proccesses run well but not the main application,
> >> > and if a make a big pause between calls the proccesses doesn't work
> >> > at a good speed. Is posible to have some threads running in an
> >> > embedded Python without making callings to it continuosly? It would
> >> > be a better idea to have another thread who's only task is to call
> >> > Python (a line that the one written before) in order to have the
> >> > Python proccesses running?
> >>
> >> why not run python in a separate thread (created by your app) and
> >> doing
> > the
> >> loop in python, like PyRun_SimpleString("while 1: mainloop()") or so?
> >> you do have a main loop in python, don't you? otherwise you wouldn't
> >> need python threads, right? :-)
> >>
> >> you could possibly also release the GIL, but don't forget to grab it
> >> again when accessing the python interpreter. i'm not sure about this
> >> one, so i would try the above.
> >>
> >> chris
> >>
> >> --
> >> Chris <cliechti at gmx.net>
> >>
> >
> >
> >
>
>
>
> --
> Chris <cliechti at gmx.net>
>




From erict at millfilm.co.uk  Wed Jun  5 08:12:28 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Wed, 05 Jun 2002 13:12:28 +0100
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org> <3cfb2695$1_1@news.uni-bielefeld.de>
Message-ID: <3CFE002C.73A706C8@millfilm.co.uk>

What did you lose using pygtk instead of wxWindow ?


Mario Wehbrink wrote:

> Am  Sat, 01 Jun 2002 14:38:46 GMT  schrieb/wrote/
> Andrei Kulakov  < ak at silmarill.org > in comp.lang.python :
>
> > Hello,
> >
> >     "hello world" type program from wxPython tutorial takes 6
> > seconds to start on my cel 366 128mb system (running Debian). Is
> > this typical? Why is it so slow? I'm looking for a graphic
> > toolkit that'd be fairly quick. I don't want to use Tk because of
>
> Hi,
>
> try pygtk. I also switched to it, after i found wxPython too slow for my
> PPro-200.
>
> http://www.daa.com.au/~james/pygtk/
>
> Mario
>
> --
> http://www.porcupinetree.com



From cmkim at shinbiro.com  Fri Jun 28 02:11:16 2002
From: cmkim at shinbiro.com (Kim Chul Min)
Date: Fri, 28 Jun 2002 15:11:16 +0900
Subject: [Q] How Can I call python function from C source?
References: <afe3ji$gvk$1@news.kreonet.re.kr> <affguf$e3cn5$1@ID-139865.news.dfncis.de>
Message-ID: <afgukt$m26$1@news.kreonet.re.kr>

Thank you for the help, Dave.

BTW, I make test1.c like what you said and when I compile it I got some
compile errors. Plz help more.

**************************************************************
$ gcc -o test1 test1.c
Undefined                       first referenced
 symbol                             in file
PyCallable_Check                    /var/tmp/ccob3Tkv.o
PyLong_AsLong                       /var/tmp/ccob3Tkv.o
PyObject_HasAttrString              /var/tmp/ccob3Tkv.o
PyObject_GetAttrString              /var/tmp/ccob3Tkv.o
Py_Exit                             /var/tmp/ccob3Tkv.o
PyImport_ImportModule               /var/tmp/ccob3Tkv.o
PyObject_CallFunction               /var/tmp/ccob3Tkv.o
Py_Initialize                       /var/tmp/ccob3Tkv.o
ld: fatal: Symbol referencing errors. No output written to main
collect2: ld returned 1 exit status
**************************************************************

Anything I mistake?
Wait for Any Idea. Plz help me.

Kim Chul Min <cmkim at shinbiro.com>

"Dave Kuhlman" <dkuhlman at rexx.com> wrote in message
news:affguf$e3cn5$1 at ID-139865.news.dfncis.de...
> Kim Chul Min wrote:
>
> > Hello there,
> >
> > I want to call python function from c source.
> >
> > for example,
> >
> > in file "main.c"
> > --------------------------------------------------------------
> > int main(void)
> > {
> >
> >     /** want to call python function add(a,b) that return a+b **/
> >
> > }
> > --------------------------------------------------------------
> >
> > in the file "add.py"
> > --------------------------------------------------------------
> > def add(a,b):
> >     return(a+b)
> > --------------------------------------------------------------
> >
> > Can I call any function that written by python from C source ?
> >
> > If possible, can anyone show me simple example like above??
> >
> > Plz,,, Help me....Thankx in advance...
> >
> > Ciao,
> > Farangan Xavio
> > cmkim at shinbiro.com
>
> Here is a simple example.
>
> You will find the following especially helpful:
>
>     http://www.python.org/doc/current/api/importing.html
>     http://www.python.org/doc/current/api/object.html
>
> /* =========================================== */
>
> #include "Python.h"
>
> int main(int argc, char ** argv)
> {
>     PyObject * module;
>     PyObject * function;
>     PyObject * result;
>     long cresult;
>     int arg1;
>     int arg2;
>
>     if (argc != 3)
>     {
>         printf("Usage: test1 <arg1> <arg2>\n");
>         exit(-1);
>     } /* if */
>
>     arg1 = atoi(argv[1]);
>     arg2 = atoi(argv[2]);
>
>     Py_Initialize();
>     module = PyImport_ImportModule("add");
>     if (PyObject_HasAttrString(module, "add"))
>     {
>         function = PyObject_GetAttrString(module, "add");
>         if (PyCallable_Check(function))
>         {
>             result = PyObject_CallFunction(function, "ii", arg1, arg2);
>             cresult = PyLong_AsLong(result);
>             printf("arg1: %d  arg2: %d  result: %d\n", arg1, arg2,
cresult);
>         }
>         else
>         {
>             printf("Error -- attribute is not a function/method\n");
>         } /* if */
>     }
>     else
>     {
>         printf("Error -- Module does not contain \"add\" function.\n");
>     } /* if */
>     Py_Exit(0);
> } /* main */
>
> /* =========================================== */
>
>
>
> --
> Dave Kuhlman
> dkuhlman at rexx.com
> http://www.rexx.com/~dkuhlman
>




From loewis at informatik.hu-berlin.de  Sat Jun 15 07:31:27 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 15 Jun 2002 13:31:27 +0200
Subject: Pronouncing '__init__'
References: <mailman.1024136779.31270.python-list@python.org>
Message-ID: <j4sn3ozskw.fsf@informatik.hu-berlin.de>

Oren Tirosh <oren-py-l at hishome.net> writes:

>  under under init under under
[...]

> Most of these pronounciations also have a shorter version with the
> trailing part omitted.

I use

  under under init

Regards,
Martin


From rdacker at pacbell.net  Mon Jun 10 22:35:07 2002
From: rdacker at pacbell.net (rdack)
Date: 10 Jun 2002 19:35:07 -0700
Subject: urllib question
Message-ID: <644f6688.0206101835.5dc793b1@posting.google.com>

this works but i don't understand it. it retrieves data from my
router.
  params = urllib.urlencode({'RC': '@D', 'ACCT' : "root", 'PSWD' :
"71:29:26", 'URL': 'admin' })
  ipurl = "http://" + iphost + "/cgi-bin/logi"
  urlfp = urllib.urlopen(ipurl, params)

i was wondering exactly what those params mean.
i have some other code that wants a user name and password and i don't
know what to give it. is the username = 'root'?
and what is the password? a string with 3 two digit numbers separated
by colons?
what is the 'URL' param mean?
somewhere to find documentation?


From BPettersen at NAREX.com  Tue Jun 18 12:11:49 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 18 Jun 2002 10:11:49 -0600
Subject: Embedding Python - time module problem
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158EDE@admin56.narex.com>

> From: Richard Townsend [mailto:richardt at edshk.demon.co.uk] 
> 
> I have been having problems embedding Python in a small C 
> program which calls a Python script which, in turn, imports 
> the time module.
> 
> This fails with the error:
> 
> "ImportError: dynamic module does not define init function (inittime)"
> 
> I have simplified this down to a minimal C program which just 
> tries to import the time module. This gives the same error.
> 
>   #include <stdio.h>
>   #include <Python.h>
> 
>   int main(int argc, char **argv) 
>   {
>     PyObject *module;
>     Py_Initialize();
>     module = PyImport_ImportModule("time");
>     if (module == NULL)
>     {
>       PyErr_Print();
>     }
>   }
> 
> If I substitute "os" or "sys" for "time" then no error 
> occurs. However "math" does give a similar error.  
> 
> If I run the interpreter on its own, I can import time and 
> math with no problems. 
> 
> I am using Python 2.2.1 on HP-UX 11i
> 
> Is there something special I need to do to be able to import 
> time and math modules when using embedded Python?

Your example works for me (WinXP). The time module should be compiled
into the Python library, so I'm not sure why your installation is trying
to import it as a dynamic module... but since it is, do you happen to
have a shared library named time somewhere on your path?

Sorry I couldn't be of more help, it's been a while since I saw an HP
box.

-- bjorn




From markus.vonehr at ipm.fhg.de  Thu Jun 27 05:04:52 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Thu, 27 Jun 2002 11:04:52 +0200
Subject: How to get rid the new line
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org>
Message-ID: <3D1AD534.A74EC9F4@ipm.fhg.de>

        f = open(filename, 'r')
        lines = f.readlines()
        line1 = lines[0] # exemplarily for first line
        line1 = line1[0:len(line1)-1]

or:
        f = open(filename, 'r')
        lines = f.readlines()
        line1 = lines[0][0:len(line1)-1] # exemplarily for first line

Markus

SiverFish schrieb:

> Anyone tell me how to skip the empty line when you read through the file
> and copy the information in to the list(no empty line )



From gumuz at looze.net  Thu Jun 13 07:39:26 2002
From: gumuz at looze.net (Guyon Morée)
Date: Thu, 13 Jun 2002 13:39:26 +0200
Subject: Bits and bytes?
References: <mailman.1023896057.21318.python-list@python.org>
Message-ID: <3d08839a$0$226$4d4ebb8e@news.nl.uu.net>

ok, really a lot of thanx for your help, this is awesome! :D

but, let me explain what i am actually trying to do here and why i think
both of your solutions won't work for me.

i am trying to implement my own sort of huffman encoding/compression.
converting asci-characters to a compressed bitstream requires me to write it
at bit-level. Bill's solution might work in some way, but my sequence of
bits aren't allways divideable by eight. a word like 'ape' (which is 3x8=24
bits) might become 1001101, which is 7 bits.... catch my drift? strings is
no option as this would have no positive effect on the 'real size' ;there's
no compression.

i hope this makes it a bit clearer for you. anyway, thanx a lot.

guyon

"Scherer, Bill" <Bill.Scherer at VerizonWireless.com> wrote in message
news:mailman.1023896057.21318.python-list at python.org...
On Wed, 12 Jun 2002, Guyon Mor?e wrote:






From storminDOTnorm at excite.com  Wed Jun 12 08:59:59 2002
From: storminDOTnorm at excite.com (Stormin Norm)
Date: Wed, 12 Jun 2002 08:59:59 -0400
Subject: filecmp.cmpfiles on directories
References: <3D06491B.302@excite.com> <Xns922B52DB6F313gmcmhypernetcom@199.171.54.215>
Message-ID: <3D0745CF.7060203@excite.com>

Thanks Gordon,
The documentation does not mention the filecmp.dircmp function.  Any 
documentation on filecmp.dircmp anywhere?? 
Do you know if there is any ability with filecmp to walk the tree 
structure. Otherwise I'll use os.path.walk and call filecmp.dircmp.

-StorminNorm.

Gordon McMillan wrote:

>Stormin wrote:
>
>>I am trying to walk a directory structure and compare directories for 
>>duplicates.  For some reason I get the filecmp to work on individual 
>>files, but not on directories.
>>
>
>>example:
>>DOES NOT WORK:
>>dups=filecmp.cmpfiles("C:\comp1","c:\comp2",common)
>>print dups,"||",common
>>result:
>>([],[],[])||
>>
>
>For each filename in common, cmpfiles will compare the files
>with that basename in the two directories. What you want
>is:
> x = filecmp.dircmp("C:\\comp1","c:\\comp2")
> print x.same_files
> print x.diff_files
>
>which will figure out "common" and call cmpfiles for you.
>
>-- Gordon
>http://www.mcmillan-inc.com/
>



From merkosh at hadiko.de  Sun Jun 30 22:22:15 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Sun, 30 Jun 2002 19:22:15 -0700
Subject: arguments to functions
References: <MPG.178955a4fc8bfc1c989680@news.rz.uni-karlsruhe.de> <Xns923DC4803F1FDcliechtigmxnet@62.2.16.82>
Message-ID: <MPG.17895cabdbc8f052989681@news.rz.uni-karlsruhe.de>

In article <Xns923DC4803F1FDcliechtigmxnet at 62.2.16.82>, cliechti at gmx.net 
says...
> Uwe Mayer <merkosh at hadiko.de> wrote in 
> news:MPG.178955a4fc8bfc1c989680 at news.rz.uni-karlsruhe.de:
> > I wanted to pass one named and a collection of unnamed (an arbitrary 
> > number) to a function:
> 
> def f(*args, **kwargs):
>     	print "positional args:", args
>     	print "keyword args:", kwargs
>     	if kwargs.has_attr("expected"):
>     	    	print "Expected was set to", kwargs["expected"]
> 
> does that help somehow? :-)
> 
> chris

yes, thanks.
I think I should rather think of "expected=None" as an initialisation 
value rather than a "named parameter".

Ciao
Uwe


From jjl at pobox.com  Sun Jun  9 17:29:25 2002
From: jjl at pobox.com (John J. Lee)
Date: Sun, 9 Jun 2002 22:29:25 +0100
Subject: Efficient python programming...
In-Reply-To: <3D02FCBF.EEA954C4@engcorp.com>
References: <mailman.1023452786.27923.python-list@python.org> <3D013FB1.3D7477EA@engcorp.com> <83y9dp43d0.fsf@panacea.canonical.org> <3D02FCBF.EEA954C4@engcorp.com>
Message-ID: <Pine.SOL.4.30.0206092228010.15771-100000@mimosa.csv.warwick.ac.uk>

On Sun, 9 Jun 2002, Peter Hansen wrote:
[...]
> What tests do, however, is vastly increase your confidence in the
> correctness of the code, and when written properly (in advance
> of each small morsel of code that is written, and having been
> shown to fail without that code) they are sufficient to deliver
[...]

This point -- that tests should fail before they pass -- is easy to forget
about.  Thanks for reminding me.


John



From phd at phd.pp.ru  Tue Jun 18 04:42:05 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 18 Jun 2002 12:42:05 +0400
Subject: unicode strings and strings mix
In-Reply-To: <Pine.LNX.4.33.0206180848090.28845-100000@rnd.onego.ru>; from rnd@onego.ru on Tue, Jun 18, 2002 at 08:51:11AM +0400
References: <Pine.LNX.4.33.0206180848090.28845-100000@rnd.onego.ru>
Message-ID: <20020618124205.E17532@phd.pp.ru>

On Tue, Jun 18, 2002 at 08:51:11AM +0400, Roman Suzi wrote:
> Is there any way to define "default" encoding so the following error
> will not happen?
> 
> UnicodeError: ASCII decoding error: ordinal not in range(128)

   http://www.python.org/cgi-bin/faqw.py?req=show&file=faq04.102.htp

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From Nixdisciplenews at hotmail.com  Thu Jun 13 22:26:46 2002
From: Nixdisciplenews at hotmail.com (Felix)
Date: Thu, 13 Jun 2002 22:26:46 -0400
Subject: Scripting with Python in Xchat
Message-ID: <3D095466.4000908@hotmail.com>

Dear Community,

                   does anyone here have experience with scripting in 
Xchat (IRC client) with Python programming language? I really need some 
sort of help considering there isn't any real documentation for this 
project.

    Your help will be greatly appreciated





From phlip_cpp at yahoo.com  Sun Jun  2 22:24:11 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 03 Jun 2002 02:24:11 GMT
Subject: SNIP IT: inspectObject(thing)
Message-ID: <adek0b$3cq@dispatch.concentric.net>

Hop Nyh:

Because groups.google.com is the world's ultimate backup server, I'm 
parking this snippet here. Given most object references, it prints out an 
"instant man page" of the value of every member, the name of every local or 
inherited method, and every __doc__ string of every method.

Some of it could be replaced by calls into the inspect.py module, and one 
could extend it using that module. Put another way I wrote this before I 
learned of inspect.py.

def printDocString(thing):
    dok = None

    try:
        dok = thing.__doc__
    except: pass  #  whatEver

    if dok == None:
        dok = ''

    print '"'+dok+'"'

    return


def _showAllMethods(klassList):
    for klass in klassList:
        for z, funk in klass.__dict__.items():
            print klass.__name__+"."+z,
            printDocString(funk)

        _showAllMethods(klass.__bases__)


#
# prints out the dir() for any object in a pretty format.
#
def inspectObject(obj):
    printDocString(obj)
    klass = obj.__class__

    try:
        print "("+klass.__name__+")"
    except(AttributeError):
        pass

    printDocString(klass)

    # TODO: call inspectObject on each member?

    for z in dir(obj):
        print "."+z, repr(getattr(obj, z))

    _showAllMethods([klass])
    return

-- 
  Phlip
       http://www.greencheese.org/LucidScheming
  --  Just think: Four billion people in the
      world have never received Spam...  --



From seeger at sitewaerts.de  Wed Jun 12 06:30:44 2002
From: seeger at sitewaerts.de (Felix Seeger)
Date: Wed, 12 Jun 2002 12:30:44 +0200
Subject: Reload an imported file
References: <ae76vb$4l6iq$1@ID-79019.news.dfncis.de> <3D071FC6.5000309@mxm.dk>
Message-ID: <ae77sk$4bbl7$1@ID-79019.news.dfncis.de>

Max M wrote:

> Felix Seeger wrote:
> 
>> I have a config file. This is a normal python file.
>> Can I reimport this file during runtime ?
>> So changing the configuration will not need a restart of my program.
> 
> 
> import myModule
> reload(myModule)
> 
> regards Max M


Mhm, sorry I have another problem now ;)
I do an "from configuration import Configu"

configuration is not known and Configu is no module.

What can I do ?

thanks
have fun
Felix


From erict at millfilm.co.uk  Thu Jun  6 10:11:18 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Thu, 06 Jun 2002 15:11:18 +0100
Subject: HttpServer help!
References: <3CFF4C14.450850D8@millfilm.co.uk> <P1JL8.312164$%u2.176575@atlpnn01.usenetserver.com> <3CFF6568.7EE613CE@millfilm.co.uk> <ERJL8.313108$%u2.202230@atlpnn01.usenetserver.com>
Message-ID: <3CFF6D86.57CF7F6D@millfilm.co.uk>

req = HTTP("10.16.3.132:2112") works
Thanks You



Steve Holden wrote:

> "Eric Texier" <erict at millfilm.co.uk> wrote in message
> news:3CFF6568.7EE613CE at millfilm.co.uk...
> > The server and the client run from the same directory.
> > and the cgi file is in the same directory.
> >
> > all the following failled.
> > req.putrequest("POST", "favquote.cgi")
> > req.putrequest("POST", "/favquote.cgi")
> > req.putrequest("POST", "./favquote.cgi")
> >
> >
> > What is the meaning of  *relative to the server root*
> >
> Well, I was kind of assuming that the server was written to run CGI scripts,
> but I think on further examination I was wrong. The "server root" is the
> directopry that the server looks in when you request a document like
> http://server:port/index.html -- it's the top-level content directory of a
> web server, which may have subdirectories, and so on.
>
> From your server code:
>
> # start the server on port 2112, requires browser URLs
> # to be addressed as http://servername:2112/pathtofile
> myServer = HTTPServer(('', 2112), myRequestHandler)
>
> This has one obvious implictaion you seem to have overlooked. Your call in
> the client
>
>     req = HTTP("10.16.3.132")
>
> is trying to connect to the default port, port 80. You may need to change
> this to
>
>     req = HTTP("10.16.3.132:2112")
>
> It looks, from the code of the server, as though *any* POST request will
> result in the same output, so the path you request maybe isn't that
> important. Although I haven't read Fred's book yet, I am guessing that the
> .putrequest() method assumes you are already connected to the server, and
> that the HTTP() object will allow you to put a port number after the IP
> address.
>
> If this isn't the case, try writing an HTML page containing a form whose
> action attribute is something like
>
>     ACTION="10.16.3.132:2112/something"
>
> and see if submitting that displays a response in the browser.
>
> regards
> -----------------------------------------------------------------------
> Steve Holden                                 http://www.holdenweb.com/
> Python Web Programming                http://pydish.holdenweb.com/pwp/
> -----------------------------------------------------------------------



From peter at engcorp.com  Sun Jun 23 19:09:10 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 23 Jun 2002 19:09:10 -0400
Subject: Python hits the spot
References: <mailman.1024772248.9399.python-list@python.org> <yppfsn3dlpco.fsf@hotlips.cs.chalmers.se> <3D1634ED.B9ACF9EC@engcorp.com>
Message-ID: <3D165516.110FD94D@engcorp.com>

Peter Hansen wrote:
> 
> P?r Kurlberg wrote:
> >
> > > It's no wonder that 1000 lines of uncommented code behave like that, taking
> > > into account that you were trying to integrate several (Tcl/Tk, Fortran, C,
> > > Python) systems in a hurry, and, I presume, not knowing most of them.
> >
> > Funny you should mention this.  On my laptop (IBM Thinkpad with 128
> > megs of ram) the following two-liner gets me a lockup (i.e. keyboard
> > and mouse completely unresponsive, powertoggle needed):
> >
> > for i in range(50*1000*1000):
> >    pass
> >
> > (I am serious.)  Happened some time ago, but I never got around to
> > filing a bug report (this thread was a good reminder.)
> 
> Given that you're asking it to allocate something like 800MB of
> memory, and not all at once but in a list that grows steadily
> in size and has to be copied over and over, I'm not at all
> surprised it appears to lock up a machine with only 128MB.

Judging by Tim's reply elsewhere, and giving this a moment more
thought, I can see the comment about 'grows steadily' is false.
Of course it knows how many elements need to be allocated ahead
of time so it doesn't actually need to grow the list.

Doesn't change anything else though...

-Peter


From jbublitzNO at SPAMnwinternet.com  Tue Jun 11 18:20:25 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Tue, 11 Jun 2002 22:20:25 GMT
Subject: Installing PyQt on Mandrake 8.2
References: <3D06689A.9040807@nospam_xs4all.nl>
Message-ID: <3D06777B.7060105@SPAMnwinternet.com>

iwk wrote:
> I need to install PyQt on Mandrake 8.2 but I've been running into some 
> trouble. PyQt doesn't seem be included in the PowerPack edition which I 
> bought so I downloaded the RPM from the contributions directory on a 
> mirrorsite but that one failed because it was linked to an older version 
> of SIP. Then I tried to install from source but ./configure failed 
> because it couldn find the Python SIP modules. Apparently these ones 
> aren 't included with Mdk8.2 either.

> Has anyone been more succesfull? What should be the proper steps to get 
> PyQt installed on Mdk 8.2?


Uninstall any rpm sip versions and download the latest sip from
http://riverbankcomputing.co.uk.

Install the new sip version, and then configure, make and
install PyQt. The PyQt version (3.2.4 for example) and sip
version (also 3.2.4) have to match. Also, the location of the
sip libraries was changed in recent versions.


Jim



From zopestoller at thomas-guettler.de  Tue Jun 25 10:49:38 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Tue, 25 Jun 2002 16:49:38 +0200
Subject: Switch from 2.1 to 2.2
References: <f29c2586.0206250600.48e0a8e3@posting.google.com>
Message-ID: <3D188302.1030109@thomas-guettler.de>

Mathieu Belanger wrote:

> Hi all!
> 
> I was trying to switch my server from python 2.1 to python 2.2 when I
> began to have some compatibility problems.
> 
> I'm a newbie in python and the following code was not written by me,
> I'm only try to make it works!
> 
> 	docType = implementation.createDocumentType('','','')
> 	doc = implementation.createDocument('',None,docType)
> 	parentNode = doc.createElementNS(None,'ACTION')
> 	parentNode.setAttribute("name", "RESULT")
> 	parentNode.setAttribute("DCId", DCSystemId)
> 	doc.appendChild(parentNode)
> 	RegSection1(doc,parentNode)
> 	FileSection1(doc,parentNode)
> 	SrvSection1(doc,parentNode)
> 	StationSection1(doc,parentNode)
> 	strDoc = StringIO()
> 	PrettyPrint(doc, strDoc)
> 
> I got an error in the PrettyPrint call. This call worked well before!!
> Whats wrong now??
> 
> Thanks for your help!


Plese send the stracktrace, too

  thomas




From geoff at galtar.co.uk  Wed Jun 26 13:57:30 2002
From: geoff at galtar.co.uk (Geoff Tarrant)
Date: Wed, 26 Jun 2002 18:57:30 +0100
Subject: printing (again!!)
Message-ID: <afcusp$66r$1@newsg2.svr.pol.co.uk>

I want to use Python as the language in a Computing course that I teach.
However, I can't find a way that is relatively straightforward to send data
to a printer in a Windows environment.  If I can't find a way around this, I
will have to revert to Delphi or VB.  Please help.  I like the idea of
sending 20 or so Python programmers to University each year.

Geoff Tarrant




From sholden at holdenweb.com  Fri Jun 28 10:00:42 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 28 Jun 2002 10:00:42 -0400
Subject: [NEWBIE]- Syntax error
References: <a7YS8.727$yY2.23825@ozemail.com.au>
Message-ID: <WYZS8.108234$mF.53566@atlpnn01.usenetserver.com>

"Alistair Campbell" <campbells4 at ozemail.com.au> wrote in message
news:a7YS8.727$yY2.23825 at ozemail.com.au...
> Hi,
> I imagine the answer to this post is very simple ... but I'm not proud.
>
> Using Python2.2 on Windows OS
>
> Anyway, when I try to run a script fom the command line in IDLE or Python
>
> >>>python name_of_script.py
>
> the interpreter complains that there is a syntax error and highlights the
> "name_of _script" part.
>
> I imagine that I have not initialised some PATH variable or other. I guess
> that i need to put something in the autoexec.bat. Can anyone suggest what?
>

When you see the ">>>" prompt, the program that's currently reading your
input *is* the Python interpreter.

You should use

    python nameofscript.py

at your shell command line prompt. You *can* execute your Python code from
the ">>>" prompt with

    import nameofscript

but then I'd have to explain why this works only once, and why thereafter
you'd need to use

    reload(nameofscript)

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From fredrik at pythonware.com  Thu Jun 20 12:17:39 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 20 Jun 2002 16:17:39 GMT
Subject: module string.py
References: <3d11bd6a.11177520@news.polito.it> <mailman.1024576767.22581.python-list@python.org> <3d11f18b.24524431@news.polito.it>
Message-ID: <DenQ8.44301$n4.10331931@newsc.telia.net>

"NOPollution" <r_proietti at yahoo.com> wrote:

> Where could I find the source code of the functions implemented in
> "string.py"?

The functions implemented in string.py are implemented in
string.py (obviously).

However, where that implementation looks something like

    def lower(s):
        return s.lower()

you need to look at the implementation of the "lower" method
of whatever "s" you want to pass to that function.  Assuming
you're using strings, the implementations can be found in

    Objects/stringobject.c

or

    Objects/unicodeobject.c

depending on the string type.

To implement "lower", the latter uses Objects/unicodectype.c
which depends on data in Objects/unicodetype_db.h, which in
turn is generated by tools/unicode/makeunicodedata.py from
the Unicode database available from the Unicode Consortium.

(are you sure it's not easier to just read the library reference,
and be done with it?)

Btw, the source code can be found here (make sure to get the
tgz file, not the installers)

    http://www.python.org/2.2.1/

</F>




From pinard at iro.umontreal.ca  Thu Jun  6 16:32:19 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 06 Jun 2002 16:32:19 -0400
Subject: Efficient python programming...
In-Reply-To: <adoa9u$vqed$1@ID-146704.news.dfncis.de>
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de>
Message-ID: <oqg000i1ws.fsf@carouge.sram.qc.ca>

[Shagshag13]

> I'm looking for any state of art, optimizing techniques, tips, that a
> python beginner couldn't know by itself, but should learn and would be
> aware to write "efficient and good looking"python's code...

There is a lot of Python code all around to look it, including all the
Python which comes with the Python distribution itself.  I guess you will
quickly get used to grasp in a glimpse which code is worth looking and
which looks less interesting.

One feels when code smells written by people having good taste and
experience.  There is practical principle that people generally who do
well in many aspect of their work, which seem good to you right now,
are likely to do equally well in the aspects which are more new to you.

Also, on the Python site, there are a few essays written by Guido van
Rossum about how one writes and optimises, and on Python in general,
and which I read soon after starting to learn Python.  I found these to
be well written, quite instructive, and very welcome at the time.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard





From mikeb at mitre.org  Sat Jun  1 20:35:01 2002
From: mikeb at mitre.org (Mike Brenner)
Date: Sat, 01 Jun 2002 20:35:01 -0400
Subject: File Iterator - Print - Double New Lines
Message-ID: <3CF96835.14EB1C61@mitre.org>

Hi Peter,

That is not what happens, though. The first three outputs (from the first call to reader) do not get newlines. The second three outputs (from the second call to reader) get double newlines. The explanation that "the print statement adds newlines after every lines it prints" would explain the double newlines the second time, but it does not explain the lack of newlines the first time. Did you run it on a windows machine?

Mike


Peter wrote:
Mike, the print statement adds newlines after every line it prints.
If you want to see the real contents of the file, use sys.stdout.write().
Alternatively, you might want to strip the newlines from each line
as you read them in.  Depends what you are trying to do.



MIKE's OUTPUT (CONTAINING HIS QUESTION):

QUESTION: How to get SINGLE SPACING?
Using Python 2.2.1, it does the same under win32all build 142 and
      running it from a DOS box without the windows extensions.
trying it without the \n
abc
trying it with the \n
a

b

c


MIKE's PROGRAM:

print "QUESTION: Why did it jump from NO LINE SPACING to DOUBLE SPACING?"
print "QUESTION: How to get SINGLE SPACING?"
print
print "Using Python 2.2.1, it does the same under win32all build 142 and
"
print "      running it from a DOS box without the windows extensions."

def writer(list):
    f=open("tmp.tmp","w")
    for item in list:
        f.write(item)
    f.close()

def reader():
    f=open("tmp.tmp","r")
    for line in f:
        print line
    f.close()

print "trying it without the \\n"
writer(["a","b","c"])
reader()

print "trying it with the \\n"
writer(["a\n","b\n","c\n"])
reader()abc
trying it with the \n
a

b

c





From akakakhel at attbi.com  Sat Jun 29 20:19:29 2002
From: akakakhel at attbi.com (Ali K)
Date: Sun, 30 Jun 2002 00:19:29 GMT
Subject: geting python files for use on web
References: <k6oS8.191997$6m5.159601@rwcrnsc51.ops.asp.att.net> <Xns9239DCB2C57A0cliechtigmxnet@62.2.16.82>
Message-ID: <l8sT8.231729$6m5.196131@rwcrnsc51.ops.asp.att.net>

how do i install the files? Do I have the people from which i am getting my
free space install the files?  Do I just upload them onto my freespace?




From pan-newsreader at thomas-guettler.de  Sun Jun 16 18:18:12 2002
From: pan-newsreader at thomas-guettler.de (Thomas Guettler)
Date: Mon, 17 Jun 2002 00:18:12 +0200
Subject: GdkImLib: Image(filename) seems to cache
Message-ID: <pan.2002.06.17.00.18.11.698272.716@thomas-guettler.de>

Hi!

I use GdkImLib.Image(filename) to load and display jpeg files:

img=GdkImLib.Image(filename)
#(1)... rotate filename with shell script calling image-magic
img=GdkImLib.Image(filename)

but the image seems to be cached. The second time I load
it, it is not rotated.

It even gets stranger: If I add img=None at (1) the image
gets rotated if I rotate it 90 or 270 degrees. If it gets
rotated 180 I still see the old picture.

Does someone know why the images seems to be cached

 thomas


From sameer_ at email.com  Wed Jun  5 11:36:06 2002
From: sameer_ at email.com (sameer)
Date: 5 Jun 2002 08:36:06 -0700
Subject: constructive critisism on pythonwin IDE
Message-ID: <6bd9f01b.0206050736.f469082@posting.google.com>

First of all, let me start of by saying that Pythonwin is a great IDE.
 Now for the constructive critisism.  The thing I hate about PythonWin
is, that despite the fact that it's a development environment, there
is a lot of caching done of modules, and the cached copy is not
checked against a potentialy modified copy.  Whenever I change the
contents of a file and try to run a module that depends on that file,
I always end up with the old copy.  I have to then import and then
reload the changed module in the interactive window for it to refresh.
 It would greatly increase my productivity, if I can specify which
modules can be cached.  I understand that there is a speed and
development time issue, but I would like a little more fine grained
control on how caching works on PythonWin, as the way it's currently
implemented now is not acceptable.


From krissepu at vip.fi  Thu Jun 20 13:33:04 2002
From: krissepu at vip.fi (Pekka Niiranen)
Date: Thu, 20 Jun 2002 20:33:04 +0300
Subject: Brackets in brackets, braces in braces
Message-ID: <3D1211CF.59CBDC1B@vip.fi>

It is possible to use regular expressions to match nested brackets, if
you FIX the amount of nesting:

>>> pattern = re.compile("(\?[^?!]+(\?[^?!]+\!)*[^?!]+\!)")
>>> Line = "?AA?BB!CC!?DD!ee?EE!ff?FF?GG!HH!"
>>> print re.findall(pattern, Line)
[('?AA?BB!CC!', '?BB!'), ('?DD!', ''), ('?EE!', ''), ('?FF?GG!HH!',
'?GG!')]

In above I am searching strings limited by ?- and ! -characters.
Only one level of nesting is allowed.

The only problem I find is that re.findall() returns empty matches too.
Maybe in next version of SRE modified findall is introduced. ;)

-pekka-

PS. How about searching line = "??BB!CC!"   ? ;)



From shagshag13 at yahoo.fr  Wed Jun  5 11:58:36 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Wed, 5 Jun 2002 17:58:36 +0200
Subject: List containing objects and subclassing...
Message-ID: <adlcer$6lqd$1@ID-146704.news.dfncis.de>

Hello,

I need an advice, i have a python list containing only object "node", and i'm wondering if this
is best to hold it in a python list (as by now) or if i should write a "nodes" object subclassing from
list and implementing some other methods i need to process on node object (contained by this list)
and that i do now from "outside", or if i should write my own container implementing __getitem__
(i also have done that in another version)...

...
[hash_i] --> [node_j] -> [node_k]...
...

Which way is better OO practice ? which is the efficient one ?
What are the pros and cons (maybe best oo vs efficient ?) ?

Thanks,

s13.








From dman at dman.ddts.net  Thu Jun 27 13:55:20 2002
From: dman at dman.ddts.net (Derrick 'dman' Hudson)
Date: Thu, 27 Jun 2002 12:55:20 -0500
Subject: How to find out DNS ?
In-Reply-To: <3D1B6186.3142.295959@localhost>
References: <3D1B6186.3142.295959@localhost>
Message-ID: <20020627175520.GA3547@dman.ddts.net>

On Thu, Jun 27, 2002 at 07:03:34PM +0200, A wrote:

| Is there a way how to find out, from Python , what primary or 
| secondary DNS I use when connecting to internet?

f = open( "/etc/resolv.conf" , "r" )
nameservers = []
for line in f.xreadlines() :
    line = line.strip()
    if line.startswith( "nameserver" ) :
        _ , ip = line.split()
        nameservers.append( ip )
f.close()
for ns in nameserver :
    print "nameserver" , ns


Of course, this only works on a UNIX-like system where
/etc/resolv.conf is used to list the nameservers.  A common setup (on
unix systems) is for some entries to be in /etc/hosts which takes
precedence over DNS lookups.  Any given application can also do its
own thing too, if it wants.  For example, the 'host' command can be
told to contact any arbitrary host and submit a DNS query to it.  Some
setups will have LDAP (or maybe NIS) also serving host information, in
which case DNS isn't involved for those hosts.

Why do you want to know what the nameserver(s) are?  Usually it is
better to not know that and just let the underlying libresolv handle
name resolution for you.  (program to the interface, not the
implementation ...)

-D

-- 

Microsoft is to operating systems & security ....
                                     .... what McDonald's is to gourmet cooking
 
http://dman.ddts.net/~dman/

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 248 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020627/464df0a4/attachment.sig>

From fperez528 at yahoo.com  Wed Jun 19 20:00:14 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 19 Jun 2002 18:00:14 -0600
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <um7Q8.44210$n4.10292249@newsc.telia.net> <aer1gr$fj1$1@peabody.colorado.edu> <mailman.1024527625.31621.python-list@python.org>
Message-ID: <aer5um$h8m$1@peabody.colorado.edu>

Gerhard H?ring wrote:

>> I've always been curious. Why isn't the standard tutorial linked to here?
>> It seems like the obvious first link to have, no?
> 
> Because it's not well suited for programming newbies?

Mmmhh, I'd disagree with that. Considering that it introduces all the data 
types and control structures one by one, nicely and slowly, I'd imagine many 
programming newbies could actually pick it up fairly well. Granted, by the 
time I read it I had been programming for many years and python was like my 
10th language, so I have a hard time judging it from a 'programming newbie' 
perspective. But I honestly think it's clear and slow-paced enough to be a 
good starting point for any newbie.

It could perhaps be included there with a simple comment like 'This tutorial 
may be a bit difficult to follow if you have zero previous programming 
experience in other languages. But if you know the basic concepts of computer 
programming, it will get you up to speed in python in half an afternoon.'

Cheers,

f.


From steve at ferg.org  Sun Jun 23 14:03:47 2002
From: steve at ferg.org (Stephen Ferg)
Date: 23 Jun 2002 11:03:47 -0700
Subject: a Tree data-structure class?
References: <b16e4ef7.0206210510.77a956b6@posting.google.com> <slrnah6chh.154.gerhard@lilith.my-fqdn.de>
Message-ID: <b16e4ef7.0206231003.4add011f@posting.google.com>

Thanks!


From candiazoo at attbi.com  Thu Jun 13 21:18:42 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Fri, 14 Jun 2002 01:18:42 GMT
Subject: ADO Command object/Parameters problem...
Message-ID: <3d09426d.313465031@netnews.attbi.com>

I am still having problems using dynamic sql statements to perform lookups and
inserts into a MS Access database and an Oracle database.  I've looked up the
error at Microsoft and the solution is to find the OLEDB provider and add
"OLEDB_SERVICES" DWORD value to the registry entry... did this (actually, it
already HAD it)... no go.  I used MakePy... no go.  The code I used works in a
VB application.  I used the errors collection from the connection object and it
only gives me a more concise version of what you see below.  Does anyone have
any idea why I cannot create a command object with parameter objects??  I am
using only adVarChar and adNumeric values... no dates, no blobs, nothing odd at
all...!!!

Has anyone else run into this (note: I have seen other entries for the same
error, which is rather generic, but not for my specific reason).

Thank you, in advance...

Mike J.

-----------------------------------------------------------------------------------------------------------------------------------------
begin updating recType[s] on Thu Jun 13 12:15:39 2002
Traceback (most recent call last):
  File "C:\Documents and Settings\mjessop\My Documents\Python Development\Catalo
g\BuildCatalog.py", line 1300, in ?
    setRecordType()
  File "C:\Documents and Settings\mjessop\My Documents\Python Development\Catalo
g\BuildCatalog.py", line 506, in setRecordType
    rstOracle, status = cmdOracle.Execute()
  File "C:\Python22\Lib\site-packages\win32com\gen_py\00000205-0000-0010-8000-00
AA006D2EA4x0x2x5.py", line 1698, in Execute
    return self._ApplyTypes_(0x5, 1, (9, 0), ((16396, 18), (16396, 17), (3, 49))
, 'Execute', '{00000556-0000-0010-8000-00AA006D2EA4}',RecordsAffected, Parameter
s, Options)
  File "C:\Python22\Lib\site-packages\win32com\client\__init__.py", line 341, in
 _ApplyTypes_
    return self._get_good_object_(apply(self._oleobj_.InvokeTypes, (dispid, 0, w
Flags, retType, argTypes) + args), user, resultCLSID)
pywintypes.com_error: (-2147352567, 'Exception occurred.', (0, 'Microsoft OLE DB
 Provider for ODBC Drivers', 'Multiple-step OLE DB operation generated errors. C
heck each OLE DB status value, if available. No work was done.', None, 0, -21472
17887), None)

HERE is an example of what I am doing... (note, I added the long() typecasting
to make sure the value was coerced into a long integer)...

sql = "select pty.orgname " + \
      "from ccc_party pty, ccc_right_rightsholder rgh, ccc_right rgt " + \
      "where pty.pty_inst = rgh.rgh_inst " + \
      "and rgh.rgt_inst = rgt.rgt_inst " + \
      "and rgt.rgt_inst = ?"

cmdOracle.ActiveConnection = cnnOracle
cmdOracle.CommandType = 1
cmdOracle.CommandText = sql

trsInst = long(rstTempDB.Fields("trs_inst").Value)
aasInst = long(rstTempDB.Fields("aas_inst").Value)
anyInst = trsInst or aasInst
        
cmdOracle.Parameters.Append(cmdOracle.CreateParameter("orgname", 200, 2, 80))
cmdOracle.Parameters.Append(cmdOracle.CreateParameter("rgt_inst", 3, 1, 4,
long(anyInst)))
rstOracle, status = cmdOracle.Execute()


From bt98 at doc.ic.ac.uk  Sun Jun  2 05:55:21 2002
From: bt98 at doc.ic.ac.uk (Benjamin Tai)
Date: Sun, 02 Jun 2002 10:55:21 +0100
Subject: Pop a list from beginning ? and memory saving...
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de>
Message-ID: <3CF9EB89.7CA7204B@doc.ic.ac.uk>

Shagshag13 wrote:

> and is there a way to pop from the beginning  of the list ? (i think i could
> use reverse but
> i don't know if this memory consuming)
>
> >>> a = [1,2,3,4,5,6,7]
> >>> while 1:
>  try:
>   f = a.pop()
>  except IndexError:
>   break
>  print f
>
> thanks,
>
> s13.

You could try a.pop(0), which pop the first instance of the list, like a FIFO.

Thanks

Ben



From Alexandre.Fayolle at logilab.fr  Thu Jun  6 09:58:35 2002
From: Alexandre.Fayolle at logilab.fr (Alexandre)
Date: Thu, 6 Jun 2002 15:58:35 +0200
Subject: [ANN] constraint-0.1
Message-ID: <20020606135835.GF26159@orion.logilab.fr>

Hello,

Logilab has released constraint-0.1.

The constraint package is an extensible Constraint Satisfaction Problem 
(CSP) solver using constraint propagation algorithms written in 100% 
pure Python (tested with Python 2.1)

You can download it from ftp://ftp.logilab.org/pub/constraint, or read a
bit more about it on http://www.logilab.org/python-logic/constraint.html, 
and discuss it on the logic-sig mailing list (see
http://lists.logilab.org/mailman/listinfo/python-logic to subscribe)

The package is still young, but we find it useful enough right now, so I 
thought I might as well publish it.

What can it do?
 * solve some finite-domain CSP
 * handle n-ary constraints

How can you extend it:
 * by implementing new Domain classes
 * by implementing new constraint classes
 * by implementing new Distributor classes
       
How well does it perform?
 * 8-queens problems solved in 15 seconds on my 1GHz PC
 * SEND+MORE=MONEY problem solved in 8 seconds on the same machine

If you're interested by constraint propagation programming in python,
please join the logic-sig mailing list. 
 
Alexandre Fayolle
-- 
LOGILAB, Paris (France).
http://www.logilab.com   http://www.logilab.fr  http://www.logilab.org
Narval, the first software agent available as free software (GPL).




From ifls at rz.tu-clausthal.de  Mon Jun  3 08:19:32 2002
From: ifls at rz.tu-clausthal.de (Lutz Schroeer)
Date: 3 Jun 2002 12:19:32 GMT
Subject: pygoogle search restricted by date
Message-ID: <adfmsk$205i$1@ariadne.rz.tu-clausthal.de>

Hi Pythonistas,

The Goolge API doc tells me that I can restrict my search to a specific date
range using "daterange:<date1>-<date2>" where date1 and date2 are Julian Day
Numbers. I put the term in the restrict part auf the doGoogleSearch function
(-> code snippet). Nevertheless I'm getting the same results.

Any hints?

Lutz

<--snip-->

import google
from mx.DateTime import *

google.setLicense('<mykey>')
today = int(now().jdn)
yesterday = today - 1

restrict_query = 'daterange:%d-%d' % (yesterday, today)

data = google.doGoogleSearch('python', restrict = restrict_query)
for x in data.results:
  print x.title, x.URL

-- 
  



From maxm at mxm.dk  Mon Jun 17 03:37:00 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 17 Jun 2002 09:37:00 +0200
Subject: I feel stoopid ... this code is to verbose
References: <3D0A1791.3060307@mxm.dk> <871yb922el.fsf@bunty.ruud.org>
Message-ID: <3D0D919C.1010101@mxm.dk>

ruud de rooij wrote:
> Max M <maxm at mxm.dk> writes:

> import locale
> locale.setlocale(locale.LC_NUMERIC, 'da_DK')
> locale.format('%.2f', 42000000.00, 1)
> 
> batteries-included-ly y'rs,


Yes this was actually one of the first things I tried, but I got a bit 
confused by the documentation::

"setlocale() is not thread safe on most systems. Applications typically 
start with a call of::

     import locale
     locale.setlocale(locale.LC_ALL, '')

This sets the locale for all categories to the user's default setting 
(typically specified in the LANG environment variable). If the locale is 
not changed thereafter, using multithreading should not cause problems.
"

So I worried that it would wreck havock with my application as I am 
using IIS, and I worried that the locale would be out of sync on 
different pages.

My misunderstanding was seeded by me reading the docs wrong. I though 
that if I set the locale to 'danish' that the built in functions str(), 
atof() etc. would change behind the scene to using different locales.

But off course it is only if I do something like::

     from locale import str
     locale.setlocale(locale.LC_NUMERIC, 'da_DK')
     str('%.2f', 42000000.00, 1)

that locale is used on str().

If I just do the below, I get no problems::

     import locale
     locale.setlocale(locale.LC_NUMERIC, 'da_DK')
     locale.str('%.2f', 42000000.00, 1)

So I worried for no reason, and will use the batteries instead of my 
home cooked function.

regards Max M



From martin at strakt.com  Mon Jun 17 09:56:02 2002
From: martin at strakt.com (Martin =?iso-8859-1?Q?Sj=F6gren?=)
Date: Mon, 17 Jun 2002 15:56:02 +0200
Subject: non blocking socket.accept()
In-Reply-To: <qeokea.b6q.ln@neo.fdns.net>
References: <qeokea.b6q.ln@neo.fdns.net>
Message-ID: <20020617135602.GA24002@strakt.com>

On Mon, Jun 17, 2002 at 01:45:05PM +0000, Zsolt Marx wrote:
> I far as I know socket.accept() waits for a client
> to ask for a connection with the server.
> 
> This command seems to be blocking, though I would
> like to limit the wait time onto a predefined interval
> and to throw an error, if there is no connection.
> 
> How can I achieve this?

You will probably be able to do this with the new timout sockets, but in
the meantime, check out the select module:

import select
if select.select([serversock], [], [], timeout):
    client,addr = serversock.accept()


Regards,
Martin

-- 
Martin Sj?gren
  martin at strakt.com              ICQ : 41245059
  Phone: +46 (0)31 7710870       Cell: +46 (0)739 169191
  GPG key: http://www.strakt.com/~martin/gpg.html




From hjwidmaier at web.de  Fri Jun 21 04:10:08 2002
From: hjwidmaier at web.de (Hans-Joachim Widmaier)
Date: 21 Jun 2002 01:10:08 -0700
Subject: Advice sought on ps module
References: <6e990e29.0206200325.355f0413@posting.google.com> <3D11BE17.C6AE67A@ipm.fhg.de>
Message-ID: <6e990e29.0206210010.26d2b887@posting.google.com>

Markus von Ehr <markus.vonehr at ipm.fhg.de> wrote in message news:<3D11BE17.C6AE67A at ipm.fhg.de>...
> Hi,
> 
> do you know the PSDraw Module in the PIL?
> http://www.pythonware.com/library/pil/handbook/psdraw.htm
> Maybe this is a solution for you.

I've looked at it, of course. But it provides nothing I can't do today
(granted, my ps module lacks most of PSDraw's functions -- drawing
lines and rectangles is easily done directly; I was more concerned
with the text and font side).

Looks like I do have to write lots of wrappers and start second
guessing. ;-(

Hans-Joachim


From kragen at pobox.com  Sun Jun  2 23:54:09 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 02 Jun 2002 23:54:09 -0400
Subject: Generating unique numbers?
References: <3CF7EBC8.4080802@verio.net> <3CF8FC1C.7060903@lindbergs.org> <adav4g$o2b$1@panix1.panix.com>
Message-ID: <834rglrp9a.fsf@panacea.canonical.org>

aahz at pythoncraft.com (Aahz) writes:
> No, if you want to trade IDs across a network, you need GUIDs (DCE
> UUIDs).  There's just no way around that.  It's not the most perfect
> system in the world, but it's better than any other generally available
> scheme.

URLs are an alternative to UUIDs that have some advantages.  They tend
to be shorter, more human-readable, more memorable, less
privacy-invasive, and there's a decentralized scheme for attaching
arbitrary data to URLs you own, specifically, the Web.

Their disadvantage is that they are reusable, i.e. that the same URL
may designate two different objects at different times.



From max at alcyone.com  Wed Jun 12 14:20:42 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 12 Jun 2002 11:20:42 -0700
Subject: Reload an imported file
References: <ae76vb$4l6iq$1@ID-79019.news.dfncis.de>
Message-ID: <3D0790FA.B730A9BE@alcyone.com>

Felix Seeger wrote:

> I have a config file. This is a normal python file.
> Can I reimport this file during runtime ?
> So changing the configuration will not need a restart of my program.

Try the reload builtin.  Note, though, that this will not change the
state of objects already created, and it won't help if you did from
module import *.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Who'd ever think it / Such a squalid little ending
\__/ The American and Florence, _Chess_
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From sholden at holdenweb.com  Wed Jun 26 11:49:59 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 26 Jun 2002 11:49:59 -0400
Subject: Blank results using mxODBC against 
References: <WS0Q8.7228$Tu6.481485@typhoon.austin.rr.com>
Message-ID: <3olS8.46473$mF.2290@atlpnn01.usenetserver.com>

"Sandeep Gupta" <sandeep182 at hotmail.com> wrote ...
> We are trying to connect to MS SQL Server from Linux, but we are only
> receiving blank values in the query results.  The number of rows and
columns
> in the result is correct, but each result cell is blank.
>
> As a test, we successfully used mxODBC against a local MySQL going through
> unixODBC.
>
> We then tried: Python2.2 --> mxODBC --> unixODBC --> inline TDS 1.6 --> MS
> SQL Server
> When I use isql with unixODBC, I am able to view the results of a simple
> select.  That same select in mxODBC returns blank result values.
>
> The python code I'm using is:
> import mx.ODBC.unixODBC
> db = mx.ODBC.unixODBC.DriverConnect('DSN=MSSQL')
>
> c = db.cursor()
>
> c.execute('select * from dbo.Source;')
>
> mx.ODBC.print_resultset(c)
>
[other information]


Try running the mxODBC test program on the data source you are expecting to
use, ands see if it returns any errors.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From cliechti at gmx.net  Thu Jun 20 16:54:58 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 20 Jun 2002 22:54:58 +0200
Subject: How to represent the infinite ?
References: <3d11ee9b$1@news.swissonline.ch> <20020620193230.53d58f8f.christophe.delord@free.fr> <3D121A29.A733609F@noaa.gov>
Message-ID: <Xns9233E98B5585cliechtigmxnet@62.2.16.82>

Chris Barker <Chris.Barker at noaa.gov> wrote in
news:3D121A29.A733609F at noaa.gov: 

> Christophe Delord wrote:
>> There is another simple solution. Just use a float that is bigger
>> than any 64-bit float. For example 10^1000. This float has a special
>> encoding meaning +oo ! 
> 
> this is part of the IEEE 754 standard for floating point computation
> (http://cch.loria.fr/documentation/IEEE754/) It is not guaranteed to
> be implimented by Python (someone please correct me if I'm wrong).
> whether this works is a function of the machine, compiler and library
> used to compile Python. I'm not sure how common it is, but if you want
> your code portable, it may not be reliabel on all machines, which is
> too bad, because IEEE 754 is pretty darn good standard.
> 
> BTW: I'd love to see literals for Inf, -Inf and NaN in Python.

do it yourself ;-)
>>> Inf = struct.unpack('f', '\x00\x00\x80\x7f')[0]
>>> Inf
1.#INF
>>> -Inf
-1.#INF
>>> NaN = struct.unpack('f', '\x7f\xff\xff\x7f')[0]
>>> NaN
1.#QNAN

don't know if it works anywhere but it does on Py2.2 win32.

BTW:
>>> struct.unpack('>f', '\x00\x00\x00\x01')
(1.4012984643248171e-045,)
>>> struct.unpack('<f', '\x00\x00\x00\x01')
(2.350988701644575e-038,)
>>> struct.unpack('@f', '\x00\x00\x00\x01')
(2.350988701644575e-038,)

so the format code '<' is same as '@', little endian which is correct on my 
intel P3 but:

>>> struct.unpack('@f', '\x00\x00\x80\x7f')
(1.#INF,)
>>> struct.unpack('<f', '\x00\x00\x80\x7f')
(3.4028236692093846e+038,)

huh? they're no longer the same for special values of floats??

chris

-- 
Chris <cliechti at gmx.net>



From steve at lurking.demon.co.uk  Tue Jun  4 18:52:41 2002
From: steve at lurking.demon.co.uk (Stephen Horne)
Date: Tue, 04 Jun 2002 23:52:41 +0100
Subject: y=if x>High: 2 elif x<Low: 1 else: 0
References: <315ea996.0205221339.1976f50e@posting.google.com> <3CEC15F0.DF78A4F@earthlink.net> <ach5nb$ubs$1@slb2.atl.mindspring.net> <3CEC1D26.AF0CE2FB@earthlink.net> <Pine.OSX.4.43.0205221906390.17977-100000@d-128-61-180.bootp.virginia.edu>
Message-ID: <qhcqfu8bji1b8ltialjhl5qriqoiolrr5m@4ax.com>

On Wed, 22 May 2002 19:12:10 -0400, Steven Majewski
<sdm7g at Virginia.EDU> wrote:

>I prefer:
>
>>>> ( 2, 1, 0 )[ [ x > hi, x < low, 1  ].index(1)]
>
>as more readable, although the lambda functional version gets a bit
>more baroque:
>
>>>> (lambda x: (2,1,0)[(lambda x: [x > hi, x < low, 1])(x).index(1)])(arg)

Hmmm

I occasionally use the following little trick...


def IF(p_Condition, p_True, p_False=lambda: None) :
  """
  Expects parameterless functions or lambdas for the p_True
  and p_False arguments
  """
  if p_Condition : return p_True ()
  return p_False ()

#...

print IF(x != 0, lambda: 5 // x)


This could easily be adapted...

print IF(x>High, lambda:2, IF(x<Low, lambda:1, lambda:0))


and if multi-condition patterns are extensively used...

def FIRST (p_Tests, p_Default) :
  """
  p_Tests : a sequence of 2-tuples, each containing
            0 : A parameterless function for the test
            1 : A parameterless function for the value
  p_Default : A parameterless function for the default
              result if all tests fail.

  Example...

  FIRST( [(lambda: x<Low,  lambda: 1),
          (lambda: x>High, lambda: 2) ], lambda: 0 )

  """
  for l_Test,l_Value in p_Tests :
    if l_Test () :
      return l_Value ()
  return p_Default ()

  if p_Lo_Test () : return p_Lo_Result ()
  if p_Hi_Test () : return p_Hi_Result ()
  return p_Mid_Result ()


Actually, I prefer to use 'None' as the fail case in this pattern so
that the test and result can be bound in one function.

def FIRST (p_Functions) :
  """
  p_Functions : a sequence of parameterless functions,
                each either returning a valid result or
                None.

  Example...

  FIRST( [(lambda: x<Low  and 1 or None),
          (lambda: x>High and 2 or None),
          (lambda: 0                   ) ] )

  """
  for l_Function in p_Functions :
    l_Result = l_Function ()
    if l_Result != None : return l_Result
  return None


Personally, I think there should be an if 'function' built in which
has similar semantics to the IF function above but without needing the
lambdas. It would be nice if it also supported more than three
parameters - each pair giving an 'if' or 'elif' condition plus result
and the final odd parameter giving the 'else' result.

This would allow something like...

  if (x < Low : 1, x > High : 2, 0)



From bl at netgroup.dk  Mon Jun  3 08:52:23 2002
From: bl at netgroup.dk (Bo Lorentsen)
Date: 03 Jun 2002 14:52:23 +0200
Subject: Embedding and threads ?
In-Reply-To: <j4wutg8shk.fsf@informatik.hu-berlin.de>
References: <mailman.1022667286.17372.python-list@python.org>
	<j44rgrnkfm.fsf@informatik.hu-berlin.de>
	<mailman.1022740196.32586.python-list@python.org>
	<j4d6vdwvyk.fsf@informatik.hu-berlin.de>
	<1022771667.689.136.camel@netgroup>
	<j4znyhzguh.fsf@informatik.hu-berlin.de> <1022830475.893.17.camel@netgroup>
	<j4znygd7iv.fsf@informatik.hu-berlin.de> <1022839266.823.66.camel@netgroup>
	<j4d6vac9b7.fsf@informatik.hu-berlin.de>
	<mailman.1023091388.3109.python-list@python.org> 
	<j4wutg8shk.fsf@informatik.hu-berlin.de>
Message-ID: <1023108743.685.88.camel@netgroup>

On Mon, 2002-06-03 at 14:19, Martin v. L?wis wrote:
> 
> So under normal conditions, on a single processor, with enough
> runnable Python threads: the processor will always be busy, and all
> threads will make progress. You can hardly ever tell the difference to
> free threading.
Ok, I think my problem is that I try to model my code around a thread
model (to keep things clear), and I got confussed about the way Python
did things, as I somehow assumed it had a more strit forward thread
model (ala : 1 thread = 1 interpreter state).

I will use Python for this project as long as what I do only need a
single thread model, and when I need more than that I'll se what to do
about it :-)

/BL





From ethanw at rediffmail.com  Sat Jun  1 21:20:12 2002
From: ethanw at rediffmail.com (Ethan)
Date: 1 Jun 2002 18:20:12 -0700
Subject: GUI Toolkit!
Message-ID: <1a52c7f1.0206011720.7c3ac7ea@posting.google.com>

Hi,
  I need to do some GUI programming with Python. I found that there
are quite a few options available for the toolkits . Apart from
TkInter, I have seen bindings for Qt, wxWindows, FOX and FLTK. I want
to pick one with the best open visual tools available. From what I
have looked I am really impressed with the screen shots of Boa
Constructor. Apparently it even generates the event procs. But I doubt
that I have seen everything. Can anyone please tell me about any other
options that I might have. I am developing on Windows.

Thanks


From kzaragoza at attbi.com  Mon Jun 17 12:45:31 2002
From: kzaragoza at attbi.com (Kris J. Zaragoza)
Date: 17 Jun 2002 11:45:31 -0500
Subject: Python and SQL Server 2000
References: <ZCuO8.10006$9b.678135@typhoon.austin.rr.com> <mailman.1024112300.2184.python-list@python.org> <slrnagn75g.263.kzaragoza@localhost.localdomain> <1bf96e7f.0206170345.515227c1@posting.google.com>
Message-ID: <slrnags4ob.cie.kzaragoza@localhost.localdomain>

In article <1bf96e7f.0206170345.515227c1 at posting.google.com>, Dave Moor wrote:
> kzaragoza at attbi.com (Kris J. Zaragoza) wrote in message news:<slrnagn75g.263.kzaragoza at localhost.localdomain>...
>> So, let me add to the original poster's question: Has anyone ever
>> tried running the Windows version of Python (either the main
>> distribution or ActiveState's) under Wine to take advantage of the
>> database drivers, etc.?
> 
> Another solution is to write a simple pair of client/server socket
> apps.  The server is passed and executes sql on the windows using COM
> or ODBC.  Then pass the results using the struct module back to the
> client.
> 
> Dave

I don't like using this kind of relay to expose a database.  There are
too many security issues to address.  You end up having to roll your
own authentication and authorization mechanisms, figure out whether
masquerade database users or use the credentials given, and figure out
how best to encrypt data if appropriate.  Too much work, and far too
error prone if alternatives are available.  That's why I asked about
the feasibility of using Wine with the Windows version of Python.
Although it's a less than ideal solution, it could work in a pinch.
Unfortunately, I haven't had the time to try it out myself. :-(

Kris


-- 
Kris J. Zaragoza       | On the face of it, Microsoft complaining about
kzaragoza at attbi.com    | the source license used by Linux is like the
                       | event horizon calling the kettle black.
                       | -- Adam Barr, article on kuro5hin.org


From theller at python.net  Mon Jun 17 09:34:02 2002
From: theller at python.net (Thomas Heller)
Date: Mon, 17 Jun 2002 15:34:02 +0200
Subject: pychecker + XEmacs
References: <3d0de483$0$3881$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <aekok8$7p118$1@ID-59885.news.dfncis.de>

"Boudewijn Rempt" <boud at valdyas.org> wrote in message news:3d0de483$0$3881$e4fe514c at dreader4.news.xs4all.nl...
> Has anyone succeeded in combining pychecker and XEmacs (or emacs)? I'm
> checking a large project I began in 1999, and which is full of the most
> atrocious errors, and I'd like to click on the line that mention the error
> and be transported there, just like when you run a python program from
> XEmacs and get a stacktrace.
>
> --
> Boudewijn Rempt | http://www.valdyas.org

python-mode.el from CVS contains pychecker support.
C-c C-w runs 'pychecker --stdlib <current-python-file>',
and shows the results in a *compilation* window.
Then you can M-x next-error or M-x previous-error
to your hearts content.

Thomas




From garth at deadlybloodyserious.com  Sun Jun 16 18:57:52 2002
From: garth at deadlybloodyserious.com (Garth T Kidd)
Date: 16 Jun 2002 15:57:52 -0700
Subject: MS Word -- finding text
References: <mailman.1024247060.12127.python-list@python.org>
Message-ID: <4c877253.0206161457.a2289c8@posting.google.com>

> To avoid this problem, I use the following conversion routine. 
> After making the necessary check for None, it attempts a quick 
> conversion str() first. When necessary, it slowly goes through 
> each character, handling the exceptions that are raised. 

Great technique! Now, I don't suppose you could identify the
characters str() isn't handling, so that someone can fix str()? :)

Regards,
Garth.


From mcfletch at rogers.com  Thu Jun  6 22:22:45 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Thu, 06 Jun 2002 22:22:45 -0400
Subject: Problem pickling class instance
References: <3D000D77.7040008@erols.com>
Message-ID: <3D0018F5.4000904@rogers.com>

The name of the top-level script's module is __main__, so your instance 
is getting stored with the class name __main__.A .  The __main__ module 
of your second script doesn't include an A class, so the pickler can't 
find the "A" class to use when it tries to load the instance.  Put A in 
it's own module, and always import it explicitly.

HTH,
Mike

Edward C. Jones wrote:
> I think the following problem may be discussed in the Python Library
> Reference, section 3.14.4 but I don't understand it.
> 
> Here is the file mod.py:
> 
>     import pickle
> 
>     def Read():
>         f = open('saved', 'r')
>     return pickle.load(f)
> 
>     class A:
>         def __init__(self):
>             pass
> 
>     if __name__ == '__main__':
>         a = A()
>     f = open('saved', 'w')
>         pickle.dump(a, f)
> 
> and the file top.py:
> 
>     import mod, pickle
> 
>     a = mod.Read()
> 
> First I run mod.py then top.py. The latter give the message:
> 
>     AttributeError: 'module' object has no attribute 'A'
> 
> Please explain the problem to me. Is there a standard work-around?
> 
> Thanks,
> Ed Jones
> 


-- 
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From peter at engcorp.com  Sun Jun 23 16:51:57 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 23 Jun 2002 16:51:57 -0400
Subject: Python hits the spot
References: <mailman.1024772248.9399.python-list@python.org> <yppfsn3dlpco.fsf@hotlips.cs.chalmers.se>
Message-ID: <3D1634ED.B9ACF9EC@engcorp.com>

P?r Kurlberg wrote:
> 
> > It's no wonder that 1000 lines of uncommented code behave like that, taking
> > into account that you were trying to integrate several (Tcl/Tk, Fortran, C,
> > Python) systems in a hurry, and, I presume, not knowing most of them.
> 
> Funny you should mention this.  On my laptop (IBM Thinkpad with 128
> megs of ram) the following two-liner gets me a lockup (i.e. keyboard
> and mouse completely unresponsive, powertoggle needed):
> 
> for i in range(50*1000*1000):
>    pass
> 
> (I am serious.)  Happened some time ago, but I never got around to
> filing a bug report (this thread was a good reminder.)  

Given that you're asking it to allocate something like 800MB of
memory, and not all at once but in a list that grows steadily 
in size and has to be copied over and over, I'm not at all 
surprised it appears to lock up a machine with only 128MB.
You don't need the for loop for this to work either: just
try "biglist = range(50000000)".  It took several minutes for
my 96MB Windows 98 machine to manage even 10 million items,
let alone 50 million!

-Peter


From peter at engcorp.com  Mon Jun 24 20:57:05 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 24 Jun 2002 20:57:05 -0400
Subject: SOLVED (Re: Python hits the spot)
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at> <af4vrl$ffe$0@216.39.172.122> <3D16DD5C.1050708@kfunigraz.ac.at> <uhdume2n1mhbba@corp.supernews.com> <3D171291.2060905@kfunigraz.ac.at> <mailman.1024950033.16332.python-list@python.org> <Xns9237E4F0DB10Dcliechtigmxnet@62.2.16.82>
Message-ID: <3D17BFE1.B55C689F@engcorp.com>

Chris Liechti wrote:
> 
> Tim Churches <tchur at optushome.com.au> wrote in
> news:mailman.1024950033.16332.python-list at python.org:
> 
> > Siegfried Gonzi wrote:
> >> It was really the laptop processor. I tried the calculation on a
> >> stationary machine (Pentium II 450 MHz and SuSE 8.0), and it works.
> >>
> >> Die Typen k?nnen jetzt was erleben.
> >
> > Could someone who understands idiomatic ?sterreichisches Deutsch
> > confirm that the above means "And sorry for being such a prat."?
> 
> it's more like a threat to the laptop manufacturer....
> 
> you have to admit that it was a nasty failure... is bad hardware your first
> thought when your program fails? on the other side he realy had a bad start
> with this thread.

I think if we were right there, "try the program on another
machine" would have come up pretty quickly.  In all the noise
I guess nobody thought of it.

-Peter


From tdelaney at avaya.com  Wed Jun  5 19:41:37 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Thu, 6 Jun 2002 09:41:37 +1000 
Subject: semi-concatenated strings
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A357@natasha.auslabs.avaya.com>

> From: Steve Holden [mailto:sholden at holdenweb.com]
>
> # The next three lines are bogus
> > > sql = string.split(sql)
> > > sql = string.join(sql, ' ')
> > > sql = string.strip(sql)
> > >
> > > rows = self.executesql(sql, (city,))
>
> Not only that, but for any reasonable database the whole schemozzle is
> completely unnecessary. SQL is a free-format language, and 
> newlines and
> spaces embedded in statements make absolutely no difference.

I agree entirely - if you can rely on the database to handle embedded
newlines in queries (and most can) then there is no need to perform the
preprocessing.

I meant to include that those lines were only there to match the original,
which did not have embedded newlines.

The main point is the preference to use a triple-quoted string to lay out
the query readably. If preprocessing is necessary, I would *still* prefer
that, and add the required preprocessing.

After all, SQL queries are often read in from files, with newlines in the
queries ...

Tim Delaney




From rjones at ekit-inc.com  Mon Jun  3 04:23:30 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Mon, 3 Jun 2002 18:23:30 +1000
Subject: How to suspend execution for at least x seconds in python ? (ie does "sleep" exist in Python ?)
In-Reply-To: <adf7t3$ki71@news.rd.francetelecom.fr>
References: <adf7t3$ki71@news.rd.francetelecom.fr>
Message-ID: <200206031823.30498.rjones@ekit-inc.com>

On Mon, 3 Jun 2002 18:03, Nicolas Torzec wrote:
> Dear all,
> I was wondering if an utility like "sleep x" exist in python ?
> Under Unix-like operating systems, the sleep utility suspends execution for
> at least x seconds under...

Look up the time module, specifically time.sleep.


   Richard





From sholden at holdenweb.com  Wed Jun  5 11:47:26 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 5 Jun 2002 11:47:26 -0400
Subject: self
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu>
Message-ID: <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com>

"Vojin Jovanovic" <vjovanov at stevens-tech.edu> wrote ...
[...]
>
> > Well, good luck with your project -- and don't spring any more
> > requirements on us, it's getting harder each time to solve your problem.
>
> Sorry, but it is not that my problem is hard.  It is that the concept of
> having
> to reference instance variables always through self. is making it hard and
> inconvenient which was my point in my original post.
> My opinion is that those points explained in FAQ are irrelevant.  The only
> one relevant is .....
>
[...]

Could you maybe use self.__dict__ as a namespace argument to the eval()?

if-tim-didn't-solve-this-i-don't-fancy-my-chances-ly y'rs  - steve
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From pyth at devel.trillke.net  Thu Jun  6 08:19:28 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 6 Jun 2002 14:19:28 +0200
Subject: introspect c-functions?
Message-ID: <20020606141928.A16359@prim.han.de>

Hello, 

is there any way to introspect the *number of arguments*
of a function implemented at c-level? the inspect 
module works only on python-functions, seemingly.

>>> import inspect
>>> inspect.getargspec(filter)
------------------------------------------------------------
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/local/lib/python2.3/inspect.py", line 621, in getargspec
    if not isfunction(func): raise TypeError, 'arg is not a Python
function'
TypeError: arg is not a Python function
>>>

Any input and comments appreciated. 

    holger




From qual at epost.de  Sun Jun 23 05:14:37 2002
From: qual at epost.de (Uwe Hoffmann)
Date: Sun, 23 Jun 2002 11:14:37 +0200
Subject: DerivVar instance has no attribute '__float__'
References: <uhb6cfbuvmmr45@corp.supernews.com>
Message-ID: <3D15917D.8020908@epost.de>

Lee Harr wrote:
> I just installed the Scientific module and am trying it out:
> 
> 
>>>>from Scientific.Functions.Derivatives import DerivVar
>>>>from math import sin

instead use:
from Numeric import sin

...


Uwe



From shagshag13 at yahoo.fr  Fri Jun  7 02:47:33 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Fri, 7 Jun 2002 08:47:33 +0200
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <20020606215349.GA13474@node1.opengeometry.net> <mailman.1023401650.18434.python-list@python.org>
Message-ID: <adpktk$17j3k$1@ID-146704.news.dfncis.de>

"gabor" <gabor at realtime.sk> a ?crit dans le message de news: mailman.1023401650.18434.python-list at python.org...

> but if you want to know about specific python issues, then you have to
> learn things like "dictionary are very well optimized etc.."

That's what i'm asking : learning and understanding... but where should i go ???

I'm wondering if there is a place to go to find "why this solution is better in this
appropriate case... ". For example, i was looking for the best pythoner's way to
remove duplicate in a list, and after reading many posts, i find at the end of a
cookbook recipe that it was :

set = {}
map(set.__setitem__, alist, [])
alist = set.keys()

But that's not trivial !!! And quite difficult to find...

> those things doesn't make your code obfuscated..

I agree :)

> [file.printsize() for file in directory]
>
> they're the same, and everyone would recommend you the last one..

Yes but i would have asked why this is the best... And i think that
your answer would have helped me to become a better pythoner...

I mean there should be a place to go where you could read generalities
like "as dictionary are very well optimized, you can use it without
trouble and for example it's best in : removing duplicates, ..." or
"if you need to do an operation on a list use map because it's
c-optimized", there are "object overhead if you...", and so on
that's not in the books, that's not in 'dive in python',
it's spreaded in Guido's essays, and posts over there... But i worry,
you could miss it if you don't spend many time to search, and say
"there should be a place for that"...

> so to summarize this: what do you want to know/learn?

What a beginner should learn to do it the pythoner way...

s13.

ps : and sorry for my poor english and typos...




From not.this at seebelow.org  Thu Jun 13 11:32:13 2002
From: not.this at seebelow.org (Grant Griffin)
Date: 13 Jun 2002 08:32:13 -0700
Subject: why not "'in' in 'in'"?
References: <mailman.1023978378.11084.python-list@python.org>
Message-ID: <aeadtt06rq@drn.newsguy.com>

In article <mailman.1023978378.11084.python-list at python.org>, "Mark says...
>
>I haven't followed this thread.  Someone else has probably already suggested
>this, but in case they haven't, it's fairly easy to define your own function
>that does what you want:
>
>def is_in(search, search_for):
>  """Return bool indicating whether search_for is in search."""
>  return search.find(search_for) >= 0
>
>subject = "i like spam, i do."
>x = "spam"
>y = "ham"
>
>print is_in(subject, x)
>print is_in(subject, y)
>

That's not a bad suggestion, Mark, but if I did that I guess I would be forever
wondering if I was testing whether the first one was in the second or the second
one was in the first.  Again, it doesn't read nearly as well as:

   if x in subject:
       ...

which leaves no doubt.

Ironically, that's already perfectly legal--so long as x has no more than one
character:

>>> subject = "i like spam, i do."
>>> x = "s"
>>> y = "h"
>>> if x in subject:
...    print "x is in subject and x has no more than one character"
...
x is in subject and x has no more than one character
>>> if y in subject:
...    print "same goes for y"
... else:
...    print "can't say the same for y"
...
can't say the same for y
>>>

reluctant-to-discard-my-hard-won-near-mastery-of-"find"-ly y'rs,

=g2

_________________________________________________________________________

Grant R. Griffin                                           g2 at dspguru.com
Publisher of dspGuru                               http://www.dspguru.com
Iowegian International Corporation                http://www.iowegian.com



From tim.one at comcast.net  Fri Jun 21 01:19:50 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 21 Jun 2002 01:19:50 -0400
Subject: How to represent the infinite ?
In-Reply-To: <20020620225355.7b8dd9db.christophe.delord@free.fr>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEADPPAA.tim.one@comcast.net>

[Christophe Delord]
> ...
> >>> 1e200**2
> Traceback (most recent call last):
>   File "<pyshell#1>", line 1, in ?
>     1e200**2
> OverflowError: (34, 'Numerical result out of range')
> >>> 1e400
> inf
>
> When the overflow occurs, the OverflowError exception is thrown.
> inf can't be the result of a computation (without patch and
> without catching exception).

It can, but it's a x-platform crapshoot as to exactly how.  For example,
under current Windows CVS, this still "works":

>>> 1e300 * 1e300
1.#INF
>>>

Here's a clumsier way <wink>:

>>> 1.6e308 + 1.6e308
1.#INF
>>>

Both of those assume the user hasn't installed and enabled the fpectl
module, whose purpose in life is to cause even this to complain.

maybe-number-crunchers-will-agree-on-something-after-i'm-dead-
    but-i'm-not-holding-my-breath-ly y'rs  - tim





From garrett at bgb.cc  Wed Jun 19 00:48:36 2002
From: garrett at bgb.cc (Don Garrett)
Date: Wed, 19 Jun 2002 04:48:36 GMT
Subject: What If.....    Strong Types
References: <3D0C0F0D.4030900@bgb.cc> <slrnagv0ue.46u.huaiyu@gauss.almadan.ibm.com>
Message-ID: <3D100C9A.10502@bgb.cc>

Huaiyu Zhu wrote:
> Don Garrett <garrett at bgb.cc> wrote:
> 
>>  I'm not suggesting any real changes to any, only proposing a thought 
>>experiment.
>>
>>  What if Python had been built from the start with strong types? And I mean 
>>strong types, not strong typing. The distinction I am trying to draw is that 
>>variables would not be typed, but that all types would have rigid interfaces.
>>
>>  Primarily, what if classes always had rigidly defined interfaces. I mean 
>>that public members had to be declared to exist, and that methods couldn't be 
>>modified on an instance. Private members could exist but would be really private.
>>
>>  All types would be classes (including ints and such). Amoung other things, 
>>I would add the concepts of Interfaces, instanceof operators.
>>
>>  Local variables would not require declaration, and would work just like 
>>today. Introspection would work, but be read-only.
>>
>>  My belief is that almost all the convenience of Python would be maintained, 
>>but that compilation, optimization, and automated code analysis would be simpler.
>>
>>  I'm just wondering if the change would be good or bad. Would it really 
>>break anything important in Python today? Would it prevent any common errors? 
>>Would it help with automated code analysis (compilation and optimization 
>>included) as much as I think it would?
> 

<Excellent discussion of seperation of interface and implementation snipped>

   Have you, by chance, ever used Sather? It was (I think development is 
stopped) an Eiffel inspired language that seperated interface and 
implementation inheritance (amoung several other cool ideas). Unfotunatly, I 
never got to use it for any large projects, so never felt that I learned it well.

   http://www.icsi.berkeley.edu/~sather/


-- 
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc



From jknapka at earthlink.net  Wed Jun 26 17:25:26 2002
From: jknapka at earthlink.net (Joseph A Knapka)
Date: Wed, 26 Jun 2002 21:25:26 GMT
Subject: Python daemon
References: <mailman.1024999953.15765.python-list@python.org> <F733C874CD014E16.148E57261FFD5E7A.AFFA4BD17B2F8D8B@lp.airnews.net> <WA%R8.21229$mF.14665@atlpnn01.usenetserver.com>
Message-ID: <3D1A3119.5A57CF4B@earthlink.net>

Steve Holden wrote:
> 
> "Cameron Laird" <claird at starbase.neosoft.com> wrote ...
> > In article <mailman.1024999953.15765.python-list at python.org>,
> > Roman Suzi  <rnd at onego.ru> wrote:
> > .
> > .
> > .
> > >> 3. restarting of the program if it fails, or the watchdog doesn't
> trigg=
> > >er
> > >
> > >Run the program from the /etc/inittab - if it terminates,=20
> > >init will restart it for you.
> > .
> > .
> > .
> > This is one of the great FMMs under Unix--the under-utilization
> > of init(1).  From everything I know, Roman's exactly right, that
> > /etc/inittab is the right way to set up a process that you truly
> > want to keep running.  However, many, MANY Unix hosts have all
> > kinds of ugly home-brewed hacks to duplicate this functionality.
> > I don't have an explanation, beyond the rather tautologous ob-
> > servation that init(1) simply isn't as widely understood as it
> > deserves to be.
> 
> You're probably right. Don't forget you need to specify "respawn" to have
> the process restarted when it terminates.
> 
> Of course, this doesn't address the issue of if the controlled process stops
> making progress but doesn't die.

Normally I'd allow wiser heads to address such issues,
but surprisingly none of them have directly answered
this question thus far, so here goes:

Just ensure that your daemon touches a file, say
/var/run/<daemonname>.doggy, every N seconds. Another
job run from cron attempts to delete
/var/run/<daemonname>.doggy every 2*N seconds. If
that fails, it does

"kill `cat /var/run/<daemonname>.pid`"

and then init restarts the daemon.

-- Joe
   "They call them the Diamond Dogs."
   -- David Bowie, "Diamond Dogs"


From fredrik at pythonware.com  Sun Jun 23 18:10:05 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 23 Jun 2002 22:10:05 GMT
Subject: PIL, Python and CGI
References: <3D1626FD.DA934B09@infineon.net>
Message-ID: <1HrR8.44835$n4.10519234@newsc.telia.net>

Sascha Ferley wrote:

>     raise ImportError, "The _imaging C module is not installed"
> ImportError: The _imaging C module is not installed
>
> Anyone have any ideas?

when running your CGI script, the Python interpreter
cannot find the _imaging extension.

chances are that sys.path isn't what you think it is.

</F>




From kalab at gmx.net  Tue Jun 25 08:07:40 2002
From: kalab at gmx.net (Gerhard Kalab)
Date: 25 Jun 2002 05:07:40 -0700
Subject: Python and Eclipse
References: <3D17FAF4.35CC9F1B@earthlink.net>
Message-ID: <bf340738.0206250407.5e293844@posting.google.com>

Joseph A Knapka <jknapka at earthlink.net> wrote in message news:<3D17FAF4.35CC9F1B at earthlink.net>...

> I am interested in having Python support in
> Eclipse <URL: http://www.eclipse.org> - interested

Have a look at http://sourceforge.net/projects/rubyeclipse.
I replaced all occurrences of "Ruby" with "Python" and it works quite
well :)

Unfortunately it's under the GPL and I'd like a python integration
based on a different license (IBM Public License or Python License).

So I started to play with the wizards in eclipse and I've got a
working python editor (only with some sort of syntax hightlighting)
for now.

Gerhard


From peter at engcorp.com  Wed Jun 19 22:07:42 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 19 Jun 2002 22:07:42 -0400
Subject: python version?
References: <mailman.1024534466.22643.python-list@python.org>
Message-ID: <3D1138EE.FDDCE5DF@engcorp.com>

"Delaney, Timothy" wrote:
> 
> > From: George Hester [mailto:hesterloli at hotmail.com]
> >
> > All fixed
> >
> [snip questionnaire]
> 
> Now I'm confused. What does that mean???

I interpreted it as "I got the point, please leave it alone now."

-Peter


From peter at engcorp.com  Sun Jun  9 09:29:02 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 09 Jun 2002 09:29:02 -0400
Subject: Socket error problem
References: <mailman.1023614413.20191.python-list@python.org>
Message-ID: <3D03581E.FA5890EF@engcorp.com>

David LeBlanc wrote:
> 
> I have this problem on windows where socket.socket(...) keeps returning
> 10022, 'Invalid argument'. It's actually dieing in the
> poplib.POP3.__init__(...) method. I have verified that host is reasonable
> (mymail.myisp.net or something like that ;)), and port is 110.
> 
> If I use the test script in poplib.__main__(...) with the correct host, user
> and password, all is well, but if I call it from this other program i'm
> trying to get working I get the above error.
> 
> I can't figure out what's wrong - any pointers?

As usual, post a traceback, or code snippets... don't make us guess
what you're doing.


From johnroth at ameritech.net  Wed Jun 19 16:58:25 2002
From: johnroth at ameritech.net (John Roth)
Date: Wed, 19 Jun 2002 16:58:25 -0400
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.>
Message-ID: <uh1s1qdcl5b1b0@news.supernews.com>

"Angela Williams" <adwilliams76 at hotmail.com> wrote in message
news:3d10ed95 at xpl-sdf1-sec1....
> I'm looking for the equivalent of the GOTO command for Python 2.2.1
running
> on Windows 98. I have programmed very little and am new to Python. I
found
> the following information on the FAQ page but it doesn't make any
sense to
> me.
>
>
> 6.26. Why no goto?
> Actually, you can use exceptions to provide a "structured goto" that
even
> works across function calls. Many feel that exceptions can
conveniently
> emulate all reasonable uses of the "go" or "goto" constructs of C,
Fortran,
> and other languages. For example:
>
>    class label: pass # declare a label
>    try:
>         ...
>         if (condition): raise label() # goto label
>         ...
>    except label: # where to goto
>         pass
>    ...
>
> This doesn't allow you to jump into the middle of a loop, but that's
usually
> considered an abuse of goto anyway. Use sparingly.
>
>
>
> Any and all help would be greatly appreciated.
>
> Thanks, Angela

There's no equivalent of a goto command in Python, for good reason,
which I won't repeat here. Prof. Djikstra did it very well in 1974,
in his letter "Goto Considered Harmful."

If you'd like to post the problem you're trying to solve, someone
will be sure to help you.

The reason for mentioning exceptions in the FAQ is that, for a
long while, exception handling was the sole programming task
where people thought they still needed a goto. That was solved
over a decade ago, with structured exception handling, and Python
simply does it the same as everyone else, only (possibly) better.

John Roth
>
>




From saint at ghost.lt  Thu Jun 20 13:03:08 2002
From: saint at ghost.lt (martynas sklizmantas)
Date: 20 Jun 2002 19:03:08 +0200
Subject: Function to add commas in numbers?
In-Reply-To: <XFMail.20020620095323.shalehperry@attbi.com>
References: <XFMail.20020620095323.shalehperry@attbi.com>
Message-ID: <1024592588.720.4.camel@naked>

hello, 

On Thu, 2002-06-20 at 18:53, Sean 'Shaleh' Perry wrote:
> and also takes up less room on your system.
> 
> Edit /etc/locale.gen and follow the directions in the comments.
> 

indeed, that was the problem. thank you very much.


best regards, 

ms


-- 
Alas, how love can trifle with itself!
		-- William Shakespeare, "The Two Gentlemen of Verona"





From cliechti at gmx.net  Wed Jun 26 22:23:18 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 27 Jun 2002 04:23:18 +0200
Subject: Accessing C++ variables from Python
References: <mailman.1025138558.21469.python-list@python.org>
Message-ID: <Xns923A2D158845Fcliechtigmxnet@62.2.16.82>

Teja Sastry <kurugant at denshi.ece.utk.edu> wrote in
news:mailman.1025138558.21469.python-list at python.org: 

> I have a c++ program storing a binary stream into a buffer, i want to 
> access that buffer from python. The buffer is actually a cPickled
> binary stream from another program loaded in to c++ 'cuz we use a
> routing layer called diffusion, and its API is in c++ so i have to put
> the variable into C++ to transport it over the network, at the
> receiving end i have another C++ diffusion API program holding the
> variable, i have to transfer it as object back to python, i'm not
> quite sure about how to do this could you please clarify 

i think passing around those buffers as python strings will be the easiest 
approach. binary or not does not matter for python-strings.

(see below)

> or is there any mailing list for python users 

you're already posting to it...
 
> P.S: I know how to use SWIG and creating shared modules for python

my experience with SWIG is that while it's powerful it doesn't do quite 
that waht I want... therefore i end up writing my extensions by hand. (when 
i tried it it generated wrapper objects en masse but i simple wanted to 
pass around strngs.)

don't know about BOOST or CXX cause i don't use C++ in my projects, but 
those get usualy mentioned for Python extensions with C++. maybe one of 
those could be very helpful for you.

http://www.boost.org/libs/python/doc/
http://cxx.sourceforge.net/

chris

-- 
Chris <cliechti at gmx.net>



From sholden at holdenweb.com  Tue Jun 18 16:24:38 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 18 Jun 2002 16:24:38 -0400
Subject: Python interface to ImageMagick
References: <aec8qd$19p$05$1@news.t-online.com>
Message-ID: <yEMP8.165651$Fp1.37569@atlpnn01.usenetserver.com>

"Achim Domma" <achim.domma at syynx.de> wrote in message
news:aec8qd$19p$05$1 at news.t-online.com...
> Hi,
>
> I just uploaded a first alpha version of my python wrapper for
ImageMagick.
> Currently there is only a windows version, which is tested on Win2K with
> ActiveState Python 2.2.1. Other version, more documentation and examples
> will follow hopefully soon. Every feedback or bugreport is wellcome,
because
> I don't have different testing environments.
>
> If you want to try it, you can download it here:
> http://www.procoders.net/imagemagick/ImageMagick-0.1a.win32.exe
>
> An working ImageMagick (www.imagemagick.org) Installation is required. The
> wrapper is linked with ImageMagick 5.4.5-1.
>

You should certainly consider renaming the installer so it looks like a
Python extension and not an ImageMagick installer. PyImageMagick...?

 regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From wurmy at earthlink.net  Fri Jun 28 00:16:15 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Fri, 28 Jun 2002 04:16:15 GMT
Subject: Plugins in Python
References: <afgeol$sgd$1@mulga.cs.mu.OZ.AU>
Message-ID: <3D1BE2DD.608@earthlink.net>

Lars Yencken wrote:
> Hello,
> 
> I'm trying to do plugins for a program I'm writing in Python, and I'm not 
> sure how exactly to do it. Basically what I want it to be able to write a 
> module which loads every python file in a directory and stores an instance 
> of each one that loaded successfully.
> 
> I tried using exec() and compile() to run each of the files in the 
> directory from my Plugins module, and had each of the files import my 
> Plugins module and store an instance of themselves there, but at the end of 
> it they weren't stored.
> 
> Any ideas?

One way to do this is making a subdirectory and using it as a package. For 
example, a structure like this:

   main.py
   plugins/
     __init__.py
     foo.py
     bar.py

If you want all the plugin modules imported, you can simply do this in 
__init__.py (import foo, import bar, etc). Then all you have to do from a 
program that uses the plugins, is import the package:

   import plugins

   plugins.bar.baz()
   # call baz() function in module bar in plugins directory

HTH,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/



From max at alcyone.com  Thu Jun 13 15:45:25 2002
From: max at alcyone.com (Erik Max Francis)
Date: Thu, 13 Jun 2002 12:45:25 -0700
Subject: why not "'in' in 'in'"?
References: <ae8dg70217g@drn.newsguy.com> <3D080482.7A7FD514@alcyone.com> <3D082A62.CF800C75@seebelow.org> <3D083298.470850D0@alcyone.com> <3D08A161.A9F3136F@seebelow.org>
Message-ID: <3D08F655.27790070@alcyone.com>

Grant Griffin wrote:

> Again, it has to do with the fact that the illegal multi-character
> "in"
> is intuitive and reads better.  Compare the following:
> 
>         if 'mortgage' in email_subject:
>            ...
> 
> to:
> 
>         if email_subject.find('mortgage') >= 0:
>            ...
> 
> I consider the former to be highly practical, if not highly pure.

The problem here is that you're declaring it more intuitive because
you've already decided a priori that you want it.  To a true Python
programming, the line

	if 'mortgage' in email_subject:

is pretty suspicious; it makes it sound like email_subject is a sequence
of keywords.

> As to the latter, I frequently made this mistake at first:
> 
>         if email_subject.find('mortage'):
>            ...
> 
> which is fairly intutitive, but doesn't work, of course, because
> "find"
> returns -1 (logical true) in the failure case!

That is just the nature of string functions; it is documented well.  Not
every aspect of Python programming can be short _and_ do what you
expect.  Some learning is required.  That learning is what tells you the
`in' operator is for membership, not subsequence comparison.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Who'd ever think it / Such a squalid little ending
\__/ The American and Florence, _Chess_
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From mdk at mdk.com  Thu Jun 20 11:17:55 2002
From: mdk at mdk.com (mdk)
Date: 20 Jun 2002 15:17:55 GMT
Subject: Function to add commas in numbers?
Message-ID: <Xns923372EFBBC5Dmdkmdkmdkmdk@130.133.1.4>

Hello,

Is there a function that will take, for example, 1234567 and return 
1,234,567?

Thanks


From andreas at kostyrka.priv.at  Sat Jun  1 02:26:12 2002
From: andreas at kostyrka.priv.at (Andreas Kostyrka)
Date: Sat, 1 Jun 2002 08:26:12 +0200
Subject: Why no open(f, "w").write()?
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8BF0A329@natasha.auslabs.avaya.com>
References: <B43D149A9AB2D411971300B0D03D7E8BF0A329@natasha.auslabs.avaya.com>
Message-ID: <200206010826.12474.andreas@kostyrka.priv.at>

Am Donnerstag, 30. Mai 2002 02:47 schrieb Delaney, Timothy:
> I can come up with a perfectly good example where
Well, replacing the semantics of a builtin operation is not a perfectly good
example IMHO. ;)
>
> 	open(FILE, 'w').write(CONTENTS)
>
> will fail to close the file in CPython.
>
> There are no guarantees that the reference to the object returned is the
> *only* reference to that object. Assume the following (perverse)
> implementation of open:
>
> __files = []
>
> def open (path, mode):
>     f = __open(path, mode)
>     __files.append(f)
>     return f
>
> or
>
> class file:
>
>     __files = []
>
>     def __init__(self, path, mode):
>         __files.append(self)
>
> Whilst I know this is not the implementation, there is nothing that
> prevents it from being like this. So you get back an object that has one
> more reference count than expected. It never gets collected.
Well, then the implementation is broken. If you allow the implementation
to collect additional references at random places, you can prove quite
easily that nothing is safe :)

Your proof is a bit like the classical error from "full induction" (is this 
the english name?) proofs in math class. If you reason with faulty premises 
you can prove anything. 
>
> You should never rely on automatic freeing of *any* resources unless it is
> guaranteed. Always explicitly free resources when you have finished with
> them.
Well, then you should del all your names too.
Oops wait, how can we be sure that del works? What if del also increments the 
refcount instead of decrementing? What if some operation before the del
incremented the refcount?

Basically what you have proven is, that if you have a faulty open() 
implementation, the open(f,t).write(d) idiom doesn't work.

Guess what. I can prove that open(f,t).write(d) wouldn't work if write is 
implemented buggy. The exact proof is left as an exercise to the reader. :)

In respect to CPython, refcounting behaviour, or more general retaining 
references in any gc regime, is a quite important part of the interface.

Andreas
-- 
Andreas Kostyrka				+43 676 4091256
andreas at kostyrka.priv.at			andreas at mtg.co.at





From jhndnn at yahoo.com  Fri Jun 14 12:02:04 2002
From: jhndnn at yahoo.com (John Dunn)
Date: 14 Jun 2002 09:02:04 -0700
Subject: Standalone embedded Win32 Python
Message-ID: <2cc96dc7.0206140801.7d43ff5e@posting.google.com>

I have embedded Python in my C++ Win32 app. What is the recommended
way of installation/distribuion? It seems the best would be to
statically link to Python - this would also allow me to customized the
Python environment that runs with my app. Unfortunately, Python
doesn't appear to support static linking on Win32.

I really don't want to run the whole Python installed when I install
my app. If I can't statically link, is it acceptable to just ship my
own local copy of the DLL? I assume I will have to edit some code so
it doesn't look to the registry to find the path to Python.

Is there any documentation on doing this? This seems to be important
to any app that has Python embedded - you can't very well tell your
customers to download the latest Python install just to run your
program.

Thanks for any help-

John Dunn
--
John Dunn
Peak Audio, a division of Cirrus Logic, Inc.
http://www.peakaudio.com


From ponderor at lycos.com  Mon Jun 24 20:27:39 2002
From: ponderor at lycos.com (Dean Goodmanson)
Date: 24 Jun 2002 17:27:39 -0700
Subject: Windows versions of Python---pros and cons?
References: <mailman.1024613669.14676.python-list@python.org> <3D138293.85892240@noaa.gov>
Message-ID: <e81be8b2.0206241627.6a3a89e6@posting.google.com>

> Now I have to figure out how to change my PATHEXT environment variable,
> which is pretty OT for this group...

A clearer answer to my hasty response in Trent's direction:

WinME: Start , Run , MSConfig .  (Also works in Win98.)


From aahz at pythoncraft.com  Fri Jun 28 15:28:52 2002
From: aahz at pythoncraft.com (Aahz)
Date: 28 Jun 2002 15:28:52 -0400
Subject: wxpython threads help needed
References: <6bd9f01b.0206281055.2b01a2a6@posting.google.com>
Message-ID: <afiddk$qso$1@panix1.panix.com>

In article <6bd9f01b.0206281055.2b01a2a6 at posting.google.com>,
sameer <sameer_ at email.com> wrote:
>
>Does anyone have an example of wxpython threads in action?  I am
>trying to create  an application with several (tabbed) panels, each of
>which is being updated on its on thread.  Basically, I don't want the
>application to freeze up, or to have long startup times as all the
>panels are being updated.  I want the startup to happen as soon as the
>main panel is built.  I would appreciate it if someone can give an
>example either from their own app or code, or a simple example cooked
>up on the fly.

Not wxPython, but if you go to my web page, you can see an example for
TKinter that should be useful for you.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From frank.sonnenburg at biosolveit.de  Thu Jun 27 12:30:19 2002
From: frank.sonnenburg at biosolveit.de (Frank Sonnenburg)
Date: Thu, 27 Jun 2002 18:30:19 +0200
Subject: C API: PyXXX_SetItem
Message-ID: <200206271630.SAA24525@post.webmailer.de>

Hi all,

am i right, that PyTuple_SetItem() and PyList_SetItem() only *transfer* 
references to objects, but don't increment reference counts?

And, in contrast, PyDict_SetItem() *does* increment the object reference 
count?


So,


        PyTuple_SetItem(tup,idx,PyInt_FromLong(123));

is perfect, but


        PyDict_SetItemString(dict,"key_name",PyInt_FromLong(123));

will lead to a memory leak. To avoid this:


        PyObject *py_int;
        py_int = PyInt_FromLong(123);
        PyDict_SetItemString(dict,"key_name",py_int);
        Py_DECREF(py_int);


I would appreciate any suggestions/corrections to this point. Thanks in 
advance

Frank Sonnenburg
-- 
____________________________________________________________________________
       _________
               / _________          Frank Sonnenburg
           Bio \ \                  Software Developer
       ________/ / olve
                 \________          BioSolveIT GmbH, An der Ziegelei 75

     I N F O R M A T I O N -        53757 Sankt Augustin, Germany
     - - T E C H N O L O G Y
                                    Tel: +49 - 2241 - 973 66 81
                                    Fax: +49 - 2241 - 973 66 88
       www.biosolveit.de            frank.sonnenburg at biosolveit.de
____________________________________________________________________________




From dom at edgereport.put_a_c_o_m_here  Thu Jun  6 22:11:14 2002
From: dom at edgereport.put_a_c_o_m_here (Domenic R. Merenda)
Date: Fri, 07 Jun 2002 02:11:14 GMT
Subject: Medium to Large Scale Python Deployments
References: <KHkL8.21383$eD2.2399923271@newssvr10.news.prodigy.com> <34b82fda.0206061639.67964684@posting.google.com>
Message-ID: <6DUL8.21673$BJ6.2510024856@newssvr10.news.prodigy.com>

   Yes, thank you.  I've been following this since 2000, and have looked at
using some of the code, but wasn't able to open source our work, and so
refrained for using theirs. :-)


--
Domenic R. Merenda
Editor, The Edge Report
http://www.edgereport.com


"Stuart Quimby" <stuq at dstoys.com> wrote in message
news:34b82fda.0206061639.67964684 at posting.google.com...
> You may be interested in the gnuenterprise project.  They are
> developing a multi-tier db independant ERP that is open-source.
> Details can be found at http://www.gnuenterprise.org.
>
> Stuart Quimby




From look at replyto.address.invalid  Tue Jun  4 01:43:55 2002
From: look at replyto.address.invalid (Greg Ewing)
Date: Tue, 04 Jun 2002 17:43:55 +1200
Subject: Python Turing machine
References: <mailman.1023068284.27231.python-list@python.org>
Message-ID: <3CFC539B.BCBFC855@replyto.address.invalid>

"Delaney, Timothy" wrote:
> 
> Yeah - the dependence on an infinitely-long tape can be a real bugger ...

well, any terminating computation only uses a finite
amount of tape, so you just need to be prepared to
buy more tape as needed.

Of course, if the universe is closed, you could
find yourself in a situation of having run out
of matter to make new tape from. Hopefully this
will have been fixed by the time Python
3000000000000000000000000000000 is released.

-- 
Greg Ewing, Computer Science Dept, 
University of Canterbury,	  
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg


From fperez528 at yahoo.com  Mon Jun  3 15:30:15 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 03 Jun 2002 19:30:15 +0000
Subject: using getopt
References: <adh37d$dj8$1@news.asu.edu>
Message-ID: <adh59t$k4g$1@peabody.colorado.edu>

David Bear wrote:

> Found a couple of frustrating things with getopt.  I am writing a filter
> that will be called by lprng -- or other spool services.  rather than
> trying to code for every possible cmd line arg I may be given, in there a
> way to use getopt to grab just a few options that I care about and through
> the rest away?
 
getopt sucks. Period. Take a look at DPyGetOpt, which implements the 
Getopt::Long perl facilities. It's very very nice. I've also heard good 
things about optik, but haven't used it myself.

If you can't find DPyGetOpt (it's a bit hard to get) drop me a line.

Cheers,

f


From SBrunning at trisystems.co.uk  Thu Jun 20 04:49:55 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 20 Jun 2002 09:49:55 +0100
Subject: GOTO w/ Python?
Message-ID: <31575A892FF6D1118F5800600846864DCBD3E8@intrepid>

> From:	John Roth [SMTP:johnroth at ameritech.net]
> There's no equivalent of a goto command in Python, for good reason,
> which I won't repeat here. Prof. Djikstra did it very well in 1974,
> in his letter "Goto Considered Harmful."
 
<http://www.acm.org/classics/oct95/>

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.




From whisper at oz.net  Fri Jun 28 15:23:47 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 28 Jun 2002 12:23:47 -0700
Subject: metaclasses vs. inheritance
In-Reply-To: <ghTS8.58749$mh.1826422@news1.telusplanet.net>
Message-ID: <GCEDKONBLEFPPADDJCOEEENLDMAA.whisper@oz.net>

<snip>
> PS - what's the antonym of meta?

Mundane I think.

This might also shed some light (from dictionary.com):
meta

<philosophy> /me't*/ or /may't*/ or (Commonwealth) /mee't*/ A prefix meaning
one level of description higher. If X is some concept then meta-X is data
about, or processes operating on, X.

For example, a metasyntax is syntax for specifying syntax, metalanguage is a
language used to discuss language, meta-data is data about data, and
meta-reasoning is reasoning about reasoning.

Dave LeBlanc
Seattle, WA USA





From krissepu at vip.fi  Mon Jun  3 16:07:04 2002
From: krissepu at vip.fi (Pekka Niiranen)
Date: Mon, 03 Jun 2002 23:07:04 +0300
Subject: Guru advice needed for mxTextTools
Message-ID: <3CFBCC68.D35A03D8@vip.fi>

I am trying to optimize a function that searches nested strings from
set of (allmost) flat files (about 2 MB each) . If I use regular
expressions, I must
fix the amount of nesting:

----- code example ----
    def make_regular_expression(prefix, suffix):
        if prefix == suffix:
            pattern = "%s%s%s" % (re.escape(prefix), '.+?',
re.escape(suffix))
        else:
            repeat = "%s%s%s%s" % ('[^', prefix, suffix,']+')
            pattern = "%s%s%s%s%s%s%s%s%s%s%s" %
('(',re.escape(prefix),repeat,

'(',re.escape(prefix),repeat,
                                              re.escape(suffix),')*',

repeat,re.escape(suffix),')')
            return pattern
---- code example ends ----

In the code above support two nested strings. If prefix is "?" and
suffix is "!" then
it will evaluate into:

>>> pattern = re.compile("(\?[^?!]+(\?[^?!]+\!)*[^?!]+\!)")
>>> Line = "?AA?BB!CC!?DD!ee?EE!ff?FF?GG!HH!"
>>> print re.findall(pattern, Line)
[('?AA?BB!CC!', '?BB!'), ('?DD!', ''), ('?EE!', ''), ('?FF?GG!HH!',
'?GG!')]

So far so good, but:

1)    Re -module returns also empty matches which I have to clean:
            pars =
filter(operator.truth,reduce(operator.add,re.findall(pattern, Line)))

2)    The file is not flat: I also need to check the contents of the
previous line. If previous line
       does not contain correct value, I do not have to run the regular
       expression on the current line:
            for i in range(1,len(lines),2):
                        test = lines[i-1].strip()
                        if (test == 'x' or test == 'y'):
                            matches = re.findall(pattern,
lines[i].strip())
                            if matches:
                                # Remove empty results with filters
                                pars =
filter(operator.truth,reduce(operator.add, matches))

3)    Amount of nesting may vary in the future

I have managed to speed up the search about 10x by using map() instead
of for -loop
and the current bottleneck is the regular expression.

I have thought of EBNF -notation that should be supported with
Simpleparse  + mxtexttools

Questions are:

1)    What is the mxtexttool tagtable for the regular expression above
with additions of unlimited nesting.
       If suffix is the same as prefix, no nesting is assumed
2)    Is it possible to parse the file without keeping the record of the
current line number since
       values to be checked are allways on odd line numbers and regular
expression is allways run
       on even line numbers. If I could read two lines at a time and
parsing them both simultaneously
       (as a single line) with mxtexttools (with lookAhead or whatever
), I could gain some speed ?
3)    Should I seek examples from XML -tools instead OR write my own
parser with C + SWIG ?


-pekka-





From sholden at holdenweb.com  Wed Jun  5 08:51:04 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 5 Jun 2002 08:51:04 -0400
Subject: Problem with urllib.urlopen()
References: <cc14131d.0206031445.5154de31@posting.google.com> <cc14131d.0206032128.5b2efcf6@posting.google.com> <cc14131d.0206041631.157f11d1@posting.google.com>
Message-ID: <WSnL8.268879$%u2.175369@atlpnn01.usenetserver.com>

"Christopher" <chris_mk at hotmail.com> wrote in message
news:cc14131d.0206041631.157f11d1 at posting.google.com...
> Okay okay, here are links that 1) are short, 2) should work, and 3)
> show my problem.
>
> link 1:
>
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=PureSearch&db=PubMed&detai
ls_term=%28%22glucagon%22%5BMeSH%20Terms%5D%20OR%20glucagon%5BText%20Word%5D
%29
>
> link2:
>
>
http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Display&db=PubMed&details_
term=%28%22glucagon%22%5BMeSH%20Terms%5D%20OR%20glucagon%5BText%20Word%5D%29
&dopt=XML&query_key=1
>
> Look at link2 in your browser, then look at what urllib.urlopen or
> .urlretrieve get.  Thanks.
>

Just to be sure, do you expect the second link (in a browser) to show "Error
handling request: no current query"? This seems to work w/Python 2.2 under
cygwin, at least. There was so much output I omitted quite a lot, replaced
by ellipses below:

>>> import urllib
>>> f =
urllib.urlopen("""http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Display&
db=PubMed&details_term=%28%22glucagon%22%5BMeSH%20Terms%5D%20OR%20glucagon%5
BText%20Word%5D%29&dopt=XML&query_key=1""")
>>> ll  = f.read()
>>> ll
'<html>\n  <head> ...
<table
cellpadding="0" cellspacing="0" width="100%"><tr><td bgcolor="#ffcccc">Error
handling request: no current query<BR></td></tr></table> ...
</body>\n</html>\n'
>>>

regards
 Steve
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From shagshag13 at yahoo.fr  Sun Jun  2 05:01:09 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Sun, 2 Jun 2002 11:01:09 +0200
Subject: Which is fastest dict or list.index() ?
References: <adaih2$vtb8c$1@ID-146704.news.dfncis.de>
Message-ID: <adcms5$vm6j0$1@ID-146704.news.dfncis.de>

Thanks !!! (i use some tricks i find in seqdict)

> If you want a skiplist version, let me know and I'll send it.
> Raymond Hettinger

I'm sorry, but i don't understand what you mean by a "skiplist version"

So here is how i do it now, i think i miss to explain that i must keep some
kind of
id (which i use as a key in another table, and i want to keep this other
table
indexed by integer instead of words to save mermory). <- thoug i don't know
if this is
a real saving...

i'm huge newbie but i feel python really pleasant, great, etc... (even if i
had lots of trouble
with memory run out... on a solaris system with  1024mo...)

i post for comments or advices as i'm not a skilled pythoner...

thanks in advance for reading and helping

class IKeys:
 def __init__(self):
  self._size = 0
  self._term = {}  # element -> int key - ex: {a : 0, b : 2, c : 1}
  self._iKeys = [] # contains elements, int key -> element - ex: [a, c, b]

 def __str__(self):
  s = 'iKey (' + "\n"
  s += str(self._term) + "\n"
  s += str(self._iKeys) + "\n"
  s += str(self._size) + " # size)\n"
  return s

 def __len__(self):
  return self._size

 def __getitem__(self, pos):
  return self._iKeys[pos]

 def index(self, element):
  return self._term[element]

 def add(self, element):
  if self._term.has_key(element):
   return self._term[element]
  self._iKeys.append(element)
  self._term[element] = self._size
  self._size += 1
  return (self._size - 1)

 def keys(self):
  return self._iKeys

 def get_size(self):
  return self._size

which i use like this :

>>> toons = IKeys()
>>> for t in ['Bugs', 'Mickey', 'Babs', 'Jerry', 'Tom', 'Babs']:
 toons.add(t)
0
1
2
3
4
2  # already in toons so we don't had it, we get id
>>> print toons
iKey (
{'Mickey': 1, 'Babs': 2, 'Tom': 4, 'Bugs': 0, 'Jerry': 3}
['Babs', 'Bugs', 'Jerry', 'Mickey', 'Tom']
5 # size)
>>> t[2]
'b'
>>> toons[2]
'Babs'
>>> toons.index('Jerry')
3
>>> tmp = toons.keys()
>>> tmp.sort()
>>> for t in tmp: # here it's sorted
 i = toons.index(t)
 print "Toons [%s] has id [%s]" % (t, i)
Toons [Babs] has id [2]
Toons [Bugs] has id [0]
Toons [Jerry] has id [3]
Toons [Mickey] has id [1]
Toons [Tom] has id [4]






From sarmstrong13 at mac.com  Sat Jun  8 12:30:26 2002
From: sarmstrong13 at mac.com (SA)
Date: Sat, 08 Jun 2002 11:30:26 -0500
Subject: Newbie OOP Question.(diff between function, module and class)
Message-ID: <B9279B52.6AEB%sarmstrong13@mac.com>

Hi Everyone-

    I'm still trying to grasp this OOP concept in Python. Can someone tell
me the difference between a function, a module, and a class? Since Python
defines them all as objects, how do they differ outside of being called from
inside the script or outside the script?

Thanks.
SA



From loewis at informatik.hu-berlin.de  Tue Jun 18 03:07:44 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Jun 2002 09:07:44 +0200
Subject: unicode strings and strings mix
References: <mailman.1024375941.29946.python-list@python.org> <slrnagthhn.5tb.gerhard@lilith.my-fqdn.de>
Message-ID: <j4ofe9t67z.fsf@informatik.hu-berlin.de>

Gerhard H?ring <gerhard at bigfoot.de> writes:

> 'x' and 'A' are in the ASCII range, so this shouldn't produce an
> exception. I also cannot reproduce it with sys.getdefaultencoding() ==
> "ascii".

These where not 'x' and 'A', but '\xd7\xc1\xd7\xc1\xd7'. Since the
article was posted in KOI8-R, Roman probably meant those bytes to
denote CYRILLIC SMALL LETTER VE and CYRILLIC SMALL LETTER A,
respectively.

Of course, when Python add strings, it can't possibly know that this
is how the byte string was meant to be interpreted, so you need to
write

unichr(0x3345) + unicode('\xd7\xc1\xd7\xc1\xd7', 'koi8-r')

The result string cannot be represented in KOI8-R, though, since it
contains SQUARE MAHHA.

Regards,
Martin


From ljd at nospam.com  Sat Jun  8 18:19:47 2002
From: ljd at nospam.com (LJD)
Date: Sat, 08 Jun 2002 22:19:47 GMT
Subject: Program mysteriously STOPS!
References: <yZuM8.146064$ux5.186314@rwcrnsc51.ops.asp.att.net>
Message-ID: <7qvM8.134377$352.6595@sccrnsc02>

Just for more info...

Running Linux 7.3
Python 1.52

Program autostarts, runs for a while, then stops!

Larry


From eric.brunel at pragmadev.com  Fri Jun 28 09:57:27 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Fri, 28 Jun 2002 13:57:27 +0000
Subject: menue windowing help needed
References: <mailman.1025258137.27591.python-list@python.org>
Message-ID: <afhijk$baj$1@wanadoo.fr>

Klaus Reinhardt wrote:
> ---------------------------------------------------------------------
> Hi
> 
> I tried the samples from wxPython and tkinter
> without succes.

Can you tell us what samples? And what happened?

> I need some scripting help for:
> A menue
> each alternative can open
> one window for print or ..

Here is a simple example with Tkinter:

----------------------------------
from Tkinter import *

def openWindow():
  wdw = Toplevel()

root = Tk()
menuBar = Menu(root)
root['menu'] = menuBar
fileMenu = Menu(root)
fileMenu.add_command(label='Open window', command=openWindow)
fileMenu.add_separator()
fileMenu.add_command(label='Quit', command=root.quit)
menuBar.add_cascade(label='File', menu=fileMenu)

root.mainloop()
----------------------------------

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From phr-n2002b at NOSPAMnightsong.com  Thu Jun 27 19:38:02 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 27 Jun 2002 16:38:02 -0700
Subject: How to get rid the new line
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org> <3D1AD534.A74EC9F4@ipm.fhg.de> <3D1B1A76.563E5394@engcorp.com> <pan.2002.06.28.00.14.48.789371.16627@linuxmail.org> <afg513$2426$1@news.idiom.com>
Message-ID: <7xlm90cmxx.fsf@ruckus.brouhaha.com>

"James T. Dennis" <jadestar at idiom.com> writes:
>  Could try this:
> 
> 	def chomp(line):
> 		if line[-1]=='\n':
> 			line=line[:-1]
> 		return line
> 
>  (or variations thereon, to satisfy other semantic requirements).

If you don't mind removing all trailing whitespace, there's also

  line = line.strip()

or more portably

   import string
   line = string.strip(line)


From alecg at inwave.com  Mon Jun 10 23:32:19 2002
From: alecg at inwave.com (Alexander Garden)
Date: Mon, 10 Jun 2002 22:32:19 -0500
Subject: Correct code/unit testing (was Re: Efficient python programming...)
In-Reply-To: <83u1oaq0jt.fsf@panacea.canonical.org>; from kragen@pobox.com on Mon, Jun 10, 2002 at 05:35:50PM -0400
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com> <834rgd5isv.fsf@panacea.canonical.org> <3D02FB3B.CF75E925@engcorp.com> <83u1oaq0jt.fsf@panacea.canonical.org>
Message-ID: <20020610223219.B1061@genesis.gnet>

On Mon, Jun 10, 2002 at 05:35:50PM -0400, Kragen Sitaker wrote:
> Yes, one of the beautiful things about programming is how it impresses
> upon each of us how fallible we are.

Amen!




From gherron at islandtraining.com  Mon Jun 17 12:56:55 2002
From: gherron at islandtraining.com (Gary Herron)
Date: Mon, 17 Jun 2002 09:56:55 -0700
Subject: tab size
In-Reply-To: <29A97D00F387D411AC7900902770E14805853F72@lcoeexc01.coelce.net>
References: <29A97D00F387D411AC7900902770E14805853F72@lcoeexc01.coelce.net>
Message-ID: <200206170956.55939.gherron@islandtraining.com>

On Monday 17 June 2002 09:05 am, Alves, Carlos Alberto - Coelce wrote:
> Hi all,
>
> Consider the following code:
> >>> for i in range(10):
>
> 	print "%d\t"%i,
>
>
> 0	1	2	3	4	5	6	7	8	9
>
>
>
> How can I change the default size of tab in formating string \t?!

You can't.  In fact the question does not even make sense in this
context!  A tab has no size.  It is a single ascii character just like
any other ascii character.

However, when a string containing tabs is DISPLAYED, then the tabs are
expaned into some definite amount of space on the screen or paper or
wherever the display is occuring.

So to answer your question, we'd need to know where you are running
this:
  unix xterm?
  dos prompt?
  IDLE?
  emacs?
  ...?

Once you get the display of tabs to work as you wish in your setting,
you would still have the problem that running the program elsewhere
would expand the tabs according to other settings.  If you want FULL
control of the output wherever the thing is run, don't use tabs.


Dr.  Gary Herron






From marklists at mceahern.com  Tue Jun 18 15:02:37 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 18 Jun 2002 14:02:37 -0500
Subject: Python to MS SQL Server
In-Reply-To: <3D0F8099.8080106@matteicos.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBKEDOLFAA.marklists@mceahern.com>

> I am interested in using Python in a MS SQL Server environment, but do 
> not see any database modules out there, especially for native versus 
> using ODBC. Is there such a module?

This is a total guess, because I haven't done this myself:

Have you tried using win32com to use ADO?

// mark
-





From bjarne_christiansen at hotmail.com  Wed Jun 26 06:48:59 2002
From: bjarne_christiansen at hotmail.com (Bjarne Christiansen)
Date: Wed, 26 Jun 2002 10:48:59 +0000
Subject: Python file upload
Message-ID: <F1448cU6xfPTeip68V80000009a@hotmail.com>

Thanks for your reply Richard!

I have changed the script to the following:
...
  item = form["filename"]
  if item.file:
    data = item.file.read()
    f = open("file1.jpg","wb")
    f.write(data)
...

But I still run into the same problem, any ideas?

~Bjarne


>From: Richard Barrett <R.Barrett at ftel.co.uk>
>To: "Bjarne Christiansen" <bjarne_christiansen at hotmail.com>,   
>python-list at python.org
>Subject: Re: Python file upload
>Date: Wed, 26 Jun 2002 09:56:33 +0100
>
>I think you will find that your calls to cgi.escape the incoming data are 
>the cause of your problem.
>
>You do not need to escape incoming multipart/form data. The documentation 
>for cgi.escape says:
>
><quote>
>Use this if you need to display text that might contain such characters in 
>HTML
></quote>
>
>which is clearly not your purpose in handling the incoming data, 
>particularly when it is binary data.
>
>At 08:08 26/06/2002 +0000, Bjarne Christiansen wrote:
>>Hi,
>>I have some problems uploading binary files though the web browser. It 
>>seems to work fine when uploading ASCII file but but binary files seems to 
>>be corrupted. The beginning of the binary file seems fine, but some data 
>>is missing....
>>
>>Here is my file upload script:
>>
>>#!c:\Python\python -d
>>
>>import cgi
>>
>>print "content-type: text/html\n\n"
>>
>>form = cgi.FieldStorage()
>>if not form:
>>  print """
>><form action="http://localhost/cgi-bin/upload.py" method="POST" 
>>enctype="multipart/form-data">
>><input type="file" name="filename">
>><input type="submit">
>></form>
>>  """
>>elif form.has_key("filename"):
>>  item = form["filename"]
>>  if item.file:
>>    data = item.file.read()
>>    print cgi.escape(data)
>>    data1 = cgi.escape(data)
>>    f = open("file1.jpg","wb")
>>    f.write(data1)
>>
>>Any help will be greatly appriciated!
>>
>>Best Regards,
>>Bjarne Christiansen
>>bjarne_christiansen at hotmail.com
>>
>>
>>_________________________________________________________________
>>MSN Photos is the easiest way to share and print your photos: 
>>http://photos.msn.com/support/worldwide.aspx
>>
>>
>>
>>--
>>http://mail.python.org/mailman/listinfo/python-list
>
>
>
>--
>http://mail.python.org/mailman/listinfo/python-list


_________________________________________________________________
Send and receive Hotmail on your mobile device: http://mobile.msn.com





From arlo.maillist at ozmaker.com  Sun Jun  9 11:28:02 2002
From: arlo.maillist at ozmaker.com (Arlo)
Date: Sun, 09 Jun 2002 23:28:02 +0800
Subject: Embed python in multi-thread environment
Message-ID: <3D037402.1070301@ozmaker.com>

hi, all

i've tried to embed python in multi-thread application( written in c++)
the flow is:

1.create socket,listen,accept.
2. create thread to handle different clients.
3. in thread func, use Python C API to import my python script, and use 
PyObject_CallObject(func,args);
    to call python function ( just passing socket descriptor which got 
from accept() to this function)

but when it just one client, it works perfect, but when more than one 
client, it core dumped.
it seems i can't call PyObject_CallObject twice at the same process?







From pedro_rodriguez at club-internet.fr  Sun Jun 30 17:12:13 2002
From: pedro_rodriguez at club-internet.fr (Pedro Rodriguez)
Date: Sun, 30 Jun 2002 23:12:13 +0200
Subject: aspect-oriented demo using metaclasses
References: <mailman.1025459319.8164.python-list@python.org>
Message-ID: <pan.2002.06.30.21.12.12.836708.5173@club-internet.fr>

[Mark]
> The latest version I posted continues to punt on staticmethod and
> classmethod.  They're currently not aspected because my type comparison
> to type(lambda x:x) will effectively filter out staticmethod and
> classmethod. I would have to change the observer interface slightly
> since staticmethods don't have have a self parameter and the first
> parameter of classmethods is a reference to the class not the instance.
> I don't see this being too hard--I plan to fix it in the next iteration.
> 
>> Even if deprecation of the types module have been discussed on the
>> devel list, I think it is preferable.
> 
> Yeah, using type(lambda x:x) seems more portable or something.
> 

Sorry, but I wasn't clear : I meant that using the types module should
be better. But honnestly I am not that sure.


[Mark]
> Hmm, you're right, my approach will do NOTHING for methods that aren't
> associated with a class.  BTW, I thought a normal method associated with
> a class whether new type or old was a bound method?  I'm not too worried
> about old style classes--although maybe I should be?  I figure if a
> requirement of using this is that it only works with new-style classes,
> that's fine with me.  Until I run into a problem, of course.  ;-)
> 

Just to be sure, let me clarify some points about functions and methods :
- a def statement introduces a function
- when a function is declare in a class we should talk of an unbound 
  method
- when we make a reference to a function from an instance we talk of a
  bounded method

>>> def f(x): pass
...
>>> class A:
...     def f(self): pass
...
>>> a = A()
>>>
>>> print f
<function f at 0x8150854>
>>> print A.f
<unbound method A.f>
>>> print a.f
<bound method A.f of <__main__.A instance at 0x8150824>>
>>>

The difference between an unbound method and a bound method is that the
latest won't need to be passed a 'self' argument because it is already
bounded to an object.

For class and static methods, I will need some time to learn how to
identify them and catalog them in my 'bestiaire'.


[Mark]
> I often use methods that aren't associated with a class and I'd like to
> aspect those as well, although, when I think about something like
> Persistence, that does seem initimately bound up with classes.  So maybe
> it depends on the aspect?

I see your points here. 

Could/Should an aspect be used for Persistence ?  I don't know, but it is 
a good remark. 

I recall that I discovered AOP after having to add some notification in
Models in an Model/View/Presenter framework and it seemed quite intrusive.

With your remark I wonder if the saving of my models (call it
Persistence/Serialization) should also be considered as a good candidate
for being aspectified. By doing so that I could keep my model simple and
externalize the way I want to implement save/load.


> 
> I should say that my motivation for pursuing a metaclass implementation
> is partly just to learn metaclasses, 

Understanding metaclasses is quite a challenge, putting them to good use
is, IMO, even more challenging (Alex Martelli recommended the reading of
the book mentionned in the 'descintro' document :
    Putting Metaclasses to Work: A New Dimension in Object-Oriented Programming,
    by Ira R. Forman and Scott H. Danforth)


> ... but also partly because I want
> something that requires no modification to the aspected class and
> minimal effort, maximum flexibility for defining join points.  What I
> like about the metaclass approach is that I just wire up every single
> class that's aspected for all events.  It's then up to the observer to
> hook into them.  Of course, I need to try to use this for something real
> and I'm sure that will expose its weaknesses/my conceptual gaps quickly.
> 

I think you should clearly separate :
1. the place where you do your 'method call interception'
2. the way you stack (and call) advices

Actually you do 1. at class definition level for all methods defined in
the class by using a metaclass, and for 2. you use an observer-like
object.

For 1. I decided to it in a different way, closer to Java implementations
(and I think that aop.py in Pythius already did it this way) :
- I created a Joinpoint class
- the constructor received an object called a callableSet and a string
  pattern
- a callableSet was anything that could contain a 'function' I wanted to
  intercept. This could be a module, a class, an instance
- I scaned the callableSet in search for attributes that were functions
  that matched the pattern
- I intercepted the matching functions with what I defined for 2.


[Pedro]
>> Too intrusive. I don't believe that you can do it dynamically, at least
>> not for classes defined at module level. They will be created with the
>> metaclass defined at compilation time.
[Mark]
> 
> This is a good point.  Here's the sort of dynamism I'm aiming for:
> 
> The modules containing classes to be aspected have no reference to the
> framework inside them.  Not even a __metaclass__ statement.
> 
> Oops, a little testing shows that may simply not work.  Hmm.  I don't
> want to have to edit the modules that contain the classes to be
> aspected, even to add a __metaclass__ declaration.  Since I can't seem
> to change that at runtime, this approach probably won't work.
> 

This is what I feared with the usage of metaclasses (at least by solely
using __metaclass__ in this way)


[Pedro]
>> Yes. Interception of raised exception is a good (and easy ;) feature.
>> Just try to go a step further with 'around' methods.
[Mark]
> 
> I avoided adding around because it doesn't seem primitive to me.  In
> other words, isn't around just before + after notification?  So I could
> add support for around simply by making it so that observers wanting
> around notification just got before and after notification?  Not sure
> what to do if the method raises an error--skip the after?
> 
No 'around' is not that simple I fear. I think you could make some
comparison with 'generators'. One of the purpose of generators is to
simplify design by not having to keep a context between invocations.
The same thing occurs with 'around' in a less trivial way. 

Consider this : you want to track some log information on a method :
- when it started
- how long it lasted

So that you have a trace like :
Method XXX called for object YYY at HH:MM for XX seconds

Question : how will you track the start time so that you'll be able to
compute the duration and produce your log line ?
[Take into account that method XXX may be called recursively, or
concurently in a multi-threaded environment]

An 'around' advice will simplify your life as a user of aop (but not as
the designer of the aspect framework ;) this is were I ended with an
implementation that seemed quite complicate and that will need some rework
- and also why I didn't dare posting its UML description ;)

Bon courage,
Pedro


From gcordova at hebmex.com  Thu Jun 27 18:44:42 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 27 Jun 2002 17:44:42 -0500
Subject: Pythoniac: Thoughts on a hardware Python processor
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233015DEC3B@mail.hebmex.com>

> 
> Whoever asked if the Intel/AMD FP stack counted: nah - it's 
> not the primary mode of the CPU like it was on the HP 3000
> and the Forth chip. BTW, speaking of the Forth chip, I think
> Forth, Inc. is still offering the IP for that if anyone wants
> to burn a few hunderd k$ on new silicon.
> 

If I recall correctly, Chuck Moore's still churning out
forth processors, with on-chip fast memory and hardware
stacks, really really cool stuff. I think there's more info
in the ColorForth site, or at least it points you in the
right direction.

> While I think that creating a Python chip is a great and 
> wonderful idea, I doubt it will or should happen:
> 
> [... snip ...]
> 
> For all these reasons and more, I think a good compiler with 
> a retargetable backend is a better expenditure of time/effort
> and would yield a result far faster then a silicon development effort.
> 
> Dave LeBlanc
> Seattle, WA USA
> 

That's what I meant!! This compiler you talk about should generate
Forth instead of the normal byte code. Hmmm... I think it's doable,
but I've yet to learn more forth.

Do you know any good resources? Online books? Good archives?
Anything?

:-)

-gus




From cunsan at lycos.com  Wed Jun 19 06:19:14 2002
From: cunsan at lycos.com (cunsan)
Date: 19 Jun 2002 03:19:14 -0700
Subject: posting message
Message-ID: <54f48551.0206190219.25d6d269@posting.google.com>

hi, 

sorry if this posting is not relevant to this mailing list. i want to
post using groups.yahoo.com/group/python-list (on the web). but it
always fails (the target email seems not exist anymore). can anyone
tell me why? i mostly read the postings on the web. i sent this
message on google. the problem is that it's not so convenient as yahoo
:(.

thx.


From brueckd at tbye.com  Wed Jun 19 18:12:25 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 19 Jun 2002 15:12:25 -0700 (PDT)
Subject: What If.....    Strong Types
In-Reply-To: <3D10D17B.6070004@bgb.cc>
Message-ID: <Pine.LNX.4.44.0206191501520.11198-100000@ns0.tbye.com>

On Wed, 19 Jun 2002, Don Garrett wrote:

> > It's sometimes hard for me to model my objects this way because of having
> > had to always do it the other (completely statically defined) way, but
> > when this approach works the solutions are often quite elegant and
> > straightforward (btw - I believe that a good portion of bugs and program
> > cruftiness is related to forcing the programmer to clearly define, up
> > front, what the complete type of the object is. In many cases this forces
> > the programmer to make decisions without enough information, in others it
> > is too constricting, and in others it's just too tedious to make the
> > programmer worry about all those details, and in all cases general
> > hackiness ensues).
> > 
[snip]

>    Don't you find that you miss the advantages of a clear definition of
> what a class is and/or does? That's more important with group
> development, but even when it's just me.... I need memory aids.

A clear definition... as in a completely specified data type? ;-) It's the 
same issue of not being required to supply any more information than you 
really have.

I know what you mean though, but it's not like your objects are going to
be completely devoid of form or description when you use the dynamic
techniques. And like any other tool, when used in excess it doesn't really
help (i.e. in reality many/most data members of most classes won't be
dynamically added and removed so it's not too big of a deal).

In the cases I've used dynamic data members + group development, 99% of
the time the dynamic members were pretty private to my class, and the one
or two that weren't were covered in the test cases, which is where people
go to look to know how to use the class anyway.

Have fun,
-Dave





From gerhard at bigfoot.de  Tue Jun  4 07:55:17 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Tue, 4 Jun 2002 13:55:17 +0200
Subject: Creating explorer link on Win32
In-Reply-To: <3CFCB72E.4010405@haftmann-online.de>
References: <3CFCB72E.4010405@haftmann-online.de>
Message-ID: <20020604115516.GB704@lilith.my-fqdn.de>

* Florian Fredegar Haftmann [2002-06-04 13:48 +0100]:
> Hi!
> 
> I'm searching for a possibility to create an explorer link (*.lnk) to a 
> file on Win32. Has anyne heard of a python module able to do that in a 
> simple way?

Mark Hammond's win32 extensions. On my homepage, under VitaminP, there
is a pure-C solution, which you can live with software explictely marked
as 'experimental'. IOW, I'd go with the win32 extensions.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From sandeep182 at hotmail.com  Mon Jun 24 17:27:30 2002
From: sandeep182 at hotmail.com (Sandeep Gupta)
Date: 24 Jun 2002 14:27:30 -0700
Subject: Pychecker and /var/tmp/python2-2.2-root/usr/lib/python2.2/.....
Message-ID: <d71cc7f6.0206241327.17ebabfa@posting.google.com>

I'm using pychecker on my application, which uses BaseHTTPServer.py
and unittest.py.  Both those files have "No method" warnings that are
displayed as:
/var/tmp/python2-2.2-root/usr/lib/python2.2/BaseHTTPServer.py and
/var/tmp/python2-2.2-root/usr/lib/python2.2/unittest.py

I want to be able to ignore the "No method" warnings in those two
files, but not in other files.

I've tried both blacklist and ignoreStandardLibrary, but neither
works:
1)before removing warnings for blacklisted modules, pychecker checks
to see if the module listed in blacklist is valid. 
/var/tmp/python2-2.2-root/usr/lib/python2.2/BaseHTTPServer.py actually
doesn't exist (I don't understand why pychecker thinks that is the
fully qualified location of the module.  I would expect the module
name to be /usr/lib/python2.2/BaseHTTPServer.py ).  Blacklisting
/usr/lib/python2.2/BaseHTTPServer.py doesn't work either since it
doesn't match /var/tmp...../BaseHTTPServer.py.
2)On my python installation, ignoreStandardLibrary only checks that
the prefix of the module is /usr/lib/python2.2.  Again, that doesn't
match.

Does anyone know why /var/tmp/python2-2.2-root gets prepended to the
file name?
How can I remove the warnings for these files?

Thanks
Sandeep Gupta


From anton at vredegoor.doge.nl  Sat Jun 22 15:06:03 2002
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Sat, 22 Jun 2002 21:06:03 +0200
Subject: [ANN] cdrom catalog scripts
Message-ID: <af2i0l$27a$1@news.hccnet.nl>

Name        : Cdrom catalog scripts.
Description : Some scripts to browse directories offline. Zip archives are
              treated as subdirectories.
Dependencies: Python distribution with idle
Tested      : On windows 2000 and 98 SE with python 2.2
Audience    : Hobby programmers with a low budget. For example I myself do
              a lot of archiving and backupping by packing files into a 
              zip archive and later on I burn these files onto a cdrom.
              There's a nice little tool called catfish to inspect 
              directory listings of cdroms but it doesn't look into 
              zipfiles. That's why I wrote this utility. Of course now 
              that its there there's no reason anymore for those 
              applications that do exactly this to hide themselves for me.
              That's another reason why I wrote this.
Warning     : This is just a first trial so don't be surprised if 
              something goes wrong. Use at your own risk, provide feedback
              to make it better.
Usage       : Just put a CD in the drive, run dirdict.py and point it to 
              the CD or any other drive or directory structure that needs
              to be stored. Have patience, it may take a while. The 
              program returns to ask for a place to store a file. Make
              sure the extension is ".zid". Then run ziptree.py to open
              the zidfile that was just created.

The cdromcatalog scripts are:

dirdict.py  : A utility to store a directory structure in a file with
              extension ".zid", Zip archives are treated as 
              subdirectories, but zipfiles inside zipfiles are not -yet-
              expanded.
ziptree.py  : A utility to view .zid files. It can be run without 
              arguments it will then ask for a filename to open, this file
              should be produced first by using dirdict.py. Alternatively
              it could be used with a file type association. The 
              commandline should then be something like this (adjust for
              the cdromcatalog directory):
              d:\python22\pythonw.exe f:\cdromcatalog\ziptree.py "%1"

License     : BSD style, use at your own risk. 
Contact     : For questions and comments I can be reached at 
              <anton at vredegoor.doge.nl>
Author      : Anton Vredegoor, july 2002
Download    : http://home.hccnet.nl/a.vredegoor/cdromcatalog/cdromcatalog.zip


From dfackrell at DELETETHIS.linuxmail.org  Thu Jun 13 13:51:26 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Thu, 13 Jun 2002 11:51:26 -0600
Subject: system command
References: <aeajuh$5b5eo$1@ID-108351.news.dfncis.de>
Message-ID: <3d08dba0_2@hpb10302.boi.hp.com>

Assumption: You are referring to the "source" command as provided by the
bash shell on some *nix.

The man page for bash suggests that you probably want:

print "source "+root+".in"
os.system("bash "+root+".in")

--
Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.

"Thor" <thor__00 at yahoo.com> wrote in message
news:aeajuh$5b5eo$1 at ID-108351.news.dfncis.de...
> Having this code
>
>   print "source "+root+".in"
>   os.system("source "+root+".in")
>
> the script does not run the root+".in" file, but copying and pasting the
> output form teh 1st line work.. what i'm doing wrong? Thanks in advance
> --
> Thor -- Stockholm -- Sverige




From m2 at plusseven.com  Mon Jun 17 10:32:06 2002
From: m2 at plusseven.com (Alex Polite)
Date: Mon, 17 Jun 2002 16:32:06 +0200
Subject: Python and SQL Server 2000
In-Reply-To: <slrnagn75g.263.kzaragoza@localhost.localdomain>
References: <mailman.1024112300.2184.python-list@python.org> <slrnagn75g.263.kzaragoza@localhost.localdomain>
Message-ID: <20020617143206.GA29257@matijek.plusseven.com>

On Sat, Jun 15, 2002 at 02:56:02PM -0500, Kris J. Zaragoza wrote:
> My first instinct on seeing this reply was to point out that win32com
> doesn't run on Red Hat Linux. :-) Then I got to thinking a bit more
> about it and realized this may actually be possible.

Guess your right. This was a long time ago.

> So, let me add to the original poster's question: Has anyone ever
> tried running the Windows version of Python (either the main
> distribution or ActiveState's) under Wine to take advantage of the
> database drivers, etc.?

Wouldn't want to try it. The socket server idea that the other guy
posted is probably the way the go.


-- 
Alex Polite
http://plusseven.com/gpg/




From sandeep182 at hotmail.com  Wed Jun 19 10:50:30 2002
From: sandeep182 at hotmail.com (Sandeep Gupta)
Date: Wed, 19 Jun 2002 14:50:30 GMT
Subject: Blank results using mxODBC against 
Message-ID: <WS0Q8.7228$Tu6.481485@typhoon.austin.rr.com>

We are trying to connect to MS SQL Server from Linux, but we are only
receiving blank values in the query results.  The number of rows and columns
in the result is correct, but each result cell is blank.

As a test, we successfully used mxODBC against a local MySQL going through
unixODBC.

We then tried: Python2.2 --> mxODBC --> unixODBC --> inline TDS 1.6 --> MS
SQL Server
When I use isql with unixODBC, I am able to view the results of a simple
select.  That same select in mxODBC returns blank result values.

The python code I'm using is:
import mx.ODBC.unixODBC
db = mx.ODBC.unixODBC.DriverConnect('DSN=MSSQL')

c = db.cursor()

c.execute('select * from dbo.Source;')

mx.ODBC.print_resultset(c)


odbcinst.ini contains
==================
[SQLServer]
Description             = SQL Server driver for Linux
Driver          = /usr/lib/odbc-i02.so
FileUsage               = 1
==================

odbc.ini contains
==================
[MSSQL]
Description             = MSSQL Test DB
Driver          = SQLServer
Server          = <ip>
Port            = 1433
User            = <user>
Pass            = <pw>
Database                = <database>
majorver    = 5
minorver    = 0
Trace           = On
TraceFile               = /tmp/DB.out
==================


Thanks
Sandeep




From debl2nononospammywhammy at bellatlantic.net  Sat Jun  1 18:24:28 2002
From: debl2nononospammywhammy at bellatlantic.net (David Lees)
Date: Sat, 01 Jun 2002 22:24:28 GMT
Subject: OverflowError: math range error ???
References: <ad8bkk$va3sb$1@ID-146704.news.dfncis.de> <ad8d7k$v8m1s$1@ID-146704.news.dfncis.de>
Message-ID: <3CF9499E.F5365921@bellatlantic.net>

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

david lees


Shagshag13 wrote:
> 
> What are "the known limits" of math modules ?
> 
> Where can i find more infos on this ?
> 
> Thanks,
> 
> s13.


From gleki at gol.ge  Mon Jun 24 05:19:09 2002
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Mon, 24 Jun 2002 11:19:09 +0200
Subject: Synchrinized chat?
Message-ID: <3D16E40D.500A5DD4@gol.ge>

Hi all!

Is somebody making a synchronized chat in Python? I don't need the
script of the chat itself, rather a learning example of a real-world
synchronized (multi)threading.


Thank you in advance.

Grtz,
Giorgi



From Johannes.Nix at mail.uni-oldenburg.de  Tue Jun  4 04:57:49 2002
From: Johannes.Nix at mail.uni-oldenburg.de (Johannes Nix)
Date: 04 Jun 2002 10:57:49 +0200
Subject: Kalman Filters, Particle Filters and SVM's
Message-ID: <671ybne7zm.fsf@mail.uni-oldenburg.de>

Are there any implementations of Kalman Filters, Particle Filters or
Support Vector Machines in (Numerical) Python ?

Especially the latter would be fine, I'm dreaming to implement a SVM
based script with the capability to distinguish spam from personal
mails.... but I'm way to lazy to write my own SVM.

Johannes



-- 
--------------------------------------------------
Johannes Nix

jnix at medi.physik.uni-oldenburg.de

AG Medizinische Physik       Fax +441 798-3902
Universit?t Oldenburg        Tel +441 798-3270
26111 Oldenburg
--------------------------------------------------


From tjreedy at udel.edu  Mon Jun 10 12:20:25 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 10 Jun 2002 16:20:25 GMT
Subject: [ANN] constraint-0.2
References: <mailman.1023468735.788.python-list@python.org> <slrnag50sd.gac.mlh@vier.idi.ntnu.no> <slrnag8to5.dr.alf@virgo.logilab.fr>
Message-ID: <dl4N8.181687$Kp.17633467@bin7.nnrp.aus1.giganews.com>

To paraphrase Magnus Lie Hetland, the general constraint-propagation
system constraint-0.2 takes 100 times as long to do the 8-queens
problem as the specific backtracking queens.py found in the standard
Python distribution: 15 versus .1 sec on comparable machines.

The is about the same comparison, for some computations, as between
generic/dynamic Python and type-specific C.  In both cases, there are
two more questions of practical interest:  which platform/method is
faster for the *programmer*?  and, is the slow method fast enough?

So, what is the relative programming speed of the general constraint
propagation system versus specific backtracking?  I suspect that some
people could whip up something like queens.py as fast or faster that
they could the (unpublished?) constraint input for the same problem.
However, I also suspect the reverse would be true for the more
realistic and useful scheduling example that Fayolle used to
illustrate and motivate constraint-0.2.
http://www.logilab.org/python-logic/documentation.html

Real world problems may have no solution as given.  On the other hand,
constraints may not be absolute.  In the scheduling example, two
workshops *can* be scheduled simultaneously, if necessary, even though
many people want to take both.  So a stepwise procedure may be needed.
(Given multiple solutions, one might also want to add 'wishes' as
pseudoconstraints to help select just one.)  So, is it equally easy,
with the two approaches, to add or drop constraints and rerun?  This
would be fairly simple with c-0.2 input but probably harder for
backtracking code unless planned for from the beginning.  Is it
equally easy to give constraints priorities and have the system
automatically either add them in order from the highest or drop them
from lowest?  Constraint-0.2 evaluates constraints in the order
listed, but I do not know if  it properly backtracks from failure to
report the most satisfying solutions.

Terry J. Reedy





From troels at 2-10.org  Tue Jun 18 04:38:49 2002
From: troels at 2-10.org (Troels Therkelsen)
Date: Tue, 18 Jun 2002 10:38:49 +0200
Subject: python version?
In-Reply-To: <S0CP8.44419$GY.13868719@twister.nyroc.rr.com>
Message-ID: <NFBBIAKAGLNEMNJKNCLBIEAKCBAA.troels@2-10.org>

> I remember once I tried to explain to my students the theory 
> behind the limit in
> calculus.  To 35 students.  Maybe 3 understood it in such a way 
> that they could
> present a solution back to me.  But most just didn't want to 
> know.  I was wrong.
> I was not teaching Mathematicians but Buisness students that just 
> needed the
> credit.  Now I am not someone who necessarily wants to code using 
> Python in ASP.
> I just want some code to find my version of Python using ASP.  
> I'm sorry if I
> have not lived up to your expectations.
> 
> --
> George Hester

The document on this URL might help explain why you feel you have not
gotten the answer to the question you thought you asked :-)

http://www.tuxedo.org/~esr/faqs/smart-questions.html

Regards,

Troels Therkelsen





From peter at engcorp.com  Sun Jun 30 12:12:53 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 30 Jun 2002 12:12:53 -0400
Subject: [Newbie] Is the text delimiter changeable?
References: <3d1f26a4.5915886@news.texas.net>
Message-ID: <3D1F2E05.93926893@engcorp.com>

Ben Fairbank wrote:
> 
> I have to process large amounts of text and the text includes numerous
> apostrophes and quotation marks.  It would be very handy if I could
> temporarily assign a substitute for Python's use of the caracters "
> and ' to delimit text strings.  If I could use, for example, the @
> character or the # character, neither of which appears in the text I
> have to work with, and if I could disable the " and ' delimiters, then
> my task would be greatly simplified.  I realize I can backslash escape
> these characters in some circumstances, but I do not think that will
> work when a user's input is a word such as "can't" or "don't."
> Anyway, short of doing something truly drastic, such as recompiling
> the system (which I have no intention of doing), is there a workaround
> that will permit this?

Do you know about Python's triple-quoted strings?

>>> a = """This isn't much of a "string" now, is it?"""
>>> b = '''This can't have "triple-quotes" inside, of course, but that's okay.'''

-Peter


From debl2Nonospammywhammy at bellatlantic.net  Sun Jun  9 02:24:54 2002
From: debl2Nonospammywhammy at bellatlantic.net (David Lees)
Date: Sun, 09 Jun 2002 06:24:54 GMT
Subject: Determination of screen resolution?
Message-ID: <3D02F4C8.C2DF2EE0@bellatlantic.net>

I would like to use tkinter and have the canvas object automatically
scaled fit the screen.  Is it possible to read the screen resolution
(800x600, 1024,768...) from somewhere.

David Lees


From cliechti at gmx.net  Mon Jun 10 19:40:35 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 11 Jun 2002 01:40:35 +0200
Subject: Iterators vs. Generators
References: <ae3a6f$9d9$1@panix1.panix.com>
Message-ID: <Xns922A1168B60A2cliechtigmxnet@62.2.16.82>

aahz at pythoncraft.com (Aahz) wrote in news:ae3a6f$9d9$1 at panix1.panix.com:

> So when would one actually write an iterator instead of a generator?
> I've been trying to think of an example and failing.

don't know, maybe when you don't like the additional loop in the generator 
(see example below)

maybe the iterator has advantages when the data in the object changes or 
has to be chaged as the object q (from below) _is_ the iteraror but it's 
_not_ the generator (but this does not matter in the example below ...).
if you pass "i = iter(q)" somewhere then with the iterator (and forget 
about q) you can still modify the original object as "i is q" (at least in 
my example), but this is not even possible with a generator.

iterators can be generated implicit by __iter__ in a "for", but a generator 
needs a factory function (of course that can be named __iter__ ...)

seems to be a matter of taste...

chris

PS: why is Queue not iterable by default?
here's a Queue for fun <wink>:
----

#!/usr/bin/env python
from __future__ import generators

import Queue, threading, time

class Q(Queue.Queue):
    def __iter__(self):
        return self
    def next(self):
        x = self.get(1)
        if x is None: raise StopIteration
        return x

class Q2(Queue.Queue):
    def __iter__(self):
        while 1:
            x = self.get(1)
            if x is None: return
            yield x
    

class Producer(threading.Thread):
    def __init__(self, queue):
        threading.Thread.__init__(self)
        self.q = queue
    def run(self):
        for n in range(10):
            time.sleep(0.2)
            self.q.put(n)
        self.q.put(None)

if __name__ == '__main__':
    #iterator
    q = Q()
    print ":", q is iter(q), type(iter(q))
    Producer(q).start()
    for x in q:
        print x
        
    #generator
    q = Q2()
    print ":", q is iter(q), type(iter(q))
    Producer(q).start()
    for x in q:
        print x

-- 
Chris <cliechti at gmx.net>



From stephen at theboulets.net  Fri Jun  7 22:31:38 2002
From: stephen at theboulets.net (Stephen Boulet)
Date: Fri, 07 Jun 2002 21:31:38 -0500
Subject: Confidence intervals, stats modules?
Message-ID: <ug2r3aac4etnef@corp.supernews.com>

Are there any stats modules to calculate things like confidence intervals, 
mean, standard deviation? Thanks.

-- Stephen


From radix at twistedmatrix.com  Mon Jun  3 08:47:21 2002
From: radix at twistedmatrix.com (Christopher Armstrong)
Date: 03 Jun 2002 08:47:21 -0400
Subject: Interrupting a continuous script
In-Reply-To: <_6JK8.1535$376.74360@newsfep1-win.server.ntli.net>
References: <_6JK8.1535$376.74360@newsfep1-win.server.ntli.net>
Message-ID: <87u1okmsva.fsf@twistedmatrix.com>

>>>>> "Julian" == Julian Gollop <julian.gollop at ntlworld.com> writes:

    Julian> I have a few programs continually processing text files on a Linux
    Julian> machine.  Occasionally I need to stop them, but I don't want to use
    Julian> a keyboard interrupt because it may cause the script to stop in the
    Julian> middle of doing something.  Does anyone know a simple way to allow
    Julian> me to stop the process by pressing a key on the keyboard - without
    Julian> stopping inside something crucial?

    Julian> Julian.

>>> import signal
>>> def handleBreak(sig, frame):
...     print "I caught a break!"
...
>>> signal.signal(signal.SIGINT, handleBreak)
<built-in function default_int_handler>
>>> while 1: pass
...
I caught a break!
I caught a break!
I caught a break!

(I pressed C-c 3 times)

Your handleBreak method should set a flag somewhere so your processing
code knows it should stop and clean up whenever it gets the chance.

In case you're considering just catching KeyboardInterrupt exception: don't.
It's not the same as using a signal handler, bacause it will still interrupt
execution and jump to your 'except' block. (with signals, after running thi
handler, it goes back to the code it was running last)
-- 
                                Chris Armstrong
                         << radix at twistedmatrix.com >>
                http://twistedmatrix.com/users/carmstro.twistd/





From phawkins at connact.com  Thu Jun 20 13:14:39 2002
From: phawkins at connact.com (Patricia J. Hawkins)
Date: 20 Jun 2002 13:14:39 -0400
Subject: GUI (PyQt) & thread problem
References: <8b5d764c.0202240451.1c4bda3b@posting.google.com> <a5c8a2$r5e$1@news.tudelft.nl> <8b5d764c.0202250020.268f4a81@posting.google.com>
Message-ID: <wk8z59eutc.fsf@mail.connact.com>

>>>>> "KN" == Kristian Nylund <krnylund at msn.com> writes:

KN> Thank you for your reply. Do you have any pointers on how to implement
KN> this kind of functionality, i.e. when something happens in the worker
KN> thread (not GUI related), the GUI needs to be updated. Is it at all
KN> possible?

Boudewijn Rempt has added a QT modification to an Activestate recipe
for using Tkinter and threads -- scroll down to see his version.  Note
that his version changes the GUI based on thread results, though the
original Tkinter example doesn't:

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/82965

-- 
Patricia J. Hawkins
Hawkins Internet Applications, LLC





From kragen at pobox.com  Thu Jun  6 14:15:55 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 06 Jun 2002 14:15:55 -0400
Subject: self
References: <mailman.1023216606.793.python-list@python.org>
Message-ID: <83g000e0is.fsf@panacea.canonical.org>

"Sean 'Shaleh' Perry" <shalehperry at attbi.com> writes:
> it sees 'a*c' and asks it own dict 'do we know a?' if so it access dict['a']
> and returns its value.  Looking at your example you seem to be implementing a
> symbolic calculator.  Most implementations I have seen involve writing a small
> language with interpreter.

It's a shame that this is necessary, since we already have a perfectly
good language with an interpreter already present.



From jubafre at zipmail.com.br  Wed Jun 19 00:38:45 2002
From: jubafre at zipmail.com.br (jubafre at zipmail.com.br)
Date: Wed, 19 Jun 2002 01:38:45 -0300
Subject: =?iso-8859-1?Q?how=20i=20use=20=27=E3=27=20in=20python=3F=3F=3F?=
Message-ID: <3D0F6623000017AA@www.zipmail.com.br>

i?m brazilian, and in my laguage(Portguese) there are many other caracters
how ?, ?, ? and a string in python doesn?t support this
how i can use??? have a import module for this???

BRASILEIROS DA LISTA COMO VCS CONSEGUEM USAR ACENTO NO PYTHON??

for exmple:
s='?'
doesn?t work?why??


Juliano Freitas
www.gebrasil.hpg.com.br



------------------------------------------
Use o melhor sistema de busca da Internet
Radar UOL - http://www.radaruol.com.br







From ballabio at mac.com  Thu Jun 27 15:38:03 2002
From: ballabio at mac.com (Luigi Ballabio)
Date: Thu, 27 Jun 2002 21:38:03 +0200
Subject: squared functions--most Pythonic way?
Message-ID: <ballabio-E54F22.21380327062002@newssrv.fastweb.it>

Hi all,
      suppose we want to define a function which, given f(x), returns 
its square. There are a number of ways which come to mind, such as:

object-oriented programming:

    def fsquare:
        def __init__(self,f):
            self.f = f
        def __call__(self,x):
            return self.f(x)**2

functional programming:

    # with an internal named function
    def fsquare(f):
        def f2(x):
            return f(x)**2
        return f2

    # with lambda
    def fsquare(f):
        return lambda x: f(x)**2

and I'm sure people can come out with others that I didn't figure but 
are obvious to any Dutchmen.

My question: is any of the above considered more Pythonic? 

Thanks,
         Luigi


From eric at enthought.com  Tue Jun  4 22:51:02 2002
From: eric at enthought.com (eric jones)
Date: Wed, 05 Jun 2002 02:51:02 GMT
Subject: weave.accelerate?
References: <c3835e5f.0206040624.7054ff3c@posting.google.com>
Message-ID: <q0fL8.71094$9z5.4586866@typhoon.austin.rr.com>

The CVS version has accelerate in it.  You can get a nightly snapshot here:

http://www.scipy.org/site_content/download_list

Take note that accelerate is still minimally capable.  It is unlikely to be
useful for real work.

eric


"Chris Fonnesbeck" <spam at fisher.forestry.uga.edu> wrote in message
news:c3835e5f.0206040624.7054ff3c at posting.google.com...
> I'm trying to locate a version of weave that includes the accelerate()
> function; the version posted on scipy.org doesnt seem to have it. Can
> anyone point me in the right direction.
>
> Thanks,
> cjf




From peter at engcorp.com  Sun Jun  9 13:02:17 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 09 Jun 2002 13:02:17 -0400
Subject: libftp.size function gives wrong results!
References: <RNKM8.560$Yw3.96921@news20.bellglobal.com>
Message-ID: <3D038A19.D991A885@engcorp.com>

Michael Davis wrote:
> 
> I'm writing a python program which will upload or download files via ftp.
> It's intended to be used to deploy a web site from a local machine to a
> server. It tries to determine which files are modified on the local machine
> by comparing file sizes and modification times.
> 
> I try to compare files using os.stat for the local file and FTP.size for the
> remote. The problem is, FTP.size is giving me incorrect results. If I use
> the ftp command line client, and look at one of my files on the server (the
> output looks like 'ls -l') then I can see that the file size, is, say,
> 1200. But the FTP.size function returns, say, 1242. 

I think you'll find the number of lines in that file is 42.  The difference
is probably between CRLF and LF line terminations in the file.  If you
transfer files with FTP and want to preserve the line endings, you have
to specify the binary type, not ascii.  Don't know how to do this with
ftplib, but from the ftp command line you would use 'type binary' or
'type ascii' to switch.

-Peter


From david_griswold1 at yahoo.com  Fri Jun 28 23:49:48 2002
From: david_griswold1 at yahoo.com (David Griswold)
Date: Sat, 29 Jun 2002 03:49:48 GMT
Subject: I'm an idiot
References: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16> <mailman.1025316799.2919.python-list@python.org>
Message-ID: <Xns923BD3E551D3Ddavidgriswold1netsca@204.127.68.17>

I had no idea it could be this compact.  Thank you for your help!

David

Carel Fellinger <carel.fellinger at iae.nl> wrote in
news:mailman.1025316799.2919.python-list at python.org: 

 and write it like this:
> 
>    g = open('c:/temp/temp1.txt', 'w')
>    for line in open('c:/temp/temp.txt'):
>        g.write(line.strip())
> 
> 



From maio at stoupa.sh.cvut.cz  Tue Jun 25 08:21:54 2002
From: maio at stoupa.sh.cvut.cz (Marian Schubert)
Date: Tue, 25 Jun 2002 12:21:54 +0000 (UTC)
Subject: Python daemon
References: <3d182dd6$1@news.mt.net.mk> <mailman.1024998812.26782.python-list@python.org>
Message-ID: <slrnahgo1j.5ir.maio@stoupa.sh.cvut.cz>

On Tue, 25 Jun 2002 10:52:49 +0100, Dave Swegen <dswegen at software.plasmon.com> wrote:
>>  4. Logging
>
>Don't know what the best python way of doing this is.

http://www.red-dove.com/python_logging.html
pretty nice logging system


CU,
maio


From cliechti at gmx.net  Sat Jun  8 13:37:46 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 8 Jun 2002 19:37:46 +0200
Subject: Q: scope of module functions
References: <adte4c$nu8$1@f1node01.rhrz.uni-bonn.de>
Message-ID: <Xns9227C812CD4B2cliechtigmxnet@62.2.16.82>

"Frank Sonnenburg" <sonnenbu at informatik.uni-bonn.de> wrote in 
news:adte4c$nu8$1 at f1node01.rhrz.uni-bonn.de:
> a little tricky problem (at least from my point of view); given module
> CALLER:
> 
> # i need "fct" and some other from module "UTILS"
> from UTILS import *
> fct()
> # END of module CALLER
> 
> 
> Now, inside fct(), how do i determine the module - in this case CALLER -
> from where fct() was called?

i think __name__ can help in this case but not in the one below.
but are you sure that you wan't t do this? a function that behaves 
differently depending on the namespace gives very hard times debugging...

a cleaner solution would be to create an class that contains fct and each 
module that uses it can create its own object.

if you can say what you want to achieve, we might give you a more specifuc 
advice.

> Even concerning cases like (continuation of example):
> 
> # Now we are in __main__
> import CALLER
> fct_in_main = CALLER.fct
> fct_in_main()


-- 
Chris <cliechti at gmx.net>



From gerhard.haering at gmx.de  Sun Jun 23 12:54:29 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sun, 23 Jun 2002 18:54:29 +0200
Subject: Newbie question (anyone programming with curses?)
In-Reply-To: <af4t70$1igl$1@pingwin.acn.pl>
References: <af4ovd$1bt3$1@pingwin.acn.pl> <ummR8.290538$352.27434@sccrnsc02> <af4t70$1igl$1@pingwin.acn.pl>
Message-ID: <20020623165429.GA600@lilith.my-fqdn.de>

* hellboy <rkr1410 at poczta.onet.pl> [2002-06-23 18:32 +0200]:
> Emile van Sebille <emile at fenx.com> wrote:
> 
> > No one has made a Windows port of the curses module. On a Windows
> > platform, try the Console module
> 
> 8<--cut--8<
> 
> So there is no way to get fast console input/output + colors in text mode
> and code portability?
> 
> btw. there actually is a windows implemetation of curses
> (http://pdcurses.sourceforge.net), but it is written in c (of course) and I
> am too inexperienced to say if there is some way of incorporating it into
> python, and how to do it....

I and somebody else failed at making the Python curses module compile
against pdcurses. The proper solution is to build ncurses on win32, then
just compile Python against it. ncurses currently works on Cygwin, and
AFAIR MS-DOS, too. A mingw32 port should thus be doable, too.

If you want curses on Windows now, you can just use the Cygwin port of
Python (install Cygwin, then just choose Python as a package during
installation).

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 31.3 ?C      Wind: 1.2 m/s




From dfackrell at DELETETHIS.linuxmail.org  Tue Jun 25 10:54:52 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Tue, 25 Jun 2002 08:54:52 -0600
Subject: How to stop a SocketServer?
References: <7xhejv9af2.fsf_-_@ruckus.brouhaha.com>        <3d1744d4$1_1@hpb10302.boi.hp.com> <mailman.1024958494.18742.python-list@python.org>
Message-ID: <3d18843d$1_3@hpb10302.boi.hp.com>

Skip,

Thanks!  This is exactly what I was looking for.

For lurkers and searchers, the Python library reference (thread objects,
currently 7.5.4) gives the following:

---
setDaemon(daemonic)
Set the thread's daemon flag to the Boolean value daemonic. This must be
called before start() is called.

The initial value is inherited from the creating thread.

The entire Python program exits when no active non-daemon threads are left.
---

Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.

"Skip Montanaro" <skip at pobox.com> wrote in message
news:mailman.1024958494.18742.python-list at python.org...
>
>     Daniel> I faced this problem with a simple chat server I wrote.  I
>     Daniel> expected that termination of the server with <ctrl-C> should
>     Daniel> have stopped it immediately, but it instead waits until some
>     Daniel> communication is received from each thread that was stopped at
a
>     Daniel> read().
>
> Just make all the non-main threads daemon threads.  I believe the main
> thread will be the one to catch the Ctrl-C.  It can then cleanup and exit.
> Assuming it's the only non-daemon thread, the application will also exit.
>
> --
> Skip Montanaro
> skip at pobox.com
> consulting: http://manatee.mojam.com/~skip/resume.html




From mikeb at mitre.org  Sat Jun  1 14:37:46 2002
From: mikeb at mitre.org (Mike Brenner)
Date: Sat, 01 Jun 2002 14:37:46 -0400
Subject: File Iterator - Print - Double New Lines
Message-ID: <3CF9147A.14926C08@mitre.org>

print "QUESTION: Why did it jump from NO LINE SPACING to DOUBLE SPACING?"
print "QUESTION: How to get SINGLE SPACING?"
print
print "Using Python 2.2.1, it does the same under win32all build 142 and "
print "      running it from a DOS box without the windows extensions."

def writer(list):
    f=open("tmp.tmp","w")
    for item in list:
        f.write(item)
    f.close()

def reader():
    f=open("tmp.tmp","r")
    for line in f:
        print line
    f.close()

print "trying it without the \\n"
writer(["a","b","c"])
reader()

print "trying it with the \\n"
writer(["a\n","b\n","c\n"])
reader()





From mgilfix at eecs.tufts.edu  Mon Jun 17 14:20:01 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Mon, 17 Jun 2002 14:20:01 -0400
Subject: Messing with GUIs via Python...
In-Reply-To: <ael553$3ai$1@coranto.ucs.mun.ca>; from kondwani@cs.mun.ca on Mon, Jun 17, 2002 at 02:34:07PM -0230
References: <ael553$3ai$1@coranto.ucs.mun.ca>
Message-ID: <20020617142001.F422@eecs.tufts.edu>

  I suggest searching the archives since this question is often
answered in this newgroup... But if you're too lazy, here's two links
to get you started:

  For basic gui apps:
    http://www.python.org/doc/current/lib/module-Tkinter.html

  For more advanced stuff:
    http://www.wxpython.org
    http://www.daa.com.au/~james/pygtk/

                    -- Mike

On Mon, Jun 17 @ 14:34, Kondwani Spike Mkandawire wrote:
> Howdy Folks is there a way to create simple GUI apps.
> using Python ( I have only used it for 4 months)...  If so
> is there a way of doing this without having it as a language
> binding of some graphics software (as OpenGL)...
> 
> Kondwani Spike Mkandawire
> CS Intern. Memorial University, Canada...
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
`-> (kondwani)

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From larooy at xtar.co.nz  Sun Jun 16 02:57:01 2002
From: larooy at xtar.co.nz (John La Rooy)
Date: Sun, 16 Jun 2002 18:57:01 +1200
Subject: I feel stoopid ... this code is to verbose
References: <3D0A1791.3060307@mxm.dk>
Message-ID: <20020616185701.147c820e.larooy@xtar.co.nz>

On Fri, 14 Jun 2002 18:19:29 +0200
Max M <maxm at mxm.dk> wrote:

> Hmm ... I am working on a problem. In Danish we have a number format 
> that looks like:
> 
> 42.000.000,00 which is 42 millions
> 
> So I need to insert dot's at every three character from end of the 
> string for the integer value of the number.
> 
> Let's discard decimal points and just focus on the meat. I have written 
> a funcion which works::
> 
> def stringSlicer(string, chunkSize=3):
>      chunkList = []
>      reverseString = list(string)
>      reverseString.reverse()
>      for i in range(0, len(string), chunkSize):
>          chunk = reverseString[i:i+chunkSize]
>          chunk.reverse()
>          chunk = ''.join(chunk)
>          chunkList.append(chunk)
>      chunkList.reverse()
>      return '.'.join(chunkList)
> 
> 
> print stringSlicer('42000000')
> 
>  >>> 40.000.000
> 
> I just find that it's a lot of code for such a little function an it 
> annoys my sense of aestetics. I have tried a lot of different approaches 
> including using zip on a list like ['','','.'], and other weird stuff :-)
> 
> I just cannot seem to find the nice 3-liner I expected from the 
> beginning. Has anybody got a better solution ? I thought somebody might 
> find it a fun exercise. Well I have...
> 
> 
> regards Max M
> 

As other people have pointed out already, the locale module is the way to go
but some people do find it a fun exercise too ;o)

(this breaks if you have more than 3 digits after the decimal point. what's
supposed to happen then anyway?)

import re
def StringSlicer(s):
	return re.sub("\\B(\d{3})(?=(\d{3})*(,\d*)?$)",".\\1",s.replace(".",","))

John


From K.Rdt at TU-Berlin.DE  Thu Jun 27 02:02:16 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Thu, 27 Jun 2002 08:02:16 +0200
Subject: background-processes with win-python?
Message-ID: <CHD61IFJFHECBNH41IC713WITQGFC.3d1aaa68@FRITZweb>

---------------------------------------------------------------------
Hi

Is there a possibility to execute system-processes
under windows-python in the background? So
I can do this
	system("plink  .. tu-berlin.de")
without waiting for the real connection.

		K at Rdt
---------------------------------------------------------------------







From mike.christie at athensgroup.com  Mon Jun 24 13:45:59 2002
From: mike.christie at athensgroup.com (Mike Christie)
Date: Mon, 24 Jun 2002 17:45:59 GMT
Subject: wxPython question: window closing event
Message-ID: <3D175AD7.9CB28305@athensgroup.com>

I have a specific and a general question about wxPython.  I have worked with
GUIs and event-handlers before, but am new to both wxPython and Python itself.

The specific question is: how do I attach code to the window closing event
that occurs when (in MS Windows systems) you click the X in the top right
corner of the screen?  It appears I need to place a "self.close(true)" there,
since exiting via an exit menu (which does call "self.close(true)") shuts
things down politely, but exiting via the X leaves the Python interpreter
still executing my script, and forces me to close Python to shut it down.

The general question is: what's the best resource for finding this sort of
thing out?  I've been working through the tutorial material at
http://wiki.wxpython.org/index.cgi/Getting_20Started which has been very
helpful indeed.  I've also made some use of the online documentation at
http://wxpython.org/onlinedocs.php which is also great.  However, for a
question like this, it seems like the right resource is something like a
categorization of events, or a reference guide for event-handling, or
something like that.   The online docs do have an "Event Handling Overview",
but I couldn't find this information there.  Is there a book or online
reference that I could use as a resource for this kind of question?

Thanks in advance for any help.

Mike


From prestomation at yahoo.com  Sun Jun 23 20:43:31 2002
From: prestomation at yahoo.com (Preston Tamkin)
Date: Mon, 24 Jun 2002 00:43:31 GMT
Subject: HTTP authentication
Message-ID: <TWtR8.86654$Am4.23833590@twister.columbus.rr.com>

hello all. I've just recently jumped on the python bandwagon, and I need
some help!
I'm trying to write a program(in python) to go out to several websites and
get a few numbers off of the page. Only problem is, I need to give the site
a username and password! It's not HTTPS, SSL, or anything like that.  I did
some packet sniffing and it is a HTTP POST command, the server then requests
the information then the client sends the u/p. I'm using raw sockets and
can't figure out how to listen for the "request" packet or excatly what to
send for the last part. The site uses ASP. It's http://boards.ign.com if you
want to look at it.  Is there an easier way to do it? Am I just simply
missing something? Thanks in advance

P.S. Me = TOTAL n00b




From steph_lar at videotron.ca  Fri Jun  7 21:56:51 2002
From: steph_lar at videotron.ca (Stephane Larouche)
Date: 7 Jun 2002 18:56:51 -0700
Subject: Problem with wxPython
Message-ID: <b7226ba.0206071756.13c3ebcd@posting.google.com>

I use wxPython 2.3.2.1 with Python 2.2.1 on multiple computers running
WinNT 4.0 SP6. On one of those computers, which has a setup that is
identical to the setup of the other computers (I think), wxPython
software crashes when a wxFrame is initialized with any of the window
styles wxMAXIMIZE_BOX, wxRESIZE_BOX or wxSYSTEM_MENU (or
wxDEFAULT_FRAME_STYLE that includes those styles). Even the demo
provided with wxPython crashes. According to DrWatson, the crash is
caused by a stack overflow during the execution of the function
wxTransformMatrix::operator-.

Have your ever seen the same problem? Is there a workaround?

Thank you for your help!

Stephane Larouche
steph_lar at videotron.ca


From aoeu at go.com  Mon Jun 17 07:06:57 2002
From: aoeu at go.com (Thinkit)
Date: 17 Jun 2002 04:06:57 -0700
Subject: List containing objects and subclassing...
References: <adlcer$6lqd$1@ID-146704.news.dfncis.de>
Message-ID: <3eeda89d.0206170306.655f6c72@posting.google.com>

"Shagshag13" <shagshag13 at yahoo.fr> wrote in message news:<adlcer$6lqd$1 at ID-146704.news.dfncis.de>...
> Hello,
> 
> I need an advice, i have a python list containing only object "node", and i'm wondering if this
> is best to hold it in a python list (as by now) or if i should write a "nodes" object subclassing from
> list and implementing some other methods i need to process on node object (contained by this list)
> and that i do now from "outside", or if i should write my own container implementing __getitem__
> (i also have done that in another version)...
> 
> ...
> [hash_i] --> [node_j] -> [node_k]...
> ...
> 
> Which way is better OO practice ? which is the efficient one ?
> What are the pros and cons (maybe best oo vs efficient ?) ?
> 
> Thanks,
> 
> s13.

You'd have problems if you subclass list with possibly adding non-node
data into it.  Shouldn't you just subclass tuple to get around this?


From mgerrans at mindspring.com  Fri Jun 21 02:50:13 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Thu, 20 Jun 2002 23:50:13 -0700
Subject: text search??
References: <mailman.1024637129.8118.python-list@python.org>
Message-ID: <aeuic2$k21$1@slb6.atl.mindspring.net>

You could either lower() the string before comparing ('WinZip'.lower() ==
'winzip'), or use regular expresions (look at the re module).




From jonathan at onegoodidea.com  Wed Jun 26 04:20:55 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Wed, 26 Jun 2002 09:20:55 +0100
Subject: Parsing strings (\n and \\)
References: <3D188006.7050202@thomas-guettler.de> <mailman.1025017712.20580.python-list@python.org> <lt0S8.45025$n4.10623651@newsc.telia.net>
Message-ID: <B93F37F7.CEA4%jonathan@onegoodidea.com>

On 25/6/2002 17:01, in article lt0S8.45025$n4.10623651 at newsc.telia.net,
"Fredrik Lundh" <fredrik at pythonware.com> wrote:

>>>> UNPARSED = "__import__('os').system('echo dream on!')"
>>>> PARSED = eval(UNPARSED, {}, {})
> dream on!

OK. Moving on from the arguments against using 'eval', how about:

>>> import re
>>> 
>>> ESCAPE_CRE = re.compile( r'\\(.)' )
>>> ESCAPE_SUBS = { 'n': '\n', 't': '\t' }
>>> 
>>> def replace_match( m ):
...     c = m.group( 1 )
...     return ESCAPE_SUBS.get( c, c )
... 
>>> def replace_escapes( s ):
...     return ESCAPE_CRE.sub( replace_match, s )
... 
>>> replace_escapes( '\\\\this is a \\test\\n' )
'\\this is a \test\n'
>>> print _
\this is a      est

>>> 

Obviously this deals only with the easiest cases, but extending it to handle
octal codes and any other special cases shouldn't be too hard (change the
regular expression and put some if...then...else magic into replace_match).

Jonathan



From hjwidmaier at web.de  Thu Jun 20 07:25:53 2002
From: hjwidmaier at web.de (Hans-Joachim Widmaier)
Date: 20 Jun 2002 04:25:53 -0700
Subject: Advice sought on ps module
Message-ID: <6e990e29.0206200325.355f0413@posting.google.com>

About 2 years ago I wrote a primitive PostScript module to help me
create various things like CD inlays or diagrams. It served that
purpose quite well, despite its primitiveness.
Since long I wanted to rewrite the whole thing from scratch and do it
(at least kind of) "right." But I'm still not so sure as which way to
go. At one time I thought I might employ ghostscript, which would have
enabled me to, e. g. get the string width for some text, so I can make
adjustments (not only to the text). Apart from that I couldn't get a
sensible communication with gs through popen2, the PostScript output
is created somewhat nonlinear, making it harder to ensure that the gs
context is correct (in that query case).
I'm convinced that's the wrong way, but the alternative doesn't look
too good either. If you want to do some optimizations, you have to
double some functionality of the ps interpreter.

Examples:
  - To calculate the with of a string, you have to read the AFM file,
add the character widths, aplly kerning, transform it with the CTM and
the fonts TM.
  - To know whether a font must be reset or is already the current
font you have to keep track of all the save/restore, gsave/grestore
and the like.

I do not want to write a "do all, end all" PostScript module. I do not
even want to give it away (I'd be simply too embarrassed).

Still I hope that some of you might give me some good advice on how to
best do it.

Thanks for any suggestion!
Hans-Joachim


From dfackrell at DELETETHIS.linuxmail.org  Fri Jun  7 11:33:25 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Fri, 7 Jun 2002 09:33:25 -0600
Subject: Detecting OS and Window Manager
References: <mailman.1023304388.9659.python-list@python.org> <ufvsco8mlhf213@corp.supernews.com> <3d00bfb5$1_2@hpb10302.boi.hp.com> <ug1h5q29aulga9@corp.supernews.com>
Message-ID: <3d00d248$1_1@hpb10302.boi.hp.com>

"Darren Winsper" <dw133 at cs.york.ac.uk.go.away> wrote in message
news:ug1h5q29aulga9 at corp.supernews.com...
> In <3d00bfb5$1_2 at hpb10302.boi.hp.com>, Daniel Fackrell wrote:
>
> > Correct me if I'm wrong, but isn't this exactly the pythonic way of
handling
> > dynamic loading of modules?
> >
> > try:
> >     import kdecore
> > except ImportError:
> >     # do something else
>
> OK, fair enough.  I did only read and write my first line of Python a
> few days ago :)

Welcome to the group, then.  We're all newbies here, for various values of
new.  <wink>

> > Of course, this will probably not do what you want in this case, as
kdecore
> > will probably load if it is installed, without regard to whether the
current
> > user is running KDE.  I don't have much experience with this particular
> > topic yet.
>
> Well, I could attempt to see if something like "kdesktop" is running,
> since if it is, the chances of running KDE are very high.  It's kludgy,
> but I can't really think of anything else, since KDEINIT would be running
> if a single kde application is running IIRC.

If you're thinking of scanning the currently running processes for the
existence of a particular process, then you won't get the results you want
there, either.  The reason for this is that there can be multiple logins for
a single user or multiple users, and KDE-related processes will exist if any
one of those logins is running KDE, but not necessarily the current one.

Maybe import your kde module(s), and then try to do something KDE specific
without KDE running and see what kind of error this causes so you will know
what to trap for?  Hopefully some other newbie (for a much less significant
value of new, AKA expert) is still following this thread to give you a
better answer.

Daniel Fackrell




From aahz at pythoncraft.com  Fri Jun 21 20:32:57 2002
From: aahz at pythoncraft.com (Aahz)
Date: 21 Jun 2002 20:32:57 -0400
Subject: Tkinter and threads
References: <3D0DCC8F.9A6D3B81@ipm.fhg.de> <wk8z58w81y.fsf@mail.connact.com>
Message-ID: <af0gjp$g67$1@panix1.panix.com>

In article <wk8z58w81y.fsf at mail.connact.com>,
Patricia J. Hawkins <phawkins at connact.com> wrote:
>
>Also be aware that the early code examples aren't intended as models,
>just as teaching tools -- you don't hit the really useful stuff till
>page 48. And the format makes it tough to flip back and forth (not to
>mention, tiny amounts of information per page drives me nuts, but
>perhaps that's just me) -- and at 84 pages, I'm not going to print it
>out to take it downstairs to an armchair and mark it up, which is what
>I want to do.  And skimming for what you don't know already is
>impossible.

<sigh>  Yeah, I know.  Last summer, I was negotiating with O'Reilly to
write a whole book on the subject (their idea).  Last November, they
canned the project.  I haven't had the time/energy to rewrite the slides
into a real tutorial, and I'm working on a completely different book now.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From mcherm at destiny.com  Thu Jun  6 09:11:52 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Thu, 06 Jun 2002 09:11:52 -0400
Subject: array, list, performance...
Message-ID: <3CFF5F98.2060406@destiny.com>

Here is my understanding of it:

    li = [x0, x1, x2, ...]

    li[n]          element access is O(1)
    li.append(x)   appending is O(1)
                   (actually, it SOMETIMES takes O(len(li)) as it
                    resizes the list. But if you grow a list by
                    continually appending, the amortized time is
                    linear (ie, constant amortized time per element)
    li.insert(0)   prepending is O(len(li))
                   (so grow the list from the end!)
    del li[0]      deleting from the front (or middle) is O(len(li))
    li.pop()       deleting from the end is O(1)
    li[n] = x      assignment to the middle is O(len(li))
    len(li)        determining length is O(1)
    li.index(x)    search is O(len(li)) (NOT efficient)
    li.sort()      sort is O(len(li)*ln(len(li)))
                   (I believe it uses quicksort, but if a python
                    function has to be executed for the compare
                    that may take some time.)

It's really just your basic auto-growing array, but with a bunch of 
clever things done to speed up common cases, and I don't know those well 
enough to comment on them.

-- Michael Chermside





From mgilfix at eecs.tufts.edu  Mon Jun 17 11:59:02 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Mon, 17 Jun 2002 11:59:02 -0400
Subject: non blocking socket.accept()
In-Reply-To: <qeokea.b6q.ln@neo.fdns.net>; from e9226414@nospam.stud4.tuwien.ac.at on Mon, Jun 17, 2002 at 01:45:05PM +0000
References: <qeokea.b6q.ln@neo.fdns.net>
Message-ID: <20020617115902.A422@eecs.tufts.edu>

  Someone else answered what you'll want to do with earlier versions
of python but if you're feeling adventurous and want to try out
the CVS version:

   sock.set_timeout(2.0)
   try:
       sock.accept()
   except socket.error, strerror:
       print strerror

                    -- Mike

On Mon, Jun 17 @ 13:45, Zsolt Marx wrote:
> I far as I know socket.accept() waits for a client
> to ask for a connection with the server.
> 
> This command seems to be blocking, though I would
> like to limit the wait time onto a predefined interval
> and to throw an error, if there is no connection.

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From d2002xx at myrealbox.com  Sat Jun 22 20:42:09 2002
From: d2002xx at myrealbox.com (d2002xx)
Date: 22 Jun 2002 17:42:09 -0700
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at>
Message-ID: <8d3f4438.0206221642.607584b9@posting.google.com>

Hello.

Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote in message news:<3D1482AE.8FA299D8 at kfunigraz.ac.at>...
> In the following I got the advice to install a real user system on my
> laptop. Okay, I bought Linux SuSE 8.0 and installed it on my laptop (256
> MB RAM, 1000 MHz Celeron). I also installed Python 2.2 (the ActivePython
> version, otherwise I couldn't install idle due to the annoying
> "free_software_garbage_2_weeks_installation_nightmare").
> 

You didn't use offical python distribution, right? It's impossible to
show the "free_software_garbage_2_weeks_installation_nightmare".

If the binary version can't work, just use source-code and compile
yourself. Python 2.2.1 works well on my computer. The difference is
that before compiling, you need to install some XXX-devel.rpm if you
want use these python modules. (such as openssl-devel)

> 1. After starting the process from within idle the complete system
> freezes after the first wavelength (after about 20 minutes). I couldn't
> kill a process or open a different session (via F2) in order to kill
> python. The last resort: turn off the power supply.
> 

NOT use unoffical python distributions, they may change some thing.
And I NEVER see similiar problems on offical version, even in beta
version.

Also, I recommand that NOT use idle, because anything using tk are
very slow, if you want an IDE, there is http://boa-constructor.sf.net 
, it doesn't complete yet, but works very well, and provides some cool
features such as UML. It uses wxPython, which is much faster than
tkinter, there is http://www.wxpython.org

> 2. I thought it is better to start the process from the command line and
> in a X terminal: python forcing.py
> The same behavior as aforementioned: last resort: power supply. That is
> really good and fine because a Linux system really is greedy for "shut
> down immediately".
> 

It may be due to X window.

> My conclusion: I will rewrite my simulation in Fortran 90 and will
> abadon Python. Python is maybe good for some small scripts but not
> appropiate for serious programming.

Your conclusion is whole wrong! Do you know how many big companies use
Zope? See http://www.zope.com/ZopeClientList

Well, I don't know much about fortran, but if you just want to do
something about math, NO python, it will be slow (at least now).


From kseehof at neuralintegrator.com  Sun Jun  9 01:09:21 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Sat, 08 Jun 2002 22:09:21 -0700
Subject: Html
In-Reply-To: <3d02c220_1@news.iprimus.com.au>
Message-ID: <EFEFIHAINIINNELLHFAIOELACIAA.kseehof@neuralintegrator.com>

Gold Fish wrote:

> Anyone can tell me how can i run the script with the argumemt 
> such as python 
> example.py -o 
> What i have to write in script.

Use sys.argv.  It should look something like this:

>>> import sys
>>> sys.argv
['example.py', '-o']

> And if i want to render the 
> script in html 
> format hov can i do that

Do you mean render the script itself, or the output of the
script?  If you mean "render the script", you need py2html.  

There are a few versions around.  Google for py2html, or
look at one of these:

http://www.norvig.com/python/py2html.html
http://www.egenix.com/files/python/py2html.py.html

If that is not what you mean, you need to be much more
specific.

- Ken Seehof






From occeanlinux at linuxmail.org  Sun Jun  2 16:56:10 2002
From: occeanlinux at linuxmail.org (Gold Fish)
Date: Mon, 03 Jun 2002 04:56:10 +0800
Subject: List
Message-ID: <20020602205610.31575.qmail@linuxmail.org>

----- Original Message -----
From: holger krekel <pyth at devel.trillke.net>
Date: Sun, 2 Jun 2002 22:05:45 +0200
To: Gold Fish <occeanlinux at linuxmail.org>
Subject: Re: List


> Gold Fish wrote:
> > Charl P. Botha wrote:
> > 
> > > In article <3cfa6ed8_1 at news.iprimus.com.au>, Gold Fish wrote:
> > >> Assume we got 2 list
> > >> list 1 = ['ABC','CDG','DAV','FE','FECA','FEAFA','FEA']
> > >> list 2 = []
> > >> how can we copy element from list 1 to list 2 say i just want to copy the
> > >> 3 element 'CDG','DAV','FE' to list2 only.
> > > 
> > > list1 = ['ABC','CDG','DAV','FE','FECA','FEAFA','FEA']
> > > list2 = list1[1:4]
> > > 
> > > Is that what you meant?
> > > 
> > No i mean if i got from list1 i wanna take 3 elements in any position to 
> > list2 but i don't know how many item in that list.
> 
> if you have a starting index 'i' then you can do
> 
>   list1[i:i+3]  
> 
> which returns you at most 3 elements, starting from position 'i'.
> 
> If this is still not what you want, then please give 
> two or three examples of what you want to achieve.
> 
>     holger
> 

What i would achieved is that I got the list of elements say that 10 elements in the list, i want to seperate this list into 2,or 3 sublists depends on user input the number. Then these subject contain the number of elements say 2, or 3 depend on user define. How can i do this. For example. 
BigList = [e1,e2,e3,e4,e5,e6,e7]
this will be divided into 4 small sublists which is sublist1,sublist2,sublist3,sublist4.These sublists will have maximum of 2 elements inside therefore, sublist1 = [e1,e2], sublist2=[e3,e4],sublist3=[e5,e6] and sublist7=[e7]
I would appreciate that you could help me this problem. I spend so much time to think about it but it's look like i stuck in the maze when trying to divide biglist into small list and small list to nano list. 
 

-- 
Get your free email from www.linuxmail.org 


Powered by Outblaze




From whisper at oz.net  Mon Jun  3 14:43:25 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 3 Jun 2002 11:43:25 -0700
Subject: Automated Email
In-Reply-To: <adg1do$10nmr5$1@ID-49758.news.dfncis.de>
Message-ID: <GCEDKONBLEFPPADDJCOEKEFKDGAA.whisper@oz.net>

Use the standard smtplib module that comes with Python. There is an example
of using it in the python doc.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Ken
> Sent: Monday, June 03, 2002 8:22
> To: python-list at python.org
> Subject: Automated Email
>
>
> How to do automated emailing when results being processed?
>
> Thanks
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From peter at engcorp.com  Sun Jun  9 17:51:14 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 09 Jun 2002 17:51:14 -0400
Subject: Correct code/unit testing (was Re: Efficient python programming...)
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com> <834rgd5isv.fsf@panacea.canonical.org> <3D02FB3B.CF75E925@engcorp.com> <Pine.SOL.4.30.0206092211170.15771-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <3D03CDD2.78C01DE1@engcorp.com>

"John J. Lee" wrote:
> 
> > Kragen Sitaker wrote:
> > > It is certainly possible to err too far in the direction of believing
> > > your code correct simply because it passes its test suite, and this is
> > > a common beginner's error.  I think it is also possible to err too far
> > > in the direction of writing and running tests instead of desk-checking
> > > code.
> 
> Isn't it true that, however much time you're going to spend, the largest
> number of bugs are going to get removed by allocating some of the time to
> both activities: testing, and staring at the code?

Perhaps, assuming you put a lot of bugs in the code, some amount of staring
will be necessary.

I believe staring is of little or no use, however, having tried it many 
times with my and others' code which had bugs and not having found those 
bugs by staring.

The nastiest bugs were found by lots and lots of work, which usually
included repeated attempts to reproduce the problem, manual runs, 
and sometimes using a debugger.  Staring helped little.  And then,
having found the bug, we look at the code and say "oh yeah, we 
should have seen that." 

Better not to put the bugs in the code in the first place, however,
so staring or those other things won't be necessary.  Writing the 
tests and the code together, tests first, in small iterative steps, 
tends not to lead to buggy code.  The payback is high compared to
staring.

-Peter


From observer at NOSPAM.space.pl  Thu Jun 13 05:29:52 2002
From: observer at NOSPAM.space.pl (Johann)
Date: Thu, 13 Jun 2002 11:29:52 +0200
Subject: How to check version of Python under CGI
References: <rbmggus7ln1uiutktf5libf4bbngu29ji7@4ax.com> <873cvrin1q.fsf@tux.ntw23.fr> <cgnggu40ethjhpqcmltnfb06d5qvuf078k@4ax.com> <t1ogguoua7b9ne9n3mb3n7rkp8mrqe2nee@4ax.com> <87y9djh6tg.fsf@tux.ntw23.fr>
Message-ID: <ndpggu8uqsmi6ppumkem4c16rt7p5i6k9o@4ax.com>

On 13 Jun 2002 11:22:03 +0200, Sylvain Thenault <thenault at nerim.net>
wrote:

>I think the best way to know if a module is available is to try to import it:

But it is difficult to import all modules. I found much better
sollution. Ready to run  CGI script at http://snakefarm.org/

Thanx to Simon :)

--
Johann


From martin at v.loewis.de  Sun Jun  9 04:14:58 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 09 Jun 2002 10:14:58 +0200
Subject: Source formatting for long expressions.
References: <mailman.1023570972.27430.python-list@python.org>
Message-ID: <m3it4srhq5.fsf@mira.informatik.hu-berlin.de>

Christopher Armstrong <radix at twistedmatrix.com> writes:

> Does anyone know of a tool that will beak up a very, very long
> expression like this into multiple lines with sane indentation?

Emacs indentation gives this:

Ref(1, Instance('twisted.internet.app.Application',
                persistenceVersion=7, services={}, udpPorts=[], gid=1000,
                connectors=[], uid=1000, tcpPorts=[(9080,
                                                    Instance('twisted.web.server.Site', sessions={},
                                                             resource=Instance('twisted.coil.web.ConfigRoot', children={},
                                                                               modules=[], widgets={'config':
                                                                                                    Instance('twisted.coil.web.AppConfiguratorPage', variables={},
                                                                                                             dispensers=Instance('twisted.coil.coil.DispenserStorage',
                                                                                                                                 dispensers={}})})))])

As you can see, it tries to align further arguments to a function
together with the opening parenthesis. I would assume that any other
auto-formatters use the same style, so you won't get any "sane"
indentation until you drop the nesting level (perhaps in favour of
method invocations).

Regards,
Martin



From whisper at oz.net  Wed Jun  5 19:48:03 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 5 Jun 2002 16:48:03 -0700
Subject: Using AT on XP to run python script - MARK HAMMOND PLEASE READ
In-Reply-To: <mchqfu448cnq5q8toh4nq7n1sura1ug72g@4ax.com>
Message-ID: <GCEDKONBLEFPPADDJCOEMEONDGAA.whisper@oz.net>

This is most curious, but I believe the commands _are_ getting executed!

I tried the following (assume time is 16:30) "at 16:32 pythonw
j:\python22\tools\idle\idle.py" (without the qoutes of course). At the
appointed time, there was disk activity, but no idle window popped up (idle
was chosen just to get something to pop up!). However, there _was_ a pythonw
process running in the task manager! I think there might be a way to attach
a console or get access to the GUI, but i'm not sure - maybe Mark Hammond
might have a clue (he being the Python Win guru :) ).

I tried the following:
#at_test.py
import sys
import time

tst = open("testfile.txt", 'a+')

tst.write(time.asctime() + "\n")

tst.close()

put into an 'at' job as: at xx:xx python j:\python22\at_test.py

I expected the output from this to be in j:\python22, but to my surprise, it
was in c:\winnt\system32 where cmd.exe resides! testfile.txt contained the
exact time the 'at' job was set to run at too!

I suspect your last example that produced no visible output did in fact work
:)

In any case, it works - now if one could just get visible output!

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Chris Stromberger
> Sent: Tuesday, June 04, 2002 17:35
> To: python-list at python.org
> Subject: Using AT on XP to run python script
>
>
> I'm having trouble getting AT on XP to run a python script for me.
> Have tried:
>
> at xx:xx "c:\python21\python.exe c:\scripts\script.py"
> at xx:xx "start c:\scripts\script.py"
> at xx:xx "c:\scripts\script.py"
>
> and get "the system cannot find the file specified" for the first two.
> The last version does nothing apparently--no error indication or
> anything.
>
> I imagine it is related to user accounts and such.  Anyone have any
> pointers?
> --
> http://mail.python.org/mailman/listinfo/python-list





From rnd at onego.ru  Wed Jun  5 17:41:27 2002
From: rnd at onego.ru (Roman Suzi)
Date: Thu, 6 Jun 2002 01:41:27 +0400 (MSD)
Subject: prototyping good OOdesign in Python?
In-Reply-To: <20020605203051.GC1455@isis.gerrietts.net>
Message-ID: <Pine.LNX.4.33.0206060126380.30758-100000@rnd.onego.ru>

On Wed, 5 Jun 2002, Geoff Gerrietts wrote:

>Quoting Roman Suzi (rnd at onego.ru):
>> On Sat, 1 Jun 2002, Sean 'Shaleh' Perry wrote:
>> 
>> Related topic is UML with Python. Does UML has everything to reflect
>> Python OO model or does it need to add features?
>> (Or maybe in order to interoperate, there is a need not to use some
>> of Python features).
>
>With UML, I generally find that it has TOO MANY features. 
>If I use
>Visio, or Rose, or even the lovely pyut announced a few months ago on
>this very list, they'll ask for information about attribute and method
>visibility, data types, etc.
>
>Since Python doesn't care about most of this stuff, modelling it is
>only useful if you expect to port the code later. And I prefer not to
>plan for an unfortunate event. ;)
>
>> And if in order to interoperate with less expressive
>> languages/systems, how to keep Python pythonic and not fall back to
>> some standard universal subset, found in many programming languages.
>
>Certainly you could use UML and thorough documentation as a hedge
>against relying too heavily on Python's dynamic nature. A startling
>percentage of the things we take for granted in python can actually be
>expressed similarly in other languages, though -- it just takes more
>work, more infrastructure, and a few more lines of code.
>
>I've written code with the thought that I'd eventually rewrite it in
>another language, and I shared many of your concerns about how well
>ideas would "port".

>For the most part, I find myself doing one of two things: either
>writing the other (generally Java) language's idioms (like Singleton,
>like stateless classes) in Python when it's not too inconvenient, or,
>when it is, just writing it off as "the burden of porting".
>
>For the most part, the really "interesting" parts of your system are
>going to port relatively cleanly, because you're doing calculations
>and computations. It's the boilerplate crap that you can skip over
>pretty effortlessly, and that's going to be the irritation of porting
>anyway.

Hmmm... Computations could be different.
For example, in python I can use a couple of .strip/.split/.join
things and get the needed format parsed.
In other languages to do the same I need to write it again and
again or find some lib to do it for me.

Or, sometimes I prefer to put things like (in one breath,
without errors):

data = """parse|me|heavy
1|2|3
2|3|4
"""
data = map(lambda x: x.split("|"), data.strip().split("\n"))

and then have fun with a "burden of porting", rewriting
the above example in C...

Or maybe I became too lazy?

>You can, of course, take special steps to avoid hard-to-port code.

What are these? Not to use Python features at all and write crap like:

data[0][0] = "parse";
data[0][1] = "me";
data[0][2] = "heavy";
....

- often found in VB programs?

My examples are data-oriented, but logic-oriented ones are 
no less funny.

Well, maybe I could automate the above with as little as:

i = 0
for d in data:
  j = 0
  for dd in d:
    print """data[%(i)s][%(j)s] = "%(dd)s";""" % vars()

# ;-)

>That's where you spend time writing foreign idioms into your Python
>code. How much time you waste doing that will largely depend on how
>quickly you need the "convert to another language" step to go.

Anyway, thank you for your ideas, Geoff!
>--G.
>
>

Sincerely yours, Roman Suzi
-- 
\_ Russia \_ Karelia \_ Petrozavodsk \_ rnd at onego.ru \_
\_ Tuesday, June 04, 2002 \_ Powered by Linux RedHat 7.2 \_
\_ "Shin: A device for finding furniture in the dark." \_





From fredrik at pythonware.com  Thu Jun 20 06:34:06 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 20 Jun 2002 10:34:06 GMT
Subject: Useful RE patterns (was: Variable Interpolation - status of PEP 215)
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <4PfQ8.44245$n4.10307683@newsc.telia.net>
Message-ID: <yciQ8.44268$n4.10312169@newsc.telia.net>

I wrote:

> # match $$ and $var and ${var}
> _dosub = sre.compile(r'\$(?:\$|(\w+)|\{([^}]*)\})').sub

As part of the python-dev discussion, I mentioned that it might
be a good idea to add a couple of standard RE patterns to the
standard library, e.g. to match things like integer literals, IP-
numbers (etc).

If I were to add a dozen (or so) patterns to the (S)RE module,
what should I pick?  What patterns do you find yourself using
over and over again?

</F>




From loewis at informatik.hu-berlin.de  Thu Jun 13 03:50:32 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 13 Jun 2002 09:50:32 +0200
Subject: How to call a function using apply with keyword args?
References: <ae9fud$u23$00$1@news.t-online.com>
Message-ID: <j4ofef7h2v.fsf@informatik.hu-berlin.de>

"Achim Domma" <achim.domma at syynx.de> writes:

> I have a function like this:
> 
> def MyFkt(A,B,C): pass
> 
> and some parameters from a config file:
> 
> params = {'A':5,'B':8,'C':9}
> 
> how can I call MyFkt with this parameters? 

I recommend

MyFkt(**params)

> I tried apply(MyFkt,[],params) 

If you want to use apply, it is

apply(MyFkt, (), params)

The arguments need to be a tuple.

Regards,
Martin



From shalehperry at attbi.com  Thu Jun 20 13:54:12 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 20 Jun 2002 10:54:12 -0700 (PDT)
Subject: application development - separating content from function
In-Reply-To: <yroQ8.260949$i6.17986369@amsnews02.chello.com>
Message-ID: <XFMail.20020620105412.shalehperry@attbi.com>

> 
> The more I have tried to solve these requirements in an elegant and
> efficient manner, the further I seem to stray from anything remotely
> resembling a viable solution. In other words, I can't see the forest for the
> trees anymore.
> 
> Anyone working on something similar? Or, just plain brilliant in all things
> Python, and ready and willing to point me in the right direction and make it
> all worth while again :-)
> 

you should try to solve part of the problem rather than all of it at once.  The
first implementation almost always reveals hidden issues.  If you try to make
this thing as full featured as you like from the beginning you will never
finish it.  Or it will be so overly complex it will crumble.

Think about issues like multi-language support when you are coding so it is
easier to add in later.  Same for user customization.  But adding all of those
choices at once leads to difficult to implement and test code.




From dunklervater at yahoo.com  Mon Jun 17 15:51:11 2002
From: dunklervater at yahoo.com (DV)
Date: 17 Jun 2002 12:51:11 -0700
Subject: How to distribute py-scripts on windows without requiring the full ActivaState package
Message-ID: <2e0f5ffe.0206171151.4305c306@posting.google.com>

Hello,

A couple of years ago, I experimented with Python on Linux, but
somehow got out of touch.

Just recently I discovered, that there is a pretty comprehensive
Windows-Version for it that even supports pure windows-stuff such as
COM or Win32API.

Yet, I found out that the ActiveState package is very large to
download and install, so this leads me to the question(s):

1)Is there some way to distribute a script foo.py by maybe bundling it
with python.exe and python.dll, delivering these three files ONLY (and
of course any .py library files that foo.py calls) and not requiring
the recipient to download and install the whole ActiveState package?


2)Are there different versions of python.exe and python.dll? What I
would find useful was, if there was ONE version of
python.exe&python.dll that only could be used for console stuff, so
you wouldn?t have such big files to distribute along with foo.py, and
another one for true GUI-stuff

3)Concerning COM and other heftier Win32 dependend stuff, or anything
that python needs a third-party dll for to use: Can the necessary dll
be bundled along with python.exe, python.dll and foo.py, just as
mentioned in 1)?

Mind, I would?t want to distribute stuff as a single EXE. I could live
with having to distribute more than one file, but I would very much
like it, if an end user receiving my script would?t have to install
the whole ActiveState package.

I could imagine that one weak point about this is, that i have read
somewhere that dlls must be registered before they are usable, but I
also read that this is possible by using rundll.exe or some other
Windows system program manually. If so, the registering process could
be done by a .BAT file.

As for the harddisk-space penalty of havin more than one python.exe
and python.dll (one for each distributed script), this is the reason
why I asked, whether a console version exists: If it did exist, you
could do the GUI by designing a HTML-page that drives it.

Sorry, if this posting appears twice under different topics, if so, it
is most likely a temporary fault of deja.com, through which I posted
this.

I would appreciate any info on this matter.

Thank you in advance,
Bye.
Franz


From eric.brunel at pragmadev.com  Fri Jun  7 12:30:56 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Fri, 7 Jun 2002 16:30:56 +0000
Subject: Program to an interface, not an implementation
References: <mailman.1023358448.9091.python-list@python.org> <adnjee$p88$1@wanadoo.fr> <mailman.1023439810.12120.python-list@python.org>
Message-ID: <adqfo2$gbc$1@wanadoo.fr>

Egbert Bouwman wrote:
> All your comments are crystal clear. Thanks.
> 
> My next problem is on nearly the same page (17) of the GoF book:
> class versus interface (== type) inheritance, but in the python context.
> 
> "In C++ and Eiffel inheritance means both interface and implementation
> inheritance" and
> "in Smalltalk inheritance means just implementation inheritance".
> I don't know any of these languages.
> 
> Now in my simplistic worldview, if you inherit from a class,
> you inherit everything, implementation and interface,
> because the implementation defines the interface as well.
> Even if all methods of the superclass only contain 'pass'.
> But I may be wrong.

I know C++ and I looked at SmallTalk a bit, and it seems the line between 
"implementation + interface" and "implementation only" inheritance is very 
thin... Apparently, it's quite directly related to strong typing and 
declarations. "Interface", in this context, seems to describe what's 
declared. So, as you noted in the beginning of your post, an "interface" is 
considered as a declared type. Since SmallTalk claims to have no types at 
all (which is also the case with Python...), they claim not to have such 
thing as an "interface", so obviously they cannot inherit it...

In fact, I don't see the point in this: as you noted, when a class inherits 
from another, it inherits everything, implementation and interface. So in 
practice, it's not really significant. As I understand it, the only 
practical consequence of this distinction *may* be in the following case:

class A:
  def m(self, i):
    pass
class B(A):
  def m(self, i, j):
    pass

In Python or SmallTalk, the preceding code is completely valid, so one 
could say that the "interface" of class A was not inherited by class B, 
since its method m has a different signature than A's method m.

Trying to do the same thing in interface-inheriting languages would not 
have the same effect at all: it would either be forbidden, or even have 
stranger consequences. In C++ or Java, it would declare a *second* method, 
also called m, that would be distinguished from the inherited one by its 
number of parameters or their type:

public class A {
  public int m(int i) {
    ...
  }
}
public class B inherits A {
  public int m(int i, int j) {
    ...
  }
}
...
o = new B();
o.m(1)      // Calls A.m
o.m(1, 2)   // Calls B.m

So, as you see, you probably can live without knowing anything about all 
this and still be able to do great stuff with whatever language you like 
:-).

However, if anybody knows a bit more than me about implementation/interface 
inheritance, I'd be glad to learn (I happen to teach OO sometimes and I'd 
like to be able to answer this question precisely if it ever gets answered 
to me ;-).
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From boud at valdyas.org  Mon Jun  3 15:40:27 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Mon, 03 Jun 2002 21:40:27 +0200
Subject: different gui-toolkits pros/cons...
References: <adfsoo$dmik4$1@hades.rz.uni-sb.de> <_rLK8.12$fd2.979@news.ecrc.de> <slrnafn583.j2i.gerhard@lilith.my-fqdn.de>
Message-ID: <3cfbc65a$0$52024$e4fe514c@dreader3.news.xs4all.nl>

Gerhard H?ring wrote:
> 
> Define "fast". PyGTK loads a lot faster than wxPython or PyQt. AFAIC,
> this is because PyQt and wxPython wrap _all_ of a very large toolkit,
> that both (wxWindows and Qt) even include a lot of non-GUI stuff. This
> might be solved by trying to split the one shared library into many
> small ones, that are loaded on demand.

This has become a bit better since PyQt was split into several modules.
The problem remains, though -- Qt is a very large library, even if you
take all the extra's out.
-- 
Boudewijn Rempt | http://www.valdyas.org


From Administrator at python.org  Thu Jun 20 14:59:44 2002
From: Administrator at python.org (Administrator at python.org)
Date: Thu, 20 Jun 2002 12:59:44 -0600
Subject: ScanMail Message: To Recipient file blocking settings matched and action taken.
Message-ID: <00dd01c2188c$a3f585a0$0e0fbb82@corp.es.com>

ScanMail for Microsoft Exchange has blocked an attachment.

Sender = Sylvain Th?nault
Recipient(s) = David LeBlanc;python-list at python.org; XML Logilab Mailing list
Subject = Re: [ANN] PyReverse 0.3
Scanning Time = 06/20/2002 12:59:43

Action on file blocking:
The attachment py2xmi.bat matches the file blocking settings. ScanMail has Deleted it. 






From emile at fenx.com  Sat Jun 29 16:08:14 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 29 Jun 2002 20:08:14 GMT
Subject: private
References: <pan.2002.06.29.17.38.51.789527.440@myself.com> <3D1DE132.4A890D9D@engcorp.com> <afl3e4$f8abc$1@ID-71831.news.dfncis.de>
Message-ID: <OsoT8.349745$352.41945@sccrnsc02>

James Kew
> Yes, of course one can circumvent private in C++. But for me the main
value
> of the public and private specifiers is that they _document_ to the
client
> of the class what it should access and what it should not.
>
> I don't -- in my admittedly very limited experience -- find Python
classes
> as immediately self-documenting. It's not obvious from the class
definition
> what is intended to be called or modfied by the client and what is
internal
> to the implementation.
>

Conventionally, any leading underscore indicates 'private - don't use'.

--

Emile van Sebille
emile at fenx.com

---------



From emile at fenx.com  Thu Jun 27 12:08:42 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 27 Jun 2002 16:08:42 GMT
Subject: How to find out operating system
References: <mailman.1025159195.9033.python-list@python.org>
Message-ID: <eMGS8.336716$cQ3.20095@sccrnsc01>

> What is the best way of  finding out the kind operating system?
> I can use os.name but if the system is Windows I would like also 
> know if the system is Windows98 or Windows ME or W2K or 
> Windows XP.

Ask Marc Andre to re-release platform.py?

[ducking...]

--

Emile van Sebille
emile at fenx.com

--------- 



From fredrik at pythonware.com  Sat Jun 29 08:06:25 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 29 Jun 2002 12:06:25 GMT
Subject: joining 2 lists together - an elegant way?
References: <pan.2002.06.28.15.03.29.911402.1787@presidency.com>
Message-ID: <5phT8.45580$n4.10822879@newsc.telia.net>

Rajarshi Guha wrote:

> Right now I do:
>
> newlist = []
> for i in l1:
>   newlist.append(i)
> for i in l2:
>   newlist.append(i)

newlist = l1 + l2

</F>




From ken at hotmail.com  Tue Jun  4 09:41:06 2002
From: ken at hotmail.com (Ken)
Date: Tue, 4 Jun 2002 23:41:06 +1000
Subject: Automated Email - What did I do wrong here?
References: <adg1do$10nmr5$1@ID-49758.news.dfncis.de> <adifmg$11i5ks$1@ID-49758.news.dfncis.de>
Message-ID: <adifsu$11ifbm$1@ID-49758.news.dfncis.de>

"Ken" <ken at hotmail.com> wrote in message
news:adifmg$11i5ks$1 at ID-49758.news.dfncis.de...
> "Ken" <ken at hotmail.com> wrote in message
> news:adg1do$10nmr5$1 at ID-49758.news.dfncis.de...
> > How to do automated emailing when results being processed?
> >
> > Thanks
> >
>
> What did I do wrong? It crash the program. Without these code, the cgi
> script works fine. Thanks
>
>     fromaddr = 'ken at hotmail.com'
>     toaddrs = 'ken at hotmail.com'
>     msg = "testing\n"
>     server = smtplib.SMTP('smtp.aol.com')
>     server.sendmail(fromaddr, toaddrs, msg)
>     server.quit()
>
oh...and I did import smtplib...




From walnut at froggy.com.au  Sat Jun  1 09:14:58 2002
From: walnut at froggy.com.au (Walnut)
Date: Sat, 01 Jun 2002 23:14:58 +1000
Subject: Migrating from PHP to Python...
References: <xWyk7.4824$v%4.492814@news20.bellglobal.com>
Message-ID: <l4ihfu0nhp1a9p53vgch3n3kcc25s8c00p@4ax.com>

On Sun, 2 Sep 2001 19:12:00 -0400, "Rick Shin" <rick at slaveware.com>
wrote:

> Hi,
> 
> I'm a PHP man looking to migrate to Python for a change of scenery. I've
> skimmed over the specs, but missed a few PHP constructs that I find
> essential.

Forget them both and go with Perl - the most flexible and powerful
language anywhere in the world today. You already know the syntax.


From sarmstrong13 at mac.com  Sat Jun  8 22:25:23 2002
From: sarmstrong13 at mac.com (SA)
Date: Sat, 08 Jun 2002 21:25:23 -0500
Subject: Newbie Question (executing Program from within)
Message-ID: <B92826C3.6BF9%sarmstrong13@mac.com>

Hi Everyone-

    I'm running Python on Mac OSX. (bsd based system) How do I call a
program from outside Python to be run from within my Python script/ In other
words if I have an executable file, say "someprogram", how would I execute
this program from within a python script.

Thanks.
SA



From jknapka at earthlink.net  Thu Jun 13 01:01:36 2002
From: jknapka at earthlink.net (Joseph A Knapka)
Date: Thu, 13 Jun 2002 05:01:36 GMT
Subject: Python GUI
References: <9ddd154e.0206122021.1b06f89a@posting.google.com>
Message-ID: <3D0818D3.52D2EF84@earthlink.net>

"H. Safarzadeh" wrote:
> 
> Hi all,
> I know that this subject is discussed in many other places, but I don't
> think it is harmful to talk about it more!
> 
>  I want to what library I should use to make GUI for my Python apps?
>  I have examined many of them, but I couldn't find what I need. Here is
> the result of some of my examinations:
> 
> -Tkinter: A nice simple lib, but really simple! It dosn't have enough
> widgets, and I do not like its look in addition!
> -wxPython: Really perfect. It has many widgets, and do many works(not
> only in GUI). But it is toooo big!(both on disk and in memory). I need a
> smaller one.
> -pyGTK and pyQt: I do not know enough about them. But I know that I do
> not like GTK look on Windows and Qt for Windows has some problems with
> its lisence(and it is tooo big, too!)
> -Jython/Swing: ....it's Jython, after all!
> -Win32 extensions: It is not portable!!!
> 
> Any comments? Other possibilities?

AnyGUI <URL: http://anygui.sourceforge.net> should
be releasing again soon. The last release supported
wx, gtk, tk, win32, Jython, and curses - you write
your code and it "magically" works unchanged on
any of those UI platforms. There were a couple of
things missing, notably canvasses (eg Tk's Canvas)
and menus; the next release may (or may not)
remedy that.

Disclaimer: I have worked on Anygui in the past,
and hopefully will have time to continue to do
so in the future...

Cheers,

-- Joe
  "I could feel a small polished stone sinking through
   the darkest waters of my heart."
   - Haruki Murakami, "Dance Dance Dance"


From ngps at vista.netmemetic.com  Tue Jun 11 22:29:31 2002
From: ngps at vista.netmemetic.com (Ng Pheng Siong)
Date: 12 Jun 2002 02:29:31 GMT
Subject: M2Crypto: select() behaves weird on SSL sockets
References: <3D04ED17.5070606@NOSPAMREMOVETHISxs4all.nl> <ae2str$b0q$1@nntp6.u.washington.edu> <3D066E77.4070901@NOSPAMREMOVETHISxs4all.nl>
Message-ID: <ae6bmb$frc$1@dahlia.singnet.com.sg>

According to Irmen de Jong  <irmen at NOSPAMREMOVETHISxs4all.nl>:
> I looked around a bit and oh boy, M2Crypto supplies a "pending"
> method on SSL sockets, that apparently returns the remaining data.
> 
> I now have a workaround; if my socket is an SSL socket, and
> there are still bytes pending (socket.pending()>0), I don't
> call select. In all other cases it is 'safe' to call select.
> 
> My code (Pyro with SSL and timeouts) now works, I'm happy :-)

Coming late into this thread, good to hear your code now works. ;-)

Just wondering: Does your code look contrived structured the way you have?
IOW, will any (simple) change(s) to M2Crypto's interfaces make it more
natural to express what you wish in this case?

Thanks. Cheers.

(I'm working hard on my start-up. Very little time for M2Crypto, alas.
Don't even have time to read mail and Usenet...)


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



From seeger at sitewaerts.de  Wed Jun 12 05:39:57 2002
From: seeger at sitewaerts.de (Felix Seeger)
Date: Wed, 12 Jun 2002 11:39:57 +0200
Subject: SyntaxWarning: name 'mailcounter' is used prior to global declaration
References: <ae74k3$49l33$1@ID-79019.news.dfncis.de>
Message-ID: <ae74tc$49l33$2@ID-79019.news.dfncis.de>

Ahh, I have it.
global was used after a mailcounter was used.

Felix Seeger wrote:

> Hi
> 
> I get this warning.
> I set mailcounter as a global var because I have to change it's value from
> a method in a class.
> This was the only way. How can I remove this warning ?
> 
> Mailcounter is defined at the beginning of the script and "global
> mailcounter" is called in the method.
> 
> 
> thanks
> have fun
> Felix



From kragen at pobox.com  Thu Jun 13 14:29:29 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 13 Jun 2002 14:29:29 -0400
Subject: What is a classmethod for?
References: <addgv30f4d@enews4.newsguy.com> <Pine.LNX.4.44.0206031559001.8474-100000@palanthas.neverending.org> <83y9dmokh0.fsf@panacea.canonical.org> <Pine.LNX.4.44.0206102306060.4174-100000@palanthas.neverending.org>
Message-ID: <833cvrm3qu.fsf@panacea.canonical.org>

Frank Tobin <ftobin at neverending.org> writes:
> On 10 Jun 2002, Kragen Sitaker wrote:
> > That's so you don't have to write this?
> > 
> > class HostPortAddress(tuple): pass
> > def HostPortAddress_from_str(s): 
> >     return HostPortAddress(s.split(':', 1))
> 
> Well, it makes sense to throw all the 'knowledge' of how to stringify and 
> de-stringify a HostPortAddress into one class.

Yes, that's reasonable.  But it seems like a small increment of
niceness to justify the conceptual complexity it adds to the language.

> > Are there times when it's more compelling?
> 
> Another example would be anytime where you have something that is close to
> a staticmethod which is a helper to the instance methods ,but needs some
> information that only the class can provide.

Helpers for instance methods can be instance methods themselves ---
there's no need for staticmethods or classmethods here.



From cliechti at gmx.net  Sat Jun 15 17:19:29 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 15 Jun 2002 23:19:29 +0200
Subject: type(socket) != SocketType
References: <3D0BAD4D.5050105@NOSPAMREMOVETHISxs4all.nl>
Message-ID: <Xns922EEDAF2FE9Ecliechtigmxnet@62.2.16.82>

Irmen de Jong <irmen at NOSPAMREMOVETHISxs4all.nl> wrote in 
news:3D0BAD4D.5050105 at NOSPAMREMOVETHISxs4all.nl:

> Hello
> 
> I have to detect if a certain socket object is a regular socket,
> or a M2Crypto SSL socket. I do this by checking type(sock),
> and then checking if this is a regular socket type. Why this way?
> because the systems that run my code do not necessarily have
> M2Crypto installed, and thus I cannot compare the socket to
> any class/type from M2Crypto, nor just go ahead and catch SSL
> exceptions (simply because they are not defined!).
> 
> There is a problem, however.
> With Python 2.2.1 on Windows, type(socket) != SocketType:
> 
> >>> from socket import *
> >>> s=socket(AF_INET,SOCK_STREAM)
> >>> type(s)
> <type 'instance'>
> >>> SocketType
> <type '_socket.socket'>
> >>> type(s) is SocketTye
> Traceback (most recent call last):
>    File "<stdin>", line 1, in ?
> NameError: name 'SocketTye' is not defined
> >>> type(s) is SocketType
> 0
> >>> type(s)== SocketType
> 0
> >>>
> 
> 
> I find this most unfortunate, because the docs on the socket module
> explicitly state that type(socket)=SocketType.

what happens to people with timoutsocket? that module repaces the 
socket.socket class with a custom one and your check will fail....

> BTW on Linux, all is fine.
> 
> Any clues how to fix things on windows?

no, sorry
maybe there is a cleverer solution that typechecking. the SSL socket has 
other/additional attributes etc right?. check for them with has_attr(). or 
use the __name__ attribute of the class. however, typechecking is probably 
not the right way to go.

chris

> Thanks a lot,
> Irmen
> 
> 



-- 
Chris <cliechti at gmx.net>



From Chris.Barker at noaa.gov  Fri Jun  7 12:43:48 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Fri, 07 Jun 2002 09:43:48 -0700
Subject: constructive critisism on pythonwin IDE
References: <6bd9f01b.0206050736.f469082@posting.google.com> <3CFE9861.5090200@skippinet.com.au>
Message-ID: <3D00E261.CA641C7E@noaa.gov>

> sameer wrote:
> > The thing I hate about PythonWin
> > is, that despite the fact that it's a development environment, there
> > is a lot of caching done of modules, and the cached copy is not
> > checked against a potentialy modified copy.  Whenever I change the
> > contents of a file and try to run a module that depends on that file,
> > I always end up with the old copy.  I have to then import and then
> > reload the changed module in the interactive window for it to refresh.

This is how Python works, it's not anything specific to PythonWin. It
works this way because a given module might be imported dozens of times
in a given Python program, and you would want that to be done fresh each
time. Many of the other Python IDEs work this way too.

> >  It would greatly increase my productivity, if I can specify which
> > modules can be cached. 

What I do is the opposite: I specify which modules can not be cached.
When a module is under develpment, I import it this way:

import modulename
reload(modulename)

that way it is guaranteed to be reloaded each time. When it's stable, I
jsut comment out the reload line.

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From BPettersen at NAREX.com  Thu Jun 13 15:27:46 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 13 Jun 2002 13:27:46 -0600
Subject: why not "'in' in 'in'"?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158ED2@admin56.narex.com>

> From: holger krekel [mailto:pyth at devel.trillke.net] 
> 
> Bjorn Pettersen wrote:
> > > From: Grant Griffin [mailto:not.this at seebelow.org]
> > > That's not a bad suggestion, Mark, but if I did that I guess 
> > > I would be forever wondering if I was testing whether the 
> > > first one was in the second or the second one was in the 
> > > first.  Again, it doesn't read nearly as well as:
> > > 
> > >    if x in subject:
> > >        ...
> > > 
> > > which leaves no doubt.
> > 
> > Ok, how about:
> > 
> > >>> class mystr(str):
> > ...   def __contains__(self, s):
> > ...     return self.find(s) >= 0
> > ...
> > >>> 'hello' in mystr('hello world')
> > 1
> > >>> 'foo' in mystr('hello world')
> > 0
> > >>>
> 
> that's fine. but usually you don't want to make up an 
> instance for every string you get from somewhere. 
> I still think that patching 'stringobject.c' to get the 
> reasonable behaviour might be worth a try for python2.3. 
> 
>     holger

I'm in total agreement, but the issue has been discussed here before on
numerous occations... nothing ever came of any of it.

-- bjorn




From kragen at pobox.com  Mon Jun 10 17:52:23 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 17:52:23 -0400
Subject: Changing TK
References: <adfv3d$jsp$1@netnews.upenn.edu> <adg3uo$p6b$1@wanadoo.fr>
Message-ID: <83elfepzs8.fsf@panacea.canonical.org>

Eric Brunel <eric.brunel at pragmadev.com> writes:
> However, both are not referenced at all in Tkinter.py, so "tweaking" it 
> would not be of any help... When you get a new tcl/tk installation, you 
> *must* recompile the interpreter, or you won't ever see any change.

Well, you must at least recompile the _tkinter extension module; you
only need to recompile the interpreter if _tkinter is statically
linked into it, which is unusual.

> But DON'T EVER replace former tcl/tk libraries by newer ones without 
> recompiling, even if they have the same name! It *may* work, but if it 
> doesn't, the tk support in your Python will be completely and definitely 
> broken...

In Linux, at least, if you add new Tcl/Tk libraries with different
version numbers, Python and Tkinter will happily continue to use the
ones they were built against.



From gerhard at bigfoot.de  Tue Jun 18 21:52:25 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 19 Jun 2002 03:52:25 +0200
Subject: tomorrow in yyyymmdd format
In-Reply-To: <XFMail.20020618183810.shalehperry@attbi.com>
References: <20020619004409.GA2583@lilith.my-fqdn.de> <XFMail.20020618183810.shalehperry@attbi.com>
Message-ID: <20020619015225.GB2739@lilith.my-fqdn.de>

* Sean 'Shaleh' Perry <shalehperry at attbi.com> [2002-06-18 18:38 -0700]:
> >> 
> >> Sure there is :-) time.mktime() can take a "malformed" time tuple and
> >> do something sensible with it. You can therefore get a time tuple from
> >> gmtime, add one to the day position and call time.mktime() on the
> >> result:
> >> 
> >> >>> x = (2002, 2, 29, 0, 31, 42, 2, 170, 0)
> >> >>> time.gmtime(time.mktime(x))
> >> (2002, 3, 1, 7, 31, 42, 4, 60, 0)
> >> >>>
> > 
> > Pure luck that this works. Don't complain if it stops workin in Python
> > 2.7 and your code breaks ;-)
> > 
> 
> actually that is a function of the mktime() function in the C library which the
> python module wraps.  Based on my reading of a man page or two this behaviour
> seems guaranteed by POSIX.  Now, the assumption made is the mktime() function
> on YOUR system follows POSIX.

Fine. I've just read the Python documentation which doesn't make this
guarantee. I'll leave it to others to check if this works on win32,
MacOS, as I'm using mxDateTime for this kind of tasks, anyway.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From dev at newdeal.ch  Sun Jun 23 14:06:37 2002
From: dev at newdeal.ch (erreur)
Date: Sun, 23 Jun 2002 20:06:37 +0200
Subject: How to represent the infinite ?
References: <3d11ee9b$1@news.swissonline.ch> <mailman.1024600723.6989.python-list@python.org> <20020620.124436.1266235189.32516@software1.logiplex.internal>
Message-ID: <3d160e45@news.swissonline.ch>

Thanks.... excelent !!


"Cliff Wells" <logiplexsoftware at earthlink.net> wrote in message
news:20020620.124436.1266235189.32516 at software1.logiplex.internal...
> In article <mailman.1024600723.6989.python-list at python.org>, "Cliff Wells"
> <logiplexsoftware at earthlink.net> wrote:
>
>
>
> > On Thu, 20 Jun 2002 17:02:33 +0200
> > erreur wrote:
> >
> >> Will somebody have an idea, to represent the infinite one?  I have
> >> variables to initialize with is +inf (or - inf).  To be sure that
> >> later, all will be smaller (or larger) than my variables. I tried to
> >> redefine the operators on an object.  It goes for Inf>10 but I do not
> >> arrive for 10>Inf (because that takes > of Int).  --------------------
> >> def __gt__(self, val):
> >>
> >>     return 1
> >> --------------------
>
>
> > Hm, for some reason, my last two posts on this thread didn't seem to go
> > through...  one more try (with a couple of little enhancements):
>
> Odd, the newsgroup -> mailing list transfer must be really slow (or down)
> today, I pulled up Pan and lo, my posts are there, they just haven't made
> it to the mailing list yet (after almost 2 hours).
>
> Or maybe I accidentally killfiled myself  ;)




From markus.vonehr at ipm.fhg.de  Tue Jun  4 02:33:43 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Tue, 04 Jun 2002 08:33:43 +0200
Subject: Using tcl/tk 8.3.4.2 with Python 2.2 on XP
References: <3CFB5994.5B2F9839@tds.net>
Message-ID: <3CFC5F47.E1D2C048@ipm.fhg.de>

Use the module hackicon to change the window icon:

http://sourceforge.net/project/showfiles.php?group_id=30460

Markus


"Edward K. Ream" schrieb:

> Hi all,
>
> This may be something trivial, but the answer doesn't seem to jump out
> at me from the FAQ or the Tkinter page at:
> http://python.org/topics/tkinter/download.html
>
> I would like to use tcl/tk 8.3.4.2 with Python/tkinter 2.2 on Windows
> XP. Apparently the Python 2.2 distribution uses tcl/tk 8.3.2.  Is there
> some way to use tcl/tk 8.3.4.2 with the binary Python 2.2 distribution
> on XP?  The thoughts I have:
>
> 1.  set an environment variable somewhere or
> 2.  just move the 8.3.4.2 dll's into the Python22/tcl folder.
>
> Neither seems likely to work well, as there may be problems with header
> compatibility, and there are more 8.3.4.2 dll's than dll's in
> Python22/tcl.
>
> I hope the next version of Python will come with tcl/tk 8.3.4 or higher
> right out of the box.  This will allow us to use wm_iconbitmap(bitmap)
> to set (at last!) the default icon for tk frames.
>
> Edward
> --------------------------------------------------------------------
> Edward K. Ream   email:  edream at tds.net
> Leo: Literate Editor with Outlines
> Leo: http://personalpages.tds.net/~edream/front.html
> --------------------------------------------------------------------



From dev at newdeal.ch  Thu Jun 20 11:02:33 2002
From: dev at newdeal.ch (erreur)
Date: Thu, 20 Jun 2002 17:02:33 +0200
Subject: How to represent the infinite ?
Message-ID: <3d11ee9b$1@news.swissonline.ch>

Will somebody have an idea, to represent the infinite one?

I have variables to initialize with is +inf (or - inf).  To be sure that
later, all will be smaller (or larger) than my variables.
I tried to redefine the operators on an object.  It goes for Inf>10 but I do
not arrive for 10>Inf (because that takes > of Int).

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

def __gt__(self, val):

    return 1

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


Not need to say to me that I can use 99999999999999 or -99999999999999 or I
do not know what.
Nor from a MAX_INT of the machine... bus Python can go higher.

Thank you for you answers (thank you to also answer by email if possible,
because I will not have any more access to the news this weekend).

PS: Sorry for my English....




From ykingma at accessforall.nl  Sat Jun  8 16:09:36 2002
From: ykingma at accessforall.nl (Ype Kingma)
Date: Sat, 08 Jun 2002 21:09:36 +0100
Subject: array, list, performance...
References: <mailman.1023391748.930.python-list@python.org> <3D0108B7.3C206C2E@accessforall.nl> <c76ff6fc.0206071806.c199f4c@posting.google.com>
Message-ID: <3D02647C.4FF9CFF7@accessforall.nl>


John Machin wrote:
> 
> Ype Kingma <ykingma at accessforall.nl> wrote in message news:
> > Multiplying the list's size by a constant factor when necessary would
> > give O(log(len(li)) * len(li)) behaviour
> 
> I'm not sure where you get this from. Here's my take:
> 
> The theoretical amortised cost is O(N), not O(log(N)*N), where N is
> the utimate size of the list.
> 
> Let F be the factor by which the list size is expanded when necessary.
> Let e be the number of expansions required to grow to size N.
> F**e is O(N).
> 
> The ith expansion involves fiddling with O(F**i) elements, with O(1)
> cost each. The total of this over e expansions is O(F**e). As above,
> F**e is O(N).

Thanks. I thought each of the e expansions would involve fiddling with
O(N) elements.
Perhaps I should post before my working day instead of after...

Regards,
Ype


From trentm at ActiveState.com  Wed Jun 19 18:46:54 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Wed, 19 Jun 2002 15:46:54 -0700
Subject: Windows versions of Python---pros and cons?
In-Reply-To: <3D110B55.4D8AD7FA@astro.cornell.edu>; from loredo@astro.cornell.edu on Wed, Jun 19, 2002 at 06:53:09PM -0400
References: <3D110B55.4D8AD7FA@astro.cornell.edu>
Message-ID: <20020619154654.C619@ActiveState.com>

[Tom Loredo wrote]
> 
> Hi folks-
> 
> I've been using Python heavily on a Solaris box and a Mac for a
> few years now.  For reasons I'll spare you from, I now am venturing
> into Windows territory (a Dell Inspiron 7500 laptop, if that matters).
> I see at least two Windows versions of Python---that at Python.org,
> and the ActiveState version.  The obvious "pro" for the ActiveState
> version is simplicity of the install (batteries included, etc.).
> But what about after that?  Are their substantial differences between
> the versions that I will bump into down the line? 

ActivePython should be 100% binary compatible with the Python installer
on python.org. If not, it is a bug. The only real "substantial"
difference, as far as your code is concerned, is the presence of some
Python extensions like PyWin32.


> In particular, I write many C extensions; can one use distutils and
> the Borland or gcc free compilers to build C extensions in both
> versions?

My understanding is that for C extensions to work they must be built
with the same compiler as that used to build the Python with which they
will run. (I may be wrong. Someone please correct me if I am.) Visual
C++ 6.0 is used to build ActivePython and a python.org's Python for
Windows. This means that you would have to build Python yourself from
source with your free compiler to be able to run your extensions built
with the free compiler.

Cheers,
Trent

-- 
Trent Mick
TrentM at ActiveState.com




From zsmith at ponymail.com.nospam  Sat Jun 15 04:50:21 2002
From: zsmith at ponymail.com.nospam (Zach)
Date: Sat, 15 Jun 2002 08:50:21 GMT
Subject: Newbie deluxe
Message-ID: <hdDO8.29984$Hn4.969427@news1.east.cox.net>

I've just started taking a look at Python (experienced in Delphi) and so far
it looks very clean and comprehensible.  My question is, what are the
fundamental differences and similarities between it and Perl?  I hear so
much about these two languages, especially in Linux circles and just
wondered what sets them apart.  Can full fledged graphical applications be
built with Python or does it lean more towards scripting?  I'm aware that it
has OOP capabilities so I assume it reaches beyond just scripting but like I
said, I've no experience with it.  I would like to broaden my coding skills
with the addition of another language so any general opinions with regards
to its usefulness are appreciated.

TIA.

Zach




From paul at soniq.net  Sun Jun 16 17:06:29 2002
From: paul at soniq.net (Paul Boehm)
Date: Sun, 16 Jun 2002 23:06:29 +0200
Subject: HT telnet interact with auto login?
In-Reply-To: <3d0cd198.17917980@news.zrz.tu-berlin.de>
References: <3d0cd198.17917980@news.zrz.tu-berlin.de>
Message-ID: <20020616210629.GA20626@soniq.net>

is this what you want?

# On Sun, Jun 16, 2002 at 06:07:30PM +0000, Klaus Reinhardt wrote:
import getpass
import sys
import telnetlib
 
HOST = "...de"
tn = telnetlib.Telnet(HOST)
 
user = '...'
password = '...'

# 2.)
tn.read_until("login: ")	# this seams to work
			# but without connection
tn.write(user + "\n")
if password:
    tn.read_until("Password: ")
    tn.write(password + "\n")
# 1.)
tn.interact()	# this is working but with manually ..





From not.this at seebelow.org  Thu Jun 13 13:09:46 2002
From: not.this at seebelow.org (Grant Griffin)
Date: 13 Jun 2002 10:09:46 -0700
Subject: why not "'in' in 'in'"?
References: <mailman.1023984257.8048.python-list@python.org>
Message-ID: <aeajkq0m4c@drn.newsguy.com>

In article <mailman.1023984257.8048.python-list at python.org>, "Mark says...
>
>> That's not a bad suggestion, Mark, but if I did that I guess I 
>> would be forever wondering if I was testing whether the first one 
>> was in the  second or the second one was in the first.
>
>Since you insist on torturing yourself, then, why not name the function:
>
>	def is_the_second_in_the_first(a, b):
>        ...
>
>;-)

Naw, you got me all wrong.  If torture was what I was after, I'd use lambda and
not name it anything at all:

   >>> subject = 'spam, eggs, sausage, spam'
   >>> eggs = 'eggs'
   >>> if lambda x, y:y.find(x) >= 0(eggs, subject):
   ...    print 'eggs is in subject'
   ...
   eggs is in subject

or-better-yet,-write-it-in-perl-<wink>-ly y'rs,

=g2

_________________________________________________________________________

Grant R. Griffin                                           g2 at dspguru.com
Publisher of dspGuru                               http://www.dspguru.com
Iowegian International Corporation                http://www.iowegian.com



From jadestar at idiom.com  Wed Jun 19 00:20:46 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 19 Jun 2002 04:20:46 GMT
Subject: Medium to Large Scale Python Deployments
References: <KHkL8.21383$eD2.2399923271@newssvr10.news.prodigy.com> <34b82fda.0206061639.67964684@posting.google.com>
Message-ID: <aep0qu$1373$1@news.idiom.com>

Stuart Quimby <stuq at dstoys.com> wrote: [top posting corrected] 
> "Domenic R. Merenda" <dom at edgereport.put_a_c_o_m_here> wrote:
>> Greetings,

>>    I am running a home-grown Enterprise Resource Planning system, written
>> entirely in Python, to manage the operations and manufacturing aspects of a
>> medium ($100,000,000 annual sales) manufacturing organization.  I would be
>> interested in hearing from other individuals who are likewise "betting the
>> farm" on Python as an ERP or MRP solution, and some of the experiences they
>> may have had.

> You may be interested in the gnuenterprise project.  They are
> developing a multi-tier db independant ERP that is open-source. 
> Details can be found at http://www.gnuenterprise.org.
> Stuart Quimby

 Notably GNUe makes extensive use of Python and PostgreSQL.



From duncan at NOSPAMrcp.co.uk  Fri Jun 14 05:07:31 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Fri, 14 Jun 2002 09:07:31 +0000 (UTC)
Subject: Date formats
References: <3D09A1B1.6040705@mxm.dk> <Xns922D5E1FF60FEduncanrcpcouk@127.0.0.1> <3D09ABEB.4000209@mxm.dk>
Message-ID: <Xns922D667FCCBF4duncanrcpcouk@127.0.0.1>

Max M <maxm at mxm.dk> wrote in news:3D09ABEB.4000209 at mxm.dk:

> Oh ... I cannot seem to find it in the docs. The closest I come is in:
> 
> 2.2.6.2 String Formatting Operations
> 
> where it says that:
> 
> "# The value conversion will use the ``alternate form'' (where defined 
> below)."
> 
In 6.9, time:
"""Additional directives may be supported on certain platforms, but only 
the ones listed here have a meaning standardized by ANSI C. 

On some platforms, an optional field width and precision specification can 
immediately follow the initial "%" of a directive in the following order; 
this is also not portable. The field width is normally 2 except for %j 
where it is 3"""

The meaning of the additional directives is not in the Python 
documentation: I used MSDN for that. Microsoft's runtime doesn't support 
field width and precision, but does use the # flag.

In case you didn't know, time.strftime calls the underlying C runtime 
strftime with a buffer 256 times longer than the format string (since there 
is no way to know how long a buffer is required).

-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From gmuller at worldonline.nl  Sun Jun 16 05:46:25 2002
From: gmuller at worldonline.nl (GerritM)
Date: Sun, 16 Jun 2002 11:46:25 +0200
Subject: Minor install annoyance on Activestate Python 2.2.1 build 222
References: <3D0BF7B4.21096694@bellatlantic.net>
Message-ID: <aehmr5$sc8$1@reader1.tiscali.nl>

"David Lees" <debl2Nonospawhammy at bellatlantic.net> schreef in bericht
news:3D0BF7B4.21096694 at bellatlantic.net...
> I just downloaded and installed the latest ActiveState build of Python
> 2.2.1 off their site and could not get it to install on my win98SE box
> even though I had previously run the windows installer.  The problem
> turns on to be that the name of the file you download is
> ActivePython-2.2.1-222.msi.exe
> Simply remove the '.exe' extension so the name is:
> ActivePython-2.2.1-222.msi and the installation goes just fine.
>
> David Lees

Quite strange, I installed ActivePython-2.2.1-222 last thursday, freshly
downloaded and did not have this (or any other) problem.

regards Gerrit
--
www.extra.research.philips.com/natlab/sysarch





From jiri at baum.com.au  Wed Jun  5 04:21:12 2002
From: jiri at baum.com.au (Jiri Baum)
Date: Wed, 5 Jun 2002 18:21:12 +1000
Subject: Graph distances, how can I speed it up?
References: <33803989.0206030235.4134d33e@posting.google.com>
Message-ID: <olhkda.q9a.ln@10.0.0.1>

Miki Tebeka:
> I'm trying to calculate distances in a graph. (The ultimate goal is to
> calculate the distances between verbs in WordNet)

> The graph is built edge after edge and each time I recalculate the
> distances matrix. However this takes ages to complete. (will finish in
> ~50H).

The matrix version of this might be more efficient, the one that solves the
whole thing as a set of simultaneous equations... I think it'd involve one
matrix inversion to begin with, and then a dot product each time you need a
distance, but I'd have to check.

Make sure you use a sparse matrix library if your matrix is sparse (and it
will be) - it can take quite some shortcuts.


Jiri
-- 
Jiri Baum <jiri at baum.com.au>           http://www.csse.monash.edu.au/~jirib
  MAT LinuxPLC project --- http://mat.sf.net --- Machine Automation Tools


From jamespp35 at yahoo.com  Fri Jun 28 15:45:25 2002
From: jamespp35 at yahoo.com (Jim Henry)
Date: Fri, 28 Jun 2002 12:45:25 -0700
Subject: incorrect upper()/lower() of UTF-8
Message-ID: <afiecp$oss$1@web1.cup.hp.com>

It appears that python does not uppercase/lowercase utf-8 strings properly:

      char
     latin-1
     utf-16
     utf-8

      ?
     F1
     00F1
     C3B1

      ?
     D1
     00D1
     C391


# create unicode/UTF-16 string with ?
>>> s = u"set description to ma?ana"
>>> s
u'set description to ma\xf1ana'                   <= correct latin-1 or
UTF-16 (F1)

# encode it as utf-8
>>> s8 = s.encode("utf8")
>>> s8
'set description to ma\xc3\xb1ana'             <= correct UTF-8 (C3B1)

# upshift the UTF-16 string
>>> s.upper()
u'SET DESCRIPTION TO MA\xd1ANA'       <= correct latin-1 or UTF-16 (D1)

# upshift the UTF-8 string
>>> s8.upper()
'SET DESCRIPTION TO MA\xc3\xb1ANA'  <= INCORRECT: should be C391

# updshift the UTF-8 string and convert back to UTF-16

>>> t = s8.upper()
>>> t
'SET DESCRIPTION TO MA\xc3\xb1ANA'
>>> u.encode("latin-1")
'SET DESCRIPTION TO MA\xf1ANA'         <= INCORRECT, will print as "MA?ANA"
instead of "MA?ANA"

Am I missing something??




From bokr at oz.net  Sun Jun  2 21:33:05 2002
From: bokr at oz.net (Bengt Richter)
Date: 3 Jun 2002 01:33:05 GMT
Subject: variable X procuct - [(x,y) for x in list1 for y in list2]
References: <1022619068.34271@newsmaster-04.atnet.at> <1022686332.297659@newsmaster-04.atnet.at> <20020531215914.15202822.larooy@xtar.co.nz>
Message-ID: <adeh0h$he6$0@216.39.172.122>

On Fri, 31 May 2002 21:59:14 +1200, John La Rooy <larooy at xtar.co.nz> wrote:

>Perhaps the translation is correct, but the example you give is a cartesian product, not a cross product.
>From my algebra book...
>
>
>In many applications of vectors to problems in geometry, physics, and engineering it is of interest to construct a vecotr in 3-space that is perpendicular to two given vectors. In this section we introduce a type of vector multiplication the facilitates this construction
>
>
>DEFINITION: If u=(u1,u2,u3) and v=(v1,v2,v3) are vectors in 3-space, then the cross product 
>is the vector defined by
>u x v = (u2v3 - u3v2 , u3v1 - u1v3 , u1v2 - u2v1)

And if i,j,k are the unit vectors, you can also write this
as the determinant of the matrix

  i  j  k
  u1 u2 u3
  v1 v2 v3

(which is easier to remember for me).

The direction of the result is perpendicular to the plane that
includes u and v, in the direction a right-hand screw would move
if turned as u rotates into v (assuming a right hand coordinate
system). Easy to see with pure x and y axis vectors:

  i  j  k
  u1 0  0
  0  v2 0

=> k*(u1*v2)  (+ in the z direction of unit vector k)

vs the same vectors but turning the other way (v x u):

  i  j  k
  0  v2 0
  u1 0  0

=> -k*(v2*u1) (also along z axis but opposite direction)

thought I'd mention, in case someone hadn't seen this before ;-)

Regards,
Bengt Richter


From storminDOTnorm at excite.com  Tue Jun 11 15:01:47 2002
From: storminDOTnorm at excite.com (Stormin)
Date: Tue, 11 Jun 2002 15:01:47 -0400
Subject: filecmp.cmpfiles on directories
Message-ID: <3D06491B.302@excite.com>

Hi python pros!

I am trying to walk a directory structure and compare directories for 
duplicates.  For some reason I get the filecmp to work on individual 
files, but not on directories.
note: I tried also with double backslash and with backslashes at the end 
of the directory name - no go. (i.e. c:\\comp1\\)

example:
DOES NOT WORK:
dups=filecmp.cmpfiles("C:\comp1","c:\comp2",common)
print dups,"||",common
result:
([],[],[])||

WORKS:
dups=filecmp.cmp("c:\comp1\testfile.txt","c:\comp1\testfile.txt")
print dups
result:
1




From maxm at mxm.dk  Thu Jun 20 19:06:48 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 21 Jun 2002 01:06:48 +0200
Subject: application development - separating content from function
References: <yroQ8.260949$i6.17986369@amsnews02.chello.com> <Xns9233ED737D952cliechtigmxnet@62.2.16.82>
Message-ID: <3D126008.9090603@mxm.dk>

Chris Liechti wrote:

> start with:
> def _(s): return s
> 
> and write all the string you want to localize later with 
> _("hello")
> _("this can be translated later by replacing the _ function")
> 
> later you can use locale or a simple dictionary lookup to translate.
> i saw that technique in "cplay" http://www.tf.hut.fi/~flu/cplay/
> 
> simple with a dict:
> translation = {'hello':'ole'}
> def _(s):
>     	return translation.get(s,s)


I have a small class I sometimes use when I need to do i18n.

class I18n:

     "Very simple i18n"

     def __init__(self, textBundle, language='uk'):
         self.language = language
         self.textBundle = textBundle

     def __call__(self, id):
         language = getattr(self, 'language', self.language)
         return self.textBundle[language].get(id,
                                     'TRANSLATION MISSING!!')

textBundle = {}

textBundle['dk'] = {}
textBundle['dk']['name'] = 'Navn'
textBundle['dk']['address'] = 'Danmark'
textBundle['dk']['date_format'] = '%d-%m-%Y %H:%M:%S'
textBundle['dk']['money_format'] = '%.2f'
textBundle['uk'] = {}
textBundle['uk']['name'] = 'Name'
textBundle['uk']['address'] = 'Denmark'
textBundle['uk']['date_format'] = '%Y.%m.%d %H:%M:%S'
textBundle['uk']['money_format'] = '%.2f'

language = 'dk'

i18n = I18n(textBundle, language)

print i18n('names')

This works to my satisfaction most of the time.



##############################################################

A pattern I frequently use is the view- Adapter/wrapper::

The idea is that the "logic" object is a python object with python types 
and methods as attributes.

To make a view of this ie. on a html page I create an adapter that takes 
the logic view as a creation parameter, and then it puts a shell/wrapper 
around the logic object to create a view without changing anything in 
the logic object.



###############################################################

## The logic object

class User:

     def __init__(self):
         self.firstName = 'Max'
         self.lastName = 'M?ller <b>Sneaky code</b>'
         self.answer = 42
         self.books = [
                 'Learning Python',
                 'Python Essential Reference',
                 'Python & XML'
         ]

## The view adapter/wrapper

# allways escape untrusted code !
import cgi

class HtmlViewAdapter:

     "this can be subclassed for several views"

     def __init__(self, obj):
         self._obj = obj

     def __getitem__(self, item):
         "Makes usage from string formatting easy"
         value = getattr(self, item)
         if callable(value):
             return str(value())
         else:
             return str(value)


class UserHtmlViewAdapter(HtmlViewAdapter):

     def firstName(self):
         return cgi.escape(self._obj.firstName)

     def lastName(self):
         return cgi.escape(self._obj.lastName)

     def answer(self):
         return self._obj.lastName

     def books(self):
         return '\n'.join(
             ['Title %s<br>' % cgi.escape(book) \
                  for book in self._obj.books])


userTemplate = """
<table>
<tr>
     <td>
         <b>First name:</b> %(firstName)s<br>
     </td>
     <td>
         <b>Last name:</b> %(lastName)s<br>
     </td>
     <td>
         <b>Answer:</b> %(answer)s<br>
     </td>
     <td>
         <b>Booklist:</b><br>
         %(books)s<br>
     </td>
</tr>
</table>
"""


user = User()

# pass the view adapter the user
userHtmlViewAdapter = UserHtmlViewAdapter(user)

# render the view
userHtmlView = userTemplate % userHtmlViewAdapter

print userHtmlView

###############################################################


The userHtmlViewAdapter can also be used for outputting directly to an 
.asp page like:

###############

<% userView = UserHtmlViewAdapter(user) %>
<%
<table>
<tr>
     <td>
         <b>First name:</b> <%= userView.firstName()%><br>
     </td>
     <td>
         <b>Last name:</b> <%= userView.lastName()%><br>
     </td>
     <td>
         <b>Answer:</b> <%= userView.answer()%><br>
     </td>
     <td>
         <b>Booklist:</b><br>
         <%= userView.books()%><br>
     </td>
</tr>
</table>
%>

This gives a good seperation of logic and view, and it's easy to see 
months later where to put the generation of view.

You can also keep as much of your code as possible in pure Python 
modules this way. Making it easier to test, debug, cross-platform etc.


regards Max M



From anoop79 at myrealbox.com  Thu Jun 13 02:44:47 2002
From: anoop79 at myrealbox.com (Anoop P B)
Date: Thu, 13 Jun 2002 12:14:47 +0530
Subject: tray icon using tkinter
Message-ID: <3D083F5F.5030209@myrealbox.com>

hi,

	is there a way that i can provide a tray icon ie. next to the clock (on 
win32) for my program (using tkinter) ? i know that there is a way of 
doing this in wxpython but i need to know whether its possible in tkinter.

thanks,
anoop





From gerhard.haering at gmx.de  Fri Jun 21 04:28:46 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 21 Jun 2002 10:28:46 +0200
Subject: compile python witout tkinter
In-Reply-To: <Pine.LNX.4.44.0206210150440.585-100000@aesaeion.com>
References: <20020621075938.GA896@lilith.my-fqdn.de> <Pine.LNX.4.44.0206210150440.585-100000@aesaeion.com>
Message-ID: <20020621082846.GA933@lilith.my-fqdn.de>

* kosh at aesaeion.com <kosh at aesaeion.com> [2002-06-21 01:55 -0600]:
> On Fri, 21 Jun 2002, Gerhard [iso-8859-15] H?ring wrote:
> 
> > * kosh at aesaeion.com <kosh at aesaeion.com> [2002-06-21 01:40 -0600]:
> > > On Fri, 21 Jun 2002, Gerhard [iso-8859-15] H?ring wrote:
> > >
> > > > * kosh at aesaeion.com <kosh at aesaeion.com> [2002-06-21 01:26 -0600]:
> > > > > On a machine I need to build python on I need to build it without tkinter.
> > > >
> > > > Why? Even if Tkinter gets compiled, it doesn't change anything for the
> > > > rest. You could even delete the tkinter modules afterwards, if you're
> > > > short on disk (quota).
> > >
> > > Because the libraries it depends on are not installed on that machine and
> > > there is no point installing them since they are not needed.
> >
> > No problem, then. It won't get built if the dependencies are not met.
> > As seen in Modules/Setup.
> 
> The problem is that it is trying to build it ayways. All the tkinter lines
> in Modules/Setup that I can find are commented out but it doesn't seem to
> have done any good.

It looks like in Python 2.1.3, Tkinter gets built with distutils in
Python's setup.py. I fear I don't have a solution to your problem, as
detect_tkinter in Python 2.1.3's setup.py looks fine to me. Maybe you
have a very strange or broken Tcl/Tk setup on your machine. As a
temporary workaround, you could try to insert a return right at the
beginning of detect_tkinter in Python's setup.py.

I sincerly hope some Tkinter expert here can come up with a better
solution and help finding the real reason behind this.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 24.5 ?C      Wind: 2.1 m/s




From jdhunter at nitace.bsd.uchicago.edu  Sat Jun 29 10:48:51 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Sat, 29 Jun 2002 09:48:51 -0500
Subject: get word base
References: <m24rfn88jp.fsf@mother.paradise.lost> <afijgq$ph0$0@216.39.172.122>
Message-ID: <m26602f8do.fsf@mother.paradise.lost>

Thanks Daniel and Bengt; excellent suggestions both.

I have installed both packages and put them into a head to head
competition.  It appears to me on a VERY LIMITED sample, that wn does
better than stemmer.  The only error (ie not what I expected) in wn
was with 'walking', which stemmer got.  So I merged them together for
the mother of all morphological root finders.  

import Stemmer, wntools
st = Stemmer.Stemmer('english')

def wnroot(w):
    return wntools.morphy(w, "noun") or \
           wntools.morphy(w, "verb") or \
           wntools.morphy(w, "adjective") or \
           wntools.morphy(w, "adverb")

def stemmer_or_wn(word):
    root = wnroot(word)
    if root != word:
        return root
    return st.stem(word)

words = ['sent', 'walking', 'thoughts', 'rakes', 'eaten', 'tried']

print 'words   : ', words
print 'stemmer : ', st.stem( words )
print 'wntools : ', [wnroot(w) for w in words]
print 'combo   : ', [stemmer_or_wn(w) for w in words]


~/python/examples $ python wordnet_demo.py
words   :  ['sent', 'walking', 'thoughts', 'rakes', 'eaten', 'tried']
stemmer :  ['sent', 'walk', 'thought', 'rake', 'eaten', 'tri']
wntools :  ['send', 'walking', 'thought', 'rake', 'eat', 'try']
combo   :  ['send', 'walk', 'thought', 'rake', 'eat', 'try']

Thanks for the help,
John Hunter


From spam at fisher.forestry.uga.edu  Tue Jun  4 10:24:56 2002
From: spam at fisher.forestry.uga.edu (Chris Fonnesbeck)
Date: 4 Jun 2002 07:24:56 -0700
Subject: weave.accelerate?
Message-ID: <c3835e5f.0206040624.7054ff3c@posting.google.com>

I'm trying to locate a version of weave that includes the accelerate()
function; the version posted on scipy.org doesnt seem to have it. Can
anyone point me in the right direction.

Thanks,
cjf


From dysfate at gmx.de  Fri Jun 14 09:36:37 2002
From: dysfate at gmx.de (Johannes Stiehler)
Date: 14 Jun 2002 15:36:37 +0200
Subject: Writing a simple linux daemon type thing...
References: <K9eO8.393$LN6.261@newsread1.prod.itd.earthlink.net> <rn3cea.rt4.ln@127.0.0.1> <yylO8.916$TP4.88596@newsread2.prod.itd.earthlink.net> <aecrfe$a4l$1@minotaurus.cip.informatik.uni-muenchen.de>
Message-ID: <aecrh5$a4l$2@minotaurus.cip.informatik.uni-muenchen.de>

Sorry, typo, I meant: redirect the output to /dev/null !!
-- 
Johannes Stiehler	
CIS der LMU M?nchen 


From shalehperry at attbi.com  Fri Jun  7 11:47:22 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 07 Jun 2002 08:47:22 -0700 (PDT)
Subject: Detecting OS and Window Manager
In-Reply-To: <3d00d248$1_1@hpb10302.boi.hp.com>
Message-ID: <XFMail.20020607084722.shalehperry@attbi.com>

> 
> Maybe import your kde module(s), and then try to do something KDE specific
> without KDE running and see what kind of error this causes so you will know
> what to trap for?  Hopefully some other newbie (for a much less significant
> value of new, AKA expert) is still following this thread to give you a
> better answer.
> 

there is no way to be sure what he wants is running.  He could do a NET_WM_*
check and see if a netwm compliant wm was running.  He could then request from
this wm a window list and see if any of those windows claim to be taskbars.

The point is, X is designed to make this hard.  One app should only use the
interfaces out there and not ask for an implementation (yep, that thread again).

Shaleh
blackbox window manager co-author




From vjovanov at stevens-tech.edu  Wed Jun  5 14:40:45 2002
From: vjovanov at stevens-tech.edu (Vojin Jovanovic)
Date: Wed, 5 Jun 2002 14:40:45 -0400
Subject: self
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu> <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com> <mailman.1023294728.2322.python-list@python.org>
Message-ID: <sWsL8.82$v_7.4260962@newnews.cc.stevens-tech.edu>

> no his questions go deeper :-)

I am glad that somebody here recognizes that.

> For every name in such a value you need to evaluate its value again.
> IOW it's recursive. The obvious solution is to intercept the
namevalue-lookups
> during evaluation of an equation. But this doesn't work for exec/eval:
> They ignore the interception points.

Correct.

Every evaluation has to go through __getattr__ method in order to access
the instance variables of the object, but that, unfortunately, can only be
done with self.
This is needed since the left hand sides of the equations must be the
variables of the instance
and not just some key values in a dictionary.

> The next best solution IMO is to preprend a namescope like 'self.'
> to every name in an expression.
> But you may have to parse the expression to
> do this.

I have been thinking about this for some time, and realized that there some
tricky cases that one has to deal with, and I am not sure that the whole
business of appending
self. will be error free in every case. For example in 'x + Geometry.x' the
correct result should
be 'self.x + self.Geometry.x' where Geometry is an instance of Geometry
class that is bound to
self.Geometry.  So one has to distinguish between two names of x in two
different places.  Maybe
stuff like this is not even so tricky.  I don't know at this point, I
haven't looked at that too
closely yet to decide the feasibility of such an approach.  I am still
trying to resolve
the problem in an easier way with whatever Python is offering to me.

> Maybe fixing exec/eval behaviour is a better idea?

Well, I don't know how to proceed in this regard. I am not familiar with the
source code.

It seems that the only two options are 1) work directly on the input string
or 2) parse the expression
and try to correctly append self. in front of the variables.

V. Jovanovic




From andreas.ames at Tenovis.com  Fri Jun 21 07:03:35 2002
From: andreas.ames at Tenovis.com (Andreas Ames)
Date: 21 Jun 2002 13:03:35 +0200
Subject: Windows versions of Python---pros and cons?
In-Reply-To: <3D1214B6.E9BB77D2@noaa.gov>
References: <3D110B55.4D8AD7FA@astro.cornell.edu>
	<aer2fn$frc$1@peabody.colorado.edu> <3D1214B6.E9BB77D2@noaa.gov>
Message-ID: <u1yb0ao6w.fsf@tenovis.com>

Hi,

Chris Barker <Chris.Barker at noaa.gov> writes:

> resorting to Py2exe and the like. What I want is something just like:
> 
> chmod +x scriptname
> and a #! line at the top

Alternatively you can use cygwin here.  It comes with it's own python port.

cheers,

andreas




From mwh at python.net  Fri Jun 14 05:33:10 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 14 Jun 2002 09:33:10 GMT
Subject: filter2
References: <mailman.1023975618.6474.python-list@python.org> <lkptyvmdfz.fsf@pc150.maths.bris.ac.uk> <mailman.1023984463.10662.python-list@python.org> <lkhek7m9c8.fsf@pc150.maths.bris.ac.uk> <slrnagii89.55q.huaiyu@gauss.almadan.ibm.com>
Message-ID: <lkd6uumci2.fsf@pc150.maths.bris.ac.uk>

huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu) writes:

> It's not that gross, but it only gives 14% speedup, though.

Oh yeah.  What you've done is kind of obvious...

> def filter2a(test, list):
>     l1 = []
>     l2 = []
>     funcs = [l1.append, l2.append]
>     map(lambda x: funcs[not test(x)](x), list)
>     return l1, l2
> 
> BTW, what's the reason that using a list for funcs is faster than using a
> tuple (as observed under Python 2.2.1 on Linux)?

Look at BINARY_SUBSCR in ceval.c... list[int] is special-cased.

Cheers,
M.

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


From spammers.do.not.bother at void.bogus  Sun Jun 23 15:05:14 2002
From: spammers.do.not.bother at void.bogus (Magnus)
Date: Sun, 23 Jun 2002 19:05:14 GMT
Subject: sys.exc_info() into string
Message-ID: <KZoR8.44805$n4.10504231@newsc.telia.net>

Hi,

in my program I am catching exceptions that I might have missed by using 
something like:

try:
        doMyStuff()
except:
        type = sys.exc_info()[0]
        syslog.syslog(syslog.LOG_EMERG, "My program: unexpected error" + type)

The thing is that I can not figure out how to make "type" into a string 
above so I can use it for e.g. syslog(). Is there an easy way of doing 
this?

Thanks,
Magnus



From loewis at informatik.hu-berlin.de  Wed Jun 12 05:07:59 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 12 Jun 2002 11:07:59 +0200
Subject: compiling error: undefined symbol: __gxx_personality_v0
References: <93523ebb.0206110732.791e745@posting.google.com>
Message-ID: <j4elfcyids.fsf@informatik.hu-berlin.de>

dubreus at csit.fsu.edu (mitsch) writes:

> gcc -shared -lc  getting_started1.o  -o getting_started1.so

You need to link with g++, not gcc.

Regards,
Martin


From hesterloli at hotmail.com  Mon Jun 24 22:47:05 2002
From: hesterloli at hotmail.com (George Hester)
Date: Tue, 25 Jun 2002 02:47:05 GMT
Subject: enlightenment (Re: python cgi + IIS)
References: <3D0F1813.3010304@gmx.de> <x0NP8.165989$Fp1.91529@atlpnn01.usenetserver.com> <3D102AF2.5010204@gmx.de> <u4rfyc4gj.fsf@online.no> <3D13522E.6010703@gmx.de>
Message-ID: <JQQR8.62208$GY.22749282@twister.nyroc.rr.com>

That's right you change it at the server not at the Web.

--
George Hester
_________________________________
"Axel Grune" <NineOfSix at gmx.de> wrote in message news:3D13522E.6010703 at gmx.de...
> > for, and so forth) give an account of all the steps you have performed
>
> This steps: http://www.e-coli.net/pyiis.html
>
> This doc tells you to change the 'App Mapping' in the properties of
> [IISAdmin] to enable the execution of python files for the entire server
> but for any reason it doesn't work.
> In the end our admin suggested to change the properties of 'Default Web
> Site' (the node above [IISAdmin]) and then it works.
>
> Thanks,
> Axel
>
>
> --
>
> please check /dev/zero for further information
>




From joe at notcharles.ca  Fri Jun  7 15:36:26 2002
From: joe at notcharles.ca (Joe Mason)
Date: Fri, 07 Jun 2002 19:36:26 -0000
Subject: Using popen interactively
References: <slrnag0fhu.6u3.joe@zork.plover.net> <adqm9k$22u6$1@nntp6.u.washington.edu>
Message-ID: <slrnag22pq.9io.joe@zork.plover.net>

In article <adqm9k$22u6$1 at nntp6.u.washington.edu>, Donn Cave wrote:
> Right.  I expect you'd do better to throw away the file objects
> popen2 makes for you, and use the pipe file descriptors directly.
> Instead of
>     infile = app.fromchild
>     ...
>         line = infile.readline()
> say
>     infile = app.fromchild.fileno()
>     ...
>         data = posix.read(infile, 8192)
>         #  split into lines if you care

Ah!  Light begins to dawn.  Thank you very much.

Joe


From m_anderson14 at hotmail.com  Tue Jun 25 15:12:40 2002
From: m_anderson14 at hotmail.com (Michael Anderson)
Date: 25 Jun 2002 12:12:40 -0700
Subject: ASP only runs once
Message-ID: <3842735a.0206251112.25609e37@posting.google.com>

Hello,

I've been using python in my asp pages for over a year now with no
problems, and just ran into a problem I can't figure out. I installed
it on a new machine and my pages only run the first time you hit them.
After that they do nothing. Pages in VBScript and JavaScript run fine.

If you edit the file and save it, it will run one time and then
nothing. Since a regular script gets compiled to a .pyc on the first
execution, I'm guessing it is somehow related. But I don't know how it
works in an asp page. Has anyone had this problem? My configuration is
Win2000 Pro, IIS 5, Python 2.1.3, win32all build 145. I also tried it
with Python 2.2 and win32all 146 but get the same results.

Thanks.

Mike


From pyth at devel.trillke.net  Tue Jun 11 20:13:03 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 12 Jun 2002 02:13:03 +0200
Subject: RfD: Automagically making a class instance 'synchronized' (draft impl.)
In-Reply-To: <20020611233740.GA28544@lilith.my-fqdn.de>; from gh_pythonlist@gmx.de on Wed, Jun 12, 2002 at 01:37:40AM +0200
References: <20020611205423.GA28113@lilith.my-fqdn.de> <20020612000425.U6609@prim.han.de> <20020611233740.GA28544@lilith.my-fqdn.de>
Message-ID: <20020612021303.V6609@prim.han.de>

Gerhard H?ring wrote:
> ...
> Btw. I've only recently started to seriously experiment with
> multithreading. Have you read my recent post
> 
> Message-ID: <slrnafpeei.p8.gerhard at developer2.opus-gmbh.net>
> with the title "Need advice on multithreading problem"
> 
> ? I find it really nifty. It's about wrapping a C library function that
> uses callbacks to Python generators. I'd welcome any comments on the
> approach I've taken there (Queue.Queue, a deferred producer thread with
> the generator in the main thread being the consumer).

I think i read it. Although i know something about bigger 
multithreaded applications (ACE/c++) and some things about python i haven't really
digged into the c-extension api so i can't be of much help there. 
But i am going to enter the wonderful world of extensions Real Soon Now (tm).

good luck anyway,

    holger




From mt at open2web.com  Thu Jun 20 22:45:49 2002
From: mt at open2web.com (Don Low)
Date: 21 Jun 2002 02:45:49 GMT
Subject: newbie string question
Message-ID: <aeu40s$9jg2s$1@ID-145503.news.dfncis.de>

There's probably a very simple explanation, but I can't see it for now, so
bear with me.

>>>pystr = 'Python'
>>>pystr = [5]
'n'
>>>pystr [2:5]
'tho'

The question is, why isn't pystr [2:5] 'thon' instead of 'tho'. Could
somebody explain?

-- 
Thanks,

Mark


From rpm1deletethis at nospamfrontiernet.net  Mon Jun 24 06:56:24 2002
From: rpm1deletethis at nospamfrontiernet.net (RPM1)
Date: Mon, 24 Jun 2002 06:56:24 -0400
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at> <af4vrl$ffe$0@216.39.172.122> <3D16DD5C.1050708@kfunigraz.ac.at>
Message-ID: <uhdume2n1mhbba@corp.supernews.com>

"Siegfried Gonzi" <siegfried.gonzi at kfunigraz.ac.at> wrote in message
news:3D16DD5C.1050708 at kfunigraz.ac.at...
> Bengt Richter wrote:
> > On Sun, 23 Jun 2002 11:05:02 +0200, Siegfried Gonzi
<siegfried.gonzi at kfunigraz.ac.at> wrote:
>
> Someone guessed that maybe my processor got too hot. Therefore: what
> about a small script of the form: processor 99% ; memory 5%.
>
> That could at least clarify whether the processor becomes to hot after
> beeing used without breaks.
>
> Writing the memory consumption script was easy, but I do not have any
> clue how to write a processor consumption script.
>

Hey!  That's why I was asking if your laptop was a Hewlitt
Packard.  I had a Hewlitt Packard Pavillion laptop that would
die when the fan kicked on.

Any chess program will keep your CPU at 100%.

GnuChess is probably on your system now.  If not, it is a free
program.  Ask around on rec.games.chess.computer if you need
help.

Patrick





From garyr at fidalgo.net  Thu Jun 27 11:21:26 2002
From: garyr at fidalgo.net (Gary Richardson)
Date: Thu, 27 Jun 2002 08:21:26 -0700
Subject: connecting embedded Python audio samples with NumPy?
References: <3D1B2603.26EA9A9C@fxtech.com>
Message-ID: <uhmbcje6c66gbe@corp.supernews.com>

I've written wrappers using CXX for a few FFTW functions that use Python
arrays to pass data to/from the library functions. I'll be happy to send you
the wrapper code and some example Python code if that would be of interest.
Gary Richardson

"Paul Miller" <paul at fxtech.com> wrote in message
news:3D1B2603.26EA9A9C at fxtech.com...
> I have a C app that reads audio data, and I'd like to pass this audio
> data in array format to Python functions so I can use it with NumPy.
>
> I have this working right now - generating a float array from the sample
> stream. But I want to put this array into a form usable by the FFT
> functions. I've read over this paper:
>
> http://www.onlamp.com/pub/a/python/2001/01/31/numerically.html
>
> Which indicates I need to put my 1D sample stream into a 2D array. Can
> anyone elaborate more on this? I'm wondering if I should provide the
> array pre-made for FFTs from my C code, or I should build the 2D array
> from the 1D array I have in Python.






From whisper at oz.net  Wed Jun 19 21:12:21 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 19 Jun 2002 18:12:21 -0700
Subject: using os.chmod in windows
In-Reply-To: <GCEDKONBLEFPPADDJCOEGEFCDKAA.whisper@oz.net>
Message-ID: <GCEDKONBLEFPPADDJCOEKEFEDKAA.whisper@oz.net>

Well, that was bone-breakingly stupid of me. The 8th group is the last
access time LOL.

The first group is the mode bits, and that means nothing happened when the
hidden attribute was set from the File Explorer - unless I caught some
refresh lag.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of David LeBlanc
> Sent: Wednesday, June 19, 2002 17:43
> To: Locke; python-list at python.org
> Subject: RE: using os.chmod in windows
>
>
> I didn't find any doc for this, so I tried the simple expedient of:
>
> # Attrtest.py
>
> import os
>
> os.stat('testfile.txt')
>
> raw_input('Change file attribute using File Explorer, then press return')
>
> os.stat('testfile.txt')
>
> When I did this, I got:
>
> (At the start, normal r/w file)
> >>> os.stat('testfile.txt')
> (33206, 0L, 9, 1, 0, 0, 26L, 1024026085, 1023320564, 1023320564)
> (Add hidden attribute using File Explorer)
> >>> os.stat('testfile.txt')
> (33206, 0L, 9, 1, 0, 0, 26L, 1024532993, 1023320564, 1023320564)
> (Add readonly in addition to hidden)
> >>> os.stat('testfile.txt')
> (33060, 0L, 9, 1, 0, 0, 26L, 1024532997, 1023320564, 1023320564)
>
> Notice that the last five digits (an int in decimal) of the 8th group have
> changed. How you parlay that into something chmod() can use, I
> have no idea.
>
> HTH,
>
> David LeBlanc
> Seattle, WA USA
>
> > -----Original Message-----
> > From: python-list-admin at python.org
> > [mailto:python-list-admin at python.org]On Behalf Of Locke
> > Sent: Wednesday, June 19, 2002 15:59
> > To: python-list at python.org
> > Subject: using os.chmod in windows
> >
> >
> > I never really learned the number codes for file permissions in UNIX. I
> > always just did "chmod +r" because it is much easier to remember. I need
> > to change a file to be 'hidden' in windows. In UNIX, this is done by
> > renaming it with a dot in front of the name. But the python os.chmod
> > command says you need to specify the number as the argument. As far as I
> > know, there is no number for 'hidden' since it is windows only. So how
> > do I do this?
> >
> > Thanks.
> >
> > --
> > http://mail.python.org/mailman/listinfo/python-list
>
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From SBrunning at trisystems.co.uk  Tue Jun 18 06:56:20 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 18 Jun 2002 11:56:20 +0100
Subject: Can python have lib to display PDF ?
Message-ID: <31575A892FF6D1118F5800600846864DCBD3B7@intrepid>

> From:	menucool at yahoo.com.cn [SMTP:menucool at yahoo.com.cn]
> 
> i want a python application to display end edit pdf.
> where are the lib pdf for python?
> thank
 
<http://www.reportlab.com/>

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.




From stefnin.nospam at yahoo.fr  Wed Jun 19 08:55:32 2002
From: stefnin.nospam at yahoo.fr (Stephane Ninin)
Date: 19 Jun 2002 12:55:32 GMT
Subject: tkinter menu (newbie question)
References: <Xns9232778FCD335stefninnospamyahoofr@213.228.0.138> <aepps0$al$1@wanadoo.fr>
Message-ID: <Xns9232967929C7Fstefninnospamyahoofr@213.228.0.138>

Eric Brunel <eric.brunel at pragmadev.com> a tapote dans
news:aepps0$al$1 at wanadoo.fr: 

Thanks for the help. 

> This doesn't suffice. With all that code, you just created a frame
> that has an attribute named menubar with the menu in it, but you
> didn't do anything to actually display the menu (there's no "magical"
> attributes in Python: if you do not do something with an attribute, it
> is only just that, an attribute...).
> 

> What you should do is to make MyApp a sub-class of Tk instead of Frame
> (this is more pratical this way), and add at the end of the
> constructor the line:
> 
> self.configure(menu=self.menubar)
>

If I want to use several of these windows,  
can still use for base class Tk, instead of Frame ? 


> And that should do the work. Oh: and don't forget to add parameter
> "self" to your "nothing" method, or you'll get a nice traceback when
> you do File->Quit...
> 

Yep... already had it before :) 
I am not completely used to this self first parameter yet.


  Thanks again.

-- 
  Stephane Ninin
stefnin.nospam at yahoo.fr
Les cons, ca ose tout... C'est meme a ca qu'on les reconnait.


From michal.cernoevic at pvt.cz  Mon Jun 24 08:28:13 2002
From: michal.cernoevic at pvt.cz (Michal Cernoevic)
Date: Mon, 24 Jun 2002 14:28:13 +0200
Subject: Compilation error - no module named os
Message-ID: <af738j$1vsg$1@news.nextra.cz>

Excuse my question, but I am unable to compile Python 2.2.1 on Linux
2.4.9-31.

I only did tar, cd Python-2.2.1; ./configure; make

and the compilation finished so:

*-s*) CC='gcc' LDSHARED='gcc -shared'
OPT='-DNDEBUG -g -O3 -Wall -Wstrict-prototypes' ./python -E ./setup.py -q
build;; \
*) CC='gcc' LDSHARED='gcc -shared'
OPT='-DNDEBUG -g -O3 -Wall -Wstrict-prototypes' ./python -E ./setup.py
build;; \
esac
Could not find platform independent libraries <prefix>
Consider setting $PYTHONHOME to <prefix>[:<exec_prefix>]
'import site' failed; use -v for traceback
Traceback (most recent call last):
  File "./setup.py", line 6, in ?
    import sys, os, getopt
ImportError: No module named os
make: *** [sharedmods] Error 1

Thanks for some ideas.

With regards Michal Cernoevic





From gerhard at bigfoot.de  Tue Jun 18 22:21:10 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 19 Jun 2002 04:21:10 +0200
Subject: procmail replacement in Python
In-Reply-To: <oqit4grq03.fsf@carouge.sram.qc.ca>
References: <20020618235901.GA2197@lilith.my-fqdn.de> <oqit4grq03.fsf@carouge.sram.qc.ca>
Message-ID: <20020619022110.GA6996@lilith.my-fqdn.de>

* Fran?ois Pinard <pinard at iro.umontreal.ca> [2002-06-18 21:55 -0400]:
> [Gerhard H?ring]
> 
> > I've recently looked into the procmail sources and run away screaming.
> 
> ROTFL!  If I remember well, the original author knew he had a peculiar
> writing style, and was writing in some prominent file (either an FAQ or
> the README, or something like) that his style was not to be discussed! :-)
> I wondered if the new maintainer revised the sources or not, I did not check.
> 
> > I'd very much prefer a Python replacement to such badly written C code.
> 
> Is it bad style?  I never went as far as asserting that it was bad style,
> because tastes may differ.  Yet in that case, I knew that tastes may differ
> very widely. :-).

I prefer Python code which I can hack very easily. Hacking well-written
C code is just much more difficult. Hacking C code that is written in a
style I immediately don't like, is something I'd like to avoid. Btw. I
have no serious reasons to dislike it, it's just gut feeling.

> Surely, `procmail' successfully addressed difficult problems, and did
> it rather solidly, with almost no bugs.

I've heard different opinions, wrt. to losing emails and efficiency. And
I've seen maildrop recommended as a better alternative multiple times.

 
> > Does anything exist that I can start from?
> 
> This has been discussed on 2000-02-07 on this list.  In particular:

Many thanks for summarizing all this. Very nice :-)

> > If I do this, I'd like to replace fetchmail with getmail as well.
> 
> What is `getmail'?

A fetchmail replacement, written in Python. fetchmail is also rumoured
to be able to lose mail. Looks like this was one of the reasons for
writing getmail.

> My experience with `procmail' as a SPAM filter told me that it may get
> frustratingly slow,

procmail isn't very efficient in my experience. That's also what the
maildrop docs say.

> when you use hundreds of filtering rules.

I have only a few rules, most of them are for putting mailing list
messages in the right folders.

> In these days, for SPAM and virus filtering, I gave in and wrote a
> Python tool which is good for me, but which would not be comprehensive
> enough for publication.

I'm currently just using spamassassin (the best spam detector around,
very configurable, but written in Perl ;-). And a few hardcoded rules
that immediately put messages from known bulkmailing software to
/dev/null. I'm also investingating Pyzor, as I found the Razor author to
be too clueless (also the razor servers aren't open-source).

> However, there are also many SPAM filtering systems
> around which have been published.  You might want to give a comprehensive
> look into all these things, before adding yet another one.  Maybe! :-)

Yup. I've solved the spam filtering problem with spamassassin. I'm only
looking into a procmail replacement.

Thanks for your answer.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From phlip_cpp at yahoo.com  Wed Jun  5 13:57:55 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 5 Jun 2002 10:57:55 -0700
Subject: SNIP IT: tk.Canvas Drag Items Around
References: <63604d2.0206031518.2da29d3b@posting.google.com>
Message-ID: <63604d2.0206050957.60a47f1@posting.google.com>

phlip_cpp at yahoo.com (Phlip) wrote in message news:<63604d2.0206031518.2da29d3b at posting.google.com>...
> Hyp Ont:
> 
> Many moons ago one Tom Vrankar posted a snippet to this forum 
> regarding dragging Tkinter Canvas items around.
> 
> I have upgraded it to actually move the target items. This also 
> demonstrates canvasx and canvasy, the subject of recent mysteries.
> 
> The snippet does not make allowances for the offset between the 
> mouse point and the anchor point of the draggee; this should be 
> easy to add. But the symptom is clicking an object and starting 
> to drag makes it "snick" over to the mouse location.

This version of the snippet drags around polygons. These are trick 
because one must update all the coords, not just the first one.

import sys, os, string, time, copy
import Tkinter
tk = Tkinter

# A Python example of drag and drop functionality within a single Tk
widget.
# The trick is in the bindings and event handler functions.
# Tom Vrankar twv at ici.net

# empirical events between dropee and target, as determined from Tk
8.0
# down.
# leave.
# up, leave, enter.

class CanvasDnD (tk.Frame):
  def __init__ (self, master):
    self.master =master
    self.loc =self.dragged =0
    tk.Frame.__init__ (self, master)
    canvas =tk.Canvas (self, width =256, height =256,
      relief =tk.RIDGE, background ="white", borderwidth =1)
    self.defaultcolor =canvas.itemcget (canvas.create_text (30, 25,
      text ="Item 1", tags ="DnD"), "fill")
    
    canvas.create_text (175, 175, text ="Item 4", tags ="DnD")
    canvas.create_text (225, 225, text ="Item 5", tags ="DnD")
    x,y = 50, 200
    canvas.create_polygon(x,y, [[x+0,y+0], [x+10,y+10], [x-10,y+10]],
fill='pink', tags = 'DnD')
    x,y = 150, 40
    canvas.create_polygon(x,y, [[x+0,y+0], [x+0,y+15], [x-15,y+10]],
fill='green', tags = 'DnD')
    x,y = 120, 120
    canvas.create_polygon(x,y, [[x+0,y+0], [x+15,y+0], [x+15,y+15],
[x+0,y+15]], fill='red', tags = 'DnD')
    x,y = 120, 170
    canvas.create_polygon(x,y, [[x+0,y+0], [x+15,y+0], [x+15,y+15],
[x+0,y+15]], fill='gray', tags = 'DnD')
    canvas.pack (expand =1, fill =tk.BOTH)
    canvas.tag_bind ("DnD", "<ButtonPress-1>", self.down)
    canvas.tag_bind ("DnD", "<ButtonRelease-1>", self.chkup)
    canvas.tag_bind ("DnD", "<Enter>", self.enter)
    canvas.tag_bind ("DnD", "<Leave>", self.leave)

  def down (self, event):
#    print "Click on %s" %event.widget.itemcget (tk.CURRENT, "text")
#    print "Click on %s" % event.widget.itemconfigure (tk.CURRENT)
    print event.widget.coords(tk.CURRENT)
    self.loc =1
    self.dragged =0
    event.widget.bind ("<Motion>", self.motion)

  def motion (self, event):
    root.config (cursor ="exchange")
    cnv = event.widget
    cnv.itemconfigure (tk.CURRENT, fill ="blue")
    xy = cnv.canvasx(event.x), cnv.canvasy(event.y)
    points = event.widget.coords (tk.CURRENT)
    anchors = copy.copy(points[:2])
    print points
    
    for idx in range(len(points)):
#        print idx, xy[idx % 2], anchors[idx % 2]
        mouse = xy[idx % 2]
        zone = anchors[idx % 2]
        points[idx] = points[idx] - zone + mouse
    
    print points
    apply(event.widget.coords, [tk.CURRENT] + points)
  
  def leave (self, event):
    self.loc =0

  def enter (self, event):
    self.loc =1
    if self.dragged == event.time:
      self.up (event)

  def chkup (self, event):
    event.widget.unbind ("<Motion>")
    root.config (cursor ="")
    self.target =event.widget.find_withtag (tk.CURRENT)
    event.widget.itemconfigure (tk.CURRENT, fill =self.defaultcolor)
    if self.loc: # is button released in same widget as pressed?
      self.up (event)
    else:
      self.dragged = event.time

  def up (self, event):
    event.widget.unbind ("<Motion>")
    if (self.target ==event.widget.find_withtag (tk.CURRENT)):
      print #"Select %s" %event.widget.itemcget (tk.CURRENT, "text")
    else:
      event.widget.itemconfigure (tk.CURRENT, fill ="blue")
      self.master.update()
      time.sleep (.1)
      print "%s Drag-N-Dropped onto %s" \
        %(event.widget.itemcget (self.target, "text"),
   event.widget.itemcget (tk.CURRENT, "text"))
      event.widget.itemconfigure (tk.CURRENT, fill =self.defaultcolor)

root =tk.Tk()
root.title ("Drag-N-Drop Demo")
CanvasDnD (root).pack()
root.mainloop()

-- 
  Phlip
            http://www.greencheese.org/MakeItSo
  --  DARE to resist drug-war profiteering  --


From oliphant at ee.byu.edu  Tue Jun 18 14:00:55 2002
From: oliphant at ee.byu.edu (Travis Oliphant)
Date: Tue, 18 Jun 2002 12:00:55 -0600
Subject: Confidence intervals, stats modules?
In-Reply-To: <mailman.1023705913.31562.python-list@python.org>
References: <ug2r3aac4etnef@corp.supernews.com> <adt0hs$3bl$1@news7.svr.pol.co.uk>
 <mailman.1023574153.24062.python-list@python.org> <3D047944.7070305@cardiff.ac.uk>
 <mailman.1023705913.31562.python-list@python.org>
Message-ID: <Pine.LNX.4.33L2.0206181159340.15024-100000@oliphant.ee.byu.edu>

>
> Assuming your are talking about a normal variate, the stats.py module
> has a function
> for the sample std deviation but not, as far as I recall, for the
> Student's t distribution,
> which is needed to calculate confieence intervals where the popuation
> std deviation is
> unknown (which is most of the time).

The scipy.stats

module in scipy  (www.scipy.org)

has Student's t distribution and many many others.

It could be used to compute confidence intervals.

Happy computing,


-Travis O.



From graham_guttocks at yahoo.co.nz  Tue Jun  4 14:22:55 2002
From: graham_guttocks at yahoo.co.nz (Graham Guttocks)
Date: Tue, 4 Jun 2002 11:22:55 -0700 (PDT)
Subject: os.path.expanduser ignores os.seteuid
Message-ID: <20020604182255.97926.qmail@web10306.mail.yahoo.com>

Why doesn't os.path.expanduser pay attention to the fact that I've
changed the process's effective user id?

As root:

Python 2.1.3 (#1, Apr 11 2002, 18:30:30) 
[GCC 2.95.3 20010315 (release) [FreeBSD]] on freebsd4
Type "copyright", "credits" or "license" for more information.
>>> import os
>>> os.seteuid(667)
>>> os.path.expanduser("~")
'/root'
>>> 

Other functions like os.getlogin() don't have this problem.

Any workarounds?


__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com




From sholden at holdenweb.com  Thu Jun  6 08:55:50 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 6 Jun 2002 08:55:50 -0400
Subject: HttpServer help!
References: <3CFF4C14.450850D8@millfilm.co.uk>
Message-ID: <P1JL8.312164$%u2.176575@atlpnn01.usenetserver.com>

"Eric Texier" <erict at millfilm.co.uk> wrote in message
news:3CFF4C14.450850D8 at millfilm.co.uk...
> I am trying to get some simple example running locally on my
> machine of http(server/client)  code from the book 'Python & Xml'
> and I am having trouble.
>
> if someone can look at the following script and tell me if
> the problem can be because I don't have the privilege and/or
> I didn't setup my system properly:
> Thanks.
>
> I am running on Linux Red Hat with python2.2.
>
>
> The client code to POST:
> """
> post.py
> """
> from httplib import HTTP
> from urllib  import quote
>
> # establish POST data
> myquote = 'This is my quote: "I think therefore I am."'
>
> # be sure not to quote the key= sequence...
> postdata = "favquote=" + quote(myquote)
>
> print "Will POST ", len(postdata), "bytes:"
> print postdata
>
> # begin HTTP request
>
> #req = HTTP("127.0.0.1")  # change to your IP or localhost
> #req = HTTP("localhost")  # change to your IP or localhost
>
> ##### MY IP address
> req = HTTP("10.16.3.132")  # change to your IP or localhost
>
>
> #req.putrequest("POST", "/c8/favquote.cgi")
> req.putrequest("POST",
> "file:/usr/people/erict/python/srcToolskit/pythonXmlBook/c8/favquote.cgi")
>
> .
Hold it right here. How is a FILE supposed to process the input you are
trying to send it? Your .putrequest() call should give the path to the CGI
script *relative to the server root*, not it's location in the server's file
store.

> .
> .
>
> ##### That the Error after the ruprequest->"#####
> Traceback (most recent call last):
>   File "post.py", line 25, in ?
>     req.putrequest("POST",
> "file:/usr/people/erict/python/srcToolskit/pythonXmlBook/c8/favquote.cgi")
>
>   File "/usr/people/erict/bin//lib/python2.2/httplib.py", line 453, in
> putrequest
>     self.send(str)
>   File "/usr/people/erict/bin//lib/python2.2/httplib.py", line 395, in
> send
>     self.connect()
>   File "/usr/people/erict/bin//lib/python2.2/httplib.py", line 379, in
> connect
>     raise socket.error, msg
> socket.error: (111, 'Connection refused')
>
Sure enough, when the client tries to create a connection, it can't.

[...]

Hope this helps.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From kragen at pobox.com  Tue Jun  4 18:03:37 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 04 Jun 2002 18:03:37 -0400
Subject: Python-URL! correction
Message-ID: <83lm9u4s7a.fsf@panacea.canonical.org>

Python-URL! this week said someone was having trouble using McMillan
Installer with Gadfly, and that Gordon had tried and failed to help
the poster.  This was wrong --- the (anonymous) person in question was
having trouble with py2exe, not the McMillan Installer.

Gordon, please accept my apologies for the error, and thank you for
correcting me.



From buzzard at urubu.freeserve.co.uk  Sat Jun  8 09:25:41 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Sat, 8 Jun 2002 14:25:41 +0100
Subject: Confidence intervals, stats modules?
References: <ug2r3aac4etnef@corp.supernews.com>
Message-ID: <adt0hs$3bl$1@news7.svr.pol.co.uk>

"Stephen Boulet" <stephen at theboulets.net> wrote in message
news:ug2r3aac4etnef at corp.supernews.com...
> Are there any stats modules to calculate things like confidence intervals,
> mean, standard deviation? Thanks.
>
> -- Stephen

http://www.nmr.mgh.harvard.edu/Neural_Systems_Group/gary/python.html





From mitnick at sobral.org  Fri Jun 21 20:40:12 2002
From: mitnick at sobral.org (KEVIN MITNICK)
Date: Sat, 22 Jun 2002 00:40:12 GMT
Subject: =?iso-8859-1?Q?Elei=E7=F5es?= Brasileiras/Brasilian Election
Message-ID: <20020622004012.11719.qmail@linux.sobral.org>

		PORTUGU?S
O que um governo corrupto ? capaz de fazer pra manter-se no poder?!
TUDO:manipula??o da imprensa, da economia, das pesquisas eleitorais,
mentiras e falsas acusa??es s?o apenas alguns dos expedientes usados
por FHC e sua equipe (de corruptos) pra manter-se no poder.
Acordem brasileiros. N?o podemos deixar que esse terrorista social
volte a eleger os seus. Basta!Vamos dar uma resposta nas urnas! Vamos
dar uma chance a esquerda. Vamos eleger LULA pra presidente!!! 

		ENGLISH
What a corrupt goverment is capable to make to keep the power itself?!
EVERYTHING:manipulation of press, of economy, of electoral research,
lies and false accusation are just some  resources used
by FHC and his team (of corrupts) to keep the power themselves.
Brasilians wake up. We can't let this social terrorist elect his
friends again. That's enough!Let's give an answer in the elections.
Let's give a chance to Left Hand. Let's elect LULA for president of Brazil. 





From dfackrell at DELETETHIS.linuxmail.org  Fri Jun  7 11:17:10 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Fri, 7 Jun 2002 09:17:10 -0600
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <20020606215349.GA13474@node1.opengeometry.net> <mailman.1023401650.18434.python-list@python.org> <adpktk$17j3k$1@ID-146704.news.dfncis.de>
Message-ID: <3d00ce76$1_2@hpb10302.boi.hp.com>

"Shagshag13" <shagshag13 at yahoo.fr> wrote in message
news:adpktk$17j3k$1 at ID-146704.news.dfncis.de...

...

> set = {}
> map(set.__setitem__, alist, [])
> alist = set.keys()
>
> But that's not trivial !!! And quite difficult to find...

The reason that this example does not seem trivial to you or to me is most
likely that we don't fully grasp the concept and usage of map().  Once we
get that behind us, I'm sure that this example will turn into something that
we use all the time without much thinking about it.

>
> > those things doesn't make your code obfuscated..
>
> I agree :)
>
> > [file.printsize() for file in directory]
> >
> > they're the same, and everyone would recommend you the last one..
>
> Yes but i would have asked why this is the best... And i think that
> your answer would have helped me to become a better pythoner...

This concept (I believe it's called "list comprehensions") is another like
the map() usage.  Obscure and arcane to those who don't fully grasp it (like
me, again), but very powerful.  I'm not sure what either of them has to do
with optimization, though, except that they are both probably very well
optimized at the underlying C level.

Daniel Fackrell




From oren-py-l at hishome.net  Sat Jun 15 06:25:19 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Sat, 15 Jun 2002 13:25:19 +0300
Subject: Pronouncing '__init__'
Message-ID: <20020615132519.A14452@hishome.net>

There has been a thread about this issue in the past. Below are some of the
proposals people made.  I'd like to have an informal vote - how do *you*
say Python's magic names when talking with other Pythoneers?

 (all my Python-related communication is written)
 double underscore init double underscore
 under under init under under
 double bar init double bar
 bar bar init bar bar
 unders init unders
 under init under
 bar init bar
 magic init
 pyinit

Most of these pronounciations also have a shorter version with the trailing 
part omitted.

	Oren





From pyth at devel.trillke.net  Wed Jun 12 12:05:51 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 12 Jun 2002 18:05:51 +0200
Subject: methods to parse big foreign ascii-file?
Message-ID: <20020612180551.Y6609@prim.han.de>

Hello,

tommorrow i want to convince an old skool C-programmer to 
use python. We are going to make a one-day project :-)

It involves reading 50MB+ files of a strange ascii-format
containing names, floats and flags all around. These items 
span multiple lines and should be grouped in lists or objects.

How would you go about it? Options that come to my mind:

- using the re module to parse line for line and 
  try to get it all right by hand. 

- come up with a grammar and let a parser do
  the hard work (which one?)

- using some module i don't yet know?

thanks for pointers or comments,

    holger




From Sascha.Ferley at infineon.net  Sun Jun 23 22:00:46 2002
From: Sascha.Ferley at infineon.net (Sascha Ferley)
Date: Sun, 23 Jun 2002 20:00:46 -0600 (MDT)
Subject: PIL, Python and CGI
In-Reply-To: <200206241138.19665.rjones@ekit-inc.com>
Message-ID: <Pine.GSO.4.33.0206231958410.27090-100000@infinity.infineon.net>


On Mon, 24 Jun 2002, Richard Jones wrote:

> On Mon, 24 Jun 2002 11:30, Sascha Ferley wrote:
> > Thanks for replying back to me, this thing has been driving me nuts for a
> > while now. Anyways, i removed the try/catch block from the Image.py lib,
> > and checked the error output:
> >
> > Traceback (most recent call last):
> >   File "/export/server/web/CPSC461/cgi-bin/interfce.py", line 15, in ?
> >     print Image.open(urlretrieve(sourcePath, '/tmp/img1')[0]).histogram()
> >   File "/usr/local/lib/python2.2/site-packages/PIL/Image.py", line 577, in
> > histogram
> >     self.load()
> >   File "/opt/sfw/lib/python2.2/site-packages/PIL/ImageFile.py", line 126,
> > in load
> >     self.load_prepare()
> >   File "/opt/sfw/lib/python2.2/site-packages/PIL/ImageFile.py", line 180,
> > in load_prepare
> >     self.im = Image.core.new(self.mode, self.size)
> > AttributeError: 'module' object has no attribute 'core'
>
> "core" is assigned just after the "import _imaging" line in Image.py. Are you
> sure there isn't another traceback we're missing? What happens if you just
> have a simple CGI script that just does "import _imaging"?

Hi Richard

Thanks again, well here is what occurs,
program/script:

import sys
import _imaging

sys.stderr = sys.stdout
print "Content-Type: text/plain"
print sys.path


The error that results:

[Sun Jun 23 19:58:32 2002] [error] (8)Exec format error: exec of
/export/server/web/CPSC461/cgi-bin/test3.py failed
[Sun Jun 23 19:58:32 2002] [error] [client 136.159.3.51] Premature end of
script headers: /export/server/web/CPSC461/cgi-bin/test3.py

After 48hours of straight coding i might not have done the best job
here:).

Sascha





From james.kew at btinternet.com  Sat Jun 29 15:56:41 2002
From: james.kew at btinternet.com (James Kew)
Date: Sat, 29 Jun 2002 20:56:41 +0100
Subject: private
References: <pan.2002.06.29.17.38.51.789527.440@myself.com> <3D1DE132.4A890D9D@engcorp.com>
Message-ID: <afl3e4$f8abc$1@ID-71831.news.dfncis.de>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3D1DE132.4A890D9D at engcorp.com...
> Rhymes wrote:
> >
> > Is there any possibility to build up a _real_ private attribute?
> > Such as C++ private data members...
>
> You realize that it's quite possible to get at private data members
> in C++, don't you?

I can't help feeling this is overly dismissive!

Yes, of course one can circumvent private in C++. But for me the main value
of the public and private specifiers is that they _document_ to the client
of the class what it should access and what it should not.

I don't -- in my admittedly very limited experience -- find Python classes
as immediately self-documenting. It's not obvious from the class definition
what is intended to be called or modfied by the client and what is internal
to the implementation.

I suspect I'm still in the crossover period from strictly declarative C/C++
and the "anything goes" Pythonic way and that, given time, I'll find my
groove and feel comfortable.

> This has been discussed about 479 times in the last decade.

I'll bet. I'll shut up about it now.

James





From jbublitzNO at SPAMnwinternet.com  Tue Jun 18 16:37:23 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Tue, 18 Jun 2002 20:37:23 GMT
Subject: Embedding Python - time module problem
References: <cIgHcXA4SyD9swmz@edshk.demon.co.uk>
Message-ID: <3D0F99EA.7060904@SPAMnwinternet.com>

Richard Townsend wrote:

> If I substitute "os" or "sys" for "time" then no error occurs. However
> "math" does give a similar error.  

> If I run the interpreter on its own, I can import time and math with no
> problems. 

> I am using Python 2.2.1 on HP-UX 11i

> Is there something special I need to do to be able to import time and
> math modules when using embedded Python?

This sounds similar to the problem I had under Linux -
sys is builtin, time and math aren't, and the interpreter
symbols needed aren't available to the time and math
libs.

If you're using something like 'dlopen' to import the
interpreter, it needs to happen with the RTLD_GLOBAL
flag set. If you're statically linking, there's probably
an equivalent linker switch. I was using KDE which has
a method for globally loading libs. libtool looked
messier, but fortunately I didn't need to get into it.

man dlopen (at least for Linux).

Jim






From steve.bond at btinternet.com  Wed Jun 19 06:20:42 2002
From: steve.bond at btinternet.com (Steve)
Date: Wed, 19 Jun 2002 10:20:42 +0000 (UTC)
Subject: newbie: help please!
Message-ID: <aepltq$gl3$1@knossos.btinternet.com>

hello,

i have python 2.2.1 installed on my mandrake 8.1 system, which seems to 
work fine.

The problem is im trying to install Tk exstensions, i have downloaded both 
tcl8.3.4 and tk8.3.4 files and installed using the ./configure, make and 
then make install commands. I then when into the setup file in the 
Python-2.2.1/Modules directory where i understand i have to enable the 
tkinter extension module, by uncommenting the approritae lines.

Lastly i run 'make' at the top of my Python-2.2.1 directory but i get this 
error:

gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include 
-DHAVE_CONFIG_H   -DWITH_APPINIT -I/usr/local/include -I/usr/X11R6/bin -c 
./Modules/_tkinter.c -o Modules/_tkinter.o
In file included from ./Modules/_tkinter.c:49:
/usr/include/tk.h:83:29: X11/Xlib.h: No such file or directory

i have the Xlib.h in the Python-2.2.1/Include directory

the tkinter part of my setup file looks like this and everything was 
installed in the default places


# *** Always uncomment this (leave the leading underscore in!):
 _tkinter _tkinter.c tkappinit.c -DWITH_APPINIT \
# *** Uncomment and edit to reflect where your Tcl/Tk libraries are:
        -L/usr/local/lib \
# *** Uncomment and edit to reflect where your Tcl/Tk headers are:
        -I/usr/local/include \
# *** Uncomment and edit to reflect where your X11 header files are:
        -I/usr/X11R6/bin \
# *** Or uncomment this for Solaris:
#       -I/usr/openwin/include \
# *** Uncomment and edit for Tix extension only:
#       -DWITH_TIX -ltix8.1.8.2 \
# *** Uncomment and edit for BLT extension only:
#       -DWITH_BLT -I/usr/local/blt/blt8.0-unoff/include -lBLT8.0 \
# *** Uncomment and edit for PIL (TkImaging) extension only:
#     (See http://www.pythonware.com/products/pil/ for more info)
#       -DWITH_PIL -I../Extensions/Imaging/libImaging  tkImaging.c \
# *** Uncomment and edit for TOGL extension only:
#       -DWITH_TOGL togl.c \
# *** Uncomment and edit to reflect your Tcl/Tk versions:
        -ltk8.3 -ltcl8.3 \
# *** Uncomment and edit to reflect where your X11 libraries are:
        -L/usr/X11R6/lib \
# *** Or uncomment this for Solaris:
#       -L/usr/openwin/lib \
# *** Uncomment these for TOGL extension only:
#       -lGL -lGLU -lXext -lXmu \
# *** Uncomment for AIX:
#       -lld \
# *** Always uncomment this; X11 libraries to link with:
        -lX11

Any help would be much appreciated!

Thanks

Steven


From hesterloli at hotmail.com  Wed Jun 19 09:42:58 2002
From: hesterloli at hotmail.com (George Hester)
Date: Wed, 19 Jun 2002 13:42:58 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com> <3D1010C3.C321F8B0@engcorp.com> <raXP8.47926$GY.15260255@twister.nyroc.rr.com> <3D107948.B3FEE4BB@engcorp.com>
Message-ID: <CT%P8.33113$uk2.13710586@twister.nyroc.rr.com>

Well I know quite a lot about limits.  Nothing about Maple.  But the reason I
mentioned is that I hoped that my students would take an interest in the subject
and learn the concept.  They were not interested but they needed the credit to
graduate.  In my case I was not necessarily interested in learning Python.  But
I did want to determine my version of Python in ASP.  I suppose I was like my
students and ya'll were like the teacher.  Yes basically I wanted it handed to
me on a platter.  What I did with the platter may have surprised you.  So just
becasue somebody wants something handed to them on a platter does not mean the
contribution has less value.

--
George Hester
_________________________________
"Peter Hansen" <peter at engcorp.com> wrote in message
news:3D107948.B3FEE4BB at engcorp.com...
> George Hester wrote:
> >
> > Yes my take on it was that you said "I needed to crawl before I walk."
Please
> > don't hold my inference against me.
>
> That's precisely what I said.  Nothing derogatory about it, except
> in your own mind.  We /all/ need to crawl before we walk, in all
> things, unless we're lucky.  I was just reminding you of this
> fact of life.
>
> If I were attempting to learn limit calculus, and simultaneously
> apply it to a difficult problem while using Maple, and I knew
> nothing about Maple and little about limit calculus, you would
> be helping me if you said "crawl before you walk... just learn
> the basics of limit calculus before you try to figure out how
> to use Maple too".  I hope I would not be so sensitive as to
> take that the wrong way.
>
> -Peter
>
> P.S.: I know little about limit calculus, and nothing about
> Maple, so please forgive the above example if it combines
> the two infeasibly.... :)




From jtt at hnc.com  Wed Jun  5 09:57:36 2002
From: jtt at hnc.com (Jim)
Date: 5 Jun 2002 06:57:36 -0700
Subject: DCOracle2 - Can't get cursor.execute to return data when SQL query involves datetimes.
References: <be0d725a.0206041139.1f8f6647@posting.google.com>
Message-ID: <be0d725a.0206050557.1592bc3b@posting.google.com>

Here is better version of the same program.  Sorry for last posting. 

! /usr/bin/env python
import DCOracle2
#*************************************************************
# Define variables and constants
#*************************************************************
TRUE = 1
FALSE = 0
FalconPwd = 'uid/pwd'
Results_count = 0
db = DCOracle2.connect(FalconPwd)
#
# Open a new cursor
#
dbCursor = db.cursor()
tran_date = '19970116124248'
dt = DCOracle2.Timestamp(int(tran_date[0:4]), int(tran_date[4:6]),
int(tran_date[6:8]), int(tran_date[8:10]), int(tran_date[10:12]),
int(tran_date[12:14]))

Results_count = DCOracle2.execute(dbCursor,'select * from
falcon.falt002d where ACCT_NBR = :1 and trn_dt = :2',"2222222222222222
  ",dt)


tabledata = dbCursor.fetchone()

print tabledata


From oren at hishome.net  Sat Jun 29 11:24:48 2002
From: oren at hishome.net (Oren Tirosh)
Date: Sat, 29 Jun 2002 18:24:48 +0300
Subject: Shell-like functionality in Python
Message-ID: <20020629182448.A24662@hishome.net>

Check out http://www.tothink.com/python/shell

It's still in a pretty early state but already provides a very convenient 
syntax for flexible execution of external commands from Python.

External commands appear as virtual methods of an object called 'cmd'.

Example:
Python 2.2 (#1, Jan 30 2002, 23:44:31)
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-98)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> from shell.all import *
>>> cmd.ls()
all.py  command.py  __init__.py

It supports output redirection in the form of iterable commands:

>>> for x in cmd.ls['-l']:
...   print 'line:', x
...
line: total 20
line: -rw-rw-r--    1 oren     oren          216 Jun 27 20:49 all.py
line: -rw-rw-r--    1 oren     oren         8259 Jun 27 20:53 command.py
line: -rw-rw-r--    1 oren     oren          136 Jun 27 20:49 __init__.py

Passing the output of one command as arguments to another:

>>> cmd.chmod('+r', cmd.find['.', '-name', '*.html'])

And some other nifty features.

	Oren





From erict at millfilm.co.uk  Fri Jun 21 12:24:43 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Fri, 21 Jun 2002 17:24:43 +0100
Subject: property type missing
Message-ID: <3D13534A.843A616A@millfilm.co.uk>

I am using "property" and I would like to retrieve their name
automatically
from the dict-proxy of my class.

I do something like:
class foo:
    ....

>>> a = foo()
>>> d = vars(a.__class__)
>>> d.items()
[('ouputType', <property object at 0x8182e74>),.....

know I would like now, to check again the "types" module the type of all
the value
to define which one are of type property.
But   types doesn't have it define. I am wondering why and if it will be
a way
around.

Thanks
Eric




From slhath at charter.net  Fri Jun 21 14:09:33 2002
From: slhath at charter.net (Scott Hathaway)
Date: Fri, 21 Jun 2002 13:09:33 -0500
Subject: ssl on windows
Message-ID: <uh6qucjo5omi5e@corp.supernews.com>

I would like to use ssl on Windows with Python 2.2.  Is this available yet?
If so, how can I go about getting it installed?  When I try to get an https
address with urllib, it tells me that https is an unknown protocol.

Thanks,
Scott




From fredrik at pythonware.com  Thu Jun 20 05:38:28 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 20 Jun 2002 09:38:28 GMT
Subject: object object
References: <3D1195BE.9B10EE00@millfilm.co.uk>
Message-ID: <oohQ8.44260$n4.10310749@newsc.telia.net>

Eric Texier wrote:

> I didn' t find a lot about class object.
> and I would  like to know what I can
> inherit from it and what is the difference
> between, class foo(object): and class foo:

the former creates a "new-style class" or subtype,
the latter a "classic class":

to learn more, start here:

    http://www.amk.ca/python/2.2/index.html#SECTION000310000000000000000
    http://www.python.org/2.2.1/descrintro.html

</F>




From cimarron+google at taylors.org  Sun Jun  9 04:21:41 2002
From: cimarron+google at taylors.org (Cimarron Taylor)
Date: 9 Jun 2002 01:21:41 -0700
Subject: Tkinter Help Please
References: <B927D335.6B54%sarmstrong13@mac.com>
Message-ID: <29e28c51.0206090021.72466d4b@posting.google.com>

Since I started working with Pmw, I've had to do this a few times
on linux machines (RedHat 7.1).  Here is the procedure I follow:

1. install X11 headers and libraries if necessary
   (rpm -i XFree86-devel-4.0.3-5.i386.rpm)

2. put tcl8.3.4.tar.Z, tk8.3.4.tar.Z, Python-2.2.1.tgz 
   and Pmw.0.8.5.tar.gz into a temporary directory

3. copy my standard Makefile and Setup.dist (see end of message)
   files into the temporary directory. 

4. su to root

5. run 'make untar tcl tk python pmw'

You might not need Pmw If you aren't doing your own Tkinter development  
(in which case you can omit it).  Pmw is not strictly required to use
Tkinter but I have found it very helpful.

Cim



My standard Makefile contains the text in between the lines of dashes:
-------------------------------------------------------------------------
all:
        @echo you need to be root, then type
        @echo make untar tcl tk python pmw 

pmw:
        mv Pmw /usr/local/lib/python2.2/site-packages

python:
        mv Python-2.2.1/Modules/Setup.dist Python-2.2.1/Modules/Setup.dist-
        cp Setup.dist Python-2.2.1/Modules
        (cd Python-2.2.1 && ./configure && make && make install)

tk:
        (cd tk8.3.4/unix && ./configure && make && make install)
        ldconfig -v

tcl:
        (cd tcl8.3.4/unix && ./configure && make && make install)

untar:
        tar xfz Python-2.2.1.tgz
        tar xfz tcl8.3.4.tar.Z
        tar xfz tk8.3.4.tar.Z
        tar xfz Pmw.0.8.5.tar.gz
-------------------------------------------------------------------------

My Setup.dist contains the text in between the lines of dashes.
It is essentially the same as the Setup.dist which comes with Python
except that the tkinter module settings have been modified appropriately.

-------------------------------------------------------------------------
# -*- makefile -*-
# The file Setup is used by the makesetup script to construct the files
# Makefile and config.c, from Makefile.pre and config.c.in,
# respectively.  The file Setup itself is initially copied from
# Setup.dist; once it exists it will not be overwritten, so you can edit
# Setup to your heart's content.  Note that Makefile.pre is created
# from Makefile.pre.in by the toplevel configure script.

# (VPATH notes: Setup and Makefile.pre are in the build directory, as
# are Makefile and config.c; the *.in and *.dist files are in the source
# directory.)

# Each line in this file describes one or more optional modules.
# Modules enabled here will not be compiled by the setup.py script,
# so the file can be used to override setup.py's behavior.

# Lines have the following structure:
#
# <module> ... [<sourcefile> ...] [<cpparg> ...] [<library> ...]
#
# <sourcefile> is anything ending in .c (.C, .cc, .c++ are C++ files)
# <cpparg> is anything starting with -I, -D, -U or -C
# <library> is anything ending in .a or beginning with -l or -L
# <module> is anything else but should be a valid Python
# identifier (letters, digits, underscores, beginning with non-digit)
#
# (As the makesetup script changes, it may recognize some other
# arguments as well, e.g. *.so and *.sl as libraries.  See the big
# case statement in the makesetup script.)
#
# Lines can also have the form
#
# <name> = <value>
#
# which defines a Make variable definition inserted into Makefile.in
#
# Finally, if a line contains just the word "*shared*" (without the
# quotes but with the stars), then the following modules will not be
# included in the config.c file, nor in the list of objects to be
# added to the library archive, and their linker options won't be
# added to the linker options, but rules to create their .o files and
# their shared libraries will still be added to the Makefile, and
# their names will be collected in the Make variable SHAREDMODS.  This
# is used to build modules as shared libraries.  (They can be
# installed using "make sharedinstall", which is implied by the
# toplevel "make install" target.)  (For compatibility,
# *noconfig* has the same effect as *shared*.)
#
# In addition, *static* reverses this effect (negating a previous
# *shared* line).

# NOTE: As a standard policy, as many modules as can be supported by a
# platform should be present.  The distribution comes with all modules
# enabled that are supported by most platforms and don't require you
# to ftp sources from elsewhere.


# Some special rules to define PYTHONPATH.
# Edit the definitions below to indicate which options you are using.
# Don't add any whitespace or comments!

# Directories where library files get installed.
# DESTLIB is for Python modules; MACHDESTLIB for shared libraries.
DESTLIB=$(LIBDEST)
MACHDESTLIB=$(BINLIBDEST)

# NOTE: all the paths are now relative to the prefix that is computed
# at run time!

# Standard path -- don't edit.
# No leading colon since this is the first entry.
# Empty since this is now just the runtime prefix.
DESTPATH=

# Site specific path components -- should begin with : if non-empty
SITEPATH=

# Standard path components for test modules
TESTPATH=

# Path components for machine- or system-dependent modules and shared libraries
MACHDEPPATH=:plat-$(MACHDEP)

# Path component for the Tkinter-related modules
# The TKPATH variable is always enabled, to save you the effort.
TKPATH=:lib-tk

COREPYTHONPATH=$(DESTPATH)$(SITEPATH)$(TESTPATH)$(MACHDEPPATH)$(TKPATH)
PYTHONPATH=$(COREPYTHONPATH)


# The modules listed here can't be built as shared libraries for
# various reasons; therefore they are listed here instead of in the
# normal order.

# This only contains the minimal set of modules required to run the 
# setup.py script in the root of the Python source tree.

posix posixmodule.c		# posix (UNIX) system calls
_sre _sre.c			# Fredrik Lundh's new regular expressions
new newmodule.c			# Tommy Burnette's 'new' module

# The rest of the modules listed in this file are all commented out by
# default.  Usually they can be detected and built as dynamically
# loaded modules by the new setup.py script added in Python 2.1.  If
# you're on a platform that doesn't support dynamic loading, want to 
# compile modules statically into the Python binary, or need to 
# specify some odd set of compiler switches, you can uncomment the 
# appropriate lines below.

# ======================================================================

# The Python symtable module depends on .h files that setup.py doesn't track
_symtable symtablemodule.c

# The SGI specific GL module:

GLHACK=-Dclear=__GLclear
#gl glmodule.c cgensupport.c -I$(srcdir) $(GLHACK) -lgl -lX11

# Pure module.  Cannot be linked dynamically.
# -DWITH_QUANTIFY, -DWITH_PURIFY, or -DWITH_ALL_PURE
#WHICH_PURE_PRODUCTS=-DWITH_ALL_PURE
#PURE_INCLS=-I/usr/local/include
#PURE_STUBLIBS=-L/usr/local/lib -lpurify_stubs -lquantify_stubs
#pure puremodule.c $(WHICH_PURE_PRODUCTS) $(PURE_INCLS) $(PURE_STUBLIBS)

# Uncommenting the following line tells makesetup that all following
# modules are to be built as shared libraries (see above for more
# detail; also note that *static* reverses this effect):

#*shared*

# GNU readline.  Unlike previous Python incarnations, GNU readline is
# now incorporated in an optional module, configured in the Setup file
# instead of by a configure script switch.  You may have to insert a
# -L option pointing to the directory where libreadline.* lives,
# and you may have to change -ltermcap to -ltermlib or perhaps remove
# it, depending on your system -- see the GNU readline instructions.
# It's okay for this to be a shared library, too.

#readline readline.c -lreadline -ltermcap


# Modules that should always be present (non UNIX dependent):

#array arraymodule.c	# array objects
#cmath cmathmodule.c # -lm # complex math library functions
#math mathmodule.c # -lm # math library functions, e.g. sin()
#struct structmodule.c	# binary structure packing/unpacking
#time timemodule.c # -lm # time operations and variables
#operator operator.c	# operator.add() and similar goodies
#_weakref _weakref.c	# basic weak reference support
#_codecs _codecsmodule.c	# access to the builtin codecs and codec registry
#_testcapi _testcapimodule.c    # Python C API test module

#unicodedata unicodedata.c    # static Unicode character database

#_locale _localemodule.c  # access to ISO C locale support


# Modules with some UNIX dependencies -- on by default:
# (If you have a really backward UNIX, select and socket may not be
# supported...)

#fcntl fcntlmodule.c	# fcntl(2) and ioctl(2)
#pwd pwdmodule.c		# pwd(3) 
#grp grpmodule.c		# grp(3)
#errno errnomodule.c	# posix (UNIX) errno values
#select selectmodule.c	# select(2); not on ancient System V

# Memory-mapped files (also works on Win32).
#mmap mmapmodule.c

# Dynamic readlines
#xreadlines xreadlinesmodule.c

# for socket(2), without SSL support.
#_socket socketmodule.c

# Socket module compiled with SSL support; you must comment out the other
# socket line above, and possibly edit the SSL variable:
#SSL=/usr/local/ssl
#_socket socketmodule.c \
#	-DUSE_SSL -I$(SSL)/include -I$(SSL)/include/openssl \
#	-L$(SSL)/lib -lssl -lcrypto

# The crypt module is now disabled by default because it breaks builds
# on many systems (where -lcrypt is needed), e.g. Linux (I believe).
#
# First, look at Setup.config; configure may have set this for you.

#crypt cryptmodule.c # -lcrypt	# crypt(3); needs -lcrypt on some systems


# Some more UNIX dependent modules -- off by default, since these
# are not supported by all UNIX systems:

#nis nismodule.c -lnsl	# Sun yellow pages -- not everywhere
#termios termios.c	# Steen Lumholt's termios module
#resource resource.c	# Jeremy Hylton's rlimit interface


# Multimedia modules -- off by default.
# These don't work for 64-bit platforms!!!
# These represent audio samples or images as strings:

#audioop audioop.c	# Operations on audio samples
#imageop imageop.c	# Operations on images
#rgbimg rgbimgmodule.c	# Read SGI RGB image files (but coded portably)


# The md5 module implements the RSA Data Security, Inc. MD5
# Message-Digest Algorithm, described in RFC 1321.  The necessary files
# md5c.c and md5.h are included here.

#md5 md5module.c md5c.c


# The sha module implements the SHA checksum algorithm.
# (NIST's Secure Hash Algorithm.)
#sha shamodule.c


# The mpz module interfaces to the GNU Multiple Precision library.
# You need to ftp the GNU MP library.  
# The GMP variable must point to the GMP source directory.
# This was originally written and tested against GMP 1.2 and 1.3.2.
# It has been modified by Rob Hooft to work with 2.0.2 as well, but I
# haven't tested it recently.

# A compatible MP library unencombered by the GPL also exists.  It was
# posted to comp.sources.misc in volume 40 and is widely available from
# FTP archive sites. One URL for it is:
# ftp://gatekeeper.dec.com/.b/usenet/comp.sources.misc/volume40/fgmp/part01.Z

#GMP=/ufs/guido/src/gmp
#mpz mpzmodule.c -I$(GMP) $(GMP)/libgmp.a


# SGI IRIX specific modules -- off by default.

# These module work on any SGI machine:

# *** gl must be enabled higher up in this file ***
#fm fmmodule.c $(GLHACK) -lfm -lgl		# Font Manager
#sgi sgimodule.c			# sgi.nap() and a few more

# This module requires the header file
# /usr/people/4Dgifts/iristools/include/izoom.h:
#imgfile imgfile.c -limage -lgutil -lgl -lm	# Image Processing Utilities


# These modules require the Multimedia Development Option (I think):

#al almodule.c -laudio			# Audio Library
#cd cdmodule.c -lcdaudio -lds -lmediad	# CD Audio Library
#cl clmodule.c -lcl -lawareaudio	# Compression Library
#sv svmodule.c yuvconvert.c -lsvideo -lXext -lX11	# Starter Video


# The FORMS library, by Mark Overmars, implements user interface
# components such as dialogs and buttons using SGI's GL and FM
# libraries.  You must ftp the FORMS library separately from
# ftp://ftp.cs.ruu.nl/pub/SGI/FORMS.  It was tested with FORMS 2.2a.
# NOTE: if you want to be able to use FORMS and curses simultaneously
# (or both link them statically into the same binary), you must
# compile all of FORMS with the cc option "-Dclear=__GLclear".

# The FORMS variable must point to the FORMS subdirectory of the forms
# toplevel directory:

#FORMS=/ufs/guido/src/forms/FORMS
#fl flmodule.c -I$(FORMS) $(GLHACK) $(FORMS)/libforms.a -lfm -lgl


# SunOS specific modules -- off by default:

#sunaudiodev sunaudiodev.c


# A Linux specific module -- off by default; this may also work on 
# some *BSDs.

#linuxaudiodev linuxaudiodev.c


# George Neville-Neil's timing module:

#timing timingmodule.c


# The _tkinter module.
#
# The command for _tkinter is long and site specific.  Please
# uncomment and/or edit those parts as indicated.  If you don't have a
# specific extension (e.g. Tix or BLT), leave the corresponding line
# commented out.  (Leave the trailing backslashes in!  If you
# experience strange errors, you may want to join all uncommented
# lines and remove the backslashes -- the backslash interpretation is
# done by the shell's "read" command and it may not be implemented on
# every system.

# *** Always uncomment this (leave the leading underscore in!):
_tkinter _tkinter.c tkappinit.c -DWITH_APPINIT \
# *** Uncomment and edit to reflect where your Tcl/Tk libraries are:
	-L/usr/local/lib \
# *** Uncomment and edit to reflect where your Tcl/Tk headers are:
	-I/usr/local/include \
# *** Uncomment and edit to reflect where your X11 header files are:
	-I/usr/X11R6/include \
# *** Or uncomment this for Solaris:
#	-I/usr/openwin/include \
# *** Uncomment and edit for Tix extension only:
#	-DWITH_TIX -ltix8.1.8.2 \
# *** Uncomment and edit for BLT extension only:
#	-DWITH_BLT -I/usr/local/blt/blt8.0-unoff/include -lBLT8.0 \
# *** Uncomment and edit for PIL (TkImaging) extension only:
#     (See http://www.pythonware.com/products/pil/ for more info)
#	-DWITH_PIL -I../Extensions/Imaging/libImaging  tkImaging.c \
# *** Uncomment and edit for TOGL extension only:
#	-DWITH_TOGL togl.c \
# *** Uncomment and edit to reflect your Tcl/Tk versions:
	-ltk8.3 -ltcl8.3 \
# *** Uncomment and edit to reflect where your X11 libraries are:
	-L/usr/X11R6/lib \
# *** Or uncomment this for Solaris:
#	-L/usr/openwin/lib \
# *** Uncomment these for TOGL extension only:
#	-lGL -lGLU -lXext -lXmu \
# *** Uncomment for AIX:
#	-lld \
# *** Always uncomment this; X11 libraries to link with:
	-lX11

# Lance Ellinghaus's modules:

#rotor rotormodule.c		# enigma-inspired encryption
#syslog syslogmodule.c		# syslog daemon interface


# Curses support, requring the System V version of curses, often
# provided by the ncurses library.  e.g. on Linux, link with -lncurses
# instead of -lcurses; on SunOS 4.1.3, insert -I/usr/5include
# -L/usr/5lib before -lcurses).
#
# First, look at Setup.config; configure may have set this for you.

#_curses _cursesmodule.c -lcurses -ltermcap
# Wrapper for the panel library that's part of ncurses and SYSV curses.
#_curses_panel _curses_panel.c -lpanel -lncurses 


# Generic (SunOS / SVR4) dynamic loading module.
# This is not needed for dynamic loading of Python modules --
# it is a highly experimental and dangerous device for calling
# *arbitrary* C functions in *arbitrary* shared libraries:

#dl dlmodule.c


# Modules that provide persistent dictionary-like semantics.  You will
# probably want to arrange for at least one of them to be available on
# your machine, though none are defined by default because of library
# dependencies.  The Python module anydbm.py provides an
# implementation independent wrapper for these; dumbdbm.py provides
# similar functionality (but slower of course) implemented in Python.

# The standard Unix dbm module has been moved to Setup.config so that
# it will be compiled as a shared library by default.  Compiling it as
# a built-in module causes conflicts with the pybsddb3 module since it
# creates a static dependency on an out-of-date version of db.so.
#
# First, look at Setup.config; configure may have set this for you.

#dbm dbmmodule.c 	# dbm(3) may require -lndbm or similar

# Anthony Baxter's gdbm module.  GNU dbm(3) will require -lgdbm:
#
# First, look at Setup.config; configure may have set this for you.

#gdbm gdbmmodule.c -I/usr/local/include -L/usr/local/lib -lgdbm


# Berkeley DB interface.
#
# This requires the Berkeley DB code, see
# ftp://ftp.cs.berkeley.edu/pub/4bsd/db.1.85.tar.gz
#
# Edit the variables DB and DBPORT to point to the db top directory
# and the subdirectory of PORT where you built it.
#
# (See http://electricrain.com/greg/python/bsddb3/ for an interface to
# BSD DB 3.x.)

# Note: If a db.h file is found by configure, bsddb will be enabled
# automatically via Setup.config.in.  It only needs to be enabled here
# if it is not automatically enabled there; check the generated
# Setup.config before enabling it here.

#DB=/depot/sundry/src/berkeley-db/db.1.85
#DBPORT=$(DB)/PORT/irix.5.3
#bsddb bsddbmodule.c -I$(DBPORT)/include -I$(DBPORT) $(DBPORT)/libdb.a



# Helper module for various ascii-encoders
#binascii binascii.c

# Fred Drake's interface to the Python parser
#parser parsermodule.c

# Digital Creations' cStringIO and cPickle
#cStringIO cStringIO.c
#cPickle cPickle.c


# Lee Busby's SIGFPE modules.
# The library to link fpectl with is platform specific.
# Choose *one* of the options below for fpectl:

# For SGI IRIX (tested on 5.3):
#fpectl fpectlmodule.c -lfpe

# For Solaris with SunPro compiler (tested on Solaris 2.5 with SunPro C 4.2):
# (Without the compiler you don't have -lsunmath.)
#fpectl fpectlmodule.c -R/opt/SUNWspro/lib -lsunmath -lm

# For other systems: see instructions in fpectlmodule.c.
#fpectl fpectlmodule.c ...

# Test module for fpectl.  No extra libraries needed.
#fpetest fpetestmodule.c

# Andrew Kuchling's zlib module.
# This require zlib 1.1.3 (or later).
# See http://www.cdrom.com/pub/infozip/zlib/
#zlib zlibmodule.c -I$(prefix)/include -L$(exec_prefix)/lib -lz

# Interface to the Expat XML parser
#
# Expat is written by James Clark and must be downloaded separately
# (see below).  The pyexpat module was written by Paul Prescod after a
# prototype by Jack Jansen.
#
# The Expat dist includes Windows .lib and .dll files.  Home page is at
# http://www.jclark.com/xml/expat.html, the current production release is
# always ftp://ftp.jclark.com/pub/xml/expat.zip.
#
# EXPAT_DIR, below, should point to the expat/ directory created by
# unpacking the Expat source distribution.
#
# Note: the expat build process doesn't yet build a libexpat.a; you can
# do this manually while we try convince the author to add it.  To do so,
# cd to EXPAT_DIR, run "make" if you have not done so, then run:
#
#    ar cr libexpat.a xmltok/*.o xmlparse/*.o
#
#EXPAT_DIR=/usr/local/src/expat
#pyexpat pyexpat.c -I$(EXPAT_DIR)/xmlparse -L$(EXPAT_DIR) -lexpat


# Example -- included for reference only:
# xx xxmodule.c

# Another example -- the 'xxsubtype' module shows C-level subtyping in action
xxsubtype xxsubtype.c
-------------------------------------------------------------------------


From gerhard at bigfoot.de  Mon Jun  3 07:09:09 2002
From: gerhard at bigfoot.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Mon, 3 Jun 2002 13:09:09 +0200
Subject: word wrap on different OS
In-Reply-To: <3CFB4994.DD35AEA2@textraeume.de>
References: <3CFB4994.DD35AEA2@textraeume.de>
Message-ID: <20020603110909.GB17944@gargamel.hqd-internal>

* Sabine Richter <sabine at textraeume.de> [2002-06-03 12:48 +0200]:
> Hello,
> 
> I am not sure if I think clear on this subject:
> 
> A simple text file has been created for example on MS Dos or MacOS. So
> the sign for a word wrap will be "\r\n" on MS Dos or "\r" on MacOS.
> Then I save the file on a windows file system. The sign for the word
> wrap will be converted to "\n". 

Normally, you want to process the file line by line. Is this the case for you,
too? Then, you can use the following pattern:

f = open("foo")
for line in f.xreadlines():
    # Get rid of any trailing whitespace, including CR and LF characters
    line = line.rstrip()
    # do something with the line now
f.close()

There is no need to read the whole file and split it at line endings, the
readlines() and xreadlines() methods of file objects are invented just for
that.
 
> And next question:
> If I write a script, which uses MyODBC and the odbc-module from win32
> extensions to connect to a MySQL-database, it can only be used on
> windows. Isn't it?

Yes. I already recommended the MySQLdb module to you, which will work fine on
Windows and Unix. If you can spare a few euros, you can get a commercial
license for mxODBC. This is an ODBC solution that will also work on Unixen.  If
you target MySQL only, MySQLdb is the way to go. Everything else is
superfluous, and doesn't have any technical advantages.

Cheers,

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From pinard at iro.umontreal.ca  Wed Jun 19 16:39:38 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 19 Jun 2002 16:39:38 -0400
Subject: procmail replacement in Python
In-Reply-To: <20020619022110.GA6996@lilith.my-fqdn.de>
References: <20020618235901.GA2197@lilith.my-fqdn.de>
	<oqit4grq03.fsf@carouge.sram.qc.ca>
	<20020619022110.GA6996@lilith.my-fqdn.de>
Message-ID: <oq660froj9.fsf@carouge.sram.qc.ca>

[Gerhard H?ring]

> I'm currently just using spamassassin (the best spam detector around,
> very configurable, but written in Perl ;-).

I wanted to peek at Spamassassin, which seems to have a good reputation.
And now, you are trying to discourage me? :-) It's really a strange
phenomenon: more I work with Python, more Perl gets unreadable :-)

> I'm also investingating Pyzor, as I found the Razor author to be too
> clueless (also the razor servers aren't open-source).

Thanks both for pointers and hints.

I'm glad that this discussion popped up `pycmail', yet the warning of its
documentation are a bit chilling...
-------------- next part --------------

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard

From ken at hotmail.com  Sat Jun  1 04:18:50 2002
From: ken at hotmail.com (Ken)
Date: Sat, 1 Jun 2002 18:18:50 +1000
Subject: How to open a HTML file when the python cgi program is executing?
References: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de> <ad95l1.i0.1@kserver.org>
Message-ID: <ad9vsm$vhjnu$1@ID-49758.news.dfncis.de>

"Sheila King" <usenet at thinkspot.net> wrote in message
news:ad95l1.i0.1 at kserver.org...
> On Sat, 1 Jun 2002 16:43:01 +1000, "Ken" <ken at hotmail.com> wrote in
> comp.lang.python in article <ad9q91$vlrsb$1 at ID-49758.news.dfncis.de>:
>
> > How do I open a HTML file when the cgi program is executing?
> >
> > The only method I can think of, which doesn't work is this:
> > <meta HTTP-EQUIV="REFRESH" CONTENT="0; URL=main.html">
>
> What do you mean by "open an HTML file when the cgi program is executing"?
>
> Are you talking about displaying the resulting HTML output.
>
> HTML output from a cgi script is simply printed to standard output.
>
> Here is a very simply cgi script (basically "Hello World"):
>
> =============(begin hello script)===============
> #!/usr/bin/env python
>
> print "Content-type: text/html\n"
> print "<HTML><HEAD><TITLE>Hello, World!</TITLE></HEAD>"
> print "<BODY><H1>HELLO, WORLD!!!</H1></BODY></HTML>"
>
> =============(end hello script)===============
>
> Change the interpreter path
>
> #!/usr/bin/env python
>
> as appropriate for your server.
>
> Save the above as something like hello.py or hello.cgi
> upload it to the cgi-bin on your server. Make sure file permissions are
set
> for executable (this will probably be 777 or 755 depending on your server
> configuation). Make sure you upload it in ASCII/TEXT mode.
>
> Then point your browser to
>
> http://www.example.com/cgi-bin/hello.py
>
> where you replace example.com with your actual domain name.
>
> This should display an HTML page with the H1 header saying "Hello,
World!!"
>
> --
> Sheila King
> http://www.thinkspot.net/sheila/
> http://www.k12groups.org/
>

No, I mean at some point in the script, I want to get out of it and use
ordinary HTML instead (main.html in this example). I don't mean the "print"
command.

Can anyone help?

Thanks




From cpbotha at i_triple_e.org  Sun Jun  2 15:27:45 2002
From: cpbotha at i_triple_e.org (Charl P. Botha)
Date: Sun, 2 Jun 2002 19:27:45 +0000 (UTC)
Subject: List
References: <3cfa6ed8_1@news.iprimus.com.au>
Message-ID: <addrjh$imq$1@news.tudelft.nl>

In article <3cfa6ed8_1 at news.iprimus.com.au>, Gold Fish wrote:
> Assume we got 2 list 
> list 1 = ['ABC','CDG','DAV','FE','FECA','FEAFA','FEA']
> list 2 = []
> how can we copy element from list 1 to list 2 say i just want to copy the 3 
> element 'CDG','DAV','FE' to list2 only.

list1 = ['ABC','CDG','DAV','FE','FECA','FEAFA','FEA']
list2 = list1[1:4]

Is that what you meant?

-- 
charl p. botha http://cpbotha.net/ http://visualisation.tudelft.nl/


From bokr at oz.net  Fri Jun  7 15:42:16 2002
From: bokr at oz.net (Bengt Richter)
Date: 7 Jun 2002 19:42:16 GMT
Subject: Medium to Large Scale Python Deployments
References: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com> <mailman.1023409393.3361.python-list@python.org> <3D003EC7.C72D27BE@engcorp.com> <3D006320.5040805@web.de> <3D00A209.EC24F671@engcorp.com>
Message-ID: <adr2ao$bei$0@216.39.172.122>

On Fri, 07 Jun 2002 08:07:37 -0400, Peter Hansen <peter at engcorp.com> wrote:

>J?rgen Hermann wrote:
>> 
>> Peter Hansen wrote:
>> > How big is that when run through pycount?
>> >
>> > (I.e. comments and blank lines don't count!)
>> 
>> Summary on "twisted"
>> ====================
>> files                  256
>> lines                51654
>> bytes              1728211
>> comments              5993
>> empty lines          10029
>> non-commentary lines 29910
>
>Thanks!  Similar results to what I've seen over about 50000 
>lines in 235 files.  Roughly 50-60% code content, and 
>almost twenty percent blank lines.
>
>I suspect such numbers (perhaps excluding lines of comments)
>are more consistent for Python code, whoever wrote it,
>than for many other languages.
>
Is the "comments" number a count of _pure_ comment lines,
# (like this one) ?

I think it would be interesting to analyze comments further,
to see what the relative comment byte count is overall,
and for lines with both code and comments,
"""like this line:""" # (this has code and comment).

Also maybe separate doc-string lines into a separate category,
since they have a dual nature.

Regards,
Bengt Richter


From jbublitzNO at SPAMnwinternet.com  Wed Jun 19 14:42:50 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Wed, 19 Jun 2002 18:42:50 GMT
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de> <aeftmj$m29$1@bob.news.rcn.net> <3d0c6d67@nntp.server.uni-frankfurt.de> <3D0CE698.1050806@SPAMnwinternet.com> <aeiu3a$4ff$1@bob.news.rcn.net> <3D0D3223.5070500@SPAMnwinternet.com> <aekrd1$4ao$1@bob.news.rcn.net> <3D0E2606.5050001@SPAMnwinternet.com> <aen6ch$d5u$1@bob.news.rcn.net> <3D0F807A.40203@SPAMnwinternet.com> <aeoco5$adc$1@bob.news.rcn.net>
Message-ID: <3D10D0A6.6030401@SPAMnwinternet.com>

David Abrahams wrote:
> "Jim" <jbublitzNO at SPAMnwinternet.com> wrote in message
> news:3D0F807A.40203 at SPAMnwinternet.com...
> 
>>David Abrahams wrote:
>>
>>>"Jim" <jbublitzNO at SPAMnwinternet.com> wrote in message


>>>That's a given with Boost.Python; with v2 conversions are automatically
>>>registered in a global repository so they can be re-used automatically
>>>across extension modules.

>>For C++ code with a lot of templates Boost.Python has an
>>advantage, particularly if you're early on the learning curve.

> Well, yes, but I wasn't talking about that. I was talking about re-usability
> of conversions. My point was that if you wrap a class X in one module, you
> can can wrap [member] functions that operate on X in another module.

%Import <module name> links you to all of the code from another
module (and it's antecedents).

>>>OK. The C++ code you're wrapping still has to get by with #ifdef, I
>>>suppose...

>>Depends on the code I suppose. Qt and KDE just issue all
>>new files - no versioning in the C++ code.

> Ah. I guess if they were wrapped with Boost.Python, the wrapping code would
> be treated the same, then?

If you want to support that many filesets - I'm on the 8th PyKDE
version (KDE2.1.1 -> KDE3.0.1) in 14 months (all still supported),
amd I'll probably do one or two more releases in the next 6 months.
Any of the versions can build against one of 3 or more different Qt
versions (Qt2.3.0 -> Qt2.3.2, or Qt3.0.0 -> Qt3.0.4). It adds up.

>>I finally went back to your website - I took your 'like an
>>IDL' comments a little too literally. IMHO, the fact that
>>Boost.Python *is* C++ is a big advantage for people who
>>want/need that. Personally, I don't want to write C++,
>>which is how I got into this in the first place.

> <smile> A lot of people seem to feel that way.

I believe they're called "Python users" :) (j/k)

>>Thin wrappers are easily implementable with sip too. In
>>contrast to the very large packages like PyQt and PyKDE,
>>they tend to be very lightweight and quick to build.

> Sorry, I think you misunderstood what I meant by "thin wrapper". A classic
> trivial example might be used to wrap a function with default args:

>     // function to wrap
>     void f(int a, char* b = expression-1, Foo const& c = expression-2);

> Thin wrappers:

>     void f1(int a, char* b) { f(a,b); }
>     void f2(int a) { f(a); }

> wrapping code:
> 
>     module("my_module")
>         .def("f", f)
>         .def("f", f1) // overloads handle
>         .def("f", f2) // default args
>     ;

> Obviously you can do arbitrary pre- and post- processing inside a thin
> wrapper function.

I think we mean the same thing. I've done thin wrappers for a
spreadsheet plugin, and basically re-architected the entire API
for Python, including things like changing the grid addressing
from numeric row and col to more typical alphanumeric
(eg (2, 6) -> "B6"). Full bindings would compile to several MB;
binding just the wrapper took about 100K plus about the same for
the wrapper. Most of the methods in the wrapper were 5 lines or less.

sip code:

      void f (int, char * = expression1, Foo const& = expression2) [1]
      void f1 (int, char *);
      void f2 (int);

[1] Might depend on what the expressions are - I've done constants
and function/method calls as defaults, but not arbitrary expressions,
so I'm not sure. It might take more than a line. OTOH, a real
wrapper could throw out the defaults for 'f' (in Python) and replace
them with f1, f2, etc.



>>>>Yes, but I don't write it - I just edit it mostly to correct
>>>>mistakes my description file generator creates and to handwrite
>>>>code where the generator has flagged that. Crude but effective (TM)
>>>>Even without automation, the process is easiest if you start by
>>>>saving the h file with a .sip extension and then (mostly) delete
>>>>the things that sip doesn't want/need, so it's still mostly
>>>>editing (sometimes A LOT of editing).

>>>That seems to be the key difference. It's not clear wether it's really
>>>an advantage, but I suppose that time will tell.

 >>I doubt that it's ever clearly an advantage.

> Really? It seems like it should lower the barrier to entry 
 > and get people> started with an iterative development
 > process, at least.

It certainly depends on the code you're generating bindings for,
your knowledge of C++, maintability requirements, lots of other
factors.

> Probably the docs are the main obstacle there. It's amazing how many
> different wrapping tools there seems to be a market for; people will try
> anything once <wink>!

There's a lot of C++ compilers too, most of which are
incompatible to some extent (certainly at the binary
level). Choice is good.

> Well, code comparisons are important in showing the system's user-interface.

Project wide code comparisons might be useful, but that's more
work than I'd want to do.

>>Yes - it looks like that should be possible. Virtual code
>>might require a little more work in generating Boost.Python
>>code automatically, but the tradeoff is probably less things
>>flagged as requiring handwritten code.

 > Not sure what you mean here about the tradeoff. Example?

Basically a parser/code generator to create input files for
sip is easier than the equivalent for Boost.Python, since for
sip input files, I don't need to know anything about a class
member except whether it's an enum, method or variable. Stuff
like 'virtual', 'static', 'const', pure virtuals or default
values is just syntactic noise - it just gets passed on to sip,
which already knows how to deal with that stuff.

Also, I can generate versioned code by comparing the code
generated for a new h file with the previous version's sip
(input) file - the same parser does both since the syntax
is nearly identical. The version state transition logic is
tricky and that's where I have problems automatically
generating input files at the moment. Just generating a
sip input file is pretty simple to automate and seems to
work well.

OTOH, Boost.Python seems to care more about whether methods
are virtual or handling default values, and the 'input' files
for Boost.Python are cpp files and the syntax is method calls
instead of declarations - all that makes a parser harder to
write IMO. There might be a different way to approach it that's
easier - some intermediate file for example, or a variation
on diff.

The tradeoff is that it would be easier to automatically
generate code for templates with Boost.Python than with
a sip input file generator and in fact I just flag that
stuff and write it manually the first time - it gets
recycled on subsequent versions. The parser/input file
generator/version generator in general seems to be easier
with sip. It's probably not a concern for most people.


Jim



From radix at twistedmatrix.com  Thu Jun  6 20:19:28 2002
From: radix at twistedmatrix.com (Christopher Armstrong)
Date: 06 Jun 2002 20:19:28 -0400
Subject: Medium to Large Scale Python Deployments
In-Reply-To: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com>
References: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com>
Message-ID: <87lm9sx7n3.fsf@twistedmatrix.com>


> I'm also curious about large amounts of Python, period. Who knows of ~50K
> Python lines-of-code (LOC) systems? 100K LOC? More?

chris at radii:~/Projects/Twisted$ find -name '*.py' | xargs wc -l 
...
  56051 total

http://twistedmatrix.com

Twisted is a framework for writing asynchronous network apps. The reason it's
so big, though, is that we pack a lifetime supply of batteries in the
box. ;-)

-- 
                                Chris Armstrong
                         << radix at twistedmatrix.com >>
                http://twistedmatrix.com/users/carmstro.twistd/





From tim.one at comcast.net  Sun Jun 23 17:38:36 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 23 Jun 2002 17:38:36 -0400
Subject: Python hits the spot
In-Reply-To: <7xfzzdhf07.fsf@ruckus.brouhaha.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEIGPPAA.tim.one@comcast.net>

[P?r Kurlberg, sees this freeze some Linux box]
> for i in range(50*1000*1000):
>    pass

[Matt Gerrans]
> You call that a bug?   Did you consider trying xrange()?

[Paul Rubin]
> Maybe not a bug, but a pretty serious misfeature IMO.

If it is, it's a serious misfeature of Linux's overcommitment policies.
Python does a malloc() sufficient for 50*1000*1000 list elements at the
start (about 200Mb), malloc doesn't return NULL, and Python believes it can
use the memory.  What would you like it to do instead?  The platform libc
and platform OS don't give any clue that they're making promises they can't
keep.  If at any point in the process this-- or any subsequent --malloc()
returned NULL, Python would raise a well-behaved MemoryError and clean up
the partial junk.

But when the OS+libc refuse to give any hint of trouble, and themselves go
nuts later, I don't accept that as being Python's problem.  As noted, more
recent versions of Linux kill the process (well, kill *some* process(es),
probably including at least this one) instead of going completely insane.
That sounds like a misfeature to me too, but it's outside of Python's
control.

python-isn't-an-operating-system-ly y'rs  - tim






From nas at python.ca  Sat Jun 29 19:52:21 2002
From: nas at python.ca (Neil Schemenauer)
Date: Sat, 29 Jun 2002 16:52:21 -0700
Subject: newbie Q: delete list few lines in file
In-Reply-To: <7fe97cc4.0206291519.1c95ce31@posting.google.com>; from xah@xahlee.org on Sat, Jun 29, 2002 at 04:19:16PM -0700
References: <7fe97cc4.0206291519.1c95ce31@posting.google.com>
Message-ID: <20020629165221.A12739@glacier.arctrix.com>

Xah Lee wrote:
> i have a folder full of individual email text file, whose ending may
> be the following:
> 
> _______________________________________________
> some mailing list
> some at list.com
> http://some.com/ttt
> 
> I want to delete them, but only if they are the last few lines in the
> file.

I would use something like the following untested code:

    import os

    marker = ('some mailing list\n'
              'some at list.com\n'
              'http://some.com/ttt\n')
    for filename in os.listdir(sys.argv[1]):
        f = open(filename)
        f.seek(-1000, 2)
        tail = f.read()
        if tail.find(marker) >= 0:
            os.unlink(filename)

The seek() call is only necessary if you're dealing with really large
files.  If you know the files are all pretty small and the script is a
"one off job" then you could leave it out and read the whole file into
memory.  

> * how to write to file inplace? (or, do i write to a new file, then
> delete original, and rename the new file?)

You need to open the file using the 'r+' or 'a+' flag.

> * as far as a newbie, i use "xreadlines" module. I wasn't able to get
> the total number of lines in a file.
> li=xreadlines.xreadlines(f)
> num=len(li)
> fails.

xreadlines returns something that conforms to the iterator protocol.
The protocol does not require the __len__ method (which len() uses).
You can probably just use readlines() instead.  If the file is really
big you need to explicitly exhaust the iterator and keep a count, eg:

    n = 0
    for line in f.xreadlines():
        n += 1

> is there a site or mailing list that collect very small trivial python
> programs for learning purposes?

Search for "Python Cookbook".

  Neil




From emile at fenx.com  Mon Jun  3 13:09:48 2002
From: emile at fenx.com (Emile van Sebille)
Date: Mon, 03 Jun 2002 17:09:48 GMT
Subject: UnboundLocalError: local variable part 2
References: <40a8aa92.0206030903.466fc3ff@posting.google.com>
Message-ID: <wpNK8.50416$0A2.40695@rwcrnsc54>

Daniel
> That last example was a bit too contrived... instead of blah='abc',
> blah is actually a list and i'm appending and popping elements so
> func1 actually truly does affect the global blah.  So a better example
> of the logic is:

There must be something else going on.  I get exactly what you expected:

def func1():
   blah.pop()

def func2():
   print blah
   blah.pop()

def main():
   func1()
   func2()
   print blah

blah=[1,2,3]
main()


[1, 2]
[1]

--

Emile van Sebille
emile at fenx.com

---------



From jadestar at idiom.com  Wed Jun 19 20:26:07 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 20 Jun 2002 00:26:07 GMT
Subject: directory
References: <mailman.1023215768.17394.python-list@python.org>
Message-ID: <aer7ev$2em8$1@news.idiom.com>

Sean 'Shaleh' Perry <shalehperry at attbi.com> wrote:

> On 04-Jun-2002 Gold Fish wrote:
>> How to read the subdirectories in the directory
>> Support i have the directory /home/ which have some subdirectories
>> /home/sub1
>> /home/sub2
>> /home/sub3
>> I am using os.listdir('/home/') to read in the list so it will be 
>> ['sub1','sub2','sub3'] now i want to read inside the sub directory sub2
>> once again i using os.listdir(os.listdir('/home/')[1]) but it came out with 
>> an errors: not a file or directory. Can anyone tell me what did i wrong. 
>> Anh how to fix it.

> notice the values given by listdir('/home/') -> 'sub1', 'sub2'.  listdir()
> expects full path to a file.  You need to do
> listdir(os.path.join('/home', 'sub1')).  More likely a better option would be
> to use os.path.walk().


 os.path.walk() uses an awkward interface.  I wrote and posted this a couple
 of months ago:

#!/usr/bin/env python2.2
## Change the shebang line to match your needs!
import os
def dirwalk(startdir=None):
        if not startdir:
                startdir="."
        if not os.path.isdir(startdir):
                raise ValueError ## Is this the right exception
        stack = [startdir]
        while stack:
                cwd = stack.pop(0)
                try:
                        current = os.listdir(cwd)
                except (OSError):
                        continue
                for each in current:
                        each = os.path.join(cwd,each)
                        if os.path.islink(each):
                                pass
                        elif os.path.isdir(each):
                                stack.append(each)
                        yield(each)

if __name__ == "__main__":
        import sys
        for i in sys.argv[1:]:
                for j in dirwalk(i):
                        print j

 Notice that it implements a generator which is generally considered
 more pythonic than os.path.walk()'s "visitor" function.

 However, some people may prefer os.path.walk() or prefer a recursive
 version of dirwalk; some might prefer to make this a class, to instantiate
 directory walkers and obviate the need for the __future__ import.




From shalehperry at attbi.com  Wed Jun  5 21:01:06 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 05 Jun 2002 18:01:06 -0700 (PDT)
Subject: Detecting OS and Window Manager
In-Reply-To: <uft6liipd6m60e@corp.supernews.com>
Message-ID: <XFMail.20020605180106.shalehperry@attbi.com>

On 05-Jun-2002 Andrew Wilkinson wrote:
> Thanks for all the replies guys.
> 
> Basically I want to have a program that hides itself in the taskbar of
> whatever the user is running. Unfortunately there isn't (or at least I don't
> know of) a portable way of doing this. I guess I'll just have to mess around
> with looking for the environment variables and then code it up for
> KDE/Gnome/whatever.
> 

environment variables won't help you, they are not guaranteed.  Also, even if
the user is running parts of GNOME/KDE they may not have a panel running.

Do you care if they see your window?  If not perhaps you could just ask for it
to be iconified.  If you do care your best bet is to read about about "netwm"
a quick google search should help.  This is the mechanism for inter client
talking in KDE, GNOME and other desktops.  Using this you can make your app as
a taskbar/panel style app and if they can accept it they will otherwise it gets
treated like any other window.

Shaleh
Blackbox maintainer




From idot at vt.edu  Wed Jun  5 23:56:16 2002
From: idot at vt.edu (Carl Banks)
Date: 5 Jun 2002 20:56:16 -0700
Subject: another near-sighted, time-wasting syntax enhancement idea
Message-ID: <e395b886.0206051956.5f447e9b@posting.google.com>

class BigLongClassThatRunsUntilEOF::

def __init__ (self):
    pass

def whatever (self):
    pass

...



Two colons after the class (or whatever) definition means the block
continues until EOF and is not indented.  Good for big, long classes
that take up a whole module.  Been suggested before, I'm sure.


-- 
CARL BANKS


From ktf at nospamathotmail.com  Sat Jun 15 01:33:00 2002
From: ktf at nospamathotmail.com (KTF)
Date: Sat, 15 Jun 2002 17:33:00 +1200
Subject: VBA to Python
Message-ID: <GjAO8.4198$iR2.398848@news.xtra.co.nz>

In moving from using Access front end forms and code connected to an Access
back end I am looking at changing to Java front end to SQL back end with a
business rules engine using Python.

These are busy forms and there is a lot of data validation (checking data
input for correct format, string length etc), if this do that's, and
directing of drop-down list content based on the entry or selection of the
previous field (if this make, populate the next list with the models of the
chosen make etc).

Is Python a good choice to manage and run the rules and is there a way of
transcribing the VBA code across to Python other than literally doing it
line by line?

Thanks,
Keith




From martin at v.loewis.de  Sun Jun  9 15:57:23 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 09 Jun 2002 21:57:23 +0200
Subject: problem with isinstance and relative/absolute import
References: <87adq8jsq1.fsf@tux.ntw23.fr> <Xns92265AA1E2FABgmcmhypernetcom@199.171.54.216> <87k7p8z8vv.fsf@tux.ntw23.fr>
Message-ID: <m3bsakb4yk.fsf@mira.informatik.hu-berlin.de>

Sylvain Thenault <thenault at nerim.net> writes:

> actually I did understood what was the problem, I was only asking if this
> behaviour should be considered as normal...

Yes, certainly.

> Since it's current in a package to mix relative (to the python file
> directory) and absolute (package) import, isn't there a risk to
> obtain a strange behaviour without being in the package directory ?

If the current directory is not inside a package, then relative
imports won't resolve through sys.path, but instead will resolve
through the package which executes the import. Hence, the imported
module will correctly show up as a submodule.

Notice that you can produce the same problem in Java, which solves
this problem with the package declaration on top of each file (so the
problem exists only if you omit the package declaration).

Guido is tempted to solve the problem by banning relative imports.

Regards,
Martin


From peter at engcorp.com  Mon Jun 17 20:36:47 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 20:36:47 -0400
Subject: Problems with importing; must do it twice?
References: <fgphgus0pv2c7hp8l27uaenjlerr97tdta@4ax.com>
Message-ID: <3D0E809F.AD9746A2@engcorp.com>

Stew LG wrote:
> 
> 
> My web host just switched hardware, and in
> the process I've been moved to Solaris and
> Python 2.1.1.
> 
> My old scripts are breaking now, at the
> very first line:
> 
> import cgi;
> 
> I think their Python install is broken,
> but I wanted to be very sure before
> I complained about it, since I've been
> acting somewhat hysterically towards the
> company's support lately - the move broke
> a few things for me.

Presumably the hosting company is escape.com?  I just
went through trouble-shooting the same problem for a
friend (on my vacation in Manhattan! :-) and although
this is out of my area of expertise, I concluded that
they had either compiled Python with the wrong switch
(asking for ssl support when they didn't have it
installed) or that they were caught by an apparent
glitch in the build process wherein it finds the header
files for OpenSSL and builds support, even though the
library itself (libssl.so, as you found) is not 
present.

Two options I guess: get them to rebuild Python
without SSL support, or get them to install the 
properly SSL libraries.

(I'd appreciate hearing when/if you get this fixed,
in case you manage to do it before my friend succeeds.)

-Peter


From achim.domma at syynx.de  Tue Jun 18 02:50:21 2002
From: achim.domma at syynx.de (Achim Domma)
Date: Tue, 18 Jun 2002 08:50:21 +0200
Subject: ImageMagick for Python incl. Source
References: <aelmhc$35m$03$1@news.t-online.com> <aelo0t$sog$1@peabody.colorado.edu>
Message-ID: <aemkmd$6oq$05$1@news.t-online.com>

"Fernando P?rez" <fperez528 at yahoo.com> wrote in message
news:aelo0t$sog$1 at peabody.colorado.edu...

> Sorry to display my ignorance here, but what on earth is a "Jamfile"? You
> should either specify that to build this on other platforms you need
'bjam'
> (non-standard in the Unix world) giving a link to it, or (more likely to
be
> useful) provide a normal Makefile for the build. I tried google(bjam) and
got
> all sorts of irrelevant links. Under Linux, 99% of the people will try to
> build this using gcc/make, so a simple makefile will probably do.

ImageMagick for Python is based on boost.python, which uses the boost build
system so ImageMagick for Python uses it also. I realy like it, because it
works on windows and unix and was much easier to learn for me than make and
autoconf. But your are right, that the webpage was not clear enough. I added
a link to the required tools.

> An unrelated question: do you know what advantages ImageMagick has over
PIL? I
> ask out of honest curiosity. I've long been an ImageMagick fan, but since
> Python has sort of 'builtin' imaging support via PIL, I wonder what the

Because I have used both mainly for resizing and converting the image
format, my experiences are somewhat restricted. ImageMagick seems to support
much more image formats. I had some problems with PIL with some png types.
In general ImageMagick looks more complete to me.

> differences are. Anyway good luck with this, I love ImageMagick so I hope
> your project can be built under Unix soon.

Thanks, I'll do my best but I usually work on windows only, so I can't test
other plattforms.

Achim




From DLNXPEGFQVEB at spammotel.com  Tue Jun 11 04:19:53 2002
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Tue, 11 Jun 2002 11:19:53 +0300
Subject: Newby: What's the problem with this string?
References: <Xns9229E662C1E0AAusYourStandingInIt@130.133.1.4>
Message-ID: <rkcbgu8j0pi93be4m2oej7q18gpigc36sg@4ax.com>

On 10 Jun 2002 13:04:18 GMT, rumours say that fishfinger
<deathtospam43423 at altavista.com> might have written:

>For some reason I can't read this string from the command line ...

[snip of multi-line string]

>Why is that? And how can I make it work properly? The method I'm using is 
>.......
>
>ttt = raw_input("a string:")

raw_input stops reading as soon as you press the Return / Enter key.
IIUC, you need to ask for more lines until some predefined condition is
met.  The following returns a multiline string (keeps asking for more
until you input a single period):

# tested python>=2.2.1
def raw_multi_input(prompt):
	prompted_input= lambda: raw_input(prompt)
	input_lines= []
	for input_line in iter(prompted_input, "."):
		input_lines.append(input_line)
	return '\n'.join(input_lines)
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')


From kp at kyborg.dk  Mon Jun 17 08:24:01 2002
From: kp at kyborg.dk (Kim Petersen)
Date: Mon, 17 Jun 2002 14:24:01 +0200
Subject: Tkinter and threads
References: <3D0DCC8F.9A6D3B81@ipm.fhg.de>
Message-ID: <3D0DD4E1.40803@kyborg.dk>

Markus von Ehr wrote:

> Hi,
> 
> I'd like to have a progress bar increasing while a calculation process
> calculates some results.
> Unfortunately there is no mainloop processing while the calculation is
> running and so the progress bar begins after the calculation.
> Another thread neither can access the Tkinter resources, how can
> I manage this problem?


sprinkle the loop with tk.update() calls, would be my best guess


> 
> I appreciate any help,
> 
> Markus
> 
> 
> 




From talon34 at hotmail.com  Tue Jun 25 22:34:54 2002
From: talon34 at hotmail.com (Talon)
Date: Wed, 26 Jun 2002 02:34:54 GMT
Subject: Installation problems DCOracle2 & W2000
Message-ID: <talon34-C09E81.19332125062002@netnews.attbi.com>

Hi all,

I am having a heck of a time getting DCOracle2 to funcion properly.  I 
am running on a W2K machine, python2.1.3 and Zope 2.5.1, both of which 
run just fine.  I installed DCOracle2 and the ZOracleDA under Zope and 
it works fine.  But standalone python scripts don't see the dco2.py 
module.  So I tried to install DCOracle2 in the python21/lib folder.  I 
got that latest version from Zope.org and it said it installed properly.  
But it doesn't run.  If you try to import it into a script, you get 
complaints that it can't find this or that module, all of which are part 
of the Zope path.

I tried the odbc package, and it works for SELECT's but has a problem 
with INSERT's.  I downloaded the mxODBC package, but haven't tried it 
yet and I've read about  the cxOracle module.  Does anyone have any 
advice?  The docs for Windows install of DCOracle2 are really poor and 
inconsistent.  I would like to stay with DCOracle2 if possible, since 
that is what is running on Zope, but I will consider anything that works.

TIA,

Mark


From pedrovl at yahoo.com  Thu Jun  6 04:43:37 2002
From: pedrovl at yahoo.com (Pedro Vale Lima)
Date: 6 Jun 2002 01:43:37 -0700
Subject: Medium to Large Scale Python Deployments
References: <KHkL8.21383$eD2.2399923271@newssvr10.news.prodigy.com>
Message-ID: <82afc355.0206060043.d26564e@posting.google.com>

"Domenic R. Merenda" <dom at edgereport.put_a_c_o_m_here> wrote in message 
> 
>    I am running a home-grown Enterprise Resource Planning system, written
> entirely in Python

Hello,

Could you tell us more information about your system (Database Server,
User Interface, Communication protocol, etc)? I am running an ABAP
based ERP but I'm pretty sure your is better :-)

regards,

Pedro Lima


From tchur at optushome.com.au  Sat Jun 22 17:22:18 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Sun, 23 Jun 2002 07:22:18 +1000
Subject: Web templating/db tool with best designer/coder separation?
References: <af2gh6$2c2$0@216.39.172.122>
Message-ID: <3D14EA8A.E1027830@optushome.com.au>

Bengt Richter wrote:
> 
> E.g., if you were hoping for an improvement over current practice,
> which do you consider to be the tool(s) to beat, and what would
> the killer improvement(s) be? TIA.

Albatross by Object Craft (see http://www.object-craft.com.au ) does
an excellent job of separating the presentation layer from the
underlying
logic - my understanding is that such separation was one of its main
design goals. However, it does not take a pedantic position on this - it
still allows Python code to be embedded in HTML templates - such
practice
is discouraged, but Albatross provides mechanisms to allow you to do it
if you insist. But more usefully, Albatross provides tags which allow
conditional processing (if-elif-else) and most usefully, iteration, to
be
placed in the HTML template. So your Python business logic code can
assemble
a Python sequence of, say, database query results, but it is the HTML
template
which iterates through that list to create the final HTML which is sent
to the
user's browser. Such a feature is essential, because otherwise the
business logic
Python code ends up needing to creating HTML output itself, which is
then merely
substituted into the HTMl template. Thus, the Albatross model is really
more like:
"all HTML-related stuff, including conditional processing and iteration,
is looked
after in the HTML templates, and all business logic stuff, divorced from
the manner
in which it will be presented, is handled by Python modules." Of course,
this is
just the well-known model-view-controller approach, but Albatross
implements this
very nicely, IMO. 
See
http://www.object-craft.com.au/projects/albatross/albatross/fig-presimpexec.html
for more information about this (and to get an idea of the depth of the
Albatross 
documentation).

I suspect that the Object Craft guys would welcome help with further
development of
Albatross. What the world really needs are more Pythonic Web
applications, rather 
than more Pythonic Web application frameworks.

Tim C




From tjreedy at udel.edu  Tue Jun  4 11:37:51 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 04 Jun 2002 15:37:51 GMT
Subject: sharing data in list sub-classes
References: <87znybyzsg.fsf@zaraza.fep.ru>
Message-ID: <j95L8.124601$Oa1.12117854@bin8.nnrp.aus1.giganews.com>

"Sergei Barbarash" <sgt at outline.ru> wrote in message
news:87znybyzsg.fsf at zaraza.fep.ru...
> Hello,
>
> Is there a way to share list data in classes that inherit the "list"
class?
>
> Example:
>
> ,----
> | class A(list):
> |     <what should I put here?>
> |
> | mem = [1,2,3,4,5]
> |
> | a1 = A(mem)
> | a2 = A(mem)

Try a2 = a1 instead

> |
> | a1[0] = 1024
> | if a2[0] == 1024:
> |     print "Everything's fine"
> `----
>
> With obsolete UserList module it works fine, because it stores the
data in
> self.data variable. Is there a "right" 2.2-ish way to do the same
thing?
>
> Thanks,
> --
> Sergei






From tim.one at comcast.net  Thu Jun  6 15:28:11 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 06 Jun 2002 15:28:11 -0400
Subject: array, list, performance...
In-Reply-To: <3CFFB86A.4114459A@accessforall.nl>
Message-ID: <BIEJKCLHCIOIHAGOKOLHIEPHDDAA.tim.one@comcast.net>

[Michael Chermside]
>     li = [x0, x1, x2, ...]
> ...
>     li.append(x)   appending is O(1)
>                    (actually, it SOMETIMES takes O(len(li)) as it
>                     resizes the list. But if you grow a list by
>                     continually appending, the amortized time is
>                     linear (ie, constant amortized time per element)

[Ype Kingma]
> Append is often enough linear in the length of the list
> that growing by appending is O(len(li) * len(li)).

For Python 2.2 Michael is correct.  For Python 2.1 and earlier, you're
correct "in theory", although in practice it was often linear-time anyway,
depending on internal details of the platform realloc.

> You can prevent that by creating a list a large enough list
> when starting, eg:
>
> li = [None] * neededSize

Yup, and that can save time regardless.





From pyth at devel.trillke.net  Thu Jun 13 19:25:42 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 14 Jun 2002 01:25:42 +0200
Subject: returning traceback strings
In-Reply-To: <Xns922D8E94C2EAcliechtigmxnet@62.2.16.82>; from cliechti@gmx.net on Fri, Jun 14, 2002 at 12:50:09AM +0200
References: <mailman.1024006339.11571.python-list@python.org> <Xns922D8E94C2EAcliechtigmxnet@62.2.16.82>
Message-ID: <20020614012542.O6609@prim.han.de>

Chris Liechti wrote:
> Matthew Boedicker <mboedick at mboedick.org> wrote in
> > 
> > This is the best workaround I was able to come up with:
> > 
> > import cStringIO
> > import traceback
> > 
> > def print_exc_str():
> >     buf = cStringIO.StringIO()
> >     traceback.print_exc(file=buf)
> >     return buf.getvalue()
> 
> try that:
>   ''.join(traceback.format_exception(*sys.exc_info()))

yeah. why not take the easy road :-)

    holger




From huaiyu at gauss.almadan.ibm.com  Fri Jun  7 18:24:13 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 7 Jun 2002 22:24:13 +0000 (UTC)
Subject: positional modifiers in python?
References: <mailman.1023460302.20078.python-list@python.org> <adqkgl$5ob$1@wanadoo.fr>
Message-ID: <slrnag2ckd.fv3.huaiyu@gauss.almadan.ibm.com>

Eric Brunel <eric.brunel at pragmadev.com> wrote:
>fuf wrote:
>>  hello everyone,
>> 
>>  does python support positional parameters in the print method? ie.
>>  something like:
>>     print "there %2$s something %1$s" % ("there", "is")
>>  it doesn't work but maybe there's some other way?
>
>There's an utterly ugly way to do this:
>
>def toPosDict(l):
>  d = {}
>  map(lambda k,v,d=d: d.update({str(k):v}), range(len(l)), l)
>  return d
>print "there %(1)s something %(0)s" % toPosDict(("there", "is"))
>
>For post-Python 2.2 users (I'm still 2.1, sorry), the "map" line may 
>certainly be replaced by:
>
>map(d.__setitem__, [str(i) for i in range(len(l))], l)
>
>but it may be less efficient than the code above (one map + one list 
>comprehension instead of a single map in the first solution). Maybe the 
>second's a little more readable (or a little less unreadable ;-).

Or if you prefer functional programming style (Only for 2.2):

def str_enum_dict(x):  return dict(zip(map(str, range(len(x))), x))
print "there %(1)s something %(0)s" % str_enum_dict(('there', 'is')) 

Huaiyu


From peter at engcorp.com  Mon Jun 24 21:51:10 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 24 Jun 2002 21:51:10 -0400
Subject: Suggestions for good programming practices?
References: <mailman.1024967851.27426.python-list@python.org>
Message-ID: <3D17CC8E.ABB72D14@engcorp.com>

Brian Quinlan wrote:
> 
> Peter Hansen wrote:
> > You _are_ returning them "in a struct".  It's spelled "tuple"
> > but other than that what's the difference?
> 
> struct members are access by name while tuple elements are accessed
> positionally.

Okay.  So why is that a problem?

Or to ask it from the other side, what's stopping you from 
returning items in an object?

class Struct:
    pass

def returnStruct():
    result = Struct()

    result.anInt = 5
    result.aString = 'foo bar baz'
    result.spam = ('eggs', 'brian')
 
    return result

-Peter


From sonnenbu at informatik.uni-bonn.de  Sat Jun  8 13:58:43 2002
From: sonnenbu at informatik.uni-bonn.de (Frank Sonnenburg)
Date: Sat, 8 Jun 2002 19:58:43 +0200
Subject: scope of module functions
References: <adte4c$nu8$1@f1node01.rhrz.uni-bonn.de> <Xns9227C812CD4B2cliechtigmxnet@62.2.16.82>
Message-ID: <adtgkh$sbs$1@f1node01.rhrz.uni-bonn.de>

"Chris Liechti" <cliechti at gmx.net> schrieb im Newsbeitrag
news:Xns9227C812CD4B2cliechtigmxnet at 62.2.16.82...
> "Frank Sonnenburg" <sonnenbu at informatik.uni-bonn.de> wrote in
> news:adte4c$nu8$1 at f1node01.rhrz.uni-bonn.de:
> > a little tricky problem (at least from my point of view); given module
> > CALLER:
> >
> > # i need "fct" and some other from module "UTILS"
> > from UTILS import *
> > fct()
> > # END of module CALLER
> >
> >
> > Now, inside fct(), how do i determine the module - in this case CALLER -
> > from where fct() was called?
>
> i think __name__ can help in this case but not in the one below.
> but are you sure that you wan't t do this? a function that behaves
> differently depending on the namespace gives very hard times debugging...
>
> a cleaner solution would be to create an class that contains fct and each
> module that uses it can create its own object.
>
> if you can say what you want to achieve, we might give you a more specifuc
> advice.
>
> > Even concerning cases like (continuation of example):
> >
> > # Now we are in __main__
> > import CALLER
> > fct_in_main = CALLER.fct
> > fct_in_main()
>
>
> --
> Chris <cliechti at gmx.net>
>

Yes, actually i thought of classes as a workaround.

I have a tool with a large amount of functions, so i grouped them into a
menu tree. My first approach was a package solution, which has nice
advantages: easy to build, unique access to any function. But there is a
main disadvantage: function names can get very long, for instance:
utils.database.sql.fetch()

Instead i want to enter:
>> import utils
>> database()
# do something in database, then switch to submenu sql
>> sql()
# now sql commands like fetch() are visible
>> val = fetch('name')
# now return to parent menu database
>> end()
# now fetch() is not available any more
.........


Anyway, thanx a lot for your quick response!

    Frank




From hst at empolis.co.uk  Tue Jun 11 05:59:08 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Tue, 11 Jun 2002 10:59:08 +0100
Subject: Sorting list (maybe stupid)
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB04F3CF@hendrix.empolisuk.com>

J?rvinen Petri wrote
> 
> Hey,
> 
> I have a problem and can't solve it:
> 
> There is lines of text like:
> 
> 2. Header2
> 1. Header1
> 102. Header102
> 
> and so on...
> 
> And I need to sort them according to the number (1,2,3,4...)
> 
> Am I missing something.. I have tried to split the lines by 
> whitespace and 
> create lists. But how can i sort the list by first element...
> 
> Thanks, Petri
> -- 

Here's a quick and dirty 2.2 program illustrating one way. It uses a regular expression to extract the leading number, creates a tuple of the leading number (zero if none) and the original text. It then sorts the list of tuples in place and prints the
second element of each tuple.

import re

r = re.compile('\d+')
afile = open('myfile.txt')
tuplelist = []
for aline in afile:
    m = r.match(aline)
    if m:
        tuplelist.append((int(m.group(0)), aline))
    else:
        tuplelist.append((0, aline))
tuplelist.sort()
for x in tuplelist:
    print x[1]

HTH

Harvey

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.




From gerhard at bigfoot.de  Wed Jun 19 03:02:25 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 19 Jun 2002 09:02:25 +0200
Subject: zlib/gzip + anydbm or ...
In-Reply-To: <aeoufa$rkk$1@news.idiom.com>
References: <aeoufa$rkk$1@news.idiom.com>
Message-ID: <20020619070225.GA9430@lilith.my-fqdn.de>

* James T. Dennis <jadestar at idiom.com> [2002-06-19 03:40 +0000]:
> 
>  Is there an obvious and simple way to use the existing standard
>  libraries to support transparently compressed dbm files?

I don't think so.

If at all, you'll need to hack dumbdbm to achieve this goal, I think
only two approaches are feasible:

- compress the whole file with zlib, this means you need to keep the
  whole file in memory
- compress only values with zlib

Option two could even be written with a simple wrapper around anydbm,
but will obviously only help if your average values are relatively
large.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 25.2 ?C      Wind: 3.4 m/s




From pyth at devel.trillke.net  Tue Jun 18 18:14:14 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 19 Jun 2002 00:14:14 +0200
Subject: python version?
In-Reply-To: <%%NP8.46690$GY.14560922@twister.nyroc.rr.com>; from hesterloli@hotmail.com on Tue, Jun 18, 2002 at 09:56:11PM +0000
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com><aegfck$6tkre$1@ID-114079.news.dfncis.de><wEPO8.16066$uk2.8447171@twister.nyroc.rr.com><3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com><gYgP8.42356$GY.12361719@twister.nyroc.rr.com><aekcug$7js8n$1@ID-114079.news.dfncis.de><9YBP8.44418$GY.13865843@twister.nyroc.rr.com><mailman.1024423525.24359.python-list@python.org> <20020619091900.70ae2b7f.larooy@xtar.co.nz> <%%NP8.46690$GY.14560922@twister.nyroc.rr.com>
Message-ID: <20020619001414.G15079@prim.han.de>

George Hester wrote:
> oops:
> 
> No George was asking for a script he could put in a ASP to determine his version
> of Python.

well. your original question was:

    What is a simple script that will tell me the version of python I am
    currently running?  IIS 5.0 Windows 2000.

Had you given the details above i guess you would have received other 
and fewer answers.  I for one tried to answer the sentence up to the question
mark.  which didn't help you.   

But no matter if people answer to your satisfaction or not. 
IMO it's bad style to be aggreesive for not beeing served exactly 
the way you expect it. 

regards,

    holger




From gumuz at looze.net  Fri Jun 21 06:53:41 2002
From: gumuz at looze.net (Guyon Morée)
Date: Fri, 21 Jun 2002 12:53:41 +0200
Subject: Socket Disconnection
Message-ID: <3d1304d0$0$224$4d4ebb8e@news.nl.uu.net>

Hi!

I have written a small and simpel server and I was wondering how do you find
out that a client has disconnected?

thnax,

Guyon




From whisper at oz.net  Sun Jun  2 16:15:23 2002
From: whisper at oz.net (David LeBlanc)
Date: Sun, 2 Jun 2002 13:15:23 -0700
Subject: Does this widget exist?
In-Reply-To: <24d8090a.0206011740.49d253e3@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEGECLDGAA.whisper@oz.net>

Yep:
MCScrolledListBox
A scrolled list with multiple columns. Supports various selection modes.
Display and selection values are sequences of tuples (no formatting or
parsing!).

http://pmwcontribd.sourceforge.net/

Populating one with files data is left as an exercise for the reader ;-)

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Corey Woodworth
> Sent: Saturday, June 01, 2002 18:40
> To: python-list at python.org
> Subject: Does this widget exist?
>
>
> Is there a Tk widget that lists files like you see in napster or
> winMX. (you know with the kb/sec and frequency buttons up top to sort
> things.)
>
> If I use the search feature in windows and I select the details view I
> get a widget very similar (except of course windows isn't written in
> python) :)
>
> Corey
> --
> http://mail.python.org/mailman/listinfo/python-list





From emile at fenx.com  Tue Jun 25 11:42:02 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 25 Jun 2002 15:42:02 GMT
Subject: Sending string var with apply sends all chars as params!???
References: <mailman.1025017993.24641.python-list@python.org>
Message-ID: <eb0S8.66230$EP.6818@sccrnsc03>

Etienne
> value = apply(formatter,(str(value)))


Try passing in a tuple, eg:

value = apply(formatter,(str(value),))

>>> def fmt(*args):
...     print args
...
>>> s = '20020625140609'
>>> apply(fmt,s)
('2', '0', '0', '2', '0', '6', '2', '5', '1', '4', '0', '6', '0', '9')
>>> apply(fmt,(s,))
('20020625140609',)
>>>

--

Emile van Sebille
emile at fenx.com

---------



From tjreedy at udel.edu  Mon Jun  3 09:59:21 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 03 Jun 2002 13:59:21 GMT
Subject: solarwolf display ?
References: <slrnafl9gv.5h4.Taekyon@macchinetta.miadimora>
Message-ID: <ZCKK8.111623$%o.10622201@bin3.nnrp.aus1.giganews.com>

"Tae kyon" <Taekyon at macchinetta.miadimora> wrote in message
news:slrnafl9gv.5h4.Taekyon at macchinetta.miadimora...
> On a Mandrake 8.2 system I compiled pygame 1.5 and solarwolf 1.5.1.
> I have a matrox millennium g200 8 Mb video ram graphics card with 3d
support
> by XFree86 4.2.0, and a SVGA non-interlaced video with HSync 30-54,
> VSync 50-90. I even get to play tuxracer at a good framerate ! But
for
> some reason 800x600 fullscreen displays will not work (some loki
games
> have to be played in a window, for instance, and nethack must be
played
> on a 1024x768 display).
>
> Unfortunately it seems that solarwolf uses this type of display.
> I know very little about python, nothing about pygame; anyway I
tried
> setting the video mode by changing the relevant options in
> pygame.display.set_mode(()) ... but nothing seems to work.
>
> Any tips ?

I believe that there is a PyGame mailing list.  A question this
specific is more likely to get a proper answer there.

TJR





From K.Rdt at TU-Berlin.DE  Sun Jun 16 15:07:37 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Sun, 16 Jun 2002 19:07:37 GMT
Subject: HT telnet interact with auto login?
References: <3d0cd198.17917980@news.zrz.tu-berlin.de> <Xns922FDF67BADDBcliechtigmxnet@62.2.16.82>
Message-ID: <3d0ce0df.21829780@news.zrz.tu-berlin.de>

On 16 Jun 2002 21:55:15 +0200, Chris Liechti <cliechti at gmx.net> wrote:

>K.Rdt at TU-Berlin.DE (Klaus Reinhardt) wrote in 
>news:3d0cd198.17917980 at news.zrz.tu-berlin.de:
>> I want to combine:
>> 
>> 1.)
>> tn.interact()     # this is working but with manually ..
>> 2.)
>> tn.read_until("login: ")     # this seams to work
>>                                        # but without connection
>> tn.write(user + "\r\n")
>> if password:
>>     tn.read_until("Password: ")
>>     tn.write(password + "\r\n")
>
>i don't understand your question... but often a '\r\n' is required
>as end of line. maybe this helps :-)
>
>chris

The \r don't help. Well, I want to have a script, which establishes
automatically a telnet-connection, so I can there do some
console-processing. My feeling is wrong, that I can give
user and password to interact() .. perhaps there is another
function ..

	Thanks Klaus


From FooWeissBarMike at hotmail.com  Mon Jun 24 11:44:49 2002
From: FooWeissBarMike at hotmail.com (Michael Weiss)
Date: Mon, 24 Jun 2002 11:44:49 -0400
Subject: Numerical Python array math with different sized arrays
Message-ID: <3d173f56$1@news.mhogaming.com>

I'm using Numeric Python's arrays. I want to perform the various math
operators between arrays (+-*/ etc...) .
While all my arrays are 1 dimensional, they may be different lengths. The
result I'd like in a mis-matched size case is the following:
>>> a = Numeric.array([1,2,3,4])
>>> b = Numeric.array([1,2])
>>> print a+b
[2,4,3,4]

I assume the answer is to resize b, and make the new elements all zero. My
question is what is the fastest way to do this. Resize() does that repeating
thing, which is not what I want (I want zeros in those new elements).

I've tried the following, is there a better way?

#####
lena = len(a)
lenb = len(b)

if lena == lenb:
   numa = Numeric.array(a)
   numb = Numeric.array(b)
elif lena > lenb:
   numa = Numeric.array(a)
   zeros = Numeric.zeros((lena-lenb,))
   numb = Numeric.concatenate((b, zeros, ))
else:
   zeros = Numeric.zeros((lenb-lena,))
   numa = Numeric.concatenate((a, zeros, ))
   numb = Numeric.array(b)

numans = numa+numb
#####





From logiplexsoftware at earthlink.net  Wed Jun 26 13:49:23 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 26 Jun 2002 10:49:23 -0700
Subject: question od default args
In-Reply-To: <o3vjhuguen6gfm8vnq7rbk5volmi80shl5@4ax.com>
References: <o3vjhuguen6gfm8vnq7rbk5volmi80shl5@4ax.com>
Message-ID: <20020626104923.45f54821.logiplexsoftware@earthlink.net>

On Wed, 26 Jun 2002 18:45:25 +0100
Gon?alo Rodrigues wrote:

> Hi,
> 
> When I want default args I usually do
> 
> def (arg = None):
>     ...
> 
> But what if I want to make None a reasonable argument too? That is, I
> want to know if the user has in fact passed an argument, and if not to
> do something special, with None (or whatever object) being a reasonable
> arg to be passed.

How about:

def foo(*args, **kwargs):
    argc = len(args) + len(kwargs)
    if argc == 0:
        print "no arguments"
        # set default values
    else:
        print argc, "arguments"
        # assign arguments to attributes or whatever

Then if the user passes None as an argument, argc will still be 1.  You can
then get the arguments from args and kwargs.

Regards,
-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From michael at damaru.com  Sun Jun  9 13:07:15 2002
From: michael at damaru.com (Michael Davis)
Date: Sun, 09 Jun 2002 13:07:15 -0400
Subject: libftp.size function gives wrong results!
References: <RNKM8.560$Yw3.96921@news20.bellglobal.com> <3D038A19.D991A885@engcorp.com>
Message-ID: <%ULM8.903$Yw3.118334@news20.bellglobal.com>

Peter Hansen wrote:

> Michael Davis wrote:
>> 
>> I'm writing a python program which will upload or download files via ftp.
>> It's intended to be used to deploy a web site from a local machine to a
>> server. It tries to determine which files are modified on the local
>> machine by comparing file sizes and modification times.
>> 
>> I try to compare files using os.stat for the local file and FTP.size for
>> the remote. The problem is, FTP.size is giving me incorrect results. If I
>> use the ftp command line client, and look at one of my files on the
>> server (the output looks like 'ls -l') then I can see that the file size,
>> is, say, 1200. But the FTP.size function returns, say, 1242.
> 
> I think you'll find the number of lines in that file is 42.  The
> difference
> is probably between CRLF and LF line terminations in the file.  If you
> transfer files with FTP and want to preserve the line endings, you have
> to specify the binary type, not ascii.  Don't know how to do this with
> ftplib, but from the ftp command line you would use 'type binary' or
> 'type ascii' to switch.

Thanks Peter,

I had already thought of that. Both my local and the remote systems are 
unix, and so don't have CRs. But you're right - the difference in size is 
the same as the number of lines, as though the ftp size command assumed 
that I wanted to convert the end-of-lines to CRLF. I had specified binary 
mode transfers. And if I do 'ftp size' on true binary files, the same 
problem happens!

Anyway, I just found out that one of the servers I need to talk to doesn't 
allow me to use 'size' anyway, so I'm going to ask a different python 
question in a different message.

Thanks again,

-- 
Michael Davis
Damaru
Custom Programming - Web Development - Database Design
http://www.damaru.com
416-540-1284


From hesterloli at hotmail.com  Wed Jun 19 16:27:08 2002
From: hesterloli at hotmail.com (George Hester)
Date: Wed, 19 Jun 2002 20:27:08 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com> <3D1010C3.C321F8B0@engcorp.com> <raXP8.47926$GY.15260255@twister.nyroc.rr.com> <3D107948.B3FEE4BB@engcorp.com> <CT%P8.33113$uk2.13710586@twister.nyroc.rr.com> <slrnah1l0i.6o5.huaiyu@gauss.almadan.ibm.com>
Message-ID: <wO5Q8.33869$uk2.14055292@twister.nyroc.rr.com>

All fixed

--
George Hester
_________________________________
"Huaiyu Zhu" <huaiyu at gauss.almadan.ibm.com> wrote in message
news:slrnah1l0i.6o5.huaiyu at gauss.almadan.ibm.com...
> George Hester <hesterloli at hotmail.com> wrote:
> > But the reason I
> >mentioned is that I hoped that my students would take an interest in the
subject
> >and learn the concept.  They were not interested but they needed the credit
to
> >graduate.  In my case I was not necessarily interested in learning Python.
But
> >I did want to determine my version of Python in ASP.  I suppose I was like my
> >students and ya'll were like the teacher.  Yes basically I wanted it handed
to
> >me on a platter.  What I did with the platter may have surprised you.  So
just
> >becasue somebody wants something handed to them on a platter does not mean
the
> >contribution has less value.
>
> Being able to communicate effectively in Usenet may be an important skill
> for your students to master.  You may want to teach them some lessons you
> have learned yourself.  Here's a mini multiple choice questionair to get you
> started:
>
> 1. Suppose you want to get an answer to a question handed down on a platter,
>    which would be the more effective question?
>    a) What is the simplest script to get Python version?
>    b) I know nothing about Python. I just want to know its version from ASP,
>    what's the easiest way to get it?
>
> 2. Suppose you are given an answer that is all Greek to you, which would be
>    the more effective response?
>    a) I don't think you guys know anything about this.
>    b) I think I need a script to do this.
>    c) This is well over my head.  I'm newbie.  Please explain the details,
>    keeping in mind that I only want to get the version from ASP.
>
> 3. If someone tells you that you are going about it the wrong way and gives
>    you some advice which sounds like a detour to you, do you judge it based
>    on
>    a) whether he is your daddy.
>    b) whether the advice has merits.
>
> 4. Once you realize that all the answers people gave you were real, some
>    much simpler than you imagined before, and that your accusations earlier
>    were largely based on arogance, do you
>    a) Say: Hey, you guys didn't gave me the answers the way I wanted, and
>    your tones were derogotory.
>    b) Go away quietly.
>    c) Say: Thanks for all the answers.  Even though I didn't ask the
>    question clearly, one of answer was exacted what I wanted.  Sorry for the
>    tone of my earlier posts, which was largely because I didn't recognize
>    the answers.
>
>
> No doubt others can add more lessons learned from this episode.
>
> Huaiyu




From usenet at minus1.de  Mon Jun 24 03:05:00 2002
From: usenet at minus1.de (Konrad Anton)
Date: Mon, 24 Jun 2002 09:05:00 +0200
Subject: mail-problem over t-online
References: <mailman.1024859369.26585.python-list@python.org>
Message-ID: <slrnahdh4s.123.usenet@fohnhab.minus1.de>

Klaus Reinhardt <K.Rdt at TU-Berlin.DE> schrieb:
>Testing python and email I adapted this script:
>	
>	host=smtplib.SMTP('mailszrz.zrz.tu-berlin.de')
>	host.sendmail(zwi_eck(From),zwi_eck(Tp),al)
>	host.quit
>
>This is working, when I'm directly connected to my
>old ISP TU-Berlin via ISDN. But the same is failing,
>when I'm connected at t-online via ADSL, with
>authentification required .. Providing my log and pass
>produces errors, too.

First, try if this also happens if you do the SMTP dialogue using
telnet.

telnet mailszrz.zrz.tu-berlin.de 25
HELO meinhostname.meinedomain.de
MAIL FROM: soundso at tu-berlin.de
RCPT TO: blabla at tu-berlin.de
DATA
Dies ist ein Test
.



    -- Konrad


-- 
 Konrad Anton <konrad at minus1.de> ; Web: www.minus1.de ; gpg: 0x22954D8A
 Tel. +49-761-881-2122, Fax +49-721-151318943.  Alice:
And I've something: such as the two looking at the directions, just the
earls of the Duchess, I didn't! 


From edream at tds.net  Sat Jun 22 22:33:00 2002
From: edream at tds.net (Edward K. Ream)
Date: Sun, 23 Jun 2002 02:33:00 GMT
Subject: ConfigParser & converting strings to lists
References: <3D152A28.2177B054@tds.net> <3D152C69.CC41ADB0@tds.net> <3D1531DD.ACF06E31@engcorp.com>
Message-ID: <3D15335C.8CB1558D@tds.net>

> > so it is strings of the form "['file1' 'file2' ...]" that must be
> > converted to a list of strings.
> 
> >>> s = "['file1' 'file2' '...']"
> >>> [x[1:-1] for x in s[1:-1].split()]
> ['file1', 'file2', '...']
> 
> Does that do it?

I'm not sure.  Here is what I did:

s = config.get(section,"recentFiles")
files = [x[1:-1] for x in s[1:-1].split()]

Is this what you intended?  This doesn't work for me: it "doubles" every
backspace.

I am already regretting trying to be clever.  It is easy enough to set
file individual filenames like this:

for i in xrange(len(files)):
  config.set(section, "file"+`i`, files[i])

and to get up to n file names like this:

try:
  for i in xrange(n):
    files.append(config.get(section, "file" + `i`))
except: pass

This is bulletproof: filenames can contain any character at all.  The
only drawback is that the config file looks like:

[recent files]
file0 = filename0
file1 = filename1
...

But this isn't so bad, is it?

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


From whisper at oz.net  Sat Jun 15 12:20:43 2002
From: whisper at oz.net (David LeBlanc)
Date: Sat, 15 Jun 2002 09:20:43 -0700
Subject: SWIG and Callbacks
In-Reply-To: <3d0b4fd6@nntp.server.uni-frankfurt.de>
Message-ID: <GCEDKONBLEFPPADDJCOEEEFLDJAA.whisper@oz.net>

No, that understanding is not wrong, but it will enable you to write python
callbacks. From reading the doc and looking at examples, writing
Boost/Python interfaces isn't all that arduous.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Michael 'Mickey' Lauer
> Sent: Saturday, June 15, 2002 7:32
> To: python-list at python.org
> Subject: Re: SWIG and Callbacks
>
>
> David Abrahams <david.abrahams at rcn.com> wrote:
> > > "Michael 'Mickey' Lauer"
> <mickey at tm.informatik.uni-frankfurt.de> wrote in
> > message news:3d0a8b6e at nntp.server.uni-frankfurt.de...
> >
> >
> >> I had a quick look at the other wrappers (SILOON, GRAD, CXX, etc.)
> >> but most look either outdated or not appropriate for this task.
> >> Do you know an (automatic) wrapper generator which has this feature?
> >>
> >> I know sip can do this for c++ libraries - can sip also be used
> >> to wrap plain c-libs without "handwriting" code?
> >
> > On most platforms, Boost.Python would work for you. The
> determining issue is
> > whether or not the C++ compiler treats "C" and "C++" linkage function
> > pointers the same way. If the appended program will compile
> with your C++
> > compiler, you're probably OK:
>
> The attached program compiles fine here. However, regarding boost.python I
> thought this was more of a c++ library for handcrafting tightly integrated
> bindings rather than an automatic wrapper generator? Is this
> understanding wrong?
>
> Yours,
>
> :M:
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From vjovanov at stevens-tech.edu  Tue Jun  4 14:35:24 2002
From: vjovanov at stevens-tech.edu (Vojin Jovanovic)
Date: Tue, 4 Jun 2002 14:35:24 -0400
Subject: self
References: <mailman.1023208220.19083.python-list@python.org>
Message-ID: <TL7L8.79$4p7.4100181@newnews.cc.stevens-tech.edu>

> Which of i, j and k are instance variables?  Which are locals?  Which are
> globals?  For all its virtues, Python isn't telepathic <wink>.

Well, now that you pointed it out, I think that Python should be telepathic.
Those i,j and k variables should have been found in class C, otherwise they
should have been searched outside of the class.  I read the FAQ page that
somebody sent me and I really think that the main reason for self is ...
"When classes were added to Python, this was (again) the simplest way of
implementing methods without too many changes to the interpreter. " which is
the main and really lousy argument.  All, other arguments are debatable.

Now let me give you the problem with self which is making my life
complicated. Consider this class.

class Foo:
 def __init__(self):
  self.__dict__['Equations']={
    'a':str(5),
    'b':'self.a+5',
    'c':'self.b*self.a'}

 def __getattr__(self,arg):
  if self.Equations.has_key(arg) == 0:
   raise AttributeError,  "Foo instance has no attribute '" + arg + "'"
  try:
   try:  #first check if this is a simple expression
    return eval(self.Equations[arg])
   except:  #if not try to execute it
    exec self.Equations[arg]
  except:   #if even that didn't work then just return undefined
   return '#U'

 def __setattr__(self,arg,val):
  self.__dict__['Equations'][arg]=str(val)

 def __delattr__(self,arg):
  del self.__dict__['Equations'][arg]

This class implements a kind of symbolic behavior with the ability to add
new equations on the fly.  For example...


>>> import foo
>>> A=foo.Foo()
>>> A.Equations
{'b': 'self.a+5', 'c': 'self.b*self.a', 'a': '5'}
>>> A.f='self.b*self.c'
>>> A.f
500
>>> A.Equations
{'f': 'self.b*self.c', 'b': 'self.a+5', 'c': 'self.b*self.a', 'a': '5'}


As you can see, equations get added on the fly and the parsing gets done as
the attribute is being called.  Now, notice that ugly self. thing in the
equations.  Instead of just typing >>>A.f='a*c' I have to add self in front
a and c to achieve the desired effect.  I could not find any other way to
implement the class in order to avoid typing self. which of course it
doesn't mean that there isn't any. If anybody has an idea how to do this
please let me know.  Otherwise, here is the situation where the self
concept is bad and having such a dynamic language like Python I think there
has to be a way out of this problem.

Why is this all relevant? Well, I am working on embedding an interpreted
language in a large engineering application and Python is  one of few
candidates.  So I have been looking for its pros and cons.  So far it passed
all the tests except the above.  Because, imagine if we required from the
user of our system that he needs to type self. or even s. in front of every
new variable that he has in his equation, it would be a big turn off.

Vin





From shalehperry at attbi.com  Thu Jun 27 16:13:51 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 27 Jun 2002 13:13:51 -0700 (PDT)
Subject: self destruction :)
In-Reply-To: <f4KS8.1326$qq1.14832@castor.casema.net>
Message-ID: <XFMail.20020627131351.shalehperry@attbi.com>

On 27-Jun-2002 Guyon Mor?e wrote:
> i've got a list of dynamicaly instantiated objects of a class of my own. how
> can i let one of the objects delete itself?
> 
> i've tried: 'del self', but somehow it didn't work. afterwards it should
> also _not_ be in the list anymore.
> 

mylist = [obj1, obj2, ....]
del obj1 # mylist still has a reference to obj1 so
         # the del simply decrements the ref count






From erict at millfilm.co.uk  Thu Jun  6 09:36:41 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Thu, 06 Jun 2002 14:36:41 +0100
Subject: HttpServer help!
References: <3CFF4C14.450850D8@millfilm.co.uk> <P1JL8.312164$%u2.176575@atlpnn01.usenetserver.com>
Message-ID: <3CFF6568.7EE613CE@millfilm.co.uk>

The server and the client run from the same directory.
and the cgi file is in the same directory.

all the following failled.
req.putrequest("POST", "favquote.cgi")
req.putrequest("POST", "/favquote.cgi")
req.putrequest("POST", "./favquote.cgi")


What is the meaning of  *relative to the server root*


Steve Holden wrote:

> "Eric Texier" <erict at millfilm.co.uk> wrote in message
> news:3CFF4C14.450850D8 at millfilm.co.uk...
> > I am trying to get some simple example running locally on my
> > machine of http(server/client)  code from the book 'Python & Xml'
> > and I am having trouble.
> >
> > if someone can look at the following script and tell me if
> > the problem can be because I don't have the privilege and/or
> > I didn't setup my system properly:
> > Thanks.
> >
> > I am running on Linux Red Hat with python2.2.
> >
> >
> > The client code to POST:
> > """
> > post.py
> > """
> > from httplib import HTTP
> > from urllib  import quote
> >
> > # establish POST data
> > myquote = 'This is my quote: "I think therefore I am."'
> >
> > # be sure not to quote the key= sequence...
> > postdata = "favquote=" + quote(myquote)
> >
> > print "Will POST ", len(postdata), "bytes:"
> > print postdata
> >
> > # begin HTTP request
> >
> > #req = HTTP("127.0.0.1")  # change to your IP or localhost
> > #req = HTTP("localhost")  # change to your IP or localhost
> >
> > ##### MY IP address
> > req = HTTP("10.16.3.132")  # change to your IP or localhost
> >
> >
> > #req.putrequest("POST", "/c8/favquote.cgi")
> > req.putrequest("POST",
> > "file:/usr/people/erict/python/srcToolskit/pythonXmlBook/c8/favquote.cgi")
> >
> > .
> Hold it right here. How is a FILE supposed to process the input you are
> trying to send it? Your .putrequest() call should give the path to the CGI
> script *relative to the server root*, not it's location in the server's file
> store.
>
> > .
> > .
> >
> > ##### That the Error after the ruprequest->"#####
> > Traceback (most recent call last):
> >   File "post.py", line 25, in ?
> >     req.putrequest("POST",
> > "file:/usr/people/erict/python/srcToolskit/pythonXmlBook/c8/favquote.cgi")
> >
> >   File "/usr/people/erict/bin//lib/python2.2/httplib.py", line 453, in
> > putrequest
> >     self.send(str)
> >   File "/usr/people/erict/bin//lib/python2.2/httplib.py", line 395, in
> > send
> >     self.connect()
> >   File "/usr/people/erict/bin//lib/python2.2/httplib.py", line 379, in
> > connect
> >     raise socket.error, msg
> > socket.error: (111, 'Connection refused')
> >
> Sure enough, when the client tries to create a connection, it can't.
>
> [...]
>
> Hope this helps.
>
> regards
> -----------------------------------------------------------------------
> Steve Holden                                 http://www.holdenweb.com/
> Python Web Programming                http://pydish.holdenweb.com/pwp/
> -----------------------------------------------------------------------



From rnd at onego.ru  Sun Jun  2 23:31:49 2002
From: rnd at onego.ru (Roman Suzi)
Date: Mon, 3 Jun 2002 07:31:49 +0400 (MSD)
Subject: prototyping good OOdesign in Python?
In-Reply-To: <XFMail.20020601131915.shalehperry@attbi.com>
Message-ID: <Pine.LNX.4.33.0206030725420.5660-100000@rnd.onego.ru>

On Sat, 1 Jun 2002, Sean 'Shaleh' Perry wrote:

>> 
>> Well, if you still follow this thread, the question is - how well
>> Python could serve as an ARCHITECTURAL prototype if it has too rich
>> OOP facilities? That is, is Python of help when trying to
>> prototype certain design? Developing in C++ looks so unnecessary hard
>> after things were done in Python... 
>> Or, it may be put this way: what discipline a Python programmer must
>> obey to allow it's prototype to be conveniently rewriten in C++?
>> 
>> The above thoughts aren't probably well-formed. But I hope
>> my concern is understood.
>> 
>
>architectural prototyping is what I usually use python for when I prototype,
>the other use is rapid class creation for behaviour testing.  You have to keep
>in mind the goal of eventually porting to C++ but it is a prototype so there is
>not a lot of code there anyway.

But, for example, C++ has 3 categories of object attributes. Do you mark
them somehow in Python, group them to be later marked as public,
protected, private? And there are lot's of details like this one.

Related topic is UML with Python. Does UML has everything to reflect
Python OO model or does it need to add features?
(Or maybe in order to interoperate, there is a need not to use some
of Python features).

And if in order to interoperate with less expressive languages/systems,
how to keep Python pythonic and not fall back to some standard universal
subset, found in many programming languages.

Sincerely yours, Roman Suzi
-- 
\_ Russia \_ Karelia \_ Petrozavodsk \_ rnd at onego.ru \_
\_ Monday, June 03, 2002 \_ Powered by Linux RedHat 7.2 \_
\_ "Bugs are Sons of Glitches!" \_





From eikeon at eikeon.com  Fri Jun  7 14:22:14 2002
From: eikeon at eikeon.com (Daniel 'eikeon' Krech)
Date: 07 Jun 2002 14:22:14 -0400
Subject: Memory leaks when  embedding python (debug-version)  ?
In-Reply-To: <lkptz3kw5v.fsf@pc150.maths.bris.ac.uk>
References: <1eed5b9.0206070157.6d281f65@posting.google.com>
	<lkptz3kw5v.fsf@pc150.maths.bris.ac.uk>
Message-ID: <uelfjt0dl.fsf@eikeon.com>

Michael Hudson <mwh at python.net> writes:
> The interned string dict?  There's some call for getting rid of that,
> but I can't remember it off hand.

It sounds like a call to release the unused strings from the interned
dict is on its way:
    http://mail.python.org/pipermail/python-dev/2002-June/025127.html

--eikeon, http://eikeon.com/





From peter at engcorp.com  Sun Jun  9 11:24:48 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 09 Jun 2002 11:24:48 -0400
Subject: Simple pychecker question
References: <mailman.1023482712.8801.python-list@python.org> <pan.2002.06.09.00.07.47.445563.6814@metaslash.com> <3D02FDE9.19198F33@engcorp.com> <pan.2002.06.09.14.48.16.902484.18868@metaslash.com>
Message-ID: <3D037340.A9689E39@engcorp.com>

Neal Norwitz wrote:
> 
> Peter Hansen wrote:
> > Neal Norwitz wrote:
> >> I personally think that superfluous use of named args are bad, but that
> >> the feature is good when used properly.  
> >
> > transmit('this is a string', timeout=5)
> >
> > My point: named arguments can increase readability and maintainability.
> > That's not superfluous, to me.
> 
> I agree that they can enhance readability.  With the example above,
> it's questionable, but that's my opinion.  Suppose the function was
> transmitWithTimeout()?  Or suppose the code looked like this:

transmitWithTimeout() would imply it always had a timeout.
(In my example it did, with the default of 1.0, but I could as
well have had timeout=0.0 for the default.)

>         timeout = 5
>         transmit('this is a string', timeout)

This would be just fine.  Quite readable, although the name
'timeout' is duplicated unnecessarily.  With the named argument,
the intention is clear but there is no need for duplication.

> There really doesn't seem to be a right way, it just depends.

Exactly, which was my point.  Why give a warning for one of two
perfectly acceptable alternatives?

-Peter


From chris.gonnerman at newcenturycomputers.net  Mon Jun 24 08:06:09 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Mon, 24 Jun 2002 07:06:09 -0500
Subject: Newbie question (anyone programming with curses?)
References: <GCEDKONBLEFPPADDJCOECELJDLAA.whisper@oz.net>
Message-ID: <003001c21b77$8af27980$0101010a@local>

----- Original Message -----
From: "David LeBlanc" <whisper at oz.net>


> Forgive me if i'm incorrect - it's been awhile since I mucked with curses.
>
> Isn't the purpose of curses to convert ANSI standard codes into whatever a
> particular terminal understands?

Um, no.  curses layers a virtual screen system on top of an abstract
terminal
interface.  There's nothing "ANSI" about it.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net





From Chris.Barker at noaa.gov  Thu Jun 20 15:27:11 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Thu, 20 Jun 2002 12:27:11 -0700
Subject: logical statements (in Python and Numeric)
References: <3D11F1B2.21587F31@lanl.gov> <aet2j4$h8c$1@peabody.colorado.edu>
Message-ID: <3D122C5D.C6985686@noaa.gov>


Fernando P?rez wrote:
> > it has to do with a using an and statement...

> > y = y + (x >= -2. and x < 0.) * (x + 6.)

> I agree that the above failure is bloody counterintuitive, to the point where
> I'd call it a bug. It's easy to illustrate with:

I don't know if it is a bug or not, but it is annoying. Python 2.1
introduced "rich comparisons", which allowed NumPy to overload >,<,== ,
etc, but you can't overload "and" and "or". I think this is because of
how "and" shortcuts but not evaluating the second operand if the first
is false, etc. In Python, a non-empty sequence is TRUE, so x and y. will
always evaluate to true if x and y are not empty, and the value returned
is teh first one that answers the question, so, for y and y not empty:

x and y #returns y
x or y  #returns x

One solution is to use the bitwise operators instead:
y = y + (x >= -2. & x < 0.) * (x + 6.)

or 
y += ( (x >= -2. & x < 0.) * (x + 6.) ) 

to use +=

This should work fine in this case, and most comon cases.

You might also want to take a look at the where() and putmask() functions:

y = y + where( (x >= -2. & x < 0.), (x + 6.), 0 )

> You may consider posting to the Numeric mailing list about this.

before you do that, look at the archives, it has been discussed a lot.

> The 'and'
> operator should either not work at all or work correctly.

I'm not sure either of these is possible without significant changes to Python.

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From eddie at holyrood.ed.ac.uk  Fri Jun  7 07:06:51 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Fri, 7 Jun 2002 11:06:51 +0000 (UTC)
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de>
Message-ID: <adq44b$obr$1@scotsman.ed.ac.uk>

"Shagshag13" <shagshag13 at yahoo.fr> writes:


> I'm looking for any state of art, optimizing techniques, tips, that a
>python beginner couldn't know by itself, but should learn and
>would be aware to write "efficient and good looking"python's code...

> (not the kind of thing which made you an obfuscated python nerd,
>but that "map" should be preferred to "for in loop" if possible, that
>dictionary are very well optimized, etc.)

A friend is fond of quoting this to me:

  Rules of Optimization:
    Rule 1: Don't do it.
    Rule 2 (for experts only): Don't do it yet.
  -- M.A. Jackson

I wasn't going to bother contributing since others have said the same thing
but the more people say the more you'll believe it :)

Especially for a beginner the important thing is to get clear and obvious
code.  Later, IF IT ISN'T FAST ENOUGH, then you think about ways to speed it
up.

There are may reasons for this, one of them is that it is often counter
intuitive exactly what the fastest approach is anyway.

Just concentrate on finding sensible algorithms as you would in any other
procedural language.

Eddie


From rch at baustatik.rwth-aachen.de  Thu Jun 20 04:29:16 2002
From: rch at baustatik.rwth-aachen.de (Rostislav Chudoba)
Date: Thu, 20 Jun 2002 10:29:16 +0200
Subject: how to configure the search path for PyBuildExt.detect_modules in setup.py
Message-ID: <aes3mr$dnr$1@nets3.rz.RWTH-Aachen.DE>

Hi everyone,

Is it possible to compile the suite Tcl, Tk, Tix, BLT and Python with 
Tkinter in nonstandard directories: /usr/lib, /usr/include 
_without_ having to modify anything in Modules/Setup?

There is a setup.py script which can detect whether there is tcl and tk in 
the directories /usr/include and /usr/lib. 
Why does this script does not use any environment variables for the library 
and include paths? 
Of course, I could edit Modules/Setup but I would have to do that for every 
time I include a new version of Python. Setting enviromnent variables would 
allow just copy & install procedure.

My goal is to have a system-independent suite of the mentioned packages to 
avoid tedious work after each linux system upgrade. 

Rostislav Chudoba




From phlip_cpp at yahoo.com  Thu Jun 13 00:47:36 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 13 Jun 2002 04:47:36 GMT
Subject: ANN: PyUnitTestBrowser 004
Message-ID: <ae9858$jp3@dispatch.concentric.net>

<posted & mailed>

Pynt Ho:

First the non-good news. I've done everything but retest under Windows. It 
could work perfectly, screw up a test over some trivial 'os.sep' issue, or 
even go boom because I replaced 'reload' with 'import imp' and the internal 
calls like 'find_module' and 'load_module'. They started totally instantly 
and aggressively working (on Linux), so I stopped asking questions.

Dox: http://flea.sourceforge.net/BrowseMe.html

Get: http://flea.sourceforge.net/browser004.zip

Spin: http://www.c2.com/cgi/wiki?PyUnitTestBrowser

Pitch: 

An alternative guitestrunner for PyUnit. Its features, in combination with 
your editor and your console, form a plug-n-play Python Test-First IDE, 
tuned for very-large scale pythonic projects, hypothetically on any 
platform.

The highlights:

 - trigger a test event when your file saves

 - incrementally test...
        - a single test case
        - a single test suite
        - a single test module
        - all test files in the current folder
        - an entire folder hierarchy, recursively

 - keyboard or mouse support to 
        - run the UI, 
        - launch tests, 
        - and launch your editor

 - your __doc__ strings displayed at test selection time

 - dynamic refresh when you add new test suites or cases

 - locate tests thru a module's 'suite' function, 
        -  or by scanning it for descendents of TestCase.

 - selectively navigate Kate, Idle or Vim to
        - test failure points
        - syntax errors
        - test suites
        - test cases
        - testee classes
        - testee methohs
        - a debugging trace statement

 - arbitrarily complex command lines now supported

 - non-heinous documentation in one handy Web page

Put them together, and you select the test in question, keep your keyboard 
focus in your editor with your digits on the home row and your hands off 
the mouse, hit Save over and over again, perceive the Green Bar, and 
achieve Flow. 

Satisfaction guaranteed or triple your money back.

-- 
  Phlip
   http://www.greencheese.org/PhilosophyBrethrenThree
  --  It's a small Web, after all...  --



From tdelaney at avaya.com  Thu Jun 13 20:43:24 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 14 Jun 2002 10:43:24 +1000
Subject: Why does Python mix OO concepts and non OO concepts for opera
	 tion s  on basic types?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A389@natasha.auslabs.avaya.com>

> From: holger krekel [mailto:pyth at devel.trillke.net]
> Delaney, Timothy wrote:
> > > From: holger krekel [mailto:pyth at devel.trillke.net]
> > > 
> > > probably. With python 1.6 many methods of the string *module* were
> > > included included directly with the *type* string ('str'). Most 
> > > python developers think that the string module should be 
> deprecated 
> > 
> > I think "most" is a little bit strong. The general feeling 
> that I get seems to be pretty much split down the middle.
> 
> maybe. IIRC GvR is in favor of deprecating it (but not any time soon).
> He does have not-to-be-underestimated influence on such issues :-)

My apologies. I misread "python developers" as "developers using python"
rather than "people developing python".

You do appear to be right concerning the python developers ... and GvR
counts somewhat ...

Tim Delaney




From dsavitsk at e-coli.net  Sun Jun 23 01:44:02 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sun, 23 Jun 2002 05:44:02 GMT
Subject: python cgi + IIS
References: <3D0F1813.3010304@gmx.de> <x0NP8.165989$Fp1.91529@atlpnn01.usenetserver.com> <Y4OP8.46694$GY.14566397@twister.nyroc.rr.com>
Message-ID: <CedR8.1714$Kt4.681454@newssrv26.news.prodigy.com>

very nice.  i don't actually use iis on that computer.  i opened it long
enough to take a screen shot, but, on that computer anyway, i use apache.

oddly enough, the e-coli.net site does run on IIS, but not on my computer.

-d


"George Hester" <hesterloli at hotmail.com> wrote in message
news:Y4OP8.46694$GY.14566397 at twister.nyroc.rr.com...
> Have you noticed this guy/gal has errors in their IIS?
>
> http://www.e-coli.net/pyiis_server.html
>
> --
> George Hester
> _________________________________
> "Steve Holden" <sholden at holdenweb.com> wrote in message
> news:x0NP8.165989$Fp1.91529 at atlpnn01.usenetserver.com...
> > "Axel Grune" <NineOfSix at gmx.de> wrote in message
> > news:3D0F1813.3010304 at gmx.de...
> > > Hi, I'm trying to get my IIS (Win2k Pro) execute my python scripts.
> > > I added the 'App Mapping' info for '.pycgi' files in the properties of
> > > IISAdmin (Computermanagement) as I read in the internet. The string
> > > for the executable file is 'D:\Python22\python.exe -u %s %s' (the path
> > > is valid).
> > > But when i try to call it within my browser, it just displays it like
a
> > > normal text file instead of executing it.
> > >
> > >
> > > the script (its called test.pycgi):
> > > #!/usr/bin/python
> > >
> > > print "Content-Type: text/plain\n\n"
> > > print "Hallo Welt"
> > >
> > >
> > > has anyone an idea?
> > >
> > You have almost certainly made a typo, or you didn't apply the changes
to
> > the directory you are trying to run the CGI from. I just followed the
steps
> > at
> >
> >     http://www.e-coli.net/pyiis.html
> >
> > to install .cgipy as a Python-executing extension, and everything
worked --
> > the following program printed out a complete list of environment
variables.
> > #
> > # Python CGI script
> > #
> > import os
> > print """Content-Type: text/html
> >
> > <html>
> > <head>
> >   <title>Python Web Page</title>
> > </head>
> > <body>
> > """
> >
> > for k, v in os.environ.items():
> >     print "%s=%s<br>\n" % (k, v)
> >
> > print """
> > </body>
> > </html>
> > """
> >
> > Personally I'd rather run CGI than use ASP under IIS, so I hope you can
get
> > it to work. Clearly at the moment IIS isn't recognising the .cgi
extension
> > as needing to run the Python interpreter.
> >
> > regards
> > -----------------------------------------------------------------------
> > Steve Holden                                 http://www.holdenweb.com/
> > Python Web Programming                http://pydish.holdenweb.com/pwp/
> > -----------------------------------------------------------------------
> >
> >
> >
> >
> >
>
>




From monsegur at mad.scientist.com  Tue Jun  4 14:10:09 2002
From: monsegur at mad.scientist.com (Xavier Monsegur)
Date: Tue, 04 Jun 2002 13:10:09 -0500
Subject: using getopt
Message-ID: <20020604181009.33890.qmail@mail.com>

I can't wait -- that would be marvelous... getopt should have been eradicated long ago.

----- Original Message -----
From: Mark Hammond <mhammond at skippinet.com.au>
Date: Tue, 04 Jun 2002 01:53:38 GMT 
To: python-list at python.org
Subject: Re: using getopt


> Fernando P?rez wrote:
> 
> 
> > getopt sucks. Period. Take a look at DPyGetOpt, which implements the 
> > Getopt::Long perl facilities. It's very very nice. I've also heard good 
> > things about optik, but haven't used it myself.
> 
> It looks like optik will be incorporated into the core Python library 
> (but possibly inside the getopt module).  However it actually appears, 
> using optik now will almost certainly make your future transition to the 
> standard library simpler.
> 
> Mark.
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 

-- 
_______________________________________________

Sign-up for your own FREE Personalized E-mail at Mail.com

http://www.mail.com/?sr=signup






From ngiven at lanl.gov  Tue Jun 25 10:59:29 2002
From: ngiven at lanl.gov (Nathan Given)
Date: Tue, 25 Jun 2002 08:59:29 -0600
Subject: Integrals with Python
Message-ID: <3D188551.6BA00DED@lanl.gov>

Hello All,

I'm somewhat familiar with MATLAB, and I'm trying to move completely
over to python... and I was wondering if there is a function out there
that will compute integrals for me...

here's what I've found so far...

http://pylab.sourceforge.net/

but I'm not sure what I need off that page...

I already have Numeric... what else do I need?

Thanks!
--
Nathan


From claird at starbase.neosoft.com  Wed Jun  5 11:41:03 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 5 Jun 2002 10:41:03 -0500
Subject: writing a chat program
References: <d4c18195.0206040840.7ece1b8c@posting.google.com>
Message-ID: <B31F43FB0BB239E3.3B6F14D29E20D79B.2DA67517FEF82F79@lp.airnews.net>

In article <d4c18195.0206040840.7ece1b8c at posting.google.com>,
Shaun Whelden <shaunw at liberateschool.com> wrote:
			.
			.
			.
>I'm writing this on a Windows machine, so curses is out of the
>question.  Also, I want this to be a text based program, so I don't
			.
			.
			.
There *are* implementations of curses for Windows; does that help?
-- 

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


From shaunw at liberateschool.com  Tue Jun  4 12:40:33 2002
From: shaunw at liberateschool.com (Shaun Whelden)
Date: 4 Jun 2002 09:40:33 -0700
Subject: writing a chat program
Message-ID: <d4c18195.0206040840.7ece1b8c@posting.google.com>

Hi all,
I'm trying to throw together a simple text-based chat program in
Python.  Here's the problem I have right now: I need a way to allow
the user to be entering in a message they want to send, while at the
same time waiting for any data that may come through the socket.  If
data does come through the socket, I need the user to be able to
continue typing while the data is printed to the screen.

I'm writing this on a Windows machine, so curses is out of the
question.  Also, I want this to be a text based program, so I don't
want to have to mess around with Tkinter or any other forms of GUI
unless I have to.  I'm pretty sure that I need to do something
involving thread.py and/or threading.py, but I've never used a thread
before in my life.  Can anyone help?

Shaun Whelden
shaunw at liberateschool.com


From BPettersen at NAREX.com  Tue Jun 18 15:23:41 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 18 Jun 2002 13:23:41 -0600
Subject: Python to MS SQL Server
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158EDF@admin56.narex.com>

> From: Mark McEahern [mailto:marklists at mceahern.com] 
> 
> > I am interested in using Python in a MS SQL Server 
> > environment, but do
> > not see any database modules out there, especially for 
> > native versus 
> > using ODBC. Is there such a module?
> 
> This is a total guess, because I haven't done this myself:
> 
> Have you tried using win32com to use ADO?

AFAIK, there is no Python database module that goes directly to MS SQL
Server db lib. ADO does work however, so that's certainly an option.
Personally, I would probably investigate e.g. mxODBC first (the
interface is much easier to use from Python).

-- bjorn




From BPettersen at NAREX.com  Thu Jun 20 18:52:35 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 20 Jun 2002 16:52:35 -0600
Subject: Windows versions of Python---pros and cons?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E201922151A8@admin56.narex.com>

> From: Chris Barker [mailto:Chris.Barker at noaa.gov] 
[snip]
> 
> On that note: does anyone know if there is a way to turn a 
> Python script into something that acts like an application on 
> Windows...without resorting to Py2exe and the like. What I 
> want is something just like:
> 
> chmod +x scriptname
> and a #! line at the top 
> 
> on Unix. I want it to work on the windows command line.

If you use the ActiveState version (not sure about the python.org
version) this should work out-of-the-box. Just type the name of the
script, you can even leave out the .py extension. Someone else will
probably tell you how it works, I just know it's got something to do
with registering the .py extension as an executable extension in the
registry...

-- bjorn




From jeremy at alum.mit.edu  Tue Jun 18 00:03:01 2002
From: jeremy at alum.mit.edu (Jeremy Hylton)
Date: 17 Jun 2002 21:03:01 -0700
Subject: urllib2 getting a webapage authenticating proxy...
References: <3D0DD287.7B221A42@a.com>
Message-ID: <b0f083db.0206172003.45d4ddbf@posting.google.com>

Adam <a at a.com> wrote in message news:<3D0DD287.7B221A42 at a.com>...
> this code doesnt work for me... i have replaced the proxy details for
> security... can anyone help... as i havent been able find any helpfull
> websites/examples for urllib2... is urllib2 what i realy want to use for
> this...
> 
> import urllib2
> 
> def get_page(url):
>  proxy =
> urllib2.ProxyHandler({'http':'http://userid:pass at proxy.com:port/'})
>  opener = urllib2.build_opener(proxy)
>  site = urllib2.urlopen(url).read()
>  print(site)
> 
> get_page('http://www.python.org/')

There is a SF patch that should fix this problem.  See
http://sourceforge.net/tracker/index.php?func=detail&aid=527518&group_id=5470&atid=305470
If it works, you might add a comment saying that it worked for you.

I expect this will be integrated in Python CVS soon, and should get
into future bug fix releases.

Jeremy


From gmcm at hypernet.com  Wed Jun 12 08:07:32 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 12 Jun 2002 12:07:32 GMT
Subject: filecmp.cmpfiles on directories
References: <3D06491B.302@excite.com>
Message-ID: <Xns922B52DB6F313gmcmhypernetcom@199.171.54.215>

Stormin wrote:

> I am trying to walk a directory structure and compare directories for 
> duplicates.  For some reason I get the filecmp to work on individual 
> files, but not on directories.

> example:
> DOES NOT WORK:
> dups=filecmp.cmpfiles("C:\comp1","c:\comp2",common)
> print dups,"||",common
> result:
> ([],[],[])||

For each filename in common, cmpfiles will compare the files
with that basename in the two directories. What you want
is:
 x = filecmp.dircmp("C:\\comp1","c:\\comp2")
 print x.same_files
 print x.diff_files

which will figure out "common" and call cmpfiles for you.

-- Gordon
http://www.mcmillan-inc.com/


From cmonahan at yahoo.com  Tue Jun 11 18:10:37 2002
From: cmonahan at yahoo.com (Collin Monahan)
Date: Tue, 11 Jun 2002 15:10:37 -0700 (PDT)
Subject: lists and sequences
Message-ID: <20020611221037.36421.qmail@web13303.mail.yahoo.com>

What is a python list? Is it a linked list or a random access structure
in contiguous memory? What type of situations should be avoided when
using it? E.g. when does n^2 time happen using it? What sorting
algorithm is used by its sort member function?

Are the answers to these questions the same for Jython? If there is a
good situation to use a Java collection from Jython because of above
considerations, how bad are the constant factors which arise from doing
so? That is, how preferable is it to stick with the python-language
structures when using Jython?

I'm not on the list so please copy me.

Thank you
Collin Monahan

__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com




From ffjhenon at club-internet.fr  Wed Jun 26 17:31:28 2002
From: ffjhenon at club-internet.fr (Fabien =?iso-8859-1?Q?H=E9non?=)
Date: Wed, 26 Jun 2002 23:31:28 +0200
Subject: continuous redirection of stdout & stderr to Tkinter Text
Message-ID: <3D1A32B0.CA7D3618@club-internet.fr>

I am using Tkinter as an editor and a launcher (under Linux) for a
raytracer (POV-RAY)

Once started, the raytracer does two things in the same time :
- it displays the image being raytraced
- it spits out information to the console (ie : the line which is being
raytraced). 
I get the image displayed but not the redirection of stdout

That is why I use the root.after function to (try to) update every 10th
second.


With the script below, I get the image displayed, but not the stdout (or
stderr).

What am I doing wrong ?

To the pipe-redirectors gurus : Help !

8<---------snip------------------------->8
from Tkinter import *
import sys, os
import popen2

root = Tk()

class StringVarFile:
    def __init__(self,stringVar,window):
        self.__newline = 0
        self.__stringvar = stringVar
        self.__window = window

    def write(self,s):
        new = self.__stringvar.get()
        for c in s:
            if c == '\n':
                self.__newline = 1
            else:
                if self.__newline:
                    new = ""
                    self.__newline = 0
                new = new+c
        self.set(new)

    def set(self,s):
        self.__stringvar.set(s)
        self.__window.update()

    def get(self):
        return self.__stringvar.get()


class mainWin:
    def __init__(self,tkRoot):
        self.tkRoot=tkRoot
        self.createWidgets()
        return None

def runpov():
    cmd='megapovplus +i1.pov +w320 +h240 +dgt +v -f +p'
    run=popen2.Popen3(cmd)
    upd()


def upd():
       resultat.insert('1.0',statusVar))
       root.after(100,upd())




mframe=Frame(root)

bouton=Button(mframe,text="Feu !", command=runpov)
bouton.pack()
statusVar = StringVar()
#resultat=Label(mframe, width=60, height=10,justify=LEFT, anchor=W,
textvariable=statusVar)
resultat=Text(mframe)
resultat.pack(fill=X,expand=Y)
mframe.pack(expand=1, fill=BOTH)

status = StringVarFile(statusVar,root)
sys.stderr=status

root.mainloop()
8<---------snap------------------------->8



From peter at engcorp.com  Wed Jun 26 23:56:09 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 26 Jun 2002 23:56:09 -0400
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com>
Message-ID: <3D1A8CD9.C8B8EC28@engcorp.com>

candiazoo at attbi.com wrote:
> 
> I assume some sort of block i/o or preallocating a large block of file space
> would be the best way to do it?  I am outputting about 700,000 records to a text
> file and my program only spits out about 20 records per second.  This is using
> standard file open and write...

You're not opening and closing the file for each record that is written, are you?

Can you post more detail on the amount of data in a record, and the format?

Why not profile the code using the profile.py module and actually see
where the problem is?

-Peter


From matt at mondoinfo.com  Sun Jun 23 21:36:40 2002
From: matt at mondoinfo.com (Matthew Dixon Cowles)
Date: Mon, 24 Jun 2002 01:36:40 GMT
Subject: Problems with Pmw
References: <QUtR8.20$U4.2298@news.uchicago.edu>
Message-ID: <slrnahctt8.1cn.matt@happy-hour.mondoinfo.com>

On Sun, 23 Jun 2002 19:35:33 -0500, Mark Rivers
<rivers at cars.uchicago.edu> wrote:

> In a previous post I mistakenly attributed one problem to Pmw.Blt, while
> subsequent testing has shown that the problem has nothing to do with Blt, it
> is a problem (?) with Pmw.
> 
> The problem is illustrated by the following short script:
> ########################################
> from Tkinter import *
> import Pmw
> t1 = Tk()
> t2 = Tk()
> l = Label(t2, text='A label'); l.pack()
> b = Button(t2, text='Press me'); b.pack()
> e = Pmw.EntryField(t2, labelpos=W, label_text='Enter something'); e.pack()
> ########################################

[. . .]

> I think it should be legal to create multiple Tk() windows in an
> application, so this seems like a serious bug?

The usual way to create another window is with the Toplevel() call.  I
don't know whether the bug is in Tkinter's allowing a second Tk() call
or Pmw's assuming that no one ever makes a second Tk() call but you
shouldn't have a problem if you use Toplevel().

Regards,
Matt



From mgilfix at eecs.tufts.edu  Mon Jun 10 19:21:25 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Mon, 10 Jun 2002 19:21:25 -0400
Subject: wxPlotCanvas documentation?
In-Reply-To: <LlaN8.52033$ks5.2050662@news2.east.cox.net>; from thehaas@binary.net on Mon, Jun 10, 2002 at 11:10:35PM +0000
References: <mailman.1023733395.5699.python-list@python.org> <LlaN8.52033$ks5.2050662@news2.east.cox.net>
Message-ID: <20020610192125.C5683@eecs.tufts.edu>

On Mon, Jun 10 @ 23:10, thehaas at binary.net wrote:
> > 3. How do I find anything in the documentation?
> 
> > My inability to find my wxButt in the wxDark with both
> > wxHands (with or without a wxFlashlight) is not
> > limited to PlotCanvas, so I'm hoping that the problem
> > is on my end.  The wxPython website, as far as I can
> > tell, only has one documentation link, and it's a link
> > to wxWindows, not the wxPython objects that wrapper
> > it.  There's a sample code section, but every last one
> > of the examples is marked "To Be Written When I Have
> > More Time."

  ^^^ This is very funny.

> I love wxPython, but the initial learning curve is quite steep.  Once you
> get past the very initial stage, though, it grows on you.

  I really agree with this statement. The initial learning curve is
a little high but then you become quite productive. My major problem
starting with wxPython is that the sizers didn't work quite the way I
expected. I suggest looking through the many tutorials about sizers
(after I bitched about them plenty, they actually went and wrote one
:) ). But if you have experience with autolayout mechanisms from
pygtk, throw them out the window before you start with wxPython and
that might help some. The wxPython ones just behave differently, and I
found that a little confusing at first.

  But once you get the hang of it, wxPython is great.

                    -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From shalehperry at attbi.com  Sat Jun  8 11:16:40 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Sat, 08 Jun 2002 08:16:40 -0700 (PDT)
Subject: Removing ^M
In-Reply-To: <200206081448.g58Empj11746@localhost.localdomain>
Message-ID: <XFMail.20020608081640.shalehperry@attbi.com>

> 
> There's a program called dos2unix that will do this. From your mail
> headers it looks to me like you have a Unix/Linux system in which case
> it may already be installed. If not do a google search on dos2unix.
> 

another possibility is fromdos (and its brother todos).

As Chris Gonnerman pointed out ^M is '\r' or 'carriage return'.  Under Windows
every line ends in '\r\n', under Unix it is just '\n' and I think Macs only use
'\r'.  This is usually hidden from you because python, the C libs, etc all know
how to translate behind the scenes.

If you have some one at work (or a friend) ftp'ing files have them set the mode
to ASCII before transfer and FTP will do the conversion for you.




From Chris.Barker at noaa.gov  Thu Jun 20 14:09:28 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Thu, 20 Jun 2002 11:09:28 -0700
Subject: How to represent the infinite ?
References: <3d11ee9b$1@news.swissonline.ch> <20020620193230.53d58f8f.christophe.delord@free.fr>
Message-ID: <3D121A29.A733609F@noaa.gov>

Christophe Delord wrote:
> There is another simple solution. Just use a float that is bigger than any 64-bit float. For example 10^1000. This float has a special encoding meaning +oo !

this is part of the IEEE 754 standard for floating point computation
(http://cch.loria.fr/documentation/IEEE754/) It is not guaranteed to be
implimented by Python (someone please correct me if I'm wrong). whether
this works is a function of the machine, compiler and library used to
compile Python. I'm not sure how common it is, but if you want your code
portable, it may not be reliabel on all machines, which is too bad,
because IEEE 754 is pretty darn good standard.

BTW: I'd love to see literals for Inf, -Inf and NaN in Python.

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From gerhard.haering at gmx.de  Fri Jun 21 07:07:37 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 21 Jun 2002 13:07:37 +0200
Subject: I'd give up Perl tomorrow if only...
In-Reply-To: <pan.2002.06.21.10.48.42.651127.4527@NOSPAM.optushome.com.au>
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au> <mailman.1024653868.4542.python-list@python.org> <pan.2002.06.21.10.48.42.651127.4527@NOSPAM.optushome.com.au>
Message-ID: <20020621110737.GA13324@lilith.my-fqdn.de>

* David Mitchell <djmitchell at NOSPAM.optushome.com.au> [2002-06-21 20:48 +1000]:
> ASCII<->EBCDIC conversion is pretty easy to implement in just about any
> language, but it's also easy to make a typo somewhere when you're typing in
> lots and lots of ASCII codes.  I know; I've done it ;->

No need to do it yourself. In Python 2.2+:

>>> u"foo".encode("ebcdic_cp_us")
'\x86\x96\x96'

Of course, if you have ascii, you'll have to construct the unicode
string first: unicode("foo", "ascii").encode(...)

> However, when speed of development is absolutely crucial, Perl can't
> be beat.

For me, nothing beats Python. I don't know Perl (I've tried to learn it,
but it was too frustrating). And I don't believe Perl were more
efficient for me. The more terse code argument doesn't count, because
I'm a fast typist :-)

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From peter at engcorp.com  Tue Jun 18 00:12:50 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 18 Jun 2002 00:12:50 -0400
Subject: Listing all available (not just loaded) modules
References: <aem9om$888fb$1@ID-140151.news.dfncis.de>
Message-ID: <3D0EB342.43DB0FE3@engcorp.com>

Michael Steuer wrote:
> 
> I've been searching the newsgroup for information on this but so far have been
> unsuccessful:
> 
> Is there a way to dynamically list all available modules (not just those that
> are already loaded) in python? By available I mean those modules, which are
> installed on the machine executing python (in the lib path(s)).

http://snakefarm.org/  ?

-Peter


From cliechti at gmx.net  Wed Jun 26 14:52:28 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 26 Jun 2002 20:52:28 +0200
Subject: printing (again!!)
References: <afcusp$66r$1@newsg2.svr.pol.co.uk>
Message-ID: <Xns9239D4CAD9DF0cliechtigmxnet@62.2.16.82>

"Geoff Tarrant" <geoff at galtar.co.uk> wrote in
news:afcusp$66r$1 at newsg2.svr.pol.co.uk: 

> I want to use Python as the language in a Computing course that I
> teach. However, I can't find a way that is relatively straightforward
> to send data to a printer in a Windows environment.  If I can't find a
> way around this, I will have to revert to Delphi or VB.  Please help. 
> I like the idea of sending 20 or so Python programmers to University
> each year. 

you can try different things.

1. use reportlab and generate PDF. save the trees, make achiving easy. its 
has many more uses, e.g on the fly PDF for CGI etc.

2. use the printing framework of a GUI. i've successfuly printed with 
wxPython. win32all has printing support too, through the standard windows 
API. win32all is a defacto standard for python on windows users anyway.

i prefer 1) as it's more versatile, platform indepent and just cool.

chris

-- 
Chris <cliechti at gmx.net>



From candiazoo at attbi.com  Mon Jun 24 21:18:49 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Tue, 25 Jun 2002 01:18:49 GMT
Subject: Help with MySQLdb insert with data extracted from odbc <-- Oracle
References: <3d13d5fb.651829078@netnews.attbi.com> <mailman.1024712368.16609.python-list@python.org> <3d147e64.694941703@netnews.attbi.com> <IKMR8.5776$mF.3188@atlpnn01.usenetserver.com>
Message-ID: <3d17c4a5.75241265@netnews.attbi.com>

In a year or two, God willing, we'll have no need for this particular program!!
:)  You are, however, absolutely right and it would probably be questioned in a
code review.  Thanks!

Mike


From guy at lightwork.co.uk  Thu Jun 27 08:31:48 2002
From: guy at lightwork.co.uk (Guy)
Date: 27 Jun 2002 05:31:48 -0700
Subject: sys.argv[0] - Returns path and filename.
Message-ID: <c07d0edc.0206270431.5f2709b4@posting.google.com>

Hi

I work on multiple platforms, at the moment, I have python installed
on my win32 box which I mainly work on, a MacOSX box and a linux box
which has a very old version installed on it.

The only major problem that I have come across at the moment is the
argv[0], on my win32 box it retruns the path and a filename, on the
mac osx and the linux only the filename are returned.

How do I get the current path of the script I am running, on the mac
osx and the linux box ? (Any work around would be great, I don't want
to have to hardcode the path into the script, or even give the path as
an argument,I want the script to know where it is.)

I've tryed searching for this but can't find it anywhere. 

Example script called arguments.py :
import sys
print sys.argv[0]

Result Win32 :
>>> D:\TestArg\arguments.py

Result Mac OSX : 
MacOSX% python arguments.py
arguments.py

Result Linux : 
linux% python arguments.py
arguments.py


From bokr at oz.net  Fri Jun 28 18:46:42 2002
From: bokr at oz.net (Bengt Richter)
Date: 28 Jun 2002 22:46:42 GMT
Subject: generate all combinations of a list (with variable output length)
References: <mailman.1025295140.24921.python-list@python.org>
Message-ID: <afip0i$8c3$0@216.39.172.122>

On Fri, 28 Jun 2002 15:09:47 -0500, "Mark McEahern" <marklists at mceahern.com> wrote:

>> i'm looking for a way to generate all possible combinations of all items
>> in a list. list = ["A","D","$","5","R"]
>
>i don't have a solution, but some questions to clarify:
>
>if you think of your list as comprising the digits, then you're asking for
>an n-length string composed of all possible digits where the digits include:
>
>  A, D, $, 5, R, and None (the representation of the latter is '')
>
>so, for n = 10, this is a possible value:
>
>  AAAAAAAAAA
>
>as well as:
>
>  [          ]
>
>omit the brackets.
>
>True?
>
>anyway, this might get you started:
>
>  http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/105962
>
>// m
>
Or, implementing your base-5 digits idea, and leaving out '':

 >>> def adD5r(n):
 ...     codes='AD$5R' # <=> 01234
 ...     out = []; m = n
 ...     while 1:
 ...         n,d = divmod(n,5)
 ...         out.append(codes[d])
 ...         if not n:break
 ...     out.reverse()
 ...     return '%d:%s' % (m,''.join(out))
 ...
 >>> [adD5r(5**n) for n in range(5)]
 ['1:D', '5:DA', '25:DAA', '125:DAAA', '625:DAAAA']
 >>> [adD5r(5**n+1) for n in range(5)]
 ['2:$', '6:DD', '26:DAD', '126:DAAD', '626:DAAAD']
 >>> [adD5r(n) for n in range(26)]
 ['0:A', '1:D', '2:$', '3:5', '4:R', '5:DA', '6:DD', '7:D$', '8:D5', '9:DR', '10:$A', '11:$
 D', '12:$$', '13:$5', '14:$R', '15:5A', '16:5D', '17:5$', '18:55', '19:5R', '20:RA', '21:R
 D', '22:R$', '23:R5', '24:RR', '25:DAA']

The smallest and biggest 10-"digit" numbers would be
 >>> adD5r(0)
 '0:A'
(I'll leave it as an exercise for the OP to prefix leading "zeroes" to desired print width
for that set of strings ;-)

 >>> adD5r(5**10-1)
 '9765624:RRRRRRRRRR'

generated analogously to
 >>> 10**10-1
 9999999999L

Regards,
Bengt Richter


From deathtospam43423 at altavista.com  Sat Jun  8 03:19:52 2002
From: deathtospam43423 at altavista.com (netvegetable)
Date: 8 Jun 2002 07:19:52 GMT
Subject: Newby: How do I strip HTML tags?
References: <Xns92278CC566FEAusYourStandingInIt@130.133.1.4> <3d00e1b1.47853937@news.t-online.de> <mailman.1023469751.21489.python-list@python.org>
Message-ID: <Xns9227ABF0D881EAusYourStandingInIt@130.133.1.4>

Andy McKay <amckay at merlintechnologies.com> wrote in 
news:mailman.1023469751.21489.python-list at python.org:

>> This is a quite straight forward function: 
> 
> This of course assumes that there are no "<" or ">" anywhere else in
> attributes etc...
> 
> I would recommend using HTML Parser to get an accurate representation, or
> even stripogram which is pretty good (note it doesnt need Zope at all),
> although the html2text isnt perfect:
> 
>>>> from stripogram import html2text html2text("<b>This is a
>>>> <i>test</i></b>") 'This is a test' 

Thanx.

Is there an "HTMLParser Tutorial for Dummies" anywhere on the web? 


-- 
netvegetable at excite.com


From gerhard.haering at gmx.de  Fri Jun 28 15:12:19 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 28 Jun 2002 19:12:19 GMT
Subject: distutils/boost question
References: <afhp5p$fqf0d$1@hades.rz.uni-sb.de> <slrnahou7h.140.gerhard@lilith.my-fqdn.de> <afi7jf$5hh$00$1@news.t-online.com>
Message-ID: <slrnahpd8j.h0.gerhard.haering@lilith.my-fqdn.de>

Achim Domma wrote in comp.lang.python:
> "Gerhard H?ring" <gerhard at bigfoot.de> wrote:
>> The easiest solution is to statically link in the Boost.Python library.
> 
> With boost.python V2 dynamic linking will be a requirement. Isn't
> there another solution?
 
On Windows, the DLL search rules are such that putting the support DLL
in the same directory as the .pyd does help. If you get that working
with distutils, please drop me a note.

On Unix, you need to put the shared library in a directory that is in
the dynamic loader's search path. /usr/lib, /usr/local/lib, ... That's
AFAIK currently not possible with distutils. It's no problem when
using your favourite Unix package format. Writing a SPEC file, for
example, is pretty straightforward, when using the documentation and
existing SPEC files as a template.

Another ugly solution is to write a little wrapper script that sets
the LD_LIBRARY_PATH environment variable appropriately before running
Python. That's the quick and dirty fix.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From andreas.ames at Tenovis.com  Thu Jun 27 09:16:01 2002
From: andreas.ames at Tenovis.com (Andreas Ames)
Date: 27 Jun 2002 15:16:01 +0200
Subject: Interactive session via tcp
In-Reply-To: <20020627120602.3659.37119.Mailman@mail.python.org>
References: <20020627120602.3659.37119.Mailman@mail.python.org>
Message-ID: <uadpg2772.fsf@tenovis.com>

Hi,

> Sorry, if this if wrong, I read your post quickly.
> Doesn't "telnet jabber_port" do what you want?

Nope, it doesn't.  My main problem is the python (interpreter) side.
And while it's possible to jabber via telnet it's not exactly what I
call fun (I've tried it).

cheers,

Andreas




From pj at engr.sgi.com  Tue Jun 18 19:46:58 2002
From: pj at engr.sgi.com (Paul Jackson)
Date: 18 Jun 2002 23:46:58 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <D7NP8.46301$GY.14487785@twister.nyroc.rr.com>
Message-ID: <aeogpi$26rrb$1@fido.engr.sgi.com>

> Don't worry I won't return here.

Thank-you, sir.

-- 
-- 
                          I won't rest till it's the best ...
                          Programmer, Linux Scalability
                          Paul Jackson <pj at sgi.com> 1.650.933.1373


From skip at pobox.com  Thu Jun 20 11:08:26 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 20 Jun 2002 10:08:26 -0500
Subject: MySQLdb question
In-Reply-To: <m2sn3inkee.fsf@mother.paradise.lost>
References: <aesi66$cgb$1@sunsite.icm.edu.pl>
        <m2sn3inkee.fsf@mother.paradise.lost>
Message-ID: <15633.61418.44706.32731@localhost.localdomain>

    John> Use the %s thingie:

    John> c.execute('SELECT spam, eggs, sausage FROM %s
    John>            WHERE price < %s', (table_name, max_price))

I'd amend that slightly:

    c.execute('SELECT spam, eggs, sausage FROM %s
               WHERE price < %%s'%table_name, (max_price,))

or

    stmt = ('SELECT spam, eggs, sausage FROM %s
            WHERE price < %%s' % table_name)
    c.execute(stmt, (max_price,))

to make the two substitution ops clearer.

For variable substitution I recommend people always use the DB API argument
quoting facility.  In this case it doesn't matter since presumably max_price
won't need escaping.  For string args it definitely would though.

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html




From steele at cs.brandeis.edu  Sat Jun  1 06:36:01 2002
From: steele at cs.brandeis.edu (Oliver Steele)
Date: 1 Jun 2002 03:36:01 -0700
Subject: extended list comprehensions
References: <f32195af.0205261316.2482d4ac@posting.google.com>
Message-ID: <92f9b8a8.0206010236.1ac5e3f2@posting.google.com>

fxj at hotmail.com (F. Jamitzky) wrote in message news:<f32195af.0205261316.2482d4ac at posting.google.com>...
> It would be great to have something like a list comprehension for the
> reduce function. It would work in the same way as the comprehension
> for the map function, which is:
> 
> [foo(x) for x in xs]      <-is the same as->       map(foo,xs)
> 
> and maybe it could be written as:
> 
> {y=y+x for x in xs}      <-would be->         reduce(operator.add,xs)
> 
> what do the experts think about that ? 
> 
> ferdinand

What about using ellipsis?  These could do foldl (reduce):
  ...+x for x in xs
  operator.add(..., x) for x in xs

This leaves an obvious extension for foldr (reduce from the right):
swap the identifier and the ellipsis:
  x+... for x in xs
  operator.add(x, ...) for x in xs

The bug in this (putting aside the question of whether it's worth
adding anything to the language) is that it suggests that ... is a way
to write curry (similar to (x+) and (+x) in Haskell), and then:
  [...+x for x in xs]
becomes ambiguous between a singleton list containing a reduction, and
a list of lambdas.  Any way to save this?


From dl at bestpriceeu.com  Sun Jun  9 09:03:35 2002
From: dl at bestpriceeu.com (Dan Larsen)
Date: Sun, 9 Jun 2002 15:03:35 +0200
Subject: newbie system() question
Message-ID: <3d035180$0$7017$bc7fd3c@news.sonofon.dk>

Hi all :o)
Im trying to run a script on a FreeBSD 4.0 system.
It works as long as I start it like "python myprogram.py", but as soon as I
run it in the background, it is stopped, and I have to make it run in the
foreground to continue i.e. "fg %1"... I know it is a call to a system(""),
that makes it stop... The system() call runs a program that outputs some
text... how do I stop it doing so? It doesn't help telling it to do stuff
like "program >& /dev/null &"... Isn't there a function, where all output
from the command, is parsed to a variable?

Thanx in advance :o)
Dan Larsen




From paul at boddie.net  Mon Jun  3 13:01:51 2002
From: paul at boddie.net (Paul Boddie)
Date: 3 Jun 2002 10:01:51 -0700
Subject: Compiling python2.2.1 on redHat 6.1 - doesn't work
References: <bb06817.0205210533.d8d5ffb@posting.google.com> <23891c90.0205220343.4aa581c8@posting.google.com> <bb06817.0205230212.7b6a388c@posting.google.com> <23891c90.0205230743.7ea59255@posting.google.com>
Message-ID: <23891c90.0206030901.26402dd0@posting.google.com>

paul at boddie.net (Paul Boddie) wrote in message news:<23891c90.0205230743.7ea59255 at posting.google.com>...
>
> matthew.russell at securetrading.com (Matt Russell) wrote in message news:<bb06817.0205230212.7b6a388c at posting.google.com>...
> >
> > has anyone else had this problem? I would be very suprised if no one
> > else out there has not tried to compile Python2.2.1 on RedHat 6.1!!
> 
> I'll try and have a look in the next few days, at least.

OK. My suggestion can be found at the following location:

http://sourceforge.net/tracker/?group_id=5470&atid=305470&func=detail&aid=563954

That's patch #563954 in the Python project at SourceForge
(http://sf.net/projects/python).

Paul


From gabor at realtime.sk  Thu Jun  6 16:40:26 2002
From: gabor at realtime.sk (gabor)
Date: 06 Jun 2002 16:40:26 -0400
Subject: array, list, performance...
In-Reply-To: <_gKL8.314196$%u2.248214@atlpnn01.usenetserver.com>
References: <3CFF5F98.2060406@destiny.com>
	<mailman.1023373030.31871.python-list@python.org> 
	<_gKL8.314196$%u2.248214@atlpnn01.usenetserver.com>
Message-ID: <1023396026.7322.4.camel@wintermute.atriaky.sk>

On Thu, 2002-06-06 at 10:20, Steve Holden wrote:
> "Christopher Armstrong" <radix at twistedmatrix.com> wrote in message
> news:mailman.1023373030.31871.python-list at python.org...
> >
> > > Michael Chermside <mcherm at destiny.com> writes::
> > >   li[n] = x      assignment to the middle is O(len(li))
> >
> > This doesn't seem right. Shouldn't that be O(1), just like
> > access-time?
> >sorry, but i don't understand...

how can you use li[n] = x to INSERT an element into a list..
IndexError: list assignment index out of range
as far as i know:
if n<len(li) then it overwrites the existing value.
else you get an "IndexError: list assignment index out of range".

am i wrong?
thanks,
gabor






From fperez528 at yahoo.com  Thu Jun 20 15:35:24 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 20 Jun 2002 13:35:24 -0600
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <aer2fn$frc$1@peabody.colorado.edu> <3D122E79.A189C72E@astro.cornell.edu>
Message-ID: <aetapt$kld$1@peabody.colorado.edu>

Tom Loredo wrote:

> Fernando P?rez wrote:
>> 
>> If you're an old unix hand and want to preserve your sanity, you have two
>> options:
>> 
>> 1- put linux on that laptop and be happy.
> 
> Actually, I've reserved a big chunk of the disk for linux.  Just gotta
> decide which version (is there a "best" one for Python?  E.g., Redhat
> uses Python, but does their reliance on an old version still cause
> problems?).  I'll stay there when I'm developing for *me* on it!

Well, I'm a big fan of Mandrake over RedHat: more oriented towards 
'workstation' users than 'server' users, more out-of-the box tools for 
scientific use (lyx, gnuplot, octave, Numeric, Lapack,...). Plus redhat made 
the most boneheaded decision on the planet with the two-version python setup, 
which you'll have to duct-tape your way around.

_My_ (redhat fanatics, keep off :) simple experience so far is: everytime I 
need to get something done on a mandrake machine, it works in under 15 
minutes. Everytime I have to deal with a redhat machine, I'm pissed off after 
two hours. But YMMV, as always.


>> 2- if (1) is not an option for some reason, put cygwin on it.
> 
> Actually, I need to develop stuff that other Win32 users can use,
> some probably distributed via the McMillan installer (which I presume
> works as well with Active State as with the python.org version).
> So I'm stuck here....

Don't know if there's a misunderstanding here: cygwin is a unix-like 
environment under Windows. So at least you'll have a terminal worthy of the 
name, plus all the command-line tools you neeed on a day to day basis (grep, 
sed, awk, etc.). I don't use it (I simply don't use windows except in rare 
cases), but I've heard it's quite good and makes life under windows at least 
bearable.
 
>> If you have to work in the normal,
>> crippled windows environment, that laptop is going to suffer a painful
>> death after being hurled through the nearest window very soon :)
> 
> I've only had it two weeks, and only the power cord to the battery
> charger has saved it!  The fact that Dell shipped it with a defective
> install of Win98 did not help its cause.

Well, if you do have to live with windows, I'd recommend at least using XP. 
It's still an annoying crippled toy, but at least it _is_ much more stable 
than Win9x, and when things crash they typically don't take down the whole 
system with them. Don't expect unix-class stability, but it won't drive you 
nuts like win9x will.

Best,

f


From peter at engcorp.com  Mon Jun  3 21:43:44 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 03 Jun 2002 21:43:44 -0400
Subject: python and windows and pydoc and favorite editor == True
References: <un0ugnxa3.fsf@online.no> <adggk2$psv$00$1@news.t-online.com> <uznycc5om.fsf@online.no>
Message-ID: <3CFC1B50.E830AC46@engcorp.com>

Syver Enstad wrote:
> 
> "J?rgen Hermann" <jh at web.de> writes:
> 
> > "Syver Enstad" <syver-en+usenet at online.no> schrieb im Newsbeitrag
> > news:un0ugnxa3.fsf at online.no...
> > > Post if there is any interest.
> >
> > OK, post. :)
> >
> >
> 
> http://home.online.no/~syver-en/software.html/files/PyDocEditorStarter.py

Looks like this should be:

  http://home.online.no/~syver-en/files/PyDocEditorStarter.py

The other 404's me.

-Peter


From d_blade8 at hotmail.com  Thu Jun 20 14:46:10 2002
From: d_blade8 at hotmail.com (Lemniscate)
Date: 20 Jun 2002 11:46:10 -0700
Subject: Problems with mayavi, pyvtk
References: <7396d2b2.0206191025.3351a888@posting.google.com> <aeqme8$avd$1@peabody.colorado.edu>
Message-ID: <7396d2b2.0206201046.4f0530fa@posting.google.com>

Hey, thanks for the info.  Here's the only problem...  I still get
errors.  Here is what happens when I try to import vtkpython...

>>> import vtkpython
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
  File "C:\Python22\Lib\site-packages\VTK\vtkpython.py", line 29, in ?
    exec "from "+kit_name+" import *"
  File "<string>", line 1, in ?
ImportError: DLL load failed: The specified procedure could not be
found.
>>> import sys
>>> sys.path
['', 'C:\\py22', 'C:\\vtk4', 'C:\\Python22\\Lib\\site-packages\\VTK',
'C:\\Program Files\\MayaVi'...


The main problem, though, is that I just want MayaVi to work from
inside python, here is one of the errors I get when I try to import it
(per the manual).  I have pyvtk and it imports with absolutely no
problems, i can generate vtk files that open in the MayaVi.exe
program, I just can't use python to interface with MayaVi (I have the
MayaVi modules, not just the binary version, btw).  I think it has
something to do with the vtk installations.  I looked and it wants
'vtk'+kit+'Python' but all the zip files on the page you gave me (for
Python 2.2) don't have the files.  I was willing to go back to Python
2.1 (his file list for those files showed the files I needed) but the
Python2.1 link is no longer valid.  I also downloaded the vtkcore and
the vtkpython from the vtk page
(http://public.kitware.com/VTK/files/release/4.0/).  Also, I added the
following to my PythonPath (in order to access the dlls):  C:\Program
Files\vtk40\bin


Thanks a bunch,

Lem

>>> import mayavi
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
  File "C:\Python22\Lib\site-packages\mayavi\__init__.py", line 5, in
?
    import Common, Base.Objects
  File "C:\Python22\Lib\site-packages\mayavi\Base\__init__.py", line
19, in ?
    import Objects
  File "C:\Python22\Lib\site-packages\mayavi\Base\Objects.py", line
20, in ?
    import Tkinter, vtkpython, math, tkColorChooser
  File "C:\Python22\Lib\site-packages\VTK\vtkpython.py", line 29, in ?
    exec "from "+kit_name+" import *"
  File "<string>", line 1, in ?
ImportError: DLL load failed: The specified procedure could not be
found.



Fernando P?rez <fperez528 at yahoo.com> wrote in message news:<aeqme8$avd$1 at peabody.colorado.edu>...
> Lemniscate wrote:
> 
> > 
> > I just downloaded and installed the latest versions of pyvtk and
> > mayavi.??However,?when?I?try?to?follow?the?instructions?at
> > http://mayavi.sourceforge.net/docs/guide/c827.html#VIZ-DATA for using
> > mayavi, I get the ImportError as below.??For?the?like?of?me,?I?can't
> > seem to figure out why I don't have any vtkpython modules anywhere
> > (doing searches on the internet for the module hasn't been a big help
> > so far
> 
> Well, google's second hit on 'vtkpython' is 
> http://basic.netmeg.net/godzilla/, which sounds like what you want. At least 
> it should be a good start.
> 
> Cheers,
> 
> f


From =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?=  Tue Jun 25 04:46:24 2002
From: =?iso-8859-5?Q?=DF=D5=DD=D3=E3=D8=DD=D8=E1=E2=D0?= at =?iso-8859-5?Q?=DC=D0=D8=DB?=.=?iso-8859-5?Q?=DD=D5=E2?=.=?iso-8859-5?Q?=DC=DA?= (=?iso-8859-5?Q?=B4=D0=DC=F8=D0=DD_=B3=2E?=)
Date: Tue, 25 Jun 2002 10:46:24 +0200
Subject: Python daemon
Message-ID: <3d182dd6$1@news.mt.net.mk>

Hello, 

I'm writing a Python daemon on Unix and I need the folowing features:
 1. going to background, diassociate from controling tty (done)
 2. watchdog, the program should never be stuck
 3. restarting of the program if it fails, or the watchdog doesn't trigger
 4. Logging


The issue 1. I solved, it's on ActiveState's site. Issue 4. is not that
hard to do I guess. Any suggestions about 2 and 3 ?



-- 
??????

   "One world, one web, one program"  -- Microsoft promotional ad
         "Ein Volk, ein Reich, ein Fuehrer"  -- Adolf Hitler



From sholden at holdenweb.com  Mon Jun  3 07:59:05 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 3 Jun 2002 07:59:05 -0400
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <06da01c208f4$d69003c0$6601a8c0@boostconsulting.com> <200206010121.g511LxX19223@pcp742651pcs.reston01.va.comcast.net> <073501c20960$23f367e0$6601a8c0@boostconsulting.com> <200206011334.g51DY1D21669@pcp742651pcs.reston01.va.comcast.net> <07d801c20970$cdebe050$6601a8c0@boostconsulting.com> <mailman.1022940122.3421.python-list@python.org> <Id5K8.171932$Po6.340972@rwcrnsc52.ops.asp.att.net>
Message-ID: <oWIK8.179611$%u2.31494@atlpnn01.usenetserver.com>

"Emile van Sebille" <emile at fenx.com> wrote ...
> <jepler at unpythonic.net> wrote:
> > IMO the following is an odder behavior:
> > >>> t
> > ([1], [2], [3])
> > >>> t[1] += [3]
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> > TypeError: object doesn't support item assignment
> > >>> t
> > ([1], [2, 3], [3])
> >
>
> Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
> >>> class myTuple(tuple):
> ...     def __setitem__(self, attr, value):
> ...             print 'in setitem:', attr
> ...
> >>> a = myTuple(([1],[2],[3]))
> >>> a
> ([1], [2], [3])
> >>> a[1] += [4]
> in setitem: 1
> >>> a
> ([1], [2, 4], [3])
> >>>
>
>
> Here's another that's differently wrong:
> >>> a = ('aaa',)
> >>> a[0]+='bbb'
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> TypeError: object doesn't support item assignment
> >>> a
> ('aaa',)
> >>>


Why do you consider either of these examples wrong? The whole point of
tuples is that if d is a dictionary, then d[t1] == d[t2] if, and only if, t1
== t2. Both of these examples are trying to make changes that will modify an
(immutable) tuple. There's every reason to treat these modifications as
illegal, since a before the fiddling is not equal to a after the fiddling,
but tuples aren't mutable so such changes are illegal.

regards
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From Joal.Heagney at student.gu.edu.au  Sun Jun 23 16:53:26 2002
From: Joal.Heagney at student.gu.edu.au (Joal Heagney)
Date: Mon, 24 Jun 2002 06:53:26 +1000
Subject: bloody freshman question regarding strings
References: <1024821351.3424.12.camel@dorian.blaue-elise.net> <mailman.1024830689.5399.python-list@python.org>
Message-ID: <3D163546.8010707@student.gu.edu.au>

Christian Guenther wrote:
> Wow, that was quick. Many thanks for the responses. I suppose 
> I have to learn a lot about pythons sysntax and way it handles 
> objects and stuff.
> 
> I always had bit of a problem understanding OO-concepts.
> 
> 	chris

The responses in this group usually are very quick. One of the best 
things about the language. :) Have you had a chance to try some of the 
tutorials available at www.python.org? If you can read more complex 
documents, the official Python documents have a section on the syntax of 
the language.

Joal Heagney/BCDragon



From gleki at gol.ge  Tue Jun  4 06:36:04 2002
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Tue, 04 Jun 2002 12:36:04 +0200
Subject: property in Py2.2
Message-ID: <3CFC9814.4761F3CB@gol.ge>

Hi all!

I wonder if someone finds time to send me a real code of how to use
properties in Py2.2

I would be much obliged, if you also send me a version of the code for
Py2.1, with __setattr__  and __getattr__.

Thx,
Giorgi

PS. Prior to post here, I have read the materials on www.python.org. I
find them very helpful.



From gerhard.haering at gmx.de  Wed Jun  5 12:40:11 2002
From: gerhard.haering at gmx.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Wed, 5 Jun 2002 18:40:11 +0200
Subject: [ANN] pyPgSQL 2.1 released
In-Reply-To: <23891c90.0206050721.49b31f50@posting.google.com>
References: <mailman.1023270308.20913.python-list@python.org> <23891c90.0206050721.49b31f50@posting.google.com>
Message-ID: <20020605164010.GB32960@gargamel.hqd-internal>

CC-ing to pypgsql-users list, maybe other users have comments on this issue,
too.

* Paul Boddie <paul at boddie.net> [2002-06-05 08:21 -0700]:
> Gerhard <gh_pythonlist at gmx.de> wrote:
> > 	
> > * Fixed the array parsing so it also works with PostgreSQL versions 7.2.x.
> 
> I doubt that this is related, but does the 'IN' operator work with
> statement parameters (bind variables) in pyPgSQL yet? I recently found
> that doing something like this...
> 
>   cursor.execute("SELECT * FROM CAGES WHERE NAME IN %s",
>     (("mammals", "apes"),))
> 
> ...gives some kind of parsing error inside pyPgSQL.

It is indeed a related issue. If you want to show which queries pyPgSQL
produces, the trick is to access a special attribute like this:

conn.conn.toggleShowQuery

This would show the following query is executed:

QUERY: DECLARE PgSQL_08120A6C CURSOR FOR select * from cages where name in '{"mammals","apes"}'

... which obviously cannot work. The reason is that if pyPgSQL encounters a
tuple or a list in the statement parameters, it will quote them suitable for
use as an ARRAY. Unfortunately, the ARRAY syntax is different from what you
would need for the IN operator in SQL.

I'm not sure if something can be done about this problem. It seems to me that
it's an either/or with ARRAY vs. your desired conversion for the IN operator.

> [...] Another thing: pyPgSQL seems to be fine with Unicode query strings
> (unlike psycopg, apparently), but it doesn't like Unicode parameter values.
> Is there some explicit encoding step I should be performing?

I've fiddled with adding Unicode support for some time now, but I didn't get
enough testers yet to make me comfortable enough with it for adding it to a
pyPgSQL release. On the Sourceforge patches section, I've uploaded a patch for
full Unicode support, including getting Unicode results back from the database,
insert Unicode strings, and automatic encoding conversions. I'll try to update
this patch today to work with the latest pyPgSQL 2.1 release and drop you a
note when it's ready.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 25.7 ?C      Wind: 4.5 m/s




From whisper at oz.net  Mon Jun 10 23:29:14 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 10 Jun 2002 20:29:14 -0700
Subject: python and shared libaray
In-Reply-To: <a263f04a.0206101820.5461339f@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEOEAFDIAA.whisper@oz.net>

Yes, you could use that in Python, but it's not a matter of just importing
the .so library. In order to use it, you'll have to create a wrapper .lib or
.so that converts between python objects and C/C++ procedure calls and data
types.

The doc that comes with Python includes information on creating Python
extensions and a couple of other good places to start are:

www.boost.org if it's C++
www.swig.org if it's C

The missing init function is one of the things you'll have to add. In there
is where you would import the C/C++ .so lib - or it gets auto-included as a
consequence of your having linked against it.

Regards,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Steven Pokrandt
> Sent: Monday, June 10, 2002 19:21
> To: python-list at python.org
> Subject: python and shared libaray
>
>
> I have a vendor supplied .so file, header file and docs.  I would like
> to use this
> .so file in python but don't have the src.  Is it possible to link
> this library ie:
>
> import filename.so
>
> when I try the python complains about no init function..  can I add
> this to the .so  without having the src?
>
> can someone point me to a howto?
>
> please send message to the following address..  spokra (AT) attbi
> (dot) com
>
> you get the meaning of the email address, and hopefully the spammers
> don't
> --
> http://mail.python.org/mailman/listinfo/python-list





From BPettersen at NAREX.com  Tue Jun  4 17:09:07 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 4 Jun 2002 15:09:07 -0600
Subject: self
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158EA7@admin56.narex.com>

> From: Vojin Jovanovic [mailto:vjovanov at stevens-tech.edu] 
> 
> I don't see how that is going to work.  In fact I tried 
> things like that before without success.  Just sticking the 
> dictionary into eval/exec is missing the point.  Every time 
> parsing is done flow needs to go through __getattr__ special 
> method to evaluate nested equations.  Without self. you can't 
> achieve that effect.  Try the approach you suggested and 
> you'll get '#U'.

Perhaps something like this?

-- bjorn


class Eval:
    def __init__(self):
        self.__dict__['dict'] = {}
        exec "from math import *" in self.dict
        
    def __call__(self, expr):
        try:
            return eval(expr, self.dict, self.dict)
        except SyntaxError:
            exec expr in self.dict, self.dict
            
    def __getattr__(self, attr):
        return self.dict[attr]
        
    def __setattr__(self, attr, val):
        self.dict[attr] = eval(val, self.dict, self.dict)

e = Eval()

print e('pi * 3')
e('x = sin(0.5)')
print e.x
e.y = 'sin(0.5)'
print e.y




From menucool at yahoo.com.cn  Tue Jun 18 23:50:16 2002
From: menucool at yahoo.com.cn (coolmenu)
Date: 18 Jun 2002 20:50:16 -0700
Subject: Can python have lib to display PDF ?
References: <mailman.1024398081.2986.python-list@python.org>
Message-ID: <a6a2957e.0206181950.7ee6fc15@posting.google.com>

Simon Brunning <SBrunning at trisystems.co.uk> wrote in message news:<mailman.1024398081.2986.python-list at python.org>...
> > From:	menucool at yahoo.com.cn [SMTP:menucool at yahoo.com.cn]
> > 
> > i want a python application to display end edit pdf.
> > where are the lib pdf for python?
> > thank
>  
> <http://www.reportlab.com/>
> 
> Cheers,
> Simon Brunning
> TriSystems Ltd.
> sbrunning at trisystems.co.uk
> 


Thank all :-)
i will try it..


From gleki at gol.ge  Mon Jun 24 05:17:19 2002
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Mon, 24 Jun 2002 11:17:19 +0200
Subject: HTTP authentication
References: <TWtR8.86654$Am4.23833590@twister.columbus.rr.com>
Message-ID: <3D16E39F.E164D5D@gol.ge>

Hi all!

I would be much obliged if someone provides an example of the working code.

Grtz,
Giorgi

Preston Tamkin wrote:

> hello all. I've just recently jumped on the python bandwagon, and I need
> some help!
> I'm trying to write a program(in python) to go out to several websites and
> get a few numbers off of the page. Only problem is, I need to give the site
> a username and password! It's not HTTPS, SSL, or anything like that.  I did
> some packet sniffing and it is a HTTP POST command, the server then requests
> the information then the client sends the u/p. I'm using raw sockets and
> can't figure out how to listen for the "request" packet or excatly what to
> send for the last part. The site uses ASP. It's http://boards.ign.com if you
> want to look at it.  Is there an easier way to do it? Am I just simply
> missing something? Thanks in advance
>
> P.S. Me = TOTAL n00b



From dkt at registriesltd.com.au  Mon Jun 24 19:30:47 2002
From: dkt at registriesltd.com.au (David K. Trudgett)
Date: Tue, 25 Jun 2002 09:30:47 +1000
Subject: SOLVED (Re: Python hits the spot)
In-Reply-To: <Xns9237E4F0DB10Dcliechtigmxnet@62.2.16.82>; from cliechti@gmx.net on Mon, Jun 24, 2002 at 10:27:55PM +0200
References: <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at> <af4vrl$ffe$0@216.39.172.122> <3D16DD5C.1050708@kfunigraz.ac.at> <uhdume2n1mhbba@corp.supernews.com> <3D171291.2060905@kfunigraz.ac.at> <mailman.1024950033.16332.python-list@python.org> <Xns9237E4F0DB10Dcliechtigmxnet@62.2.16.82>
Message-ID: <20020625093047.A1568@regdp-02.intint.registriesltd.com.au>

On Monday 2002-06-24 at 22:27:55 +0200, Chris Liechti wrote:

> >> It was really the laptop processor. I tried the calculation on a
> >> stationary machine (Pentium II 450 MHz and SuSE 8.0), and it works.
> >> 
> 
> you have to admit that it was a nasty failure... is bad hardware your first 
> thought when your program fails?  

Should be an object lesson in that: software engineers need to keep an
eye on hardware issues, including potential failures. It is actually
more common than one would think that 100% CPU for extended periods
can cause hardware failure. Reason: most PCs are engineered and tested
under light load conditions such as those that 95% of users subject
their machines. Many servers, however, run at or near 100% CPU for
extended periods of time, with near continuous disk activity and so
on. Therefore, server class machines are engineered and tested to a
higher standard, and are less prone to such failures (although all
hardware fails eventually).

I remember that I could never do serious image processing on an "old"
Intel Pentium 166 MMX machine because after five minutes it would
become overheated with buzzers sounding to remind me of the fact! It
had the regulation heat sink and CPU fan but still overheated. Go
figure.

Classic symptom of heat failure: machine reboots, stops responding, or
specific components (such as a video card) fail after X minutes from
power on, or X minutes of high CPU usage.

Memory faults can be even harder to spot because they often cause
bizarre symptoms that seem unrelated to memory issues. I once spent
several hours diagnosing and fixing a random "CDROM" fault, including
swapping in a replacement drive and data cable, only to find that the
problem went away after replacing a faulty SIMM card.


David Trudgett






From znu at znu.dhs.org  Fri Jun  7 02:08:14 2002
From: znu at znu.dhs.org (ZnU)
Date: Fri, 07 Jun 2002 02:08:14 -0400
Subject: Method call from string
References: <znu-39AB22.01311007062002@news.fu-berlin.de> <3D004DB1.62BEB68A@engcorp.com>
Message-ID: <znu-CBAE4B.02081407062002@news.fu-berlin.de>

In article <3D004DB1.62BEB68A at engcorp.com>,
 Peter Hansen <peter at engcorp.com> wrote:

> ZnU wrote:
> > 
> > Say I've got a variable with the value 'do_something', and I have an an
> > object that happens to have a do_something method. How can I call that
> > method based on the value of the variable? Preferably without using
> > eval() or similar.
> 
> >>> var = 'do_something'
> >>> class A:
> ...   def do_something(self):
> ...      print 'ooga booga'
> ...
> >>> a = A()
> >>> getattr(a, var)
> <bound method A.do_something of <__main__.A instance at 0x007B0110>>
> >>> f = getattr(a, var)
> >>> f()
> ooga booga

Now that I actually think things though, that makes perfect sense. I 
should have been able to figure that one out myself. Thanks.

-- 
Right now the Internet gives an awfully good imitation of providing superhuman
intelligence capability, both in terms of the total hardware that is out there
and the fact that the 'net has all these human-equivalent peripheral devices
called "users" that can be appropriated in a distributed way to attack problems.
                    --Vernor Vinge


From altis at semi-retired.com  Sun Jun 23 14:26:56 2002
From: altis at semi-retired.com (Kevin Altis)
Date: Sun, 23 Jun 2002 11:26:56 -0700
Subject: The best GUI library in python
References: <8d3f4438.0206220457.b820418@posting.google.com>
Message-ID: <_moR8.392$Fb7.37368@news.uswest.net>

PythonCard uses wxPython as its underlying GUI toolkit, so it should be fast
enough for your needs.

PythonCard comes with over 30 samples that you can learn from and modify for
your own purposes as well as tools for editing source code and building
layouts and menus.

http://www.pythoncard.org/

The documentation page can help you get started quickly.

http://pythoncard.sourceforge.net/documentation.html

ka

"d2002xx" <d2002xx at myrealbox.com> wrote in message
news:8d3f4438.0206220457.b820418 at posting.google.com...
> Hello!!
>
> I want to learn GUI in python, who would tell me which GUI library is
> the best? (don't tell me Tkinter, it's too slow......)




From plugin at supercable.es  Sun Jun 23 14:19:23 2002
From: plugin at supercable.es (Ugo García)
Date: Sun, 23 Jun 2002 20:19:23 +0200
Subject: Redirecting StdOut
Message-ID: <1024856420.145085@seux119>

Hi all. I'm making a game and I'm using Python as the script language. I'm
including a console to communicate with Python. I can interactively send
commands to the interpreter but I can't obtain the response from it. Is
there any way to redirect stdout and stderr in order to print its contain to
the screen?

Thank in advance
--ugo




From shalehperry at attbi.com  Wed Jun 26 13:22:58 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 26 Jun 2002 10:22:58 -0700 (PDT)
Subject: Problems using samples
In-Reply-To: <337619fa.0206260913.41e08940@posting.google.com>
Message-ID: <XFMail.20020626102258.shalehperry@attbi.com>

On 26-Jun-2002 Raphael Ribeiro wrote:
> 1 x = 8
> 2 y = 4
> 3 v = x - y
> 4
> 5 print v
> 6 if x = y:
> 7     print "Eles s?o o mesmo valor"
> 8     else:
>               print "Eles n?o s?o iguais , manja?"
> 
> this is my code, it says there's an error in line 6, but i can't find
> what **** error is that, can anyone help me?

you are not allowed to do an assignment inside of a control statement.

if foo = 2: # invalid

while line = readline(): # invalid

did you mean to use:

if x == y? if not the python idiom is:

x = y
if x:
  # do something




From peter at engcorp.com  Wed Jun 19 22:05:57 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 19 Jun 2002 22:05:57 -0400
Subject: Threading question
References: <mailman.1024532605.1320.python-list@python.org>
Message-ID: <3D113885.13815FCF@engcorp.com>

David LeBlanc wrote:
> 
> with a thread sitting in a loop processing intermittant transactions from a
> buffer, how does one wait if reading an empty buffer does not block - how do
> you relinquish control to other threads/processes?

As Aahz points out, use Queue.  If that's not acceptable,
the usual approach is to do a time.sleep() of some size
when nothing is available.  Depending on how much CPU
you wish to waste versus how long a latency you are 
willing to suffer, you might use anything from 
time.sleep(0.01) to time.sleep(10) ...

-Peter


From robin at jessikat.fsnet.co.uk  Sat Jun 29 18:49:18 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 29 Jun 2002 23:49:18 +0100
Subject: yield equivalent in C/JavaScript?
References: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk>
 <7xy9cxhlvx.fsf@ruckus.brouhaha.com>
Message-ID: <w3xdKUAuljH9EwgH@jessikat.fsnet.co.uk>

In article <7xy9cxhlvx.fsf at ruckus.brouhaha.com>, Paul Rubin <phr-
n2002b at NOSPAMnightsong.com> writes
>There's no native way to do it in C; in some enviroments you might be
>able to simulate it with longjmp madness or threads.
>
>In Javascript the only approach I can think of is make a "virtual
>machine" that executes javascript statements using JS's eval function,
>and implement something like threads or coroutines in the virtual
>machine.  I did something like that once--it wasn't as bad as it
>might sound, but it may not be worth it for your application.
>You can just organize your generator functions like a state machine
>instead, saving their locals and yielding location in some structure
>returned to the caller.  I haven't examined any Python generator
>bytecode but I suspect that's how it works.  By being limited to
>"simple" generators (can't yield except to one's direct caller)
>it's enough to just save the local variables and "return" address
>in the state object, rather than a full continuation including
>the call stack.
the generators here are things  associated with compression or
decompression methods. The complexity is such that a severe
reformulation will be tough. I understand the concept, but I think it's
easier to just try and consume the whole input at one go and pass it on.
-- 
Robin Becker


From spenrose at intersight.com  Thu Jun  6 19:22:08 2002
From: spenrose at intersight.com (Sam Penrose)
Date: Thu, 6 Jun 2002 16:22:08 -0700
Subject: Medium to Large Scale Python Deployments 
Message-ID: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com>

[Domenic R. Merenda raises the topic of Python used in large contexts]

I'm also curious about large amounts of Python, period. Who knows of 
~50K Python lines-of-code (LOC) systems? 100K LOC? More?





From jbublitzNO at SPAMnwinternet.com  Sun Jun 16 15:27:38 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Sun, 16 Jun 2002 19:27:38 GMT
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de> <aeftmj$m29$1@bob.news.rcn.net> <3d0c6d67@nntp.server.uni-frankfurt.de>
Message-ID: <3D0CE698.1050806@SPAMnwinternet.com>

Michael 'Mickey' Lauer wrote:
> David Abrahams <david.abrahams at rcn.com> wrote:

>>With Boost.Python, what you write looks like a kind of IDL (interface
>>description language), which just happens to be built out of C++ source
>>code. Lots of jobs (function argument type checking, conversion, arity
>>checking) are done automatically, but the library doesn't decide which
>>elements to wrap. Unlike SWIG and a few others, there's no attempt to parse
>>your source code. Instead, the compile-time introspection capabilities of
>>C++ are exploited to automatically build wrappers from function and member
>>function pointers.

> I see. So it seems in order to get a nice object oriented Python interface,
> I can 1.) use SWIG to parse the header file and automatically produce a wrapper
>       1.1.) and then use this low-level wrapper to write some high-level python classes around it or
>       2.) use Boost.Python and do the two steps in one by writing a "handcrafted" interface?

> How would sip come into this picture? Is it more like SWIG for C++ or more like Boost.Python?
> (Apart from not beeing concerned about the total lack of documentation ;)

Basically, sip requires that you create a description file (.sip file)
from the h files you want to bind. While there isn't any documentation
as such, the entire process is pretty well described in a recent
thread on the PyKDE mailing list concerning libkdegames, and you
can use PyQt and PyKDE as examples.

I looked at Boost a while ago and it looks very nice. I've recommended
it to people basically because is *has* documentation and for a
single project with a small number of files would probably be more
efficient to use than sip, given the difficulties in learning sip and
setting up the configure/makefile framework.

On the other hand, I maintain PyKDE (I should point out I'm not the sip
author however), and for an actively maintained project with nearly a
dozen libraries and several hundred files, sip seems preferable. It's
possible to automate a lot of the *.sip file generation, including the
versioning (both PyQt and PyKDE use a single set of description files
that cover ALL supported versions).

As far as the example posted earlier in the thread, it doesn't look
like it would be a problem for sip, other than exposing the member
variables publicly seems like poor design and could lead to problems
in Python bindings if not used carefully.

The difficulty with sip is that while it's a very good tool for
general purpose use, it's maintained primarily for PyQt/PyKDE
and there hasn't been a lot of time available to generate
documentation or tools to go with it. It is open source, however,
and I'm sure plenty of support would be available for anyone
wanting to do work in the areas of docs or tools, or wanting to
use sip to generate bindings..


Jim



From tdelaney at avaya.com  Mon Jun 24 20:23:06 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 25 Jun 2002 10:23:06 +1000
Subject: GOTO w/ Python?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A3BE@natasha.auslabs.avaya.com>

> From: Daniel Fackrell [mailto:dfackrell at DELETETHIS.linuxmail.org]
> 
> After skimming the linked page to get a feel for the CAME 
> FROM statement, is
> it conceptually different from short procedures?  Does the 
> only difference
> lie in the fact that for compiled languages one of the 
> instances will be
> inline?  I do understand that it would have some effect on 
> variable scoping,
> but this can be easily handled in other ways.
> 
> So what's the benefit that might outweigh the opacity of the 
> code created?

I hope you're not serious ... :)

COME FROM is a joke. Literally. It is a parody of GOTO whose only redeeming
feature is that it is not GOTO.

Tim Delaney




From cliechti at gmx.net  Sun Jun 23 14:36:02 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 23 Jun 2002 20:36:02 +0200
Subject: Redirecting StdOut
References: <1024856420.145085@seux119>
Message-ID: <Xns9236D1FF016AAcliechtigmxnet@62.2.16.82>

"Ugo Garc?a" <plugin at supercable.es> wrote in
news:1024856420.145085 at seux119: 

> Hi all. I'm making a game and I'm using Python as the script language.
> I'm including a console to communicate with Python. I can
> interactively send commands to the interpreter but I can't obtain the
> response from it. Is there any way to redirect stdout and stderr in
> order to print its contain to the screen?

sure. just assign a file like object to sys.stderr and stdout.

class DummyFileForWrites:
    	def write(self, text):
    	    	#write 'text' to GUI here

sys.stdout = sys.stderr = DummyFileForWrites()

so you can do the same for sys.stdin but then you would need read, 
readline, and readlines.

chris

-- 
Chris <cliechti at gmx.net>



From jepler at unpythonic.net  Mon Jun 17 10:52:53 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Mon, 17 Jun 2002 09:52:53 -0500
Subject: Getting pixel values under cursor
In-Reply-To: <3eeda89d.0206170256.4ba75ec@posting.google.com>
References: <adrhgi$mh$1@nntp1.jpl.nasa.gov> <3eeda89d.0206170256.4ba75ec@posting.google.com>
Message-ID: <20020617145252.GC8814@unpythonic.net>

I don't know that Tk offers the general capability to find the color
under a pixel.  However, it is possible to find the color at a
particular coordinate within a photo image.  In tcl/tk,
"<image name> get x y" returns the RGB triple .. This doesn't seem to be
provided in Tkinter.Image, so I derive a class here.

from Tkinter import *

class ImagePlus(Image):
    def get(self, x, y):
	return tuple(map(int, self.tk.splitlist(
				self.tk.call(self.name, "get", x, y))))

def update_it(evt):
    x = evt.x
    y = evt.y
    rgb = p.get(x,y)
    print `rgb`
    l.configure(text="#%02x%02x%02x" % rgb)

t = Tk()
p = ImagePlus("photo", master=t, file="dilbert2001114622012.gif")
l = Label(t)
x = Label(t)
x.configure(image=p)
x.bind("<Motion>", update_it)

l.pack()
x.pack()
t.mainloop()

Jeff




From K.Rdt at TU-Berlin.DE  Tue Jun 25 09:48:46 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Tue, 25 Jun 2002 13:48:46 GMT
Subject: mail-problem over t-online
References: <mailman.1024866209.9154.python-list@python.org>
Message-ID: <1103_1025012926@127.0.0.1>

Am Sun, 23 Jun 2002 23:01:56 +0200 hat Klaus Reinhardt <K.Rdt at TU-Berlin.DE> geschrieben:
---------------------------------------------------------------------
Hi
With putty-tunnel!
		K at Rdt
---------------------------------------------------------------------





From toddg at math.utexas.edu  Mon Jun 10 12:03:18 2002
From: toddg at math.utexas.edu (Todd Gillespie)
Date: Mon, 10 Jun 2002 16:03:18 +0000 (UTC)
Subject: Embed python in multi-thread environment
References: <mailman.1023636372.9719.python-list@python.org>
Message-ID: <ae2ik6$dvi$1@geraldo.cc.utexas.edu>

Arlo <arlo.maillist at ozmaker.com> wrote:
: i've tried to embed python in multi-thread application( written in c++)
: the flow is:

First thing, I recommend you look at a successful implementation:
http://pywx.idyll.org/



From donn at u.washington.edu  Tue Jun 25 12:57:37 2002
From: donn at u.washington.edu (Donn Cave)
Date: 25 Jun 2002 16:57:37 GMT
Subject: Suggestions for good programming practices?
References: <mailman.1025013693.19616.python-list@python.org>
Message-ID: <afa7e1$eeq$1@nntp6.u.washington.edu>

Quoth brueckd at tbye.com:
| On 25 Jun 2002, Chris Liechti wrote:
(someone said)
| > >> * Avoid exec, execfile, eval, and input.

| This view is a overly extreme. Rather than teaching people to fear certain 
| features like eval/exec, it's better to explain the risks so that they can 
| make informed decisions as to when it's wise to use them.

It's not overly extreme!  It's just about extreme enough.

| So... rather than teaching "avoid W!", let's say "be careful with W 
| because of X, Y, and Z". I still wouldn't use eval/exec on code posted 
| through a web form, for example, but there are times when they are very 
| useful and I can use them in good confidence because I understand their 
| risks.

But you weren't going to be deterred by that pronouncement anyway.

"Avoid exec, execfile, eval, and input" is good advice.  Taken at face
value, it doesn't necessarily absolutely prohibit their use - if I said
"avoid walking in the road", you could reasonably reasonably assume I'm
saying something like "walk on the sidewalk when possible".

Someone whose software engineering skills have been honed by years
of wrestling with ugly code monsters will apply a different perspective
to that advice.  A 1-week newbie could do worse than to follow that
advice religiously.

	Donn Cave, donn at u.washington.edu


From shagshag13 at yahoo.fr  Wed Jun 12 05:29:30 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Wed, 12 Jun 2002 11:29:30 +0200
Subject: parse tar file with python
Message-ID: <ae7498$4jven$1@ID-146704.news.dfncis.de>

hello,

i 'm not sure this is the right place to ask, but maybe somebody had already do some kind of python code on this :

i had hundred thousand of small files, my program is very slow. I'm thinking that i loose time waiting for opening/closing, and i
think this could be best if i "tar" my files by blocks.

But in this case i don't know how to check for the end of a file and how to get its name (both in a tar archive) ...
(i don't want to tar/untar with python, just to be able to do a sequential reading of a tar file to get its content)

do you have an idea ?

thanks in advance,

s13.




From kragen at pobox.com  Sun Jun  2 22:38:03 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 02 Jun 2002 22:38:03 -0400
Subject: Why no open(f, "w").write()?
References: <mailman.1022702940.7544.python-list@python.org> <slrnafagek.64g.huaiyu@gauss.almadan.ibm.com> <slrnafcsor.12v.msoulier@pmerd071.ca.nortel.com> <20020531225504.248c92ce.larooy@xtar.co.nz> <m2lm9zh39j.fsf@mycroft.actrix.gen.nz>
Message-ID: <83sn45rss4.fsf@panacea.canonical.org>

Paul Foley <see at below> writes:
> GC is just an optimization to allow reuse of
> memory used by objects that the program will never access again,
> because you don't really have infinite memory.
> 
> Correct optimizations don't change the behaviour of correct code.
> 
> [Therefore, if the behaviour of your code does change depending on
> whether or not the optimization is done, either the optimization is
> in error or your code is.

So all finalizers should be considered bugs?  I think that's a
supremely unpragmatic approach to programming.



From jjl at pobox.com  Tue Jun  4 17:10:34 2002
From: jjl at pobox.com (John J. Lee)
Date: Tue, 4 Jun 2002 22:10:34 +0100
Subject: Cookie question
In-Reply-To: <adikmf$1100j6$1@ID-49758.news.dfncis.de>
References: <adikmf$1100j6$1@ID-49758.news.dfncis.de>
Message-ID: <Pine.SOL.4.30.0206042206340.10073-100000@mimosa.csv.warwick.ac.uk>

On Wed, 5 Jun 2002, Ken wrote:
[...]
> If I have many different cgi scripts checking UserID this way, are they all
> refering to the same cookie?

They are if and only if the browser sends back the same cookie as last
time.  I don't think I understand what you're getting at.

> If it is so, how come my other scripts can't read userid correctly? What did
> I do wrong?

Can you be more specific?

> Also, where does the cookie get saved in my computer (using Win2000 Pro)? I
> don't seem to find any cookie inside the Tempory Internet Files folder.

Why should they get saved in there?  Firstly, this is Python, not IE, that
you're programming.  Secondly, you're talking about the server end of the
cookie protocol, and it's the client's responsibility to save cookies --
that's the whole point!

Am I misunderstanding you?


John



From pyth at devel.trillke.net  Wed Jun 12 08:20:15 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 12 Jun 2002 14:20:15 +0200
Subject: Python Virtual File System: A spike test
In-Reply-To: <3D070F27.9050608@rogers.com>; from mcfletch@rogers.com on Wed, Jun 12, 2002 at 05:06:47AM -0400
References: <3D070F27.9050608@rogers.com>
Message-ID: <20020612142015.X6609@prim.han.de>

Hi Mike,

[you wrote]
> Purpose:
> 
> A class for representing file system paths which supports both 
> method-based operations, and abstraction through object encapsulation to 
> eventually allow for Virtual File System operations (for instance, 
> transparently accessing the contents of zip files).
> 
> Status:
> 
> An initial implementation (spike test) exploring the viability of the 
> approach.  Implementation provides:
> 
> 	equality testing
> 	parent/child/ancestor/descendent retrieval/testing
> 	filesystem-role testing (file/dir/root)
> 	root retrieval and root comparisons
> 	"walking" (callback-based recursive iteration)
> 	listing (returning Paths which include their parent directory)
> 	fragment determination (i.e. c:\\temp\\test -> "c:\\", "temp", "test"
> 	recursive directory-tree removal (including files)
> 	open( ) method for path specifiers

Interesting. I have written a similar 'filename' class. 
It is not based on the str-type but accepts strings in many places. 
Also the filename class offers all of the methods found in 'os' or 'os.path'
under the exact name. I think this is important to make using it easy. 
Many people already know and use the stdlib-methods.

One of the nice features is that you can do stuff like:

fn=filename('/usr/src')
cfilter=filters.OR(filters.hasext('.cpp'), filters.hasext('c'))
mfilter=filters.rmatch('.*/[Mm]akefile.*')
recursefilter=rignore('.*/CVS')

for makefiles, cppfiles in fn.filterwalk([mfilter,cfilter], recursefilter):
    # work with list of makefiles
    # work with cppfiles

'filterwalk' is a generator which honors a recursion filter and one or
more filename-filter.  It allows to recurse once and get back categories of files
(as lists). No further checking like 'islink' etc. needed. You makeup
your filters and call filterwalk and afterwards work with the lists.

> Future:
> 
> If there is interest, I am planning on creating a new SourceForge 
> project with the eventual goal of creating a virtual file system package 
> with a consistent interface to most of the major file-and-directory 
> systems (zip files, FTP sites, local file systems) and hooks for adding 
> new systems.

Having an abstraction for 'Names'/URLs and working from there would 
be very interesting IMO. At the moment we only work with (relatively
untyped) strings which places the burden of calling the right methods
(urllib/httplib/urllib2/os.path/ftplib/webdav...) on the programmer
(who just has a string).  

> Request:
> 
> If you're interested in working on such a project (or even using it), 
> let me know, so I can decide whether to devote any more resources to it.
> You can find the module here:
> 
> http://members.rogers.com/mcfletch/programming/filepath.py

i am interested but we should discuss if our approaches/viewpoints 
work well together :-) 

cheers,

    holger




From aahz at pythoncraft.com  Thu Jun 13 13:09:58 2002
From: aahz at pythoncraft.com (Aahz)
Date: 13 Jun 2002 13:09:58 -0400
Subject: Threading tutorial wanted.
References: <mailman.1023954137.22915.python-list@python.org> <slrnagh9r0.4f1.gerhard@lilith.my-fqdn.de>
Message-ID: <aeajl6$iv8$1@panix1.panix.com>

In article <slrnagh9r0.4f1.gerhard at lilith.my-fqdn.de>,
Gerhard =?iso-8859-15?Q?H=E4ring?=  <gerhard at bigfoot.de> wrote:
>
>Aahz has written a tutorial that you were already referred to. It's
>useful info, but nowhere near complete. Maybe you can write a tutorial
>yourself once you've mastered threading? ;-) Maybe there's a Python
>book that covers multithreading in depth? I'd be interested which one
>this is if so.

I was negotiating with O'Reilly to write such a book, but even though
threading is a popular topic here, it's pretty clearly a marginal book
project and various external events (including the economy) consipired
to kill the book (at least for now -- and I'm currently working on a
different book that will keep me busy until next year).
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From whisper at oz.net  Thu Jun  6 15:57:14 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 6 Jun 2002 12:57:14 -0700
Subject: Using AT on XP to run python script - MARK HAMMOND PLEASE READ
In-Reply-To: <fa7108b0.0206061106.339d547b@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEMEBNDHAA.whisper@oz.net>

I just discovered that you can use /interactive to cause a scheduled task to
show output. I wonder if this would work for your email problem...

at xx:xx /interactive python j:\python22\tools\idle\idle.py

This worked as expected and idle came up!

Dave LeBlanc
Seattle, WA USA





From mike at radiomag.com  Tue Jun 18 02:52:39 2002
From: mike at radiomag.com (Michael Steuer)
Date: Tue, 18 Jun 2002 16:52:39 +1000
Subject: Listing all available (not just loaded) modules
References: <mailman.1024378702.24878.python-list@python.org>
Message-ID: <aemlao$85r2t$1@ID-140151.news.dfncis.de>

> Idle, PythonWin and Boa (at least) do this, with PythonWin being the most
> detailed - it shows classes and methods within the modules. The version of
> Pythonwin I have does not show the contents of the site-packages tree, but
> that would be easily fixable.

They "do this" and how? Don't tell me you mean the dir() command ...

Cheers,

Mik.e




From shalehperry at attbi.com  Fri Jun 21 12:32:06 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 21 Jun 2002 09:32:06 -0700 (PDT)
Subject: Newbie-Question: Oberserving an directory
In-Reply-To: <3d1352a7@news.piro.net>
Message-ID: <XFMail.20020621093206.shalehperry@attbi.com>

On 21-Jun-2002 Marcus Klein wrote:
> Hi there,
> 
> I want to observe an directory to check if a new file has been put in, if 
> that happens I want to start an event.
> 
> Does anyone have a good idea how to start, I've tried it with signal, but 
> either I am too stupid or it is the complete wrong way :-/
> 

under UNIX the only way to accomplish this is to periodically poll the
directory with stat.  There is no signal you can watch for.  I can not speak
for either Mac or Windows.




From tdelaney at avaya.com  Mon Jun 10 22:33:49 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 11 Jun 2002 12:33:49 +1000
Subject: Correct code/unit testing (was Re: Efficient python programmi
	ng...)
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A370@natasha.auslabs.avaya.com>

> From: Peter Hansen [mailto:peter at engcorp.com]
> 
> Perhaps, assuming you put a lot of bugs in the code, some 
> amount of staring will be necessary.
> 
> I believe staring is of little or no use, however, having 
> tried it many times with my and others' code which had bugs and
> not having found those bugs by staring.

Staring in and of itself is not of much use ... however, simply getting
another set of eyes on the code is. I have found that in the majority of
cases, one of two things will occur:

1. The new set of eyes will immediately see the thing you've been failing to
see for the past 2 hours ...

OR

2. While explaining to the new set of eyes enough for them to understand the
problem, *you* will see what you've been failing to see for the past 2 hours
...

Any time I find myself staring at something and making no progress, I get
someone else to look at it. Likewise, I make myself available to others.

Again, this requires egoless programming - it doesn't work if you get upset
when someone else finds your boneheaded mistake; or if you are unwilling to
point out your own boneheaded mistake in front of someone else; or if you
are unwilling to point out someone else's boneheaded mistake because *they*
will get upset.

Tim Delaney




From mruiz at safa.es  Fri Jun 21 07:22:00 2002
From: mruiz at safa.es (Manuel Ruiz)
Date: Fri, 21 Jun 2002 11:22:00 GMT
Subject: CR+LF problem
References: <3d11bc33.1129927479@news.mad.ttd.net> <12q3hukse2cepcm9tpern92jc71a5vrhvi@4ax.com>
Message-ID: <3d1309ff.1215379262@news.mad.ttd.net>

Hi,

char * txt = "print 123\r\n" is only a sample, I am storing scripts in
a database, I put the script into the database with \n but when I get
it, It has \r\n as end of line, I have assumed this is a database
problem.

The rare thing is that PyRun_Simple() function can't execute this kind
of python code using \r\n as end of line, however I saved that code to
file and ran "python2 myfile.py" (with \r\n as end of line too) and it
ran fine, very fine.

Where is the diferrence between the api of python and the interprete
itself. Because first one can't execute this code and the second one
can do it ?

Regards
Manuel Ruiz

On Thu, 20 Jun 2002 17:38:43 +0300, Christos "TZOTZIOY" Georgiou
<DLNXPEGFQVEB at spammotel.com> wrote:

>On Thu, 20 Jun 2002 11:33:55 GMT, rumours say that mruiz at safa.es (Manuel
>Ruiz) might have written:
>
>>char * txt = "print 123\r\n";
>>result = PyRun_Simple( txt, Py_file_input, PyDict_New(), NULL );
>
>did you try
>	char *txt= "print 123\n";
>by any chance?
>I am not very sure if I do help, but in C, newlines are just "\n".
>"\r\n" is by convention the text line terminator of VMS IIRC, MS-DOS &
>window derivatives etc.
>
>>this code fails to exec txt, I got a SyntaxError exception,
>>however I if save to file the statement "print 123\r\n" and try to run
>>it ussing python interprete (version 2.2.1) It run fine.
>>
>>Can anyone tell me anything about this rare behaviour
>
>-- 
>TZOTZIOY, I speak England very best,
>Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')



From pyth at devel.trillke.net  Thu Jun  6 20:30:10 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 7 Jun 2002 02:30:10 +0200
Subject: self
In-Reply-To: <jpKL8.83$HV1.4324388@newnews.cc.stevens-tech.edu>; from vjovanov@stevens-tech.edu on Thu, Jun 06, 2002 at 10:34:09AM -0400
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu> <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com> <mailman.1023294728.2322.python-list@python.org> <sWsL8.82$v_7.4260962@newnews.cc.stevens-tech.edu> <mailman.1023323409.9676.python-list@python.org> <jpKL8.83$HV1.4324388@newnews.cc.stevens-tech.edu>
Message-ID: <20020607023010.B16359@prim.han.de>

Vojin Jovanovic wrote:
> I don't think you tested your code, because I can't get the output you
> specified.  In fact it seems that you broke your own code.

of course i tested it! Do you really think i am making up
results which can be proven wrong in ten seconds? Why such a
harsh reaction without bothering to correct the obvious 'typo'?
I don't actually know what went wrong while pasting the 
code ('Namerror' instead of 'NameError'). 

Anyway, here is the paste-typo-corrected version:

class Evaluator:
    def __init__(self, eqs={}):
        self.__dict__['equations']=eqs
    def __getattr__(self, name):
        value = self.equations[name]
        if type(value)!=str:
            return value   # if it's not a string, just give it back
        scope={}
        while 1:
            try:
                return eval(value, scope)
            except NameError, n:
                var=str(n).split("'")[1]
                scope[var]=getattr(self, var)
    def __setattr__(self, name, value):
        self.equations[name]=value

class Berta:
    a=5

>>> e=Evaluator({'a':3})
>>> e.berta=Berta()
>>> e.g='berta.a*a'
>>> e.g
15

regards,

    holger




From kzaragoza at attbi.com  Sat Jun 15 15:42:59 2002
From: kzaragoza at attbi.com (Kris J. Zaragoza)
Date: 15 Jun 2002 14:42:59 -0500
Subject: ADO Command object/Parameters problem...
References: <3d09426d.313465031@netnews.attbi.com>
Message-ID: <slrnagn6d1.263.kzaragoza@localhost.localdomain>

In article <3d09426d.313465031 at netnews.attbi.com>, candiazoo at attbi.com wrote:
> Has anyone else run into this (note: I have seen other entries for the same
> error, which is rather generic, but not for my specific reason).

I've never seen this specific error before when using ADO from Python
or any other language.  The exact error text below may be a red
herring or otherwise not accurately describe the problem.  Looking
quickly at the code you provided, however, a few things smell fishy.
Comments below.

> HERE is an example of what I am doing... (note, I added the long() typecasting
> to make sure the value was coerced into a long integer)...
> 
> sql = "select pty.orgname " + \
>       "from ccc_party pty, ccc_right_rightsholder rgh, ccc_right rgt " + \
>       "where pty.pty_inst = rgh.rgh_inst " + \
>       "and rgh.rgt_inst = rgt.rgt_inst " + \
>       "and rgt.rgt_inst = ?"

First of all, you are performing one more join than needed for what
you are selecting.  If you need to restrict your record set based on
the rgt_inst field, you only need to join ccc_party and
ccc_right_rightsholder as the latter already has the foreign key
reference which is exactly the field you are restricting on.  In other
words, you only need to use:

select pty.orgname
from ccc_party pty, ccc_right_rightsholder rgh
where pty.pty_inst = rgh.rgh_inst
and rgh.rgt_inst = ?

Or, using the more modern join syntax

select pty.orgname
from ccc_party pty
join ccc_right_rightsholder rgh
	on rgh.rgh_inst = pty.pty_inst and rgh.rgt_inst = ?

> cmdOracle.ActiveConnection = cnnOracle
> cmdOracle.CommandType = 1
> cmdOracle.CommandText = sql
> 
> trsInst = long(rstTempDB.Fields("trs_inst").Value)
> aasInst = long(rstTempDB.Fields("aas_inst").Value)
> anyInst = trsInst or aasInst
>         
> cmdOracle.Parameters.Append(cmdOracle.CreateParameter("orgname", 200, 2, 80))
> cmdOracle.Parameters.Append(cmdOracle.CreateParameter("rgt_inst", 3, 1, 4,
> long(anyInst)))
> rstOracle, status = cmdOracle.Execute()

Here's another point that confused me.  Why are you adding two
parameters when you only have one parameter placeholder?  The orgname
field is what will be returned in the record set rstOracle.  You don't
need to add it as a parameter.

I don't know how helpful this has been.  If you need further
clarification or help, you may want to provide a more accurate or
complete piece of code.

Kris


-- 
Kris J. Zaragoza       | On the face of it, Microsoft complaining about
kzaragoza at attbi.com    | the source license used by Linux is like the
                       | event horizon calling the kettle black.
                       | -- Adam Barr, article on kuro5hin.org


From sandskyfly at hotmail.com  Tue Jun  4 14:24:12 2002
From: sandskyfly at hotmail.com (Sandy Norton)
Date: 4 Jun 2002 11:24:12 -0700
Subject: Kalman Filters, Particle Filters and SVM's
References: <671ybne7zm.fsf@mail.uni-oldenburg.de>
Message-ID: <b03e80d.0206041024.20403058@posting.google.com>

> Support Vector Machines in (Numerical) Python ?
> 
> Especially the latter would be fine, I'm dreaming to implement a SVM
> based script with the capability to distinguish spam from personal
> mails.... but I'm way to lazy to write my own SVM.

Here's a couple of alternatives:

- python SVM using orange machine learning framework (Aleks Jakulin)
  http://ai.fri.uni-lj.si/~aleks/orng/

or 

- you can use jython and tap the Java Weka Machine Learning Framework
  which also has an SVM implementation
  http://www.cs.waikato.ac.nz/ml/weka/

This is fun stuff... enjoy!

Sandy Norton


From mdk at mdk.com  Thu Jun 20 11:21:51 2002
From: mdk at mdk.com (mdk)
Date: 20 Jun 2002 15:21:51 GMT
Subject: Fill a file with some text.
Message-ID: <Xns9233739A9FCB8mdkmdkmdkmdk@130.133.1.4>

Hello,

I need to fill a file with differnt amounts of some text, either just the 
letter x or random, doesn't matter.  But I need the size to be, for 
example, 6070784 bites.

What is the best and or fastest way to do this?

Thanks.


From rs at onsitetech.com  Thu Jun 13 16:53:44 2002
From: rs at onsitetech.com (Robb Shecter)
Date: 13 Jun 2002 20:53:44 GMT
Subject: another near-sighted, time-wasting syntax enhancement idea
References: <e395b886.0206051956.5f447e9b@posting.google.com> <Xns92265AD72BE0cliechtigmxnet@62.2.16.82>
Message-ID: <3D0905FF.7050003@onsitetech.com>

Chris Liechti wrote:

> i don't see any advantages for that...

I agree.  That code looks pretty horrific. :-)  Looks like someone's 
been doing lots of C++ programming?  (Just guessing)





From chris_mk at hotmail.com  Wed Jun  5 13:00:48 2002
From: chris_mk at hotmail.com (Christopher)
Date: 5 Jun 2002 10:00:48 -0700
Subject: Problem with urllib.urlopen()
References: <cc14131d.0206031445.5154de31@posting.google.com> <cc14131d.0206032128.5b2efcf6@posting.google.com> <Pine.SOL.4.30.0206042228330.10073-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <cc14131d.0206050900.ae14ecf@posting.google.com>

Yeah, me too! ;-)  What happens is that a browser takes the user to
the correct page but the urllib functions return a page that says an
error has occured during the search.  Thanks.




"John J. Lee" <jjl at pobox.com> wrote in message news:<Pine.SOL.4.30.0206042228330.10073-100000 at mimosa.csv.warwick.ac.uk>...
> On 3 Jun 2002, Christopher wrote:
> > There were some problem with posting the whole link so I'll break each
> > one up...
>  [...lengthy URLs snipped...]
> > chris_mk at hotmail.com (Christopher) wrote in message news:<cc14131d.0206031445.5154de31 at posting.google.com>...
>  [...]
> > > You should, like I said, get the XML version of the original links
> > > (with some extra information, which is why I want to interface with
> > > the XML site, btw).  Now, if you do an interface with Python instead
> > > of your browser you get a response of an error page instead.  Why is
> > > that and what can I do to change it?  If you have any information, I
> > > would appreciate it ver much.  Thanks.
> 
> What was the error?  Sorry, too lazy to follow the urls.
> 
> 
> John


From gnufnork at fuckspam.hetdigitalegat.nl  Mon Jun 10 22:51:12 2002
From: gnufnork at fuckspam.hetdigitalegat.nl (Wladimir)
Date: Tue, 11 Jun 2002 04:51:12 +0200
Subject: python and shared libaray
References: <a263f04a.0206101820.5461339f@posting.google.com>
Message-ID: <pan.2002.06.11.02.51.11.623512.25803@fuckspam.hetdigitalegat.nl>

On Tue, 11 Jun 2002 04:20:44 +0200, Steven Pokrandt wrote:

> I have a vendor supplied .so file, header file and docs.  I would like
> to use this
> .so file in python but don't have the src.  Is it possible to link this
> library ie:
> 
> import filename.so
> 
> when I try the python complains about no init function..  can I add this
> to the .so  without having the src?

I don't think that shared object is meant to be imported in python,
concluding from your post it's just a generic one with  C header files
that define the interface. To use it in python you'd have to make a
wrapper, I'm afraid, that's not too easy a job.


Wladimir


From imcmeans at home.com  Sun Jun 30 22:58:21 2002
From: imcmeans at home.com (Ian McMeans)
Date: Mon, 01 Jul 2002 02:58:21 GMT
Subject: Is this a __slot__ bug?
References: <pan.2002.06.30.00.21.54.133520.5016@vodeotron.ca>
Message-ID: <hzPT8.53165$vo2.2811682@news2.telusplanet.net>

I wasn't able to find a good documentation of slots on the python website.
Do you have a link handy? I guess I'm just bad at searching, I couldn't find
a description in the first few pages of google either.

"Steve Menard" <steve.menard at vodeotron.ca> wrote in message
news:pan.2002.06.30.00.21.54.133520.5016 at vodeotron.ca...
> OK, this is probably me not understanding something, but then again, you
> never know.
>
> First things first, I love __slots__. For the paranoid programmer I am,
> knowing I will never again assing to a variable with a typo and then
> wonder ehat the hell is going wring is a godsend.
>
> However, I also love(need) private variables (those __X). It seems they
> dont work with slots. try the following simple example under ptyon 2.2 on
> linux :
>
> class O(object) :
> __slots__ = [
> '__a'
> ]
>
> def __init__(self) :
> object.__init__(self);
> self.__a = 3
> o = O()
>
> I get thoe following error :
> AttributeError: 'O' object has no attribute '_O__a'
>
> Now, I know I can easily work around this by manually obfuscating the
> private variables names in the __slots__, but I would rather not. Feels
> wrong somehow to rely on what should be a hidden way of generating
> private variables.
>
> So if this is by design, I will not agree but I will do the workaround.
> If not, well then seems I founda bug :)
>
> Steve




From gerhard at bigfoot.de  Fri Jun 28 11:02:59 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 28 Jun 2002 15:02:59 GMT
Subject: Is Python growing?
References: <mailman.1025267736.21565.python-list@python.org>
Message-ID: <slrnahoucn.140.gerhard@lilith.my-fqdn.de>

Mark McEahern wrote:
> [...] What's so
> beautiful about Python is that all I had to do was fire up the interactive
> interpreter and do this to remind myself of what map does:
> 
>  >>> print map.__doc__

help(map) in 2.2.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From whisper at oz.net  Tue Jun 11 09:54:52 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 11 Jun 2002 06:54:52 -0700
Subject: How to play a .wav sound?
In-Reply-To: <ae4ekq$gtg$1@news.idiom.com>
Message-ID: <GCEDKONBLEFPPADDJCOEMECFDIAA.whisper@oz.net>

Perhaps the "Snack" sound extension can help - it's primarily Tcl, but has a
python binding too. It is cross-platform.

http://www.speech.kth.se/snack/

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of James T. Dennis
> Sent: Tuesday, June 11, 2002 2:08
> To: python-list at python.org
> Subject: Re: How to play a .wav sound?
>
>
> Erlend J. Leiknes <nookieNO_SP_AM at online.no> wrote:
>
> > Think this should work... not tested.
>
>  I doesn't.  I doesn't in principle (cat foo.wav > /dev/dsp doesn't work
>  for me) and the code is wrong in at least one spot:
>
> ...
> >         sdata = f.read(1024)
> >         while sdata:
> >             d.write(sdata)
> ...
>
>  that would only copy 1K of data
>
> > "Bruno Bellamy" <bellamy at neverland.net> wrote in message
> > news:ad8jo7$ohl$1 at wanadoo.fr...
> >> I guess the question has probably been asked many times, sorry for
> >> bothering...
> >> But I searched here and there, and I couldn't find how to do
> that, simply
> >> play a .wav sound from a python program using Tkinter.
>
> >> Moreover, I need to do that in a program that would run similarly under
> >> Linux and Windows.
>
> >> I saw there's a python module called wave, but it seems it can
> only read
> >> .wav files. I coulnd't find a way to actually play the sound.
> Maybe it's
> >> hidden somewhere?
>
> >> If anybody can help me, that would be cool. :)
>
>
>  I was stumped for a little while last night as well.  Unfortunately
>  I don't have a simple working example, yet; but it looks like
>  pysolsoundserver.so and the PySol (Python Solitaire) sources might
>  be a good start.  Apparently PyGame also has some sophisticated
>  sound handling.  There doesn't seem to be a way to do it (easily)
>  with the standard libraries, though.
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From whisper at oz.net  Mon Jun 10 01:40:40 2002
From: whisper at oz.net (David LeBlanc)
Date: Sun, 9 Jun 2002 22:40:40 -0700
Subject: How to capture stdout into a string?
In-Reply-To: <ae185e$77j$1@slb3.atl.mindspring.net>
Message-ID: <GCEDKONBLEFPPADDJCOEKEMADHAA.whisper@oz.net>

No, it's GTSL - Grep the Source Luke :->

David LeBlanc
Seattle, WA USA 

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Matt Gerrans
> Sent: Sunday, June 09, 2002 20:59
> To: python-list at python.org
> Subject: Re: How to capture stdout into a string?
> 
> 
> > :-)  Actually, I assumed you already had RTFM....
> > Still helpful sometimes to read the source.
> 
> Maybe you guys can coin (or have coined) a new acronym: RTFS
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From kseehof at neuralintegrator.com  Mon Jun 10 10:03:00 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Mon, 10 Jun 2002 07:03:00 -0700
Subject: Does Python need a '>>>' operator?
In-Reply-To: <Pine.GSO.4.33.0206101315030.4022-100000@techst02.technion.ac.il>
Message-ID: <EFEFIHAINIINNELLHFAIEENBCIAA.kseehof@neuralintegrator.com>

Beni Cherniavsky <cben at tx.technion.ac.il> wrote:
> On 2002-06-10, Ken Seehof wrote:
> > Unfortunately, I don't see an easy way to clean up this blemish:
> > ...
> ... This is correct behaviour ....etc.

Of course I listed the three cases together for contrast:
>>> 0xffff
65535
>>> 0xffffffff
-1
>>> 0xfffffffff
68719476735L

Obviously, only the second case is incorrect.

- Ken Seehof





From radix at twistedmatrix.com  Tue Jun 11 13:55:12 2002
From: radix at twistedmatrix.com (Christopher Armstrong)
Date: 11 Jun 2002 13:55:12 -0400
Subject: XML/Source Persistence (was: Source formatting for long expressions.)
In-Reply-To: <20020611132003.I5683@eecs.tufts.edu>
References: <mailman.1023570972.27430.python-list@python.org>
	<m3it4srhq5.fsf@mira.informatik.hu-berlin.de>
	<874rgcxy99.fsf@twistedmatrix.com>
	<mailman.1023668414.25769.python-list@python.org>
	<6bd9f01b.0206110543.168b0f83@posting.google.com>
	<878z5lx4vl.fsf@twistedmatrix.com>
	<20020611132003.I5683@eecs.tufts.edu>
Message-ID: <874rg9wvi7.fsf_-_@twistedmatrix.com>

>>>>> "Michael" == Michael Gilfix <mgilfix at eecs.tufts.edu> writes:

    Michael> On Tue, Jun 11 @ 10:32, Christopher Armstrong wrote:
    >> XML is a terrible format for object persistence, due to the lack of
    >> robust mappings. Twisted's XML uses attributes in some cases, but that
    >> is only possible when you have string:string mappings. In *any* other
    >> case, you have to resort to children nodes, with such ugly syntax as::

[snip..]
    >> *That* is why I don't just use XML.

    Michael>   I'm not disagreeing with you because I don't think you're wrong
    Michael> in this case and well, XML isn't for everything. I think XML can
    Michael> work here just as well but it's just more work. I'm not sure about
    Michael> the lack of robustness of the mappings - you just have to define
    Michael> them and can use a DTD to verify that your mappings are
    Michael> intact. But you have to write extra code for the mappings that

I meant the mappings that attributes provide. Creating mappings with multiple
tags is overbearingly verbose, no matter if you formalize it or not. (Am I
understanding you correctly?) Perhaps I should've said "flexible" rather
than "robust". BTW, ONX (http://www.seairth.com/web/onx/onx.html) provides
a better type of mapping: string:arbitrary object. It's not
arbitrary object:arbitrary object, but it's a bit better.

    Michael> isn't quite worth it in such a domain specific case. XML costs
    Michael> extra for being general.  Er, you don't have any between
    Michael> version/backwards-compat issues though do you? In which case, have
    Michael> you thought those out thoroughly?  Forgive me if it isn't
    Michael> relavent. Not sure exactly where these will be used.

Versioning the persistent data is pretty much irrelevant wrt the format used --
and yes, I have thought about it. Marmalade right now uses it's own versioned
state-getters/setters (jellyToDOM_X where X is the version number), but AOT
doesn't have its own interface -- it just uses __getstate__/__setstate__ like
pickle. I'm probably going to write a mixin class for versioned state that
supports methods like __getstate_X__ where X is the version of the persisted 
state.

Note that Twisted has a class for versioning state that's orthogonal to
persisted format: twisted.persisted.styles.Versioned. This allows you
to define 'upgradeToVersionX' methods that will be called in order to
upgrade an object to the current version of the code. The only reason
to use versioned state getters/setters as opposed to this is if you want
the format of the persistent state to change for cosmectic reasons.


-- 
                                Chris Armstrong
                         << radix at twistedmatrix.com >>
                http://twistedmatrix.com/users/carmstro.twistd/





From pyth at devel.trillke.net  Thu Jun 13 18:09:50 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 14 Jun 2002 00:09:50 +0200
Subject: dictionary: keys() and values()
In-Reply-To: <3D09160A.2030402@NOSPAMREMOVETHISxs4all.nl>; from irmen@NOSPAMREMOVETHISxs4all.nl.trillke.net on Fri, Jun 14, 2002 at 12:00:42AM +0200
References: <m2elfbuhn1.fsf@mother.paradise.lost> <yu991ybbdldd.fsf@europa.research.att.com> <3D09160A.2030402@NOSPAMREMOVETHISxs4all.nl>
Message-ID: <20020614000950.L6609@prim.han.de>

Irmen de Jong wrote:
> Andrew Koenig wrote:
> > From the library reference, 2.2.7, footnote 3:
> > 
> > If keys() and values() are called with no intervening modifications to the
> > dictionary, the two lists will directly correspond.  This allows the creation
> > of (value, key) pairs using zip():  "pairs = zip(a.values(), a.keys())".
> 
> Why would you want to do that if you have .items()?

i almost fell for the same question untill i looked harder<wink>.

order-sometimes-matters-ly y'rs, Holger




From mwh at python.net  Wed Jun 12 13:04:02 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 12 Jun 2002 17:04:02 GMT
Subject: JIT Complier for python?
References: <8d3f4438.0206111905.7934664b@posting.google.com> <20020612120936.W6609@prim.han.de> <mailman.1023897196.15652.python-list@python.org>
Message-ID: <lk7kl4xwd1.fsf@pc150.maths.bris.ac.uk>

Skip Montanaro <skip at pobox.com> writes:

>     holger> d2002xx wrote:
>     >> Is there any JIT Compiler for Python?
> 
>     holger> you might want to take a look at
> 
>     holger> http://homepages.ulb.ac.be/~arigo/psyco/
> 
> I think the Psyco project on Sourceforge might be more up-to-date:
> 
>     http://sourceforge.net/projects/psyco

Same thing (as in both of those pages link to each other).

Cheers,
M.

-- 
  People think I'm a nice guy, and the fact is that I'm a scheming,
  conniving bastard who doesn't care for any hurt feelings or lost
  hours of work if it just results in what I consider to be a better
  system.                                            -- Linus Torvalds


From cliechti at gmx.net  Fri Jun 14 16:10:52 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 14 Jun 2002 22:10:52 +0200
Subject: PythonWin - an operational question
References: <3D09F28B.19AB5929@sympatico.ca>
Message-ID: <Xns922DE20C740ABcliechtigmxnet@62.2.16.82>

"Colin J. Williams" <cjw at sympatico.ca> wrote in
news:3D09F28B.19AB5929 at sympatico.ca: 
> I am using PythonWin with WinXP.
> <p>I would like to be able to tile the Editor and Interactive windows,
> but use the editor area for all programs viewed.&nbsp; Is there some
> simple way to arrange this as the default set up?
> <p>Tiling doesn't seem to have a registry setting and doesn't do quite
> what I would like.
> <p>There are registry settings which permit one to place the
> Interactive and Editor windows on the screen.&nbsp; Is this the way to
> go? <p>Colin W</html>

goto options->general options and check "dockable window" in the 
"interactive window" section.
i think you must restart the editor to make it work. however this might be 
close to that what you  want.

chris

PS: please try to post text only, no HTML. many people will apreciate it.

-- 
Chris <cliechti at gmx.net>



From pyth at devel.trillke.net  Sun Jun 16 14:04:39 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 16 Jun 2002 20:04:39 +0200
Subject: Snort alert tail...
In-Reply-To: <1024248055.1899.3.camel@localhost.localdomain>; from logiplexsoftware@earthlink.net on Sun, Jun 16, 2002 at 10:20:54AM -0700
References: <AB0P8.107$W91.886@news1.global-one.fi> <1024248055.1899.3.camel@localhost.localdomain>
Message-ID: <20020616200439.J15079@prim.han.de>

Cliff Wells wrote:
> On Sun, 2002-06-16 at 05:51, Jan-Eric wrote:
> > HEllo !
> > I'm trying to write a dynamic firewall script in python that scans the Snort 
> > alert file like 'tail -f' and takes action based on the infomation it gets 
> > from that file. But I can't get the 'tail' function to work.It reads the 
> > file, but any new information that Snort is writing to the file doesn't 
> > show up to the script.
> > 
> > ex.
> > file = open('/var/log/snort', 'r')
> > while 1:
> >         file = file.read()
> >         print file
> > ....
> 
> 
> import os, stat
> import time
> 
> def snort(pathname):
>     size = os.stat(pathname)[stat.ST_SIZE]
>     
>     while 1:
>         lastsize = size
>         size = os.stat(pathname)[stat.ST_SIZE]
>         if size > lastsize: # there's new data
>             f = open(pathname, 'r')
>             f.seek(lastsize)
>             print f.readlines()
>             f.close()
>         else:
>             time.sleep(2)

i'd recommend checking for 'size<lastsize' also and setting lastsize=0
in this case. Otherwise a rewritten/truncated file gets you bad
behaviour.

regards,

    holger




From mt at open2web.com  Thu Jun 20 23:40:04 2002
From: mt at open2web.com (Don Low)
Date: 21 Jun 2002 03:40:04 GMT
Subject: newbie string question
References: <aeu40s$9jg2s$1@ID-145503.news.dfncis.de> <3D129BB7.C7EA15D2@engcorp.com>
Message-ID: <aeu76k$9brq1$1@ID-145503.news.dfncis.de>

>    0   1   2   3   4   5    <-- subscript index
>  +---+---+---+---+---+---+
> "| P | y | t | h | o | n |"
>  +---+---+---+---+---+---+
>  0   1   2   3   4   5   6  <-- slice index
> 
> Your slice went from 2 to 5, so it got only the "tho" part of the string.
> Here's what you got:
> 
>  +---+---+---+
> | t | h | o |
>  +---+---+---+
>  2   3   4   5
> 
> Other languages confuse the indices used when slicing things (as in
> the common substring() method) and when subscripting things (when you 
> take just a single item, such as pystr[4] which is "o" in the above).
> 
> Python avoids up this confusion by realizing they are not the same
> thing.  (Of course, this can confuse newcomers who don't realize 
> what happened, but at least the pain happens for you only once. :-)
> 
So if I understand, pystr[3] refers to the subscript index, but pystr[3:]
refers to the slice index. pystr[2:5] means slice at slice index 2 and 5,
not start at subscript index 2 and slice at slice index 5.

I just want to understand this once and for all.


-- 
Thanks,

Mark


From morton at dennisinter.com  Thu Jun 13 00:33:16 2002
From: morton at dennisinter.com (damien  morton)
Date: 12 Jun 2002 21:33:16 -0700
Subject: Something faster then sgmllib for sucking out URLs
References: <mailman.1023913936.26808.python-list@python.org>
Message-ID: <4abd9ce7.0206122033.67f31e42@posting.google.com>

Alex Polite <m2 at plusseven.com> wrote in message news:<mailman.1023913936.26808.python-list at python.org>...
> I'm working on a webspider to fit my sick needs. The profiler
> tells me that about 95% of the time is spent in sgmllib. I use sgmllib
> solely for extracting URLs. I'm looking for a faster way of doing
> this. Regular expressions, string searches? What's the way to go? I'm
> not a python purist. Calling some fast C program with the html as
> argument and getting back a list of URLs would be fine by me.

Ive got a ittle webspider that can max out my cable connection.

You can download it from www.bitfurnace.com/python/crawler.py

Uses regexes to grab anything that looks like a url.


From gerhard at bigfoot.de  Tue Jun  4 11:55:20 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 4 Jun 2002 15:55:20 GMT
Subject: Need advice on multithreading problem
References: <slrnafpeei.p8.gerhard@developer2.opus-gmbh.net> <adih99$rkm$1@panix1.panix.com>
Message-ID: <slrnafpo8j.15o.gerhard@developer2.opus-gmbh.net>

In article <adih99$rkm$1 at panix1.panix.com>, Aahz wrote:
> Gerhard H?ring <gerhard at bigfoot.de> wrote:
>>
>>The problem is this: I have a C library with a function of the following
>>interface:
>>
>>    def sqlite_exec(conn, sql, callback, arg1):
>>
>>The only important parameter here is 'callback', it is a callback
>>function that you need to give to sqlite_exec. The callback function
>>will be called with the items sqlite_exec returns. If the callback
>>functions returns a value other than zero, sqlite_exec will stop its
>>processing.
>>
>>On the Python side, I wanted to force this C library call into a
>>generator.  Which I did using a thread that's spawned for a new
>>sqlite_exec call and Queue.Queue.
> 
> Is SQLite thread-safe?

If compiled with the right options, it is.

> Do you provide a thread-local conn?

No, but I can guarantee that the connection will only be accessed by one thread
at a time. I believe the current code already does guarantee that.

> Does sqlite_exec() return a brand-new cursor?

As you can see in the code, it doesn't return anything. All it does is call the
callback function. The implementation of my callback function writes to a
queue, and the fetchone/fetchmany/fetchall methods of Cursor read from the
queue.

> If the answer to any question is "no", your code is probably broken (haven't
> actually read it).

Thanks for commenting, but I did knew these basics of multithreading already.

Gerhard


From sholden at holdenweb.com  Tue Jun 25 11:02:15 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 25 Jun 2002 11:02:15 -0400
Subject: Python daemon
References: <mailman.1024999953.15765.python-list@python.org> <F733C874CD014E16.148E57261FFD5E7A.AFFA4BD17B2F8D8B@lp.airnews.net>
Message-ID: <WA%R8.21229$mF.14665@atlpnn01.usenetserver.com>

"Cameron Laird" <claird at starbase.neosoft.com> wrote ...
> In article <mailman.1024999953.15765.python-list at python.org>,
> Roman Suzi  <rnd at onego.ru> wrote:
> .
> .
> .
> >> 3. restarting of the program if it fails, or the watchdog doesn't
trigg=
> >er
> >
> >Run the program from the /etc/inittab - if it terminates,=20
> >init will restart it for you.
> .
> .
> .
> This is one of the great FMMs under Unix--the under-utilization
> of init(1).  From everything I know, Roman's exactly right, that
> /etc/inittab is the right way to set up a process that you truly
> want to keep running.  However, many, MANY Unix hosts have all
> kinds of ugly home-brewed hacks to duplicate this functionality.
> I don't have an explanation, beyond the rather tautologous ob-
> servation that init(1) simply isn't as widely understood as it
> deserves to be.


You're probably right. Don't forget you need to specify "respawn" to have
the process restarted when it terminates.

Of course, this doesn't address the issue of if the controlled process stops
making progress but doesn't die.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From alf at logilab.fr  Tue Jun 11 05:00:05 2002
From: alf at logilab.fr (Alexandre Fayolle)
Date: Tue, 11 Jun 2002 09:00:05 +0000 (UTC)
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Jun 10)
References: <EE3CDB256E0A231C.577F9FCCC77B96EC.B451C214671092A5@lp.airnews.net>
Message-ID: <slrnagbf0l.7hd.alf@virgo.logilab.fr>

Hello,

Sorry to interrupt, but I have to correct a misleading statement in the
latest Dr. Dobb's Python URL. 

Kragen Sitaker a ?crit :

> Programming Libraries:
> 
>     Logilab's 'constraint', a constraint satisfaction problem
>     programming environment for Python, which lets you program in
>     Prolog style in Python, has released version 0.2.
      ^^^^^^^^^^^^
This is not true. For programming in Prolog style in Python, you can use
Christophe Delord's Pylog (http://christophe.delord.free.fr/en/pylog/).
This will give you facts, and rules, as you would have using SWI-Prolog
or GnuProlog. 

The constraint module won't let you do variable unification as you would
do it in prolog. It takes a set of variables with their domains and a
set of constraints on the values of these variables to find all possible
simultaneous assignments to these variable that satisfies the
constraints. 

Cheers,

Alexandre Fayolle
-- 
LOGILAB, Paris (France).
http://www.logilab.com   http://www.logilab.fr  http://www.logilab.org
Narval, the first software agent available as free software (GPL).


From quinn at upchuck.ugcs.caltech.edu  Thu Jun 27 01:47:40 2002
From: quinn at upchuck.ugcs.caltech.edu (Quinn Dunkan)
Date: 27 Jun 2002 05:47:40 GMT
Subject: Suggestions for good programming practices?
References: <mailman.1024970193.27729.python-list@python.org>
Message-ID: <slrnahl9ns.fd3.quinn@upchuck.ugcs.caltech.edu>

On Mon, 24 Jun 2002 18:55:22 -0700, David LeBlanc <whisper at oz.net> wrote:
>They may be returned in a tuple, but they're PUT in discrete variables and
>not kept in a tuple that I have to programmatically unpack.
>
>Language lawyer should be shot, _then_ heard.

So you'd prefer whole threads like this:


> > AAAAAAAAAAAAAAAAAAAAAaaaaaaaaaaaahhhhhhhhhhhhhhhhhhhhhhhh
> > 
> > .
> > 
> > .
> > 
> > .
> > 
> > *gurgle*
> 
> *hurk* ... *thud*

Yaaaaaaahhhhhh!!!!!  *wheeze*  SHRIEK!!


From sholden at holdenweb.com  Tue Jun 18 07:50:53 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 18 Jun 2002 07:50:53 -0400
Subject: Writing a simple linux daemon type thing...
References: <K9eO8.393$LN6.261@newsread1.prod.itd.earthlink.net> <rn3cea.rt4.ln@127.0.0.1> <yylO8.916$TP4.88596@newsread2.prod.itd.earthlink.net> <aecrfe$a4l$1@minotaurus.cip.informatik.uni-muenchen.de> <mailman.1024063578.5866.python-list@python.org>
Message-ID: <VaMP8.165345$Fp1.57388@atlpnn01.usenetserver.com>

"Michael Gilfix" <mgilfix at eecs.tufts.edu> wrote ...
>   Wouldn't it be nice if some daemonizing code and basic log support
> was built into SocketServer? SocketServer aims at simplifying
> the infrastructure for people who don't need a full-blown server
> app. Might also be a nice addition for smaller projects who don't want
> to deal with this sort of thing... Might make an interesting patch
> IMHO.
>

That wouldn't be quite such a good idea as it sounds. The SocketServer code
is only really intended as a demonstration of basic techniques, and
certainly hasn't been coded as industrial-strength. There would probably be
aa lot more work in the patch than you anticipate.

Any daemon process needs to ensure it

1. Detaches from its controlling terminal, using safe alternatives to stdout
and stderr, and

2. Creates a new process group so it is invulnerable to signals received by
the triggering process.

SocketServer and their derivatives don't do either of these things IIRC

regards
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From op73418 at mail.telepac.pt  Wed Jun 26 12:04:32 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 26 Jun 2002 17:04:32 +0100
Subject: question on properties
Message-ID: <p4pjhuc69jjqs0vvfkkmqd37rr52cq9qv3@4ax.com>

Hi,

Suppose in a class Klass I have a property prop. Suppose also that I
have

class AnotherKlass(Klass):
    ...

I can override property prop in AnotherKlass by just redefining it, but
what if I want to acess the base class prop how do I do it?

TIA and all the best,
Gon?alo Rodrigues


From kragen at pobox.com  Mon Jun  3 12:38:05 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 03 Jun 2002 12:38:05 -0400
Subject: HTML links in a GUI tkinter
References: <mailman.1022986323.16196.python-list@python.org> <83elfprs1x.fsf@panacea.canonical.org> <C7JK8.180067$%u2.89342@atlpnn01.usenetserver.com>
Message-ID: <83y9dwqpw2.fsf@panacea.canonical.org>

"Steve Holden" <sholden at holdenweb.com> writes:
> "Kragen Sitaker" <kragen at pobox.com> wrote in message
> news:83elfprs1x.fsf at panacea.canonical.org...
> > Sorry about the empty reply --- in Python 2.x, the webbrowser module
> > has a webbrowser.open() method that does what you want.
>
> So, all Aurel needs to know now is how to run wxPython widgets inside
> tkinter programs...

Why would Aurel want to do that?  webbrowser.open() doesn't use
wxPython.  At least, I don't have wxPython installed, and it works for
me.



From ark at research.att.com  Fri Jun  7 14:32:07 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 7 Jun 2002 14:32:07 -0400 (EDT)
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <15616.64407.863059.729351@12-248-41-177.client.attbi.com>
	(message from Skip Montanaro on Fri, 7 Jun 2002 13:29:43 -0500)
References: <yu99vg8v6orb.fsf@europa.research.att.com>
        <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
        <200206071747.g57HlbH01057@europa.research.att.com>
        <15616.62634.635544.905055@12-248-41-177.client.attbi.com>
        <200206071804.g57I4fp01212@europa.research.att.com>
        <15616.63342.559215.644086@12-248-41-177.client.attbi.com>
        <200206071815.g57IFjg01303@europa.research.att.com> <15616.64407.863059.729351@12-248-41-177.client.attbi.com>
Message-ID: <200206071832.g57IW7n01369@europa.research.att.com>

Skip> Dunno.  Try "find . -name 'struct*'.  It should be in your build/lib*
Skip> directory (where "*" matches your configuration name).

Did that; didn't find anything obvious.

Before I do anything else, I'm going to try rebuilding from scratch
using GCC 2.95.something to be sure I haven't changed anything else
that might have broken it.






From irmen at NOSPAMREMOVETHISxs4all.nl  Thu Jun 13 18:00:42 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Fri, 14 Jun 2002 00:00:42 +0200
Subject: dictionary: keys() and values()
References: <m2elfbuhn1.fsf@mother.paradise.lost> <yu991ybbdldd.fsf@europa.research.att.com>
Message-ID: <3D09160A.2030402@NOSPAMREMOVETHISxs4all.nl>

Andrew Koenig wrote:
> From the library reference, 2.2.7, footnote 3:
> 
> If keys() and values() are called with no intervening modifications to the
> dictionary, the two lists will directly correspond.  This allows the creation
> of (value, key) pairs using zip():  "pairs = zip(a.values(), a.keys())".

Why would you want to do that if you have .items()?

Irmen



From phr-n2002b at NOSPAMnightsong.com  Sat Jun 29 16:15:02 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Jun 2002 13:15:02 -0700
Subject: private
References: <pan.2002.06.29.17.38.51.789527.440@myself.com> <3D1DE132.4A890D9D@engcorp.com> <afl3e4$f8abc$1@ID-71831.news.dfncis.de>
Message-ID: <7x3cv5j0zd.fsf@ruckus.brouhaha.com>

"James Kew" <james.kew at btinternet.com> writes:
> Yes, of course one can circumvent private in C++. But for me the main value
> of the public and private specifiers is that they _document_ to the client
> of the class what it should access and what it should not.

C++ may have been a bad example since the process's whole address
space is exposed to all parts of the program.  In Java though, private
methods and variables are enforced.

Actually someone should mention the rexec/Bastion module in Python,
which gives a kludgy way to enforce privacy within a process.  It
makes a sandbox-like mechanism where an object can enforce access
controls against a caller provided that the caller runs wrapped in a
special restricted environment.  


From bokr at oz.net  Sat Jun 29 12:38:24 2002
From: bokr at oz.net (Bengt Richter)
Date: 29 Jun 2002 16:38:24 GMT
Subject: question od default args
References: <o3vjhuguen6gfm8vnq7rbk5volmi80shl5@4ax.com> <YynS8.13571$p56.4313990@newsb.telia.net> <aff4ap$aqm$2@news.wrc.xerox.com> <6NhT8.45584$n4.10824585@newsc.telia.net>
Message-ID: <afknq0$nla$0@216.39.172.122>

On Sat, 29 Jun 2002 12:32:02 GMT, "Fredrik Lundh" <fredrik at pythonware.com> wrote:

>Mark Jackson wrote:
>
>> More quickly:
>>
>>     def myfunc(arg=[]):
>>         if arg is myfunc.func_defaults[0]:
>>             print "no argument"
>>         else:
>>             ...
>
>while the func_ attributes are documented, a quick peek in
>the standard library should tell you that it's best to treat them
>as an implementation detail.
>
>the less such stuff you depend on in your production code,
>the more portable it gets...
>
Perhaps accessing "implementation details" should depend on importing
a module that enables access one way or another? E.g.,

    import __version_2_2_implementation_details__

could serve to signal that Guido et al didn't really
mean to define a portable abstraction for production use?

Regards,
Bengt Richter


From shalehperry at attbi.com  Wed Jun 26 13:48:56 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 26 Jun 2002 10:48:56 -0700 (PDT)
Subject: question od default args
In-Reply-To: <o3vjhuguen6gfm8vnq7rbk5volmi80shl5@4ax.com>
Message-ID: <XFMail.20020626104856.shalehperry@attbi.com>

On 26-Jun-2002 Gon?alo Rodrigues wrote:
> Hi,
> 
> When I want default args I usually do
> 
> def (arg = None):
>     ...
> 
> But what if I want to make None a reasonable argument too? That is, I
> want to know if the user has in fact passed an argument, and if not to
> do something special, with None (or whatever object) being a reasonable
> arg to be passed.
> 

>>> def func(arg1, *args):
...   if len(args):
...     print "user gaves us %d args" % (len(args) + 1,)
...   else:
...     print "only have arg1"
... 
>>> func('sean')
only have arg1
>>> func('sean', 'perry')
user gaves us 2 args

this means you need to throw an exception if you only want 3 args and the user
gives you 4 (or more).




From nospam at bigfoot.com  Tue Jun 18 07:23:14 2002
From: nospam at bigfoot.com (Gillou)
Date: Tue, 18 Jun 2002 13:23:14 +0200
Subject: Is there a Text to HTML conversion module?
References: <mailman.1023132485.29235.python-list@python.org> <adgn01$d18$1@slb5.atl.mindspring.net> <3CFBDF95.92099C9E@alcyone.com> <aempf6$1nq2$1@news.idiom.com>
Message-ID: <aen4n1$35b$1@norfair.nerim.net>

Did not test but...

Zope has got a StructuredText package that can be used "standalone" (outside
of a Zope site).

www.zope.org

Otherwise, here's a simple snip that will help if you want to do it by
yourself : It changes the markup from the original plain text and replaces
against entities and changes http, ftp and mail markups against hyperlinks.
Can be improved of course.

import string, re
# Subtitute HTML markup chars by entities
text = string.replace(text, '&', '&amp;')
text = string.replace(text, '<', '&lt;')
text = string.replace(text, '>', '&gt;')
text = string.replace(text, '\n', '</p>\n<p>')
# Make hyperlinks from HTTP or FTP URLs
pat = re.compile(r'((ftp|http)://[\w-]+(?:\.[\w-]+)*(?:/[\w\-\.?=]*)*)')
text = pat.sub('<a href="\\1">\\1</a>', text)
# Make hyperlinks from email addresses
pat = re.compile(r'([\w-]+(?:\.[\w-]+)*@[\w-]+(?:\.[\w-]+)*)')
text = pat.sub('<a href="mailto:\\1">\\1</a>', text)

HTH

--Gilles

"James T. Dennis" <jadestar at idiom.com> a ?crit dans le message news:
aempf6$1nq2$1 at news.idiom.com...
> Erik Max Francis <max at alcyone.com> wrote:
> > Andrew Dalke wrote:
>
> >> <wink/>
>
> >>>(sorry... couldn't resist)
> >> Neither could I.  :)
> > Ah, but that's XML, not HTML.  :-)
>
>  Of course it might be some future extension to XHTML
>



From syver-en+usenet at online.no  Wed Jun  5 07:41:13 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: 5 Jun 2002 04:41:13 -0700
Subject: manipulating child processes under Windows
References: <lcptz7zw0m.fsf@gaffa.mit.edu> <uadqb3q9v.fsf@online.no> <lcy9duhksg.fsf@gaffa.mit.edu>
Message-ID: <d7226884.0206050341.61fa2ab0@posting.google.com>

Douglas Alan <nessus at mit.edu> wrote in message news:<lcy9duhksg.fsf at gaffa.mit.edu>...
> Syver Enstad <syver-en+usenet at online.no> writes:
> 
> > Hmm, my first reaction is to think in terms of window handles in
> > some way, there's a FindWindow function in win32 (and win32all) that
> > let's you search for a window by name, I'll have to check the msdn
> > library to find out how to get the owning process handle from a
> > window handle.
> 
> What if there were more than one process with the same name.  How
> would I tell which one is the right one?

The FindWindow function doesn't search for the ProcessName it uses the
window title, which is often unique. Hope that helps.


From irmen at NOSPAMREMOVETHISxs4all.nl  Thu Jun 27 18:05:16 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Fri, 28 Jun 2002 00:05:16 +0200
Subject: OT: timeout with SSL sockets? (was Re: How to stop a SocketServer?
 )
References: <7xhejv9af2.fsf_-_@ruckus.brouhaha.com>        <3d1744d4$1_1@hpb10302.boi.hp.com>        <3D1757DF.776A428F@engcorp.com>        <3d176a0c$1_3@hpb10302.boi.hp.com>        <3D17BF72.2E1CCF40@engcorp.com> <mailman.1024970132.27318.python-list@python.org>
Message-ID: <3D1B8C1C.2020802@NOSPAMREMOVETHISxs4all.nl>

Skip Montanaro wrote:
>     Peter> By the way, look into "timeoutsocket.py" which is a plug-in
>     Peter> replacement for the standard socket.py and which might do the
>     Peter> trick for you.
> 
> Note socket objects grow a settimeout() method in 2.3.  If you get Python
> from CVS you can have it today:
[...] snip

That is nice indeed. Will this also work with (M2Crypto) SSL sockets?
I've had some struggles with M2Crypto SSL sockets, timeouts and select...

Thanks
Irmen de Jong



From dan at eevolved.com  Mon Jun  3 13:43:58 2002
From: dan at eevolved.com (Daniel Parisien)
Date: Mon, 03 Jun 2002 13:43:58 -0400
Subject: encoding and decoding repr'd version of strings
Message-ID: <1YNK8.9622$cE6.591635@wagner.videotron.net>

Hi 
is there an codec that bundles with python so you can transform strings 
into the repr'd (escaped) version and back? If not, how easy would it be to 
code? I don't want to use eval(...) because I don't trust the source of the 
string and I might mistakenly eval an expression like "' '*(2**30)" which 
would allocate about 1 GB of data (eek!)

Thanks,
Dan


From sarmstrong13 at mac.com  Mon Jun  3 03:10:11 2002
From: sarmstrong13 at mac.com (SA)
Date: Mon, 03 Jun 2002 02:10:11 -0500
Subject: Is there a Text to HTML conversion module?
Message-ID: <B9208083.5EF6%sarmstrong13@mac.com>

Does anyone know of a module that converts a text file to html?

I started writing my own, but decided I better make sure I'm not
re-inventing the wheel here.

Thanks.
SA



From marklists at mceahern.com  Fri Jun 28 16:20:47 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 28 Jun 2002 15:20:47 -0500
Subject: Introspection: determining parent classes?
In-Reply-To: <3D1CC1CA.70004@onsitetech.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBAEANLHAA.marklists@mceahern.com>

isinstance will do the trick.

class A:pass
class B(A):pass
class C(B):pass

c = C()
print isinstance(c, A)
1

If you need to traverse the bases:

c.__class__.__bases__

// m
-





From sheershion at mailexpire.com  Fri Jun 21 09:30:50 2002
From: sheershion at mailexpire.com (Robert Amesz)
Date: Fri, 21 Jun 2002 13:30:50 -0000
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.>
Message-ID: <Xns923496C17397Frcamesz@amesz.demon.nl>

Angela Williams wrote:

> I'm looking for the equivalent of the GOTO command for Python
> 2.2.1 running on Windows 98. I have programmed very little and am
> new to Python. I found the following information on the FAQ page
> but it doesn't make any sense to me.

Pleas don't go looking for GOTOs. If you find yourself needing those, 
in 99.9% of the cases there's something not quite right with your code. 
Usually, by splitting a large function/method into several smaller ones 
the problem will go away by itself. The resulting code will also be 
clearer and more maintainable.

(If after that you're still having trouble, post a page or so of your 
code here. This is a helpful group, and I'm sure somebody will give you 
some points on how to improve your code.)

The most goto-like contructs in Python are 'return' (which exits from a 
function/method before the end is reached) and 'break' (which exits a 
loop, continuing with statement followin that loop).

In a pinch you can indeed raise some kind of exception, but you really 
should use those for things like errors.


Robert Amesz


From python at rcn.com  Wed Jun 26 02:54:37 2002
From: python at rcn.com (Raymond Hettinger)
Date: Wed, 26 Jun 2002 02:54:37 -0400
Subject: buffer() survey
Message-ID: <afbojd$hds$1@bob.news.rcn.net>

I'm interested in learning knowing whether the builtin function buffer() is
much used and how it is being used.

If anyone has this in their own real-world code, I would be interesting in
hearing from you.  Also, it would be helpful to know whether use of the
array or mmap modules was considered.


Raymond Hettinger




From aahz at pythoncraft.com  Wed Jun 12 13:57:16 2002
From: aahz at pythoncraft.com (Aahz)
Date: 12 Jun 2002 13:57:16 -0400
Subject: postscript processing in python
References: <ae7t0a$f4t$1@news.asu.edu>
Message-ID: <ae821s$mi9$1@panix1.panix.com>

In article <ae7t0a$f4t$1 at news.asu.edu>, David Bear  <david.bear at asu.edu> wrote:
>
>I know these are tall orders but I've written a filter in python that takes 
>a postscript stream generated by an IBM mainframe and uses ghostscript to 
>convert it to pdf.  The problem is ghostscript seems rather picky about 
>some strange things that the mainframe puts in the postscript stream -- so 
>I was hoping to do some cleanup on it and conversion without having to 
>depend on ghostscript.  

If you're already handling postscript, try using ReportLab to handle
creating the PDF: http://www.reportlab.com/
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From gcordova at hebmex.com  Thu Jun 13 17:18:15 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 13 Jun 2002 16:18:15 -0500
Subject: Type subclassing: bug or feature
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233012631BC@mail.hebmex.com>

> >
> > Consider the following code:
> >
> > class MyStr(str):
> >     def contains(self, value):
> >         return self.find(value) >= 0
> >
> > s = MyStr("hello, world!")
> > s = s.capitalize()
> > if s.contains('Hello'):
> >     print "Found it!"
> >
> > It fails with an AttributeError when it calls s.contains(), because
> > s.capitalize() returned a str instead of a MyStr.  Anyone 
> want to take a
> > whack at defending this as the correct behavior?
> > --
> > Aahz (aahz at pythoncraft.com)           <*>
> > http://www.pythoncraft.com/
> 
> Looks like a bug to me. How can self be something else then 
> what it is?
> Should MySocket return a socket, thus stripping off any additional
> functionality added by my derived class?
> 
> David LeBlanc
> Seattle, WA USA
> 

Hmmm...

maybe all methods of builtin types should do:

	return self.__class__(return_value)

when returning a new object (like str methods).

-gustavo




From ak at silmarill.org  Sat Jun  1 23:16:41 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Sun, 02 Jun 2002 03:16:41 GMT
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org> <332c7c11.0206011639.a396f84@posting.google.com>
Message-ID: <slrnafj3ib.86f.ak@ak.silmarill.org>

In article <332c7c11.0206011639.a396f84 at posting.google.com>, Benno wrote:
> Andrei Kulakov <ak at silmarill.org> wrote in message
> news:<slrnafhmoa.389.ak at ak.silmarill.org>...
>> Hello,
>> 
>>     "hello world" type program from wxPython tutorial takes 6
>> seconds to start on my cel 366 128mb system (running Debian). Is
>> this typical? Why is it so slow? I'm looking for a graphic
>> toolkit that'd be fairly quick. I don't want to use Tk because of
>> the looks.
> 
> I was using wxPython for a while. The main advantage being that it was
> going to be cross platform, however I found its speed too slow. (Not startup
> speed, which is fairly irrelevant to me, but just general responsiveness.)
> I'm not sure if this was just wxWindows itself or the python bindings.
> 
> I have started using pygtk <http://www.daa.com.au/~james/pygtk/> (easily
> apt-getable). The only problem I've had with this is the lack of docs.
> http://www.moeraki.com/pygtktutorial/gtk-tut.html, seems to be the best 
> available, followed by reading the source.
> 
> Cheers,
> 
> Benno

Thanks, I think that's exactly what I'm gonna have to do.

-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From eric.brunel at pragmadev.com  Mon Jun  3 14:08:21 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 3 Jun 2002 18:08:21 +0000
Subject: Changing TK
References: <adfv3d$jsp$1@netnews.upenn.edu>
Message-ID: <adg3uo$p6b$1@wanadoo.fr>

Salim Zayat wrote:
> Hey all.  I have a quick question about something.  If I were to install
> a newer version of Tcl/Tk on my computer, would Tkinter.py recoginze it
> automatically, or do I need to tweak the Tkinter.py code itself?

On what kind of platform are you? Things will be a lot different if you're 
on Unix or Windows...

In fact, if your tcl/tk installation changes, there's two things that must 
be taken into account: the dynamic libraries used at run-time, and the 
tcl/tk API header files used when compiling the Python interpreter. 
However, both are not referenced at all in Tkinter.py, so "tweaking" it 
would not be of any help... When you get a new tcl/tk installation, you 
*must* recompile the interpreter, or you won't ever see any change.

If you're on Windows, you probably didn't compile your interpreter 
yourself. If you want to do it, you'll have to download the interpreter's 
sources. I can't help you any further here, since I didn't ever compile an 
interpreter on Windows. You'll have to reference your new tcl/tk 
installation somewhere, but I don't know how it can be done.

On Unix, you should already have the interpreter's sources. So go to the 
subdirectory "Modules" of your Python installation directory, edit the file 
named "Setup", change the part for the "_tkinter" module to reference the 
new location for your tcl/tk installation, recompile the interpreter and 
voila!

But DON'T EVER replace former tcl/tk libraries by newer ones without 
recompiling, even if they have the same name! It *may* work, but if it 
doesn't, the tk support in your Python will be completely and definitely 
broken...

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From sanjay2kind at yahoo.com  Wed Jun 26 07:45:28 2002
From: sanjay2kind at yahoo.com (sanjay)
Date: 26 Jun 2002 04:45:28 -0700
Subject: how to remove HTML  attributes from web pages ( HTML parseing)
Message-ID: <63170f57.0206260345.2b61b7ec@posting.google.com>

Hi,

 New to python and doing one application. i would like to  web page
content
after removing specific html tag, attribute etc. Like taking out
color,bgcolor and <font>,<p> tag from the web pages. Which one will be
helpful : regular expression  or HTMLParser  helpfully.


Thx,


From jepler at unpythonic.net  Tue Jun 25 19:45:54 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Tue, 25 Jun 2002 18:45:54 -0500
Subject: bug in os.getgroups?
In-Reply-To: <hh3cvbja9p.fsf@nightshade.la.mastaler.com>
References: <1025033152.17806.TMDA@nightshade.la.mastaler.com> <20020625202442.GG4435@unpythonic.net> <hh3cvbja9p.fsf@nightshade.la.mastaler.com>
Message-ID: <20020625184554.A7559@unpythonic.net>

On Tue, Jun 25, 2002 at 03:53:06PM -0600, Jason R. Mastaler wrote:
> Jeff Epler <jepler at unpythonic.net> writes:
> 
> > This is liely to be because /tmp is "sticky":
> 
> /tmp was a bad choice for my example.  The problem is the same in a
> non-sticky directory such as /home.

This is still a matter of Unix permissions.  The action of removing a file
from a directory depends on the permission to modify the directory, not
permission to modify the file.

Jeff




From erict at millfilm.co.uk  Thu Jun 20 14:52:09 2002
From: erict at millfilm.co.uk (Eric Texier)
Date: Thu, 20 Jun 2002 19:52:09 +0100
Subject: Proper number alignment
References: <aet5hu$9leln$1@ID-98166.news.dfncis.de>
Message-ID: <3D122459.2E338453@millfilm.co.uk>

You have a function

newstring1 = string.rjust(astring 1,numberofblankSpace)
newstring2 = string.rjust(astring 1,numberofblankSpace)

finalLine = newstring1 + newstring2

mdk wrote:

> I am using \t\t (tabs) to align these numbers (disregard the math):
>
> Total RAM                2,141,056
> Available RAM            12,934,032
> Used RAM                 1,024
>
> But they are left justified and I need them right justified.
>
> Any ideas?
>
> Thanks.



From karczma at info.unicaen.fr  Thu Jun 20 09:37:31 2002
From: karczma at info.unicaen.fr (Jerzy Karczmarczuk)
Date: Thu, 20 Jun 2002 15:37:31 +0200
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com> <lkznxqjqx2.fsf@pc150.maths.bris.ac.uk> <3D11A21D.9010809@mxm.dk> <uh3e1usajle42c@news.supernews.com>
Message-ID: <3D11DA9B.163A7BD0@info.unicaen.fr>

John Roth after Michael Hudson after Max M.:

> > Goto's are simple ;-)
> > def goto(lable):
> >      lable()
> > def start():
> >
> >      print 'starting'
> >      goto(lable2)
> >
> > def lable1():
> >
> >      print 'at lable 1'
> >      goto(stop)

etc.

> Those aren't goto's in the classical sense. For a real goto,
> you need to think assembler: if it's a statement anywhere,
> you can branch to it, and who cares about the execution
> history! Having the compiler manage your flow control is
> for sissys!

Gentlemen, what are you talking about?
(I confess I didn't follow this thread, so forgive me if I
try to open a non-existent door)

Before, John Roth suggested using structured exceptions as GOTOs,
and Michael Hudson wrote something like that:

> There are also technical difficulties implementing it (the
> blockstack), or I'd have probably done goto functionality in
> bytecodehacks :)


Well...

1. The "solution" of Max M. is erroneous for two reasons.
   A. A procedure call stacks the return address, so a few hundreds
      such gotos, and the system blows up.
      Unfortunately Python does not optimize tail calls, I WONDER WHY;
      for a member of the Functional Programming Church this is
      a mortal sin, and I am serious.
   B. The code chunks linked in such a way provide different variable
      scopes, so there is little sense in calling it a "branching".

2. "Structured exception" is something quite heavy, it may be heavier
   than the setjump/longjump contraption, and may also be useless to
   those who simply would love to do some Python programming ? la
   Fortran. 

3. No, in order to speak about genuine gotos *you don't need to think
   assembler*.
   What you need is to understand correctly the notion of continuation
   (something quite standard - again - in the Functional Programming
   world). This is a high-level model of "flat" branching, but may be
   used to more exquisite constructions, such as backtracking.

   I don't believe that there are any technical difficulties in the
   implementation of gotos at the low level. At the high level this
   is then a question of the language structure, as seen by Van Rossum,
   the lack of gotos was a conscious decision.


Jerzy Karczmarczuk
Caen, France


From tismer at tismer.com  Sat Jun 29 13:46:16 2002
From: tismer at tismer.com (Christian Tismer)
Date: Sat, 29 Jun 2002 19:46:16 +0200
Subject: EuroStackless slides online
Message-ID: <3D1DF268.8040706@tismer.com>

Hi all,

as promised, I've uploaded the slides from my Stackless
talk to the stackless site. It can be downloaded here:

http://www.stackless.com/#EuroSlides

The page has also been updated to reflect the recent
evalution. Some code concerning thread pickling
will be added, soon.

EuroPython was just -- great!
I'm sure this will be repeated next year
or even earlier.

cheers - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From edream at tds.net  Sun Jun 23 06:12:23 2002
From: edream at tds.net (Edward K. Ream)
Date: Sun, 23 Jun 2002 10:12:23 GMT
Subject: ConfigParser & converting strings to lists
References: <3D152A28.2177B054@tds.net>
Message-ID: <3D159F06.2154D3F3@tds.net>

To complete this topic, let me restate the problem and JohnJacob's
elegant solution.

Suppose files is a list of strings, and we have written a configuration
section like this:

  config = ConfigParser.ConfigParser()
  config.set("recent files", "recentFiles, files)

Then the elegant way of retrieving files _as a list_ is:

  files = eval(config.get("recent files", "recentFiles"))

Thanks to all who replied.

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


From K.Rdt at TU-Berlin.DE  Sun Jun 16 14:07:30 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Sun, 16 Jun 2002 18:07:30 GMT
Subject: HT telnet interact with auto login?
Message-ID: <3d0cd198.17917980@news.zrz.tu-berlin.de>

Hi!

I want to combine:

import getpass
import sys
import telnetlib

HOST = "...de"
tn = telnetlib.Telnet(HOST)

user = '...'
password = '...'

1.)
tn.interact()	# this is working but with manually ..
2.)
tn.read_until("login: ")	# this seams to work
			# but without connection
tn.write(user + "\n")
if password:
    tn.read_until("Password: ")
    tn.write(password + "\n")

TiA for each tip
	Klaus



From NKluge at optonline.net  Wed Jun  5 01:45:01 2002
From: NKluge at optonline.net (Nicola Kluge)
Date: Wed, 05 Jun 2002 05:45:01 GMT
Subject: self
References: <mailman.1023208220.19083.python-list@python.org> <TL7L8.79$4p7.4100181@newnews.cc.stevens-tech.edu> <mailman.1023224648.434.python-list@python.org>
Message-ID: <xzhL8.21696$tK.5425664@news02.optonline.net>

This is an interesting idea.  However, your approach doesn't handle the dot
notation.
For example,
If object.x was = to 3 and g was = to 4 I need  that
e.f='object.x*g' gives 12.
Just having e=Evaluator({'a':'3', 'b':'a*17', 'c':
'b*a','object.x':'3','g':'4'})
doesn't work.  While parsing, 'object.x' is not taken as one symbol in your
approach.
I need to have object capabilities in the equations.

Vojin Jovanovic



"holger krekel" <pyth at devel.trillke.net> wrote in message
news:mailman.1023224648.434.python-list at python.org...
> Vojin Jovanovic wrote:
> > Now let me give you the problem with self which is making my life
> > complicated. Consider this class.
> >
> > class Foo:
> >  def __init__(self):
> >   self.__dict__['Equations']={
> >     'a':str(5),
> >     'b':'self.a+5',
> >     'c':'self.b*self.a'}
> >
> > (...)
>
> To get rid of 'self' you might like to try this:
>
> class Evaluator:
>     def __init__(self, eqs={}):
>         self.__dict__['equations']=eqs
>     def __getattr__(self, name):
>         begin={}
>         while 1:
>             try:
>                 return eval(self.equations[name], begin)
>             except NameError, n:
>                 var=str(n).split("'")[1]
>                 begin[var]=getattr(self, var)
>     def __setattr__(self, name, value):
>         self.equations[name]=value
>
> >>> e=Evaluator({'a':'3', 'b':'a*17', 'c': 'b*a'})
> >>>
> >>> e.c
> 153
> >>> e.f='a*c'
> >>> e.f
> 459
>
> I don't really think that 'execing' is a good idea here
> but i don't know your complete usecase.
>
> With a little more care it should be possible to
> catch cyclic equations.
>
> have fun,
>
>     holger
>
>




From michael at damaru.com  Sun Jun  9 13:10:30 2002
From: michael at damaru.com (Michael Davis)
Date: Sun, 09 Jun 2002 13:10:30 -0400
Subject: How to capture stdout into a string?
Message-ID: <2YLM8.905$Yw3.119357@news20.bellglobal.com>

Hi,

I'm using libftp. The FTP.dir() function gets the remote directory listing 
and prints it to standard out. How can I capture that in a string?

If I were using an external process, I know I could open a pipe to it. But 
it's not an external process.

I'm sure there's a tricky way of doing it by getting the file descriptor for 
stdout and reading from it directly, but I was hoping to avoid going that 
low-level. I will if I must though.

Thanks in advance,
-- 
Michael Davis
Damaru
Custom Programming - Web Development - Database Design
http://www.damaru.com
416-540-1284


From opus at value.net  Sun Jun 30 06:09:09 2002
From: opus at value.net (Opus)
Date: Sun, 30 Jun 2002 03:09:09 -0700
Subject: Python needs better error reporting
In-Reply-To: <20020630095830.C20310@prim.han.de>
References: <3D1E5376.32716.DF5C94@localhost>; from opus@value.net on Sun, Jun 30, 2002 at 12:40:22AM -0700
Message-ID: <3D1E7655.29975.167945D@localhost>

Interesting idea.  Would these statistics be built from normal usage of the 
language or from that person's usage?

Actually, just to do the statistics on the language would be very interesting.

On 30 Jun 2002 at 9:58, holger krekel wrote:

> Opus wrote:
> 
> > At the most, I might suggest that there be an option to run just the parser, 
> > with more verbose output.  Maybe even a different parser would be in order.  
> > Idea for anyone????
> 
> i'd suggest a heuristic approach.  Compute statistics for which token
> usually comes after a group of others. Just look at AST-structures not
> values.  This way you can suggest a hint to the user what he might have
> done wrong.  I think it's not feasible and too verbose to tell the user
> what is actually correct.  A simple "Maybe you forgot ..." would cover
> 90% of the cases, i guess.
> 
>     holger


--Opus--

The more absurd the belief, the more deeply it must be held,
the more aggressively it must be promoted and angrily defended
if the <person> is to see himself as right and sane.
                            - Unknown

--------------------------------------------------------
Get added to my Humor list:
mailto:opus at value.net?subject=ADD_HUMOR
Get added to my Neat list:
mailto:opus at value.net?subject=ADD_NEAT
Get my PGP public key:
mailto:opus at value.net?subject=PSEND&body=send%20PublicKEY.asc
Visit My Home Page:
http://value.net/~opus/







From emile at fenx.com  Fri Jun 28 07:53:39 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 28 Jun 2002 11:53:39 GMT
Subject: Be the first on your block to register your computer.
References: <mailman.1025211444.21440.python-list@python.org> <afg1ks$e6lj5$1@ID-71831.news.dfncis.de> <3D1B6E69.1090204@earthlink.net> <IXWS8.368185$o66.975748@news-server.bigpond.net.au>
Message-ID: <77YS8.22903$Uu2.4194@sccrnsc03>

Neil Hodgson
>    I suspect that a Python interpreter could be built out of a set of
> Outlook message processing rules although I wouldn't be the one to
volunteer
> to do it whereas I will volunteer to do the Excel port if it should
ever be
> needed.
>

I volunteer to do the Outlook messaging implementation, if it ever
should be needed.  ;-))

--

Emile van Sebille
emile at fenx.com

---------



From marklists at mceahern.com  Tue Jun 18 06:45:16 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 18 Jun 2002 05:45:16 -0500
Subject: why len(list) instead of list.len()  ?
In-Reply-To: <EhBP8.1$Qn1.1124@news.ecrc.de>
Message-ID: <NCBBLFCOHHDIKCAFGCFBAECKLFAA.marklists@mceahern.com>

[Markus Jais]
> I way wondering why one has to write
>
> list = [2, 3, 5]
> print len(list)
>
> instead of list.len()

<yawn>  Imagine for one moment that you are not the first person to wonder
this.  Why not search the newsgroup's archives for the umpteen times this
topic has come up, peruse them, and if you find a new angle on the
discussion, please, share it with us.

Cheers,

// mark

-





From gscott at peregrine.com  Thu Jun  6 11:30:06 2002
From: gscott at peregrine.com (Gordon Scott)
Date: Thu, 6 Jun 2002 08:30:06 -0700 
Subject: Program to an interface, not an implementation
Message-ID: <D54BF26DB5A7D511B36B00508B9596A402F8EAE2@sdgcaexch2.peregrine.com>

I found it a little bit helpful to think of 
'program to an interface, not an implementation' as 'treat all objects the
same'

   The Composite pattern of DP is a good example.  The example they give is
of a 
grouping/collection/tree of various GUI widgets.  This might be a common
method
of keeping track of and displaying various windows on a panel.  
   When a widget needs to be updated, various things have to happen for
different 
widgets, a Button needs a button update, an edit field needs an edit update
etc...
One way of doing this is to step though the tree, find out what type of
widget we 
are dealing with and call the appropriate draw function

	for widget in tree:
		if widget.type == button.type:
			widget.drawButton()
		elif widget.type == edit.type:
			widget.drawEdit()
		etc..

Of course this is a poor way to do things, and makes adding new types of
widgets to
the tree extremely difficult.  By programming to an interface, you treat all
objects
the same way by calling the same function regardless of the object.  This is
shown in 
the Composite example by having all widgets implement the Graphic interface
which defines
a single Draw() function that specific classes override.  For C++ or Java
this is done by
having the widgets either subclass the parent Graphic class, or (java only)
implementing a
Graphic interface.  Then the tree does not know whether or not it has
Buttons or Edit fields
all, it is dealing with Graphic objects, so all the tree needs to know is
the Graphic.Draw()
function.  This way you can add or modify specific widgets to the tree
pretty easily.
In python the updated code for drawing all the objects in the Composite
might be

	for graphic in tree:
		graphic.Draw()

MUCH simpler than the above.

   Now comes the beautiful part, Python's dynamic typing.  When the GoF
refers to an 'interface'
it is an abstract term that means 'a set of functions an object implements',
which isn't necessarily 
a concrete thing like a Java Interface is.  
   So for two Python objects to share the same interface, (assume the
Graphic interface) above, they
simply have to have the functions that a Graphic object has, they do not
need to be subclassed from
a common parent or related in any way!!!!  When you say object.Draw() python
simply looks to see if 
that object has a Draw() function, if it does great.  This way you can keep
radically different objects
in your Composite tree and treat them the same way regardless of what they
are, no casting or subclassing
required!  Because of it's dynamic typing Polymorphism is a snap in Python.

  You can see this quite frequently in many Python programs that operate
with 'file-like' objects,
ie objects that have read() and write() operations, and these can be
sockets, files, or StringIO's which are not part of any common class
hierarchy yet the programs work normally with any of those.

Hope that helps...


-----Original Message-----
From: Egbert Bouwman [mailto:egbert at bork.demon.nl]
Sent: Thursday, June 06, 2002 3:12 AM
To: python-list at python.org
Subject: Program to an interface, not an implementation


Hello,
Can someone shed some light on the GoF adagium:
'program to an interface, not an implementation'
and especially what it means for programming in Python ?
GoF is not easy.
egbert
-- 
Egbert Bouwman - Keizersgracht 197 II - 1016 DS  Amsterdam - 020 6257991
========================================================================


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




From emile at fenx.com  Sat Jun 15 10:28:14 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 15 Jun 2002 14:28:14 GMT
Subject: Pronouncing '__init__'
References: <mailman.1024150580.2457.python-list@python.org>
Message-ID: <2aIO8.81731$pw3.3985@sccrnsc03>

Mark McEahern
> But I think saying, for example, the str method, in the context of a
user
> class, is clear enough to indicate __str__.  If you're talking about
len(),
> you probably just say "len."  If you're talking about __len__, you
probably
> say, "the len method of x" or just "the len method."  Or "the magic
len
> method."
>

I agree.  I often think in terms of the instances, so I also use 'it's
[ imul | repr | str] method.  When I need to spell it out though, my
choice wasn't in the list of options:  underscore underscore r-e-p-r [
underscore underscore ]


--

Emile van Sebille
emile at fenx.com

---------



From marklists at mceahern.com  Fri Jun 28 11:31:06 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 28 Jun 2002 10:31:06 -0500
Subject: aspect-oriented demo using metaclasses
Message-ID: <NCBBLFCOHHDIKCAFGCFBIEPMLGAA.marklists@mceahern.com>

Hi, I have tried and failed several times to understand metaclasses.  This
morning, I trudged up the learning curve a little further than I have before
and I wanted to share what I discovered.  I've been intrigued by the idea of
using metaclasses as a way to implement aspect-oriented programming (AOP) in
Python.  To me, the promise of AOP is you should ideally write your classes
in complete ignorance of the aspects that you want to plug in later.
Tracing is probably the simplest example of an aspect I can think of.
Persistence is a more complicated aspect (to me, anyway).  So I figured I'd
start with tracing.

I owe much to previous threads on AOP, in particular Pedro Rodriguez's
non-metaclass implementation:


http://groups.google.com/groups?selm=pan.2002.01.13.11.22.54.248401.9857%40c
lub-internet.fr

This is written in the style of a tutorial, but admittedly, it's not very
polished.  I would appreciate any and all comments, criticism, feedback,
suggestions, amplifications, discussion, etc.

**

A simple AOP framework in Python using metaclasses

Let's start with a simple class:

<code>
class foo:

    def bar(self, *args, **kwargs):
        pass

f = foo()
f.bar()
</code>

As expected, this doesn't do anything.  Suppose I want to be notified
before and after the bar method is called.  (The wording of that is
purposefully vague.)  And I don't want to have to modify the foo class
at all in order to do this.  I want to intervene in the creation of
this class (not just instances of the class) and insert pre and post
methods for each function.

Let's start with a simple metaclass definition that doesn't do
anything:

<code>
class aspect(type):

    def __init__(cls, name, bases, dict):
        super(aspect, cls).__init__(name, bases, dict)

__metaclass__ = aspect

class foo:

    def bar(self, *args, **kwargs):
        pass

f = foo()
f.bar()
</code>

Again, this doesn't have any noticeable effect.  Let's insert some
print statements to see what's happening though:

<code>
class aspect(type):

    def __init__(cls, name, bases, dict):
        super(aspect, cls).__init__(name, bases, dict)
        for k,v in dict.items():
            print "type(%s) --> %s" % (k, type(v))

__metaclass__ = aspect

print "before class foo:"

class foo:

    def bar(self, *args, **kwargs):
        pass

print "before f = foo()"

f = foo()
f.bar()
</code>

<output>
before class foo:
type(__module__) --> <type 'str'>
type(bar) --> <type 'function'>
before f = foo()
</output>

This shows us that the aspect metaclass' __init__ gets called as a
result of declaring our foo class, before any instance's of foo are
created.

We want to iterate through the class's dict and do something for each
method (I'm not worried about distinguishing staticmethod and
classmethod type methods for now):

<code>
class Aspect(type):

    def __init__(cls, name, bases, dict):
        super(Aspect, cls).__init__(name, bases, dict)
        for k,v in dict.items():
            if type(v) == type(lambda x:x):
                print "%s is a function." % k

__metaclass__ = Aspect

class foo:

    def bar(self, *args, **kwargs):
        print "real bar(%s, %s)" % (args, kwargs)

f = foo()
f.bar("a", "b", foo="c")
</code>

<output>
bar is a function.
real bar(('a', 'b'), {'foo': 'c'})
</output>

Rather than importing types and using types.FunctionType, I just
compare the type of the item to the type of an anonymous function (via
lambda).  Is that cool or what?  ;-)

Also, notice that I replaced pass in the body of bar() with a simple
print statement that tells me how the real bar() is being called.
That will be useful later.

For each method (or function), I want to wrap that method so that I
get notified when it's called.  So I created a wrapped_method class
that is somewhat (exactly?) like a field descriptor (e.g., property):

<code>
class wrapped_method(object):

    def __init__(self, cls, method):
        self.class_name = cls.__name__
        self.method = method
        self.method_name = method.func_name
        self.observers = []

    def __call__(self, *args, **kwargs):
        self.before(*args, **kwargs)
        self.method(self, *args, **kwargs)
        self.after(*args, **kwargs)

    def before(self, *args, **kwargs):
        print "before %s.%s(%s, %s)" % (self.class_name, self.method_name,
                                        args, kwargs)

    def after(self, *args, **kwargs):
        print "after %s.%s(%s, %s)" % (self.class_name, self.method_name,
                                       args, kwargs)

class Aspect(type):

    def __init__(cls, name, bases, dict):
        super(Aspect, cls).__init__(name, bases, dict)
        for k,v in dict.items():
            if type(v) == type(lambda x:x):
                setattr(cls, k, wrapped_method(cls, v))

__metaclass__ = Aspect

class foo:

    def bar(self, *args, **kwargs):
        print "real bar(%s, %s)" % (args, kwargs)

f = foo()
f.bar("a", "b", foo="c")
</code>

<output>
before foo.bar(('a', 'b'), {'foo': 'c'})
real bar(('a', 'b'), {'foo': 'c'})
after foo.bar(('a', 'b'), {'foo': 'c'})
</output>

Now we're starting to get somewhere!  The next step is to create a
framework where multiple observers can plug into the before and after
events for any given method call.  Rather than taking more baby steps
to get there, this is the complete implementation I have so far:

<code>
import sys

class wrapped_method(object):

    def __init__(self, cls, method):
        self.class_name = cls.__name__
        self.method = method
        self.method_name = method.func_name
        self.observers = []

    def __call__(self, *args, **kwargs):
        self.before(*args, **kwargs)
        self.method(self, *args, **kwargs)
        self.after(*args, **kwargs)

    def before(self, *args, **kwargs):
        for o in self.observers:
            o.before(self.class_name, self.method_name, *args, **kwargs)

    def after(self, *args, **kwargs):
        for o in self.observers:
            o.after(self.class_name, self.method_name, *args, **kwargs)

    def notify_me(self, observer):
        self.observers.append(observer)

class VirtualClassError(Exception):pass

class Observer:

    def __init__(self, *args, **kwargs):
        if self.__class__ is Observer:
            raise VirtualClassError(self.__class__.__name__)

    def before(self, class_name, method_name, *args, **kwargs):
        pass

    def after(self, class_name, method_name, *args, **kwargs):
        pass

class Trace(Observer):

    def __init__(self, filename=None):
        self.filename = filename

    def write(self, prefix, class_name, method_name, *args, **kwargs):
        cls = class_name
        s = "%s: %s.%s(%s, %s)" % (prefix, cls, method_name, args, kwargs)
        if not self.filename:
            f = sys.stdout
        else:
            f = file(self.filename)
        f.write(s)
        f.write("\n")
        if self.filename:
            f.close()

    def before(self, class_name, method_name, *args, **kwargs):
        self.write("before", class_name, method_name, *args, **kwargs)

    def after(self, class_name, method_name, *args, **kwargs):
        self.write("after", class_name, method_name, *args, **kwargs)

class Aspect(type):

    def __init__(cls, name, bases, dict):
        super(Aspect, cls).__init__(name, bases, dict)
        for k,v in dict.items():
            if type(v) == type(lambda x:x):
                setattr(cls, k, wrapped_method(cls, v))

__metaclass__ = Aspect

class foo:

    def bar(self, *args, **kwargs):
        print "real bar(%s, %s)" % (args, kwargs)

def main():
    foo.bar.notify_me(Trace())
    f = foo()
    f.bar("a", "b", foo="c")

if __name__ == "__main__":
    main()
</code>

<output>
before: foo.bar(('a', 'b'), {'foo': 'c'})
real bar(('a', 'b'), {'foo': 'c'})
after: foo.bar(('a', 'b'), {'foo': 'c'})
</output>

The output is the same, but it's actually being generated by an
instance of the Trace class.

************
Observations
************

Obviously, Trace is designed so that I can specify a filename when
creating an instance of it and that would use that file rather than
sys.stdout.

Observers should probably be able to register for particular events
(i.e., just before).  They should be able to unregister.  It should
also be possible to register for some classes, some methods, and not
others.  What is a good way to write the rules for enrolling
in notifications?  (And what's the word for this that AspectJ uses?)

I explicitly pass self from the wrapped_method instance to the actual
wrapped method.  However, this is not really the instance of the
class, it's the instance of the wrapped_method object.  How do I pass
the actual instance to the wrapped method--whether explicitly or
implicitly?

I should probably explicitly set the metaclass for wrapped_method,
Trace, Observer, and even VirtualClassError to "type" (the default
metaclass) to avoid weird infinite loops where the members of the
aspect framework are themselves being aspected.

I need to think more about how this would be used.  My vague
impression of AspectJ is that you use it from an IDE, specifying
different conditions for filtering classes and methods (what's the
word for that again?).  When you compile with aspect support turned
on, the generated code has the aspects weaved in.  What are the
equivalent developer use cases for weaving aspects into Python code?
The whole point it seems to me is that when I author the classes to be
aspected, I shouldn't have to do a thing--no subclassing, no method
definition.  I just write my classes, omitting entirely any
consideration of the future aspects to be weaved in.  I mean, that's
the whole point, right?  To simplify the code.

In the example, foo doesn't know diddly about the fact that it's being
traced.  And we could turn that tracing on or off simply by changing
the module-level __metaclass__ variable.

I should probably and try:...except:pass to the framework so that it
doesn't generate errors.  During development of the framework itself,
of course, I want to know about errors.

Observer is an interface-type base class.  It doesn't provide
implementation.  It could easily be omitted, but it servers to
document what Trace and other potential observers need to look like.

It would be easy to add aspects for error handling.  This is a very
quick and dirty example for error notification:

In wrapped_method, I would change __call__ like this:

    def __call__(self, *args, **kwargs):
        self.before(*args, **kwargs)
        try:
            self.method(self, *args, **kwargs)
        except Exception, e:
            self.on_error(e, *args, **kwargs)
            raise
        self.after(*args, **kwargs)  Rather than using sys.stdout, I could
have specified a file.



and add on on_error() method to wrapped_method:

    def on_error(self, error, *args, **kwargs):
        for o in self.observers:
            o.on_error(error, self.class_name, self.method_name,
                       *args, **kwargs)

Observer:

    def on_error(self, error, class_name, method_name, *args, **kwargs):
        pass

and Trace:

    def on_error(self, error, class_name, method_name, *args, **kwargs):
        self.write("error %s" % error, class_name, method_name, *args,
                   **kwargs)

-





From goodger at users.sourceforge.net  Mon Jun  3 23:51:14 2002
From: goodger at users.sourceforge.net (David Goodger)
Date: Tue, 04 Jun 2002 03:51:14 GMT
Subject: Is there a Text to HTML conversion module?
In-Reply-To: <B9208083.5EF6%sarmstrong13@mac.com>
References: <B9208083.5EF6%sarmstrong13@mac.com>
Message-ID: <B921B17E.23FEC%goodger@users.sourceforge.net>

SA wrote:
> Does anyone know of a module that converts a text file to html?

Try reStructuredText, part of Docutils (URLs below).

-- 
David Goodger  <goodger at users.sourceforge.net>  Open-source projects:
  - Python Docutils: http://docutils.sourceforge.net/
    (includes reStructuredText: http://docutils.sf.net/rst.html)
  - The Go Tools Project: http://gotools.sourceforge.net/



From BPettersen at NAREX.com  Tue Jun 11 15:59:27 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Tue, 11 Jun 2002 13:59:27 -0600
Subject: Hap debugger released
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158EC5@admin56.narex.com>

> From: nealj [mailto:mwarden22 at hotmail.com] 
> 
> Version 2.1 of the Hap python debugger (and ide) has just 
> been released to sourceforge.  This version improves 
> performance, stability and adds a number of useful features 
> such as an interactive window, watch tooltips and more.
> 
> We've been using it in house on our next product for six 
> months now and everyone is very happy with it.
> 
> Try it out at http://sourceforge.net/projects/hapdebugger

This looks really good (I know a couple of people who're going to be
pleased :-)

A couple of minor issues I noticed:

 - the color scheme is awful (the Idle/PythonWin scheme is
   much less jarring on the eyes :-)
 - I like the MSVC key bindings, but I was surprised that
   Shift-F11 wasn't bound to "step out of".
 - I couldn't step over a list comprehension (instead I was
   taken through each iteration of the for 'loop').

All in all, great work however!

-- bjorn




From kragen at pobox.com  Sat Jun  8 22:01:31 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 08 Jun 2002 22:01:31 -0400
Subject: Efficient python programming...
References: <mailman.1023452786.27923.python-list@python.org> <3D013FB1.3D7477EA@engcorp.com>
Message-ID: <83y9dp43d0.fsf@panacea.canonical.org>

Peter Hansen <peter at engcorp.com> writes:
> Roman Suzi wrote:
> > To prove that code is correct is very difficult (practically impossible).
> > Unittests are there to prove that you have errors, not the other way
> > around.
> 
> I disagree, if the tests are written first (and run, proving that
> they will fail when the code is wrong or missing).

I think you are disagreeing because you are ignorant of formal methods.

> The tests are executable requirement specifications.  Running
> them "proves" (yes, not 100%... nothing could) that the code
> meets the specs.  The specs could be wrong of course, but that
> doesn't mean the code is broken...

Typically specs specify the behavior of the code in an infinite set of
situations.  Tests test the behavior of the code in a finite set of
situations, one situation per test.

You said "nothing could", but you are mistaken.  Computer programs are
sufficiently amenable to mathematical analysis that it is possible to
prove rigorously that a particular routine meets a particular spec,
100% of the time, if the spec is phrased in sufficiently rigorous
terms.

This has the following drawbacks:
- specs can be wrong (but if they're shorter than the code, they'll
  probably have fewer bugs)
- proofs can be wrong (but at least they make you think about the
  assumptions you made when writing the code)
- writing the proofs takes a long time, so long that it's practical
  only for small programs
- your proof is dependent upon your specification of the program's
  environment as well as upon your requirements specification; if your
  program depends on a lot of other complicated software, your proof
  may not be worth much

This is important not because rigorously proving a program correct is
a worthwhile activity --- maybe it is, but I'm not convinced --- but
because proving a program correct is a more rigorous version of what
good programmers do all the time in writing good code.  You should
learn a little bit about it, at least in its lighter-weight forms like
cleanroom programming.



From loredo at astro.cornell.edu  Fri Jun 28 15:38:34 2002
From: loredo at astro.cornell.edu (Tom Loredo)
Date: Fri, 28 Jun 2002 15:38:34 -0400
Subject: C API for Python's RNG?
Message-ID: <3D1CBB3A.E1C43D8A@astro.cornell.edu>

Hi folks-

I'm doing some Monte Carlo calculations, and to speed things up
I'm dumping parts of it into an extension.  I'll be needing random
numbers both in the Python code and in the extension.  I'd like
to use Python's RNG in the extension so I can easily maintain
the state of the whole calculation if necessary.  So I need to
know the C API for Python's RNG.  I don't see it documented in the
standard documentation.  I'll dig in the source to try to puzzle
it out, but I thought I should ask publicly:  is this documented
anywhere?  If not, is there anything I should be careful about
that may not be obvious?  How stable is the API from version
to version of Python?

I suppose I could switch to the Numeric Random module, but the
C API for that isn't documented either, so I have the same questions
about that.

Thanks,
Tom Loredo


From fredrik at pythonware.com  Wed Jun 19 18:13:46 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 19 Jun 2002 22:13:46 GMT
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.>
Message-ID: <um7Q8.44210$n4.10292249@newsc.telia.net>

Angela Williams wrote:
> I'm looking for the equivalent of the GOTO command for Python 2.2.1 running
> on Windows 98. I have programmed very little and am new to Python.

like most modern programming languages, Python doesn't have
a GOTO command.  for a brief explanation of why, and what you
should use instead, check the "branching" chapter in Alan Gauld's
"Learning to Program" tutorial:

    http://www.freenetpages.co.uk/hp/alan.gauld/

(if you're new to programming, I suggest working through the
entire tutorial)

you can find more Python tutorials here:

    http://www.python.org/doc/Newbies.html

</F>




From phr-n2002b at NOSPAMnightsong.com  Thu Jun 20 03:39:44 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 20 Jun 2002 00:39:44 -0700
Subject: Variable Interpolation - status of PEP 215
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <aermlt$in5$1@panix1.panix.com>
Message-ID: <7xu1ny4cvz.fsf@ruckus.brouhaha.com>

aahz at pythoncraft.com (Aahz) writes:
> Peter already told you what you really wanted to know, but for the
> benefit of other people, PEP 215 is dead and PEP 292 is currently the
> subject of a LARGE thread on python-dev.

I didn't know 215 was dead.  A shame.  I thought it was a good idea.

292 seems ok, maybe clumsier than 215 but an improvement over the %
mess inherited from C.

What's the deal with the curly braces in 292?  That's not discussed
at all in the PEP.  I don't see what's wrong with

   print "My name is $name".sub()

(which is similar to what 215 allowed) instead of

   print "My name is ${name}".sub()

I understand that preferences can run either way, but I think PEP 292
should face this question instead of ignoring it.


From s1336one562 at bach.sun.ac.za  Thu Jun  6 17:40:16 2002
From: s1336one562 at bach.sun.ac.za (Hugo van der Merwe)
Date: Thu, 06 Jun 2002 23:40:16 +0200
Subject: distutils and distributing "scripts"
References: <adl4e4$19ep$1@news.adamastor.ac.za> <nAqL8.10168$fG3.343408@news2.ulv.nextra.no> <mCqL8.10171$fG3.343358@news2.ulv.nextra.no>
Message-ID: <adoks1$2fqg$1@news.adamastor.ac.za>

That already helps though, thanks. I'll go without .pyw for now. But I 
still wonder how .pyw files, and all scripts, for that matter, are 
distributed. bdist does place scripts in /usr/bin as one would want, but 
one would want to be able to build a bdist from an sdist, which one cannot 
if the scripts aren't included. Or, does one usually, in the case of a .py 
script, add it both to py_modules and script lists?

Hugo

Erlend J. Leiknes wrote:

> oh, sorry, didnt read your question good enough...
> 
> =(
> 
> "Erlend J. Leiknes" <nookieNO_SP_AM at online.no> wrote in message
> news:nAqL8.10168$fG3.343408 at news2.ulv.nextra.no...
>> do not call the files .pyw, just use py
>>
>> when you then use py2exe, do the following
>> python setup.py py2exe -w
>>
>> the -w means --window or --windows (dont remember), anyway, this will use
>> the pythonw.exe.


From aahz at pythoncraft.com  Mon Jun 17 20:39:58 2002
From: aahz at pythoncraft.com (Aahz)
Date: 17 Jun 2002 20:39:58 -0400
Subject: If you use Python -OO is it a permanent condition??
References: <rXpP8.32471$ks6.25445@nwrddc02.gnilink.net> <aelamd$fo1$1@panix1.panix.com> <BTsP8.57000$uX3.41630@nwrddc01.gnilink.net> <aelu0a$7lc$02$1@news.t-online.com>
Message-ID: <aelvgu$4hr$1@panix1.panix.com>

In article <aelu0a$7lc$02$1 at news.t-online.com>,
Georg Mischler  <schorsch at schorsch.com> wrote:
>Jeff Sasmor wrote:
>>
>> 'Fascinating'
>> I'm sure there's a logical reason... somehow. Anyway, I found
>> that if I renamed all the .pyo files to .pyc everything works fine.
>> I guess that's because there's little difference between them. It's
>> too bad that there's no way to switch optimized mode on and
>> off programmatically - would be useful for apps that use compile().
> 
>It would be enough to simply look for the respective "other"
>bytecode file type as well, if nothing else has been found
>when trying to import a module. The -O setting only has any
>real relevance when freshly compiling from *.py, after all.
>Some small amount of added convenience from the side of the
>interpreter would be highly appreciated in this point.

File a bug on SourceForge.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From sjmachin at lexicon.net  Fri Jun  7 22:06:52 2002
From: sjmachin at lexicon.net (John Machin)
Date: 7 Jun 2002 19:06:52 -0700
Subject: array, list, performance...
References: <mailman.1023391748.930.python-list@python.org> <3D0108B7.3C206C2E@accessforall.nl>
Message-ID: <c76ff6fc.0206071806.c199f4c@posting.google.com>

Ype Kingma <ykingma at accessforall.nl> wrote in message news:
> Multiplying the list's size by a constant factor when necessary would 
> give O(log(len(li)) * len(li)) behaviour

I'm not sure where you get this from. Here's my take:

The theoretical amortised cost is O(N), not O(log(N)*N), where N is
the utimate size of the list.

Let F be the factor by which the list size is expanded when necessary.
Let e be the number of expansions required to grow to size N.
F**e is O(N).

The ith expansion involves fiddling with O(F**i) elements, with O(1)
cost each. The total of this over e expansions is O(F**e). As above,
F**e is O(N).

The O(1) cost each is theoretical. The operating system's memory
allocator may exhibit worse-than-linear behaviour.

Cheers,
John


From joerg.baumann at stud.informatik.uni-erlangen.de  Wed Jun  5 06:55:37 2002
From: joerg.baumann at stud.informatik.uni-erlangen.de (=?ISO-8859-1?Q?J=F6rg?= Baumann)
Date: Wed, 05 Jun 2002 12:55:37 +0200
Subject: Fastest way to read / procsess / write text?
References: <3CFD5229.2060803@excite.com>
Message-ID: <adkqn9$bm7$1@rznews2.rrze.uni-erlangen.de>

Brent Miller wrote:

...
> This works beautifully most of the time (it's super fast), except when I
> pipe large files (>50megs) to it and then it usually dies half way
> through complaining of memory errors because it ran out of ram.
...

Since you alter only texte inside single lines, you can process the file 
line by line, instead of loading it into memory as a whole.
(See Python docs for xreadlines(), xranges, ...)

#!/usr/bin/python2

import sys, re

try:
  outfilename = sys.argv[1]
except:
  raise "usage: %s <output file>"%sys.argv[0]

infile = sys.stdin
outfile = open(outfilename, "w")

for line in infile.xreadlines():
  outfile.write(re.sub('[.](?=\d\d\d\d\d\d)|[.](?=\w+[ 
][<|>])|[.](?=\w+[:])|[ ](?!0x)', '\t', line))




From jknapka at earthlink.net  Tue Jun 25 12:30:02 2002
From: jknapka at earthlink.net (Joseph A Knapka)
Date: Tue, 25 Jun 2002 16:30:02 GMT
Subject: Python and Eclipse
References: <3D17FAF4.35CC9F1B@earthlink.net> <81CC01E58BF2A99B.DFCD5EC6E4654907.32D731380951B9CA@lp.airnews.net>
Message-ID: <3D189A8A.577C8186@earthlink.net>

Cameron Laird wrote:
> 
> In article <3D17FAF4.35CC9F1B at earthlink.net>,
> Joseph A Knapka  <jknapka at earthlink.net> wrote:
> >Hi folks,
> >
> >I am interested in having Python support in
> >Eclipse <URL: http://www.eclipse.org> - interested
> >enough to contribute code, or to tackle the
> >task myself, if need be. Of course, I'd like
> >the Python support to be at least as good as the
> >Java support, which is a pretty tall order,
> >'cause the Java support in Eclipse is awfully
> >damn good --- it nearly allows me to *enjoy*
> >writing Java code :-)
> >
> >Anyway, I know some other folks have expressed
> >interest in this as well, and I'd like to hook
> >up with them. If you're working on Python+Eclipse
> >and you've got code in any state of completion,
> >I'm willing to alpha- or beta-test, fix stuff,
> >add new stuff, etc.
>                         .
>                         .
>                         .
> Can you say a few more words about Eclipse?  It ... well, I'm
> wary.  I understand its marketing more than its technology,
> and on hostile days, I think it's *only* marketing.  Maybe you
> can make it more appealing.  "[It] nearly allows me to *enjoy*
> writing Java ..." is certainly a good start.

I don't know about the marketing angle. I've
been using Visual Age for the past few years,
and IMO, after having used Eclipse for just
a couple of weeks, it is a great deal better
than Visual Age as an IDE. True, it doesn't
do visual UI layout, but I must say that VA's
UI builder never caused me anything but
grief anyhow. Someone could add a UI builder
to Eclipse if they wanted. Eclipse does everything
else as well as or better than VA - syntax-
aware editing, including context-sensitive
completion (definitely better than VA's!),
incremental compilation, debugging, edit-
compile-and-continue in the debugger...

Eclipse feels more solid than either
Visual Age or MS Visual Studio - I have had
no crashes or anomalous behavior while
using Eclipse rather intensively over the
past couple of weeks, save for an episode
yesterday that I believe was triggered by
a low-VM condition caused by another
process. Eclipse nevertheless exited
cleanly and saved my work, which is something
I have *not* come to expect from Visual Age.

Eclipse does like to have RAM to play in -
the Eclipse JVM on my NT laptop is currently
about 60MB large. Of course, Visual
[Age|Studio] have similar RAM demands.

Eclipse is open source, so no matter what
IBM decides, interested parties could pick
it up and run if necessary. It is designed
to be a platform upon which to build various
IDEs. It seems to be quite well documented
for IDE extension developers, although I
have not actually written any such extension
yet, so the perceived documentation quality
may be illusory.

I have not tried JBuilder. All my Java
work prior to Eclipse was done either in
Visual Age or at the command line, and
I strongly prefer Eclipse over either of
those other methods. For the price, it's
hard to beat.

-- Joe
   "This is Radio Clash / using audio ammunition."
   -- The Clash, "Radio Clash"


From not.this at seebelow.org  Thu Jun 13 17:14:58 2002
From: not.this at seebelow.org (Grant Griffin)
Date: 13 Jun 2002 14:14:58 -0700
Subject: why not "'in' in 'in'"?
References: <mailman.1023989567.18847.python-list@python.org>
Message-ID: <aeb20i01psk@drn.newsguy.com>

In article <mailman.1023989567.18847.python-list at python.org>, "Bjorn says...
>
...
>Ok, how about:
>
>>>> class mystr(str):
>...   def __contains__(self, s):
>...     return self.find(s) >=3D 0
>...
>>>> 'hello' in mystr('hello world')
>1
>>>> 'foo' in mystr('hello world')
>0
>>>>

Well, sure--but where's the fun in that?

Actually, I don't much like this approach for a couple of reasons.  First, you
then have to remember to wrap each string literal inside the name of your new
class, a practice which is both tedious and ugly.  It isn't worth it.

Second, although Python's newfound ability to subclass from built-in types
allows one to change nearly anything one doesn't like about the way the type
works, I generally think that's going in the wrong direction because it makes
your code work different than it reads.

Instead, I find the ability to subclass from built-in types to be useful
primarily to _add_ behavior to a type.  The most common example is some sort of
specialized list or dictionary which provides new methods appropriate only for
its special use.  These cases are, of course, simply a more elegant way of
implementing what we used to implement less elegantly via an explicit list
member, e.g., "self.list".

just-because-guido-allows-it-doesn't-mean-you-should-use
   -it-injudicious-ly y'rs,

=g2

_________________________________________________________________________

Grant R. Griffin                                           g2 at dspguru.com
Publisher of dspGuru                               http://www.dspguru.com
Iowegian International Corporation                http://www.iowegian.com



From phd at phd.pp.ru  Thu Jun 20 06:54:53 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 20 Jun 2002 14:54:53 +0400
Subject: Useful RE patterns (was: Variable Interpolation - status of PEP 215)
In-Reply-To: <yciQ8.44268$n4.10312169@newsc.telia.net>; from fredrik@pythonware.com on Thu, Jun 20, 2002 at 10:34:06AM +0000
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <4PfQ8.44245$n4.10307683@newsc.telia.net> <yciQ8.44268$n4.10312169@newsc.telia.net>
Message-ID: <20020620145453.B25214@phd.pp.ru>

On Thu, Jun 20, 2002 at 10:34:06AM +0000, Fredrik Lundh wrote:
> As part of the python-dev discussion, I mentioned that it might
> be a good idea to add a couple of standard RE patterns to the
> standard library, e.g. to match things like integer literals, IP-
> numbers (etc).
> 
> If I were to add a dozen (or so) patterns to the (S)RE module,
> what should I pick?  What patterns do you find yourself using
> over and over again?

   IP address (but it is required to test that every nibble is < 256)
   URL
   e-mail address (yes, I know very good how hard it is)
   identificator (letters+digits, starting with a letter)
   empty line
   line consisted only with whitespaces
   comment line, starting with #, may have whitespaces before #

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From threeseas at earthlink.net  Tue Jun 25 23:43:14 2002
From: threeseas at earthlink.net (Timothy Rue)
Date: Wed, 26 Jun 2002 03:43:14 GMT
Subject: Reminder of a python Project.
References: <2218.941T781T12224051threeseas@earthlink.net> <3D193728.663CA774@engcorp.com>
Message-ID: <1708.941T2398T14045676threeseas@earthlink.net>

On 25-Jun-02 22:38:16 Peter Hansen <peter at engcorp.com> wrote:
>Timothy Rue wrote:
>>
>> First off, it's ok that I got such a high level of resistance to my
>> initial posting here about this project.

>Obviously the level was not high enough.  Sorry, we'll try harder.

>> $81 Billion = 31% of software development gets cancelled before complete
>> $59 Billion = 53% of software development has cost over-runs of 189%
>> 16% success - project success and failure ratio
>[...]
>> My 1997 estimate of software failures all around (total of industry and
>> consumer level development and use):  Low estimate of 300 Billion.

>Your information is out of date.  Agile software development processes
>developed since 1997 have improved the success rate of software development
>projects to the point where this is no longer an issue.  It is now possible
>to deliver software with a 96% success rate, for less cost than previous
>methods, with greater sustainability for the development teams, and at
>a much higher quality level than appeared possible in the past.

>I'm afraid there is no longer any need for an autocoder.

>-Peter


Geee, maybe you should tell that to NIST

http://www.computerworld.com/printthis/2002/0,4814,72245,00.html

Hmmm, I coulda sworn I put that in the post.

You wouldn't know what happened to it would you?



---
*3 S.E.A.S - Virtual Interaction Configuration (VIC) - VISION OF VISIONS!*
   *~ ~ ~      Advancing How we Perceive and Use the Tool of Computers!*
Timothy Rue      What's *DONE* in all we do?  *AI PK OI IP OP SF IQ ID KE*
Email @ mailto:timrue at mindspring.com      >INPUT->(Processing)->OUTPUT>v
Web @ http://www.mindspring.com/~timrue/  ^<--------<----9----<--------<



From findler_lambda at yahoo.com  Fri Jun  7 10:12:10 2002
From: findler_lambda at yahoo.com (Robert Hanlin)
Date: 7 Jun 2002 07:12:10 -0700
Subject: About writing Jython tutorial...
Message-ID: <c427d639.0206070612.4ea8ce0c@posting.google.com>

Hi all,

I'm indoctrinating friends into the dark, beautiful world of Jython. 
But I find there is no good online documentation, and I only have an
easy time because I happen to own the O'Reilly Jython Essentials book.
 (For example, when you use jythonc to compile classes, if something
is missing on the classpath you're more likely to get a null pointer
exception than an appropriate class not found exception.)

If I happened to write a walkthrough, is there a website where it
would be welcome?  If not, maybe I should make a weblog...


Thanks!
findler


From michael at belgacom.net  Wed Jun 19 08:00:34 2002
From: michael at belgacom.net (Michael Anckaert)
Date: Wed, 19 Jun 2002 14:00:34 +0200
Subject: how i use =?iso-8859-1?Q?'=E3?=
	=?iso-8859-1?Q?'?= in python???
In-Reply-To: <3D0F6623000017AA@www.zipmail.com.br>; from jubafre@zipmail.com.br on Wed, Jun 19, 2002 at 01:38:45AM -0300
References: <3D0F6623000017AA@www.zipmail.com.br>
Message-ID: <20020619140034.A2009@carpathia.anckaert.net>

On Wed, Jun 19, 2002 at 01:38:45AM -0300, jubafre at zipmail.com.br wrote:
> i?m brazilian, and in my laguage(Portguese) there are many other caracters
> how ?, ?, ? and a string in python doesn?t support this
> how i can use??? have a import module for this???
> 
> BRASILEIROS DA LISTA COMO VCS CONSEGUEM USAR ACENTO NO PYTHON??
> 
> for exmple:
> s='?'
> doesn?t work?why??
> 

It works for me. I tried the ? and ? characters and they both work out
fine if I want to asign a string these variables. Maybe there is a
problem with your editor?
What exactly is the error message you get?

LINUX: The Choice of a GNU Generation...
----------------------------------------------
Michael Anckaert
manckaert at belgacom.net
xantor at linux.be
xantor at xantor.tk
http://www.xantor.tk
Jabber: xantor at jabber.org
If you use envelopes, why not use encryption?
OpenPGP: 0xC3300BEC
----------------------------------------------
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020619/0d60cdfa/attachment.sig>

From tismer at tismer.com  Fri Jun 21 06:37:51 2002
From: tismer at tismer.com (Christian Tismer)
Date: Fri, 21 Jun 2002 12:37:51 +0200
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com> <lkznxqjqx2.fsf@pc150.maths.bris.ac.uk> <3D11A21D.9010809@mxm.dk> <uh3e1usajle42c@news.supernews.com> <3D11DA9B.163A7BD0@info.unicaen.fr> <uh446elev7af68@news.supernews.com> <3D12EEAA.9C3A3D29@info.unicaen.fr>
Message-ID: <3D1301FF.3070201@tismer.com>

Jerzy Karczmarczuk wrote:
> John Roth comments my observation concerning the relation between gotos
> and continuations:

...

>>Continuations, on the other hand, have been discussed, and may
>>very well wind up in Python some day. See the discussion around
>>a variant called 'stackless.'

Guys, I'm watching you!  :-)

> Yeah, I know it. I follow the Mission Impossible Software Team (Tismer
> et al.) for some time. I am interested in scientific computing, and I 
> believe that generators, coroutines etc. are essential for the simu-
> lation. So, the idea to use continuations was for me always very
> appealing (even if sometimes appalling...). Still, the notion of a
> full continuation, as implicit in the letter of GvR
> http://www.stackless.com/continuations.guido.van.rossum.html,
> the machinery which makes a snapshot of the actual state, and which
> is implemented in Scheme through call/cc, is not necessarily the
> very essence of the concept itself, which is simpler. My turn to send
> you to some references. I suggest the Andrew Appel's book 'Compiling
> with Continuations'.

In fact, Guido isn't correct with copying the full execution
stack. The continuations implementation which I did is completely
different in that sense and tries to copy the minimum necessary
to survive.

Appel's book can really be recommended if you like to learn
a new way of thinking.

Still, I'm not planning to support full continuations again.
The immutability of continuations is a hard requirement,
and I didn't find an application yet where mutable tasklets
didn't suffice.

ciao - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From whisper at oz.net  Sat Jun  8 03:40:26 2002
From: whisper at oz.net (David LeBlanc)
Date: Sat, 8 Jun 2002 00:40:26 -0700
Subject: no module named windowsc?
In-Reply-To: <3b738180.0206072145.1c4a655b@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEKEIBDHAA.whisper@oz.net>

The windows.py in \python22\lib\site-packages is _not_ the example
windows.py that is mentioned in the FXPy doc. It seems you have to download
and unpack the source distribution in order to get the examples and local
documentation. Then try running "python windows.py" in the
fxpy-1.0.5\examples directory and I think you'll find things will work.

BTW, I believe that under Windows 2000 Pro, you can actually just type
"windows" in the example dir and it will run - a nice feature of Windows
2000 long since overdue. If that doesn't work, let me know and I think I can
tell you how to make it work that way.

Regards,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Mr Blah
> Sent: Friday, June 07, 2002 22:46
> To: python-list at python.org
> Subject: no module named windowsc?
>
>
> I installed python 2.2 for windows on my windows 2000 professional
> machine and also installed fxpy for python 2.2.
>
> When I try to run any of the example programs that come with fxpy I
> get the same error - a complaint about windowsc:
>
> C:\Python22\FXPy>python windows.py
> Traceback (most recent call last):
>   File "windows.py", line 2, in ?
>     import windowsc
> ImportError: No module named windowsc
>
> Searching on windowsc I can't seem to come up with much of anything.
> Can anyone give me a clue about what might be going on?
>
> Mr Blah
> blah at calico-consulting.com
> --
> http://mail.python.org/mailman/listinfo/python-list





From max at alcyone.com  Sun Jun 23 20:59:00 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 23 Jun 2002 17:59:00 -0700
Subject: HTTP authentication
References: <TWtR8.86654$Am4.23833590@twister.columbus.rr.com>
Message-ID: <3D166ED4.FBE9E2C8@alcyone.com>

Preston Tamkin wrote:

> hello all. I've just recently jumped on the python bandwagon, and I
> need
> some help!

Derive from the urllib.FancyURLopener class, and override the
proper_use_password method to return the (user, password) 2-tuple you
want:

	http://www.python.org/doc/current/lib/urlopener-objs.html

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ See the son in your bad day / Smell the flowers in the valley
\__/ Chante Moore
    Bosskey.net: Aliens vs. Predator 2 / http://www.bosskey.net/avp2/
 A personal guide to Aliens vs. Predator 2.


From ianb at colorstudy.com  Thu Jun 20 14:11:33 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 20 Jun 2002 13:11:33 -0500
Subject: Variable Interpolation - status of PEP 215
In-Reply-To: <7xu1ny4cvz.fsf@ruckus.brouhaha.com>
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com>
	<aermlt$in5$1@panix1.panix.com>  <7xu1ny4cvz.fsf@ruckus.brouhaha.com>
Message-ID: <1024596693.605.174.camel@lothlorien.colorstudy.net>

On Thu, 2002-06-20 at 02:39, Paul Rubin wrote:
> What's the deal with the curly braces in 292?  That's not discussed
> at all in the PEP.  I don't see what's wrong with
> 
>    print "My name is $name".sub()
> 
> (which is similar to what 215 allowed) instead of
> 
>    print "My name is ${name}".sub()

The PEP allows either, just like the shell.  It just uses the second
form in most of the examples.

  Ian






From loewis at informatik.hu-berlin.de  Sun Jun  9 07:00:39 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 09 Jun 2002 13:00:39 +0200
Subject: xml.* getting the hang...
References: <MPG.176d47ced6b87f6d989681@news.rz.uni-karlsruhe.de>
Message-ID: <j4elfg4syw.fsf@informatik.hu-berlin.de>

Uwe Mayer <merkosh at hadiko.de> writes:


First you write

> i figured that xml.sax and xml.parsers.expat is where to look for. 
> problem number first:

and then

> - in order to build an xml dom object i have to get started on a 
> Document(); 

If you want to build a DOM, I recommend not to use sax or expat
directly. Instead, you should use the minidom build,
xml.parse.minidom.parse.

> and this is the problem: i couldn't find anything; with one
> exception, i can create xml documents using
> xml.dom.minidom.Document(), if this is the regular case then i don't
> quite see how xml.sax.* and xml.dom.minidom relate to eachother.

They relate to each other through xml.dom.minidom.parse, which uses
xml.dom.pulldom.parse, which either uses the parser given as a
parameter, or uses xml.sax.make_parser, which either uses a parser
specified in an environment variable, or uses xml.sax.expatreader,
which uses xml.parsers.expat.

> Secondly i found the package xml.dom.pulldom which seems to be just what 
> i'm looking for, however the documentation is very poor on these classe. 
> what, f.e. is a "documentFactory"?

pulldom was originally not designed for "public" use, so I recommend
you ignore it.

> I don't want to do xml validating, so xml.parsers.expat could be used, 
> too. So now i don't know what to use at all:

You should use xml.dom.minidom.parse.

Regards,
Martin


From vinay_sajip at yahoo.co.uk  Tue Jun  4 20:19:11 2002
From: vinay_sajip at yahoo.co.uk (Vinay Sajip)
Date: 4 Jun 2002 17:19:11 -0700
Subject: ANN: Logging Module v0.4.5 released
Message-ID: <2e37dc1.0206041619.787971f1@posting.google.com>

A new version of the proposed Python standard logging module (as per
PEP 282) has been released. You can get all the information from

http://www.red-dove.com/python_logging.html

There are "download" and "recent changes" links at the top of that
page. The new stuff includes some bug fixes, better support for
class-based filtering and logging, more documentation of the
configuration file format, an example hand-coded configuration file
for those people who can't use the GUI configurator, an example Filter
for regular-expression match-based filtering, and more!

As always, your feedback is most welcome (especially bug reports,
patches and suggestions for improvement).

Cheers

Vinay Sajip
Red Dove Consultants Ltd.

Changes since the last version:
=================================
Fixed bug which caused problem if no args to message (suggested by
Hye-Shik Chang).
Fixed bug in _fixupParents (thanks to Nicholas Veeser) and added
log_test19.py as a test case for this bug.
Added getMessage to LogRecord (code was moved here from
Formatter.format).
Applied str() to record.msg to allow arbitrary classes to determine
the formatting (as msg can now be a class instance).
Table of Contents added to python_logging.html, the section on Loggers
updated, and the logconf.ini file section annotated.
Added log_test20.py which demonstrates how to use class instances to
provide alternatives to numeric severities as mechanisms for control
of logging.
Added log_test21.py which builds on log_test20.py to show how you can
use a regular expression-based Filter for flexible matching similar to
e.g. Protomatter Syslog, where you can filter on e.g. "a.*" or "*.b"
or "a.*.c".
_levelNames changed to contain reverse mappings as well as forward
mappings (leveltext->level as well as level -> leveltext). The reverse
mappings are used by fileConfig().
fileConfig() now more forgiving of missing options in .ini file -
sensible defaults now used when some options are absent. Also, eval()
is used less when interpreting .ini file contents - int() and dict
lookup are used in more places.


From mgerrans at mindspring.com  Sat Jun 15 03:52:37 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Sat, 15 Jun 2002 00:52:37 -0700
Subject: I feel stoopid ... this code is to verbose
References: <3D0A1791.3060307@mxm.dk> <3d0a240f.15785671@news.t-online.de>
Message-ID: <aeeroi$5nu$1@slb3.atl.mindspring.net>

> strfmt = lambda x:"".join(map(lambda c,l,y=len(x):(l==3) and c+"," or
> (not (l-3)%3 and c+"." or c),list(x),range(len(x),0,-1)))

Needs a little more tweaking; try this:

strfmt(str(-10000))




From david_griswold1 at yahoo.com  Fri Jun 28 21:00:22 2002
From: david_griswold1 at yahoo.com (David)
Date: Sat, 29 Jun 2002 01:00:22 GMT
Subject: I'm an idiot
Message-ID: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16>

OK, I am the first to admit it.  I am an idiot.  I have RTFM on this over 
and over, and I can still not figure out what I am doing wrong.

I think the intent of the code is obvious, but just to clarify, I want to 
read every line in a file and write those lines back out to another file, 
but with leading and training space removed.  I also want to have some 
elegant way to determine that I have reached the end of the file and 
break out of the loop.

And just to explain my stupidity, my reference langauge is BASIC.  I 
could have written this in BASIC in a minute, but I need to learn 
something new.  Any help would be appreciated.

David


f=open('c:\\temp\\temp.txt', 'r')
g=open('c:\\temp\\temp1.txt', 'w')
while 1:
    try:
        s=f.readline
        g.write(s.split())
    except IOError:
        break

g.close
f.close


From phd at phd.pp.ru  Thu Jun 20 04:02:52 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 20 Jun 2002 12:02:52 +0400
Subject: procmail replacement in Python
In-Reply-To: <20020620014021.A4575@lifebook>; from dmitri.gouliaev@telkel.net on Thu, Jun 20, 2002 at 01:40:22AM -0500
References: <oqelf3s7n4.fsf@carouge.sram.qc.ca> <NCBBLFCOHHDIKCAFGCFBAEGILFAA.marklists@mceahern.com> <20020619181953.E4127@phd.pp.ru> <20020619150411.GB11789@lilith.my-fqdn.de> <20020620014021.A4575@lifebook>
Message-ID: <20020620120252.G22899@phd.pp.ru>

On Thu, Jun 20, 2002 at 01:40:22AM -0500, Dmitri I GOULIAEV wrote:
> > > On Wed, Jun 19, 2002 at 09:08:53AM -0500, Mark McEahern wrote:
> > > >   http://www.qcc.sk.ca/~charlesc/software/getmail-2.0/
> > > 
> > >    "A POP3 mail retriever..." Hence, it is not fetchmail replacement, as
> > > fetchamil also does IMAP, and does it very well.
> > 
> > I don't need to retrieve emails from an IMAP server [...]
> 
> CMIIamW, I thought Oleg was talking about the description of getmail, not about your particular needs.

   Exactly.

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From whisper at oz.net  Wed Jun 19 19:56:05 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 19 Jun 2002 16:56:05 -0700
Subject: Dispatching a Win32 COM in the background?
In-Reply-To: <3D1117D8.9AAEF2F4@engcorp.com>
Message-ID: <GCEDKONBLEFPPADDJCOEOEEODKAA.whisper@oz.net>

Hmmm... and the run method of Background could be sitting on one end of a
Queue (Queue is thread-safe: see your python doc for details) that's
buffering the asyncronous calls on the speacher. The main task then just
stuffs a talk request into the queue and the Background processes it when it
gets to it.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Peter Hansen
> Sent: Wednesday, June 19, 2002 16:47
> To: python-list at python.org
> Subject: Re: Dispatching a Win32 COM in the background?
>
>
> "Steven M. Castellotti" wrote:
> >
> > 2) Make the call inside a thread which produces the same effect.
>
> This might work (I don't know anything about COM though).  Try putting
> your own code inside the run method of a thread:
>
> import threading
> class Background(threading.Thread):
>     def __init__(self):
>         threading.Thread.__init__(self)
>
>     def run(self):
>         print 'Starting'
>         pass   # your own code goes here
>         print 'Done'
>
> >>> b = Background()
> >>> b.start()
> Starting
> <possibly a pause here?>
> Done
>
> If at this point you get a >>> prompt back right away, but the
> speech generation is carrying on "in the background" then you're
> all set.  Make sure you do this with code you've already tested
> outside of the thread, so you'll know that any failure is not
> the fault of your own code...
>
> -Peter
> --
> http://mail.python.org/mailman/listinfo/python-list





From jones at jones.jones  Fri Jun 14 08:44:14 2002
From: jones at jones.jones (Jones)
Date: Fri, 14 Jun 2002 12:44:14 GMT
Subject: Writing a simple linux daemon type thing...
References: <K9eO8.393$LN6.261@newsread1.prod.itd.earthlink.net> <rn3cea.rt4.ln@127.0.0.1>
Message-ID: <yylO8.916$TP4.88596@newsread2.prod.itd.earthlink.net>

"Radovan Garabik" <spam at melkor.dnp.fmph.uniba.sk> wrote in message
news:rn3cea.rt4.ln at 127.0.0.1...
> Jones <jones at jones.jones> wrote:
>  : Hi everyone,
>  : I hope this is a simple question for you guys.  I'm new to Python and
Linux.
>  : I'm wondering if Python (and this group) can help me.
>  : I want a program to always be running in the background that, connects
to
>  : PostgreSQL every few seconds, queries a table, and for each record it
finds,
>  : shells out to the system to execute program.
>
>  : My research indicates that "nohup" might do the trick?  So would it be
as
>  : simple as writing the program in Python and then puting something like
>  : "nohup python myprog" in the Linux startup sequence (wherever that may
be)?
>
> yes
> (python myprog &
>  would be probably better)

Thank you for you reply.  Just one more question.  Doing what you suggest
was my first thought.  But I thought I read somewhere last night that using
"&" was not a good idea..... I have no idea why (it sounds good to me), but
I guess I just want to understand the difference between running it as a
"daemon" and just running it the background... if there is one.  Or maybe
someone has a URL with a good discussion on this someone could point me to.

Thanks.




From ianb at colorstudy.com  Wed Jun 12 18:59:24 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 12 Jun 2002 17:59:24 -0500
Subject: Wanted:  Potential Python Authors
In-Reply-To: <db3fgu40sbnt3pno5v1i2h9lnmpcoti2bk@4ax.com>
References: <ae7gke$mhb$1@ccis12.baylor.edu> 
	<db3fgu40sbnt3pno5v1i2h9lnmpcoti2bk@4ax.com>
Message-ID: <1023922764.511.29.camel@lothlorien.colorstudy.net>

On Wed, 2002-06-12 at 17:27, Lothar Scholz wrote:
> >- Expert programmer
> No, definitely not. Most programmers are not trained on writing
> tutorials or books - and normally they are really bad.
> Look more for someone trained as technical writer.

I don't think programmers deserve that reputation -- at least when it
comes to writing for an audience they can sympathize with (people who
want to learn to program).  An expert programmer has already created an
effective internal model and metaphor for programming, which is not
shared by technical writers who are not experienced programmers (or
experienced in the appropriate domain).

Also, a good writer will not just tell how to write an effective
program, but how to write a *good* program.  In order to do that, you
have to be able to write good programs.

You also should have something to say that's not just a rephrasing of
the documentation -- new content, new perspective, something.

  Ian






From quinn at lira.ugcs.caltech.edu  Sat Jun  8 23:51:12 2002
From: quinn at lira.ugcs.caltech.edu (Quinn Dunkan)
Date: 9 Jun 2002 03:51:12 GMT
Subject: libpython as a shared library?
References: <20020604143402.GB22993@hal.mediasupervision.de> <Pine.OS2.4.32.0206051931120.1678-100000@tenring.andymac.org> <mailman.1023348398.27744.python-list@python.org> <j4u1og6bvt.fsf@informatik.hu-berlin.de>
Message-ID: <slrnag5k5g.17b.quinn@lira.ugcs.caltech.edu>

On 06 Jun 2002 10:37:42 +0200, Martin v. L?wis <loewis at informatik.hu-berlin.de>
wrote:
>Gregor Hoffleit <gregor at hoffleit.de> writes:
>
>> If I compile the sources with -fPIC, but put them in a static library,
>> there's nearly no speed difference to non-PIC code in a static library.
>> My best guess is that the speed penalty is caused by the indirection
>> step that's necessary when calling routines from a shared lib; but then,
>> that's no real explanation for a loss of 30% speed!
>
>Can you elaborate a bit what you measured to get that speed difference?
>
>If it involves startup time, the time for locating libpython surely
>involves part of startup time. Also, it might be necessary to perform
>many relocations to libpython at run time that might not be necessary
>for the application.

I hope 2.3 doesn't dynamicly link libpython by default.  Not really because
of the performance issue, if there is one, but that it seems silly to introduce
the annoyance and fragility of shared libraries when all it would do is save
a few K for programs which embed python (and I have yet to see any of those).


From schorsch at schorsch.com  Mon Jun 17 20:13:22 2002
From: schorsch at schorsch.com (Georg Mischler)
Date: Tue, 18 Jun 2002 02:13:22 +0200
Subject: If you use Python -OO is it a permanent condition??
References: <rXpP8.32471$ks6.25445@nwrddc02.gnilink.net> <aelamd$fo1$1@panix1.panix.com> <BTsP8.57000$uX3.41630@nwrddc01.gnilink.net>
Message-ID: <aelu0a$7lc$02$1@news.t-online.com>

Jeff Sasmor wrote:

> 'Fascinating'
> I'm sure there's a logical reason... somehow. Anyway, I found
> that if I renamed all the .pyo files to .pyc everything works fine.
> I guess that's because there's little difference between them. It's
> too bad that there's no way to switch optimized mode on and
> off programmatically - would be useful for apps that use compile().
 
It would be enough to simply look for the respective "other"
bytecode file type as well, if nothing else has been found
when trying to import a module. The -O setting only has any
real relevance when freshly compiling from *.py, after all.
Some small amount of added convenience from the side of the
interpreter would be highly appreciated in this point.


-schorsch

-- 
Georg Mischler  --  simulations developer  --  schorsch at schorsch.com
+schorsch.com+  --  Lighting Design Tools  --  http://www.schorsch.com/


From russblau at hotmail.com  Wed Jun 19 17:50:00 2002
From: russblau at hotmail.com (Russell Blau)
Date: Wed, 19 Jun 2002 17:50:00 -0400
Subject: Extending long
References: <69310ed7.0206191220.3bbf68d8@posting.google.com> <m3ofe7q902.fsf@mira.informatik.hu-berlin.de>
Message-ID: <aequal$9cbp7$1@ID-76829.news.dfncis.de>

"Martin v. Loewis" <martin at v.loewis.de> wrote in message
news:m3ofe7q902.fsf at mira.informatik.hu-berlin.de...
> estama at dblab.ntua.gr (Freeman Stopjohn) writes:
>
> > class bitnum(long):
> >     def getbit(self,bitn):
> >         if self & (1 << bitn) >0:
> >             return 1
> >         else:
> >             return 0
> >     def setbit(self,bitn):
> >         self=self ^ ( 1 << bitn )
> > -----
> >
> >  The getbit method works ok.
> >  The setbit doesn't (it doesn't change the number).
> >
> >  What am i doing wrong?
>
> You are assigning to self. self is a parameter of setbit, just like
> bitn. Assignments to formal parameters have no effect.
>
> What you want to do cannot be done; long objects are immutable. I
> recommend you implement a wrapper, rather than inheriting.

How about:

    def setbit(self,bitn):
        return bitnum(self ^ ( 1 << bitn ))

This will not modify the bitnum() instance in place, since (as you
pointed out) such an instance is immutable, but it will return a new
instance with the desired value.

--
I don't actually have a hotmail account; but I do have one on excite.com
if you really want to get in touch with me.





From sholden at holdenweb.com  Fri Jun 28 08:41:25 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 28 Jun 2002 08:41:25 -0400
Subject: Pythoniac: Thoughts on a hardware Python processor
References: <mailman.1025206656.12053.python-list@python.org>
Message-ID: <COYS8.106024$mF.22577@atlpnn01.usenetserver.com>

"David LeBlanc" <whisper at oz.net> wrote in message
news:mailman.1025206656.12053.python-list at python.org...
[...]
>
> Ahh! Haven't you heard? Forth came in Fifth and the Forth chip is dead. In
> fact I don't think anyone is still offering a stack based computer (only
HP
> did AFAIK, the HP 3000).
>

Well, there was the English Electric KDF9, circa 1960. Even though it did
have instructions to load and store stack elements from/to memory, all
operations were performed on the hardware stack.

nothing-new-under-the-HP-ly y'rs  - steve
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From djc at object-craft.com.au  Tue Jun 25 06:51:07 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 25 Jun 2002 20:51:07 +1000
Subject: Web templating/db tool with best designer/coder separation?
References: <af2gh6$2c2$0@216.39.172.122> <slrnah9mj3.rdk.jon+usenet@snowy.squish.net> <3D159696.9010209@mxm.dk> <m3r8iye24k.fsf@ferret.object-craft.com.au> <23891c90.0206240735.3395d1f1@posting.google.com>
Message-ID: <m3r8ivr5r8.fsf@ferret.object-craft.com.au>

>>>>> "Paul" == Paul Boddie <paul at boddie.net> writes:

>>  Whenever you build any toolkit you have to decide on the set of
>> problems that you wish to solve with the toolkit.  The smaller the
>> problem domain the more useful you can make the toolkit.

Paul> Indeed. Some kinds of applications may be addressed well by a
Paul> particular toolkit, and since you might spend most of your time
Paul> constructing those applications for customers, for example, it
Paul> would definitely be worthwhile abandoning "complete freedom of
Paul> expression" for something which gets the job done quickly,
Paul> effectively and cheaply. This is arguably where many
Paul> toolkits/frameworks fall down - by trying to be everything to
Paul> everyone, they become as complicated as the language they're
Paul> written in and don't always offer much more than the bare
Paul> language in doing a particular job. Consequently, people get the
Paul> temptation to write "rival frameworks" from scratch.

Touche.  :-)

- Dave

-- 
http://www.object-craft.com.au


From deltapigz at telocity.com  Wed Jun 26 16:00:59 2002
From: deltapigz at telocity.com (Adonis)
Date: Wed, 26 Jun 2002 16:00:59 -0400
Subject: printing (again!!)
References: <afcusp$66r$1@newsg2.svr.pol.co.uk>
Message-ID: <3d1a1d77$1_5@nopics.sjc>

or if your using windows NT/2K/XP you can remap your printer using the NET
command given the user access to the lpt port and can simply call it as: x =
open("lpt1:", "w") x.write("blah" x.close() etc.

that way it would be easier to corss platform.

if on win95/98/ME then go as stated on other posts resort to either win32all
or a gui toolkit.

hope this helps.

Adonis

also, if your interested in the remapping the DOS lpt port google "DOS
printing"

"Geoff Tarrant" <geoff at galtar.co.uk> wrote in message
news:afcusp$66r$1 at newsg2.svr.pol.co.uk...
> I want to use Python as the language in a Computing course that I teach.
> However, I can't find a way that is relatively straightforward to send
data
> to a printer in a Windows environment.  If I can't find a way around this,
I
> will have to revert to Delphi or VB.  Please help.  I like the idea of
> sending 20 or so Python programmers to University each year.
>
> Geoff Tarrant
>
>




From catunda at pobox.com  Tue Jun 25 10:54:25 2002
From: catunda at pobox.com (Marco Catunda)
Date: Tue, 25 Jun 2002 11:54:25 -0300
Subject: Thread Memory Leak
References: <pan.2002.06.25.12.37.12.535288.17223@pobox.com> <mailman.1025010754.9180.python-list@python.org>
Message-ID: <pan.2002.06.25.14.54.24.716404.18707@pobox.com>

The second line "listThread = []" in the loop do it., don't it?

-- Marco Catunda

On Tue, 25 Jun 2002 10:11:08 -0300, Skip Montanaro wrote:

>     Marco> while 1:
>     Marco>     listThread = []
> 
>     Marco>     for i in range(50):
>     Marco>         listThread.append( threading.Thread( None, runthread

> You never remove any thread objects from your listThread list.  It fills
> up with completed threads.


From syver-en+usenet at online.no  Thu Jun 27 14:17:45 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Thu, 27 Jun 2002 18:17:45 GMT
Subject: How to find out operating system
References: <mailman.1025162014.16302.python-list@python.org> <3D1B1A37.CF516C4C@engcorp.com>
Message-ID: <uwuska8ny.fsf@online.no>

Peter Hansen <peter at engcorp.com> writes:

> David LeBlanc wrote:
> > 
> > You might try os.environ:
> > 
> > os.environ['WINOS'] -> WIN2000
> > 
> > I don't know if WINOS is defined on '98 or ME, but it should be good
> for NT,
> 
> > 2K and XP
> 
> It is not.

Neither is it defined on win2k pro:

KeyError: WINOS

-- 

Vennlig hilsen 

Syver Enstad


From michael at damaru.com  Mon Jun 10 23:16:36 2002
From: michael at damaru.com (Michael Davis)
Date: Mon, 10 Jun 2002 23:16:36 -0400
Subject: newbie question - python blocking on int()?
References: <4LaN8.2488$Vr2.586019@news20.bellglobal.com> <Xns922A149F19723cliechtigmxnet@62.2.16.82>
Message-ID: <eWdN8.3255$Ju2.865859@news20.bellglobal.com>

Chris Liechti wrote:

> Michael Davis <michael at damaru.com> wrote in
> news:4LaN8.2488$Vr2.586019 at news20.bellglobal.com:
>> I'm writing a specialized ftp client. I'm parsing the output of ftp.dir,
>> which gives me a string like this (call it str):
>> 
>> -rw-r--r--   1 fred       527 Jun  4 22:58 report.php
>> 
>> and I'm getting the various parts like this:
>> 
>> details = string.split( str )
>> permissions = details[0]
>> size = details[3]
>> name = details[7]
>> debug( "added remote file size %5d: %s" % (int(size), name) )
>> 
>> This works. But when I replace the 3rd line with this:
>> 
>> size = int( details[3] )
>> 
>> python hangs on that line. Why?
> 
> not when i try it... what do you mean by "hangs" does it takes forever or
> do you see an error message? if its the later, then you might have
> assigned something to "int" somewhere else, hiding the builtin function.

Thanks for replying. Actually, it's the former - when I say 'hangs', I mean 
that the program appears to stop running at that point. If I write:
print "one"
size = int( details[3] )
print "two"

it prints "one" but not "two".

As far as the other points go, luckily I'm in control of all the file names, 
and I can ensure that none has spaces. Spaces in filenames were always a 
terrible idea, I think. And I didn't call the string 'str', I called it 
'line', which I don't think is a reserved word or function name.

Cheers,
Michael

> 
> note that you should not use "str" as a name as that is a builtin
> object/function that can be very handy. e.g. in python 2.2+ you could have
> written "details = str.split(line)"
> (or even "details = line.split()" which i prefer)
> 
> also note that split() might not be safe for every case. e.g when the user
> name contains a space (if possible) and surely when the filename has a
> space in it.
> 
> chris
> 

-- 
Michael Davis
Damaru
Custom Programming - Web Development - Database Design
http://www.damaru.com
416-540-1284


From hst at empolis.co.uk  Fri Jun  7 06:06:58 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Fri, 7 Jun 2002 11:06:58 +0100
Subject: question on pattern
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220BF0@hendrix.empolisuk.com>

Ulrich Pfisterer wrote:
> 
> I want to do a pattern match on a string that spans multiple 
> lines (c++ 
> style comments, re.match('(/\*.*?\*/)', filestring)
> If I use the re.findall(pattern, string) method it only finds those 
> matches that are on a single line. If I use re.match(pattern, string, 
> DOTALL) it spans multiple lines but only finds the first match.
> Any help on this one would be greatly appreciated.
> 
This works:
>>> import re
>>> r = re.compile('/\*.*?\*/', re.DOTALL)
>>> t = '/* a \nmulti-line\ncomment*/   /*another comment*/'
>>> r.findall(t)
['/* a \nmulti-line\ncomment*/', '/*another comment*/']
>>>

HTH

Harvey

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.




From whisper at oz.net  Thu Jun 13 00:37:30 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 12 Jun 2002 21:37:30 -0700
Subject: Python GUI
In-Reply-To: <9ddd154e.0206122021.1b06f89a@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEAELODIAA.whisper@oz.net>

Fox and FLTK are other options.

http://www.fox-toolkit.org/
http://www.fltk.org/

Neither of these has as wide a following as the GUI's you've already
evaluated.

I would stick with Tkinter if disk and memory usage are that critical to
you - it's look is pretty close to that of the platform it's running on. For
additional widgets, check out the Python Megawidgets.
http://pmw.sourceforge.net/

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of H. Safarzadeh
> Sent: Wednesday, June 12, 2002 21:21
> To: python-list at python.org
> Subject: Python GUI
>
>
> Hi all,
> I know that this subject is discussed in many other places, but I don't
> think it is harmful to talk about it more!
>
>  I want to what library I should use to make GUI for my Python apps?
>  I have examined many of them, but I couldn't find what I need. Here is
> the result of some of my examinations:
>
> -Tkinter: A nice simple lib, but really simple! It dosn't have enough
> widgets, and I do not like its look in addition!
> -wxPython: Really perfect. It has many widgets, and do many works(not
> only in GUI). But it is toooo big!(both on disk and in memory). I need a
> smaller one.
> -pyGTK and pyQt: I do not know enough about them. But I know that I do
> not like GTK look on Windows and Qt for Windows has some problems with
> its lisence(and it is tooo big, too!)
> -Jython/Swing: ....it's Jython, after all!
> -Win32 extensions: It is not portable!!!
>
> Any comments? Other possibilities?
>
> Thanks.
> --
> http://mail.python.org/mailman/listinfo/python-list





From trentm at ActiveState.com  Fri Jun 21 14:19:53 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Fri, 21 Jun 2002 11:19:53 -0700
Subject: Windows versions of Python---pros and cons?
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E201922151A8@admin56.narex.com>; from BPettersen@NAREX.com on Thu, Jun 20, 2002 at 04:52:35PM -0600
References: <60FB8BB7F0EFC7409B75EEEC13E201922151A8@admin56.narex.com>
Message-ID: <20020621111953.B29506@ActiveState.com>

[Bjorn Pettersen wrote]
> > From: Chris Barker [mailto:Chris.Barker at noaa.gov] 
> [snip]
> > 
> > On that note: does anyone know if there is a way to turn a 
> > Python script into something that acts like an application on 
> > Windows...without resorting to Py2exe and the like. What I 
> > want is something just like:
> > 
> > chmod +x scriptname
> > and a #! line at the top 
> > 
> > on Unix. I want it to work on the windows command line.
> 
> If you use the ActiveState version (not sure about the python.org
> version) this should work out-of-the-box. Just type the name of the
> script, you can even leave out the .py extension. Someone else will
> probably tell you how it works, I just know it's got something to do
> with registering the .py extension as an executable extension in the
> registry...

As Mark Hammond mentioned, this is done by adding .py to your PATHEXT
env. var. ActivePython does that (and .pyc, .pyo, .pyw, .pys as well).

Trent

-- 
Trent Mick
TrentM at ActiveState.com




From jadestar at idiom.com  Sun Jun 16 14:59:43 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 16 Jun 2002 18:59:43 GMT
Subject: getgrouplist() ?
References: <mailman.1024149739.29018.python-list@python.org>
Message-ID: <aein6v$2iva$1@news.idiom.com>

Mark McEahern <marklists at mceahern.com> wrote:

>> I need a way to get a back a list of supplemental group ids for a
>> specified uid.  Something like getgrouplist() in my C standard lib.  I
>> don't see a convenient method for this, have I missed it?
>> 
>> The os module contains getgroups(), but that doesn't allow me to
>> specify a uid; it uses only the uid of the current process.

> For what it's worth, the grp module is getting some improvements for 2.3:

>   http://www.python.org/dev/doc/devel/whatsnew/node8.html


 Here's a simple function to build a dictionary of lists, keyed by
 username and containing groupnames for primary GID and any 
 supplemental group memberships:

#!/usr/bin/env python2.2
import grp,pwd

def usersgroups():
	results = {} 				# dict of results
	groups = grp.getgrall()		# list of group entries 
	for i in pwd.getpwall():	# for each user account 
		user = i[0]; gid = i[3]
		# start each user's group list with his/her primary group:
		results[user] =  [grp.getgrgid(gid)[0]]
		# scan groups for this user: 
		for g in groups:
			if user in g[3]: results[user].append(g[0])
	return results 

if __name__=='__main__':

	for i,j in usersgroups().items():
		print i, ":",
		for x in j: 
			print x,
		print 

 It's deliberately simplistic.  I probably could have done something
 fancy with list comprehensions or maps or something, but this 
 should be easy to read.




From tim.one at comcast.net  Fri Jun 21 12:18:18 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 21 Jun 2002 12:18:18 -0400
Subject: RE strings (was: Variable Interpolation - status of PEP 215)
In-Reply-To: <3D132ADC.D04B52EA@engcorp.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEBMPPAA.tim.one@comcast.net>

[unattributed]
> The implementation could choose to cache the result of
> re.compile at module level, thus
>   - speeding up the 'naive' re usage
>   - freeing the programmer from doing the caching in code

[Peter Hansen]
> Any reason the re.match() method could not do the same, internally?

It already does, and always has.  sre increased the internal cache size over
pre, and introduced separate caches for "search" and "replace" uses of
regexp literals.





From gcordova at hebmex.com  Thu Jun 27 15:53:49 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 27 Jun 2002 14:53:49 -0500
Subject: Pythoniac: Thoughts on a hardware Python processor
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233015DE6BA@mail.hebmex.com>

> 
> Ahh! Haven't you heard? Forth came in Fifth and the Forth 
> chip is dead. In fact I don't think anyone is still offering
> a stack based computer (only HP did AFAIK, the HP 3000).
> 
> Dave LeBlanc
> Seattle, WA USA
> 

Which is truly a shame; it's more because of commercial
and market pressures, and nothing to do with true
throughput or technical merit.

Let's see if Python can be ported to Forth, like it was
to Java. If it can, *then* python'll run at hardware speed,
don't you think?

Anyhow, as I said, it's a shame that Forth has been lambasted
so much that it's not really possible to find much info
and code on the net. It's actually a quite nice tool.

:-/

-gus




From kragen at pobox.com  Wed Jun 26 14:14:32 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 26 Jun 2002 14:14:32 -0400
Subject: M2Crypto: select() behaves weird on SSL sockets
References: <3D04ED17.5070606@NOSPAMREMOVETHISxs4all.nl> <ae2str$b0q$1@nntp6.u.washington.edu> <3D066E77.4070901@NOSPAMREMOVETHISxs4all.nl> <ae6bmb$frc$1@dahlia.singnet.com.sg> <3D0790E1.9080904@NOSPAMREMOVETHISxs4all.nl> <831ybavprx.fsf@panacea.canonical.org> <3D0919AF.7050903@NOSPAMREMOVETHISxs4all.nl>
Message-ID: <83it45lxfb.fsf@panacea.canonical.org>

Irmen de Jong <irmen at NOSPAMREMOVETHISxs4all.nl> writes:
> Kragen Sitaker wrote:
>  > Irmen, why aren't you using asyncore?
> 
> My system (Pyro) is built around threads all the way.
> Switching to asyncore/medusa requires a big restructuring of
> my code, and that wasn't part of the plan for the next release of Pyro.

There are many pieces of middle ground between complete threading (two
threads per connection) and complete event-loop-drivenness with just
one thread --- as you're obviously aware, since you're using select()
in a threaded program.

What I guess you're not aware of is that asyncore supports the middle
ground (middle colors?) as well as the event-loop-driven end of the
spectrum.  For example, Zope has an asyncore thread that handles I/O,
but actually processes requests and produces results in some child
threads.


From garrett at bgb.cc  Wed Jun 19 01:11:22 2002
From: garrett at bgb.cc (Don Garrett)
Date: Wed, 19 Jun 2002 05:11:22 GMT
Subject: What If.....    Strong Types
References: <3D0C0F0D.4030900@bgb.cc> <3D0E8C26.7BF0772D@engcorp.com>
Message-ID: <3D1011F1.3060205@bgb.cc>

Peter Hansen wrote:
> Don Garrett wrote:
> 
> 
> It takes work to define those interfaces.  In my opinion, unnecessary,
> inconvenient, and unreadable work...  And Python does have private
> members, of course, for the right definition of private.

   I don't think it would take MUCH work to define those interfaces. The only 
addition work would be a line per member variable. However, any extra work is 
some extra work.

   The other limitation (can't delete, add or change out methods) limits what 
you can do, but doesn't require you to do any more work. Unless, of course, 
it's working around the limitations.

> Preventing modifying methods on an instance would make some of my 
> automated testing really difficult, if not impossible.

   Hum... that's the only specific example I've ever run across where someone 
would need to switch out a method implementation at run time in production 
code. Depickling and several other very nice features probably depend on it, 
but they are core libraries, not general code. I draw a distinction there.

   I suggest that you could create a testing subclass that overrides the 
methods in questions with testing versions of the methods.

   Would that work as well? As easily? In C++ or Java I'd have said that a 
testing class would be to much trouble, in Python it means one line plus the 
new methods that you would have to write anyway. You also have to be smart 
about how class instantiation is done (the bigger problem).

   Another possibility is to say that all methods are function pointers and 
can be replaced with another function pointer, as long as the signiture 
matches up well enough. That would still leave a fixed interface and make the 
main change that I think is needed for compilation. I think.

> I don't understand the "must be declared to exist" part.  If they are declared
> but not defined, they don't exist anyway.  If they are defined, then they
> already exist and a declaration is redundant.  Or were you using some other
> definitions of "define" and "declare"?

   I mean that the class definition specifies all members that instances of 
that class would contain. All instances would contain all of those members, 
and only those members until they are destroyed, even if they only hold a 
value of None.

> Fork a branch of Python and experiment!

   Eventually, I may to exactly that! Unfortunatly, I'm spening most of my 
time trying to pay the morgage (and Python helps me do that, so I feel that I 
do owe time to the Python community), so I spend time trying to decide what I 
would be doing if I really had the time. ;>

> I suppose, since Python is flexible and adding all these things to the
> language would tend to make it more restrictive and, well, brittle, you 
> could probably call the language "Stick". :-)
> 
> -Peter




From fcorneti at libero.it  Tue Jun 11 08:45:09 2002
From: fcorneti at libero.it (Fabio Corneti)
Date: Tue, 11 Jun 2002 12:45:09 GMT
Subject: Defining a method final
References: <6glN8.35376$86.884025@twister1.libero.it> <mailman.1023796095.1427.python-list@python.org>
Message-ID: <phmN8.36164$K4.893500@twister2.libero.it>

I would like to define a method for a class which has
not to be overriden. Of course, since the developer of
this project are two, this could be established by
convention, but if the project will grow consistently 
another developer could override the method by mistake,
causing unpleasant side effects.

-- 
Fabio Corneti


From sholden at holdenweb.com  Wed Jun 19 12:18:09 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 19 Jun 2002 12:18:09 -0400
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <yXBP8.44417$GY.13865423@twister.nyroc.rr.com>
Message-ID: <kQjQ8.213355$Fp1.28047@atlpnn01.usenetserver.com>

"George Hester" <hesterloli at hotmail.com> bad-temperedly top-posted in reply
to sensible advice from Peter Hansen...
> Listen you aren't my daddy.  If you do not want to help constructively
then
> please get off my boat.
>
Sound advice, since it is clearly about to sink. You were advised to get to
know Python locally before trying to write ASP scripts in it. If you choose
to ignore that advice you are making a rod for your own back.

still-i-suppose-it-*is*-your-back-ly y'rs  - steve
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From phd at phd.pp.ru  Wed Jun 19 08:03:49 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 19 Jun 2002 16:03:49 +0400
Subject: procmail replacement in Python
In-Reply-To: <oqit4grq03.fsf@carouge.sram.qc.ca>; from pinard@iro.umontreal.ca on Tue, Jun 18, 2002 at 09:55:40PM -0400
References: <20020618235901.GA2197@lilith.my-fqdn.de> <oqit4grq03.fsf@carouge.sram.qc.ca>
Message-ID: <20020619160349.L4127@phd.pp.ru>

On Tue, Jun 18, 2002 at 09:55:40PM -0400, Fran?ois Pinard wrote:
> [Gerhard H?ring]
> 
> > I've recently looked into the procmail sources and run away screaming.
> 
> ROTFL!  If I remember well, the original author knew he had a peculiar
> writing style, and was writing in some prominent file (either an FAQ or
> the README, or something like) that his style was not to be discussed! :-)

   When I first saw the procmail code, I thought it was generated by some
code obfuscator :)

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From shalehperry at attbi.com  Thu Jun 27 15:45:24 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 27 Jun 2002 12:45:24 -0700 (PDT)
Subject: squared functions--most Pythonic way?
In-Reply-To: <ballabio-E54F22.21380327062002@newssrv.fastweb.it>
Message-ID: <XFMail.20020627124524.shalehperry@attbi.com>

> 
> functional programming:
> 
>     # with an internal named function
>     def fsquare(f):
>         def f2(x):
>             return f(x)**2
>         return f2
> 

this seems to be the most pythonic (lambda is not the preferred approach
usually).

Of course someone will mention taking this to the next step with a compose()
function which take f(x) and g(x).




From peter at engcorp.com  Thu Jun 27 10:00:22 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 27 Jun 2002 10:00:22 -0400
Subject: How to get rid the new line
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org> <3D1AD534.A74EC9F4@ipm.fhg.de>
Message-ID: <3D1B1A76.563E5394@engcorp.com>

Markus von Ehr wrote:
> 
>         f = open(filename, 'r')
>         lines = f.readlines()
>         line1 = lines[0] # exemplarily for first line
>         line1 = line1[0:len(line1)-1]
> 
> or:
>         f = open(filename, 'r')
>         lines = f.readlines()
>         line1 = lines[0][0:len(line1)-1] # exemplarily for first line

I believe this is unsafe.  The final line may not be terminated with \n.

-Peter


From gleki at gol.ge  Thu Jun 13 04:22:20 2002
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Thu, 13 Jun 2002 10:22:20 +0200
Subject: py2exe 4 linux?
Message-ID: <3D08563C.33D041E3@gol.ge>

Hi all!

Is there anything like py2exe for Linux?

I have never tried to find such a tool as
chmod +x
with the 1st line of
#!/usr/bin/python
suffices for POSIX systems.

Anyway, maybe there's the tool for linux?

Thx,

Giorgi



From ods at fep.ru  Thu Jun 20 07:06:34 2002
From: ods at fep.ru (Denis S. Otkidach)
Date: Thu, 20 Jun 2002 15:06:34 +0400 (MSD)
Subject: Useful RE patterns (was: Variable Interpolation - status of PEP
 215)
In-Reply-To: <20020620145453.B25214@phd.pp.ru>
Message-ID: <Pine.LNX.4.44L.0206201501290.1553-100000@ods.pp.ru>

On Thu, 20 Jun 2002, Oleg Broytmann wrote:

OB> > If I were to add a dozen (or so) patterns to the (S)RE
OB> module,
OB> > what should I pick?  What patterns do you find yourself
OB> using
OB> > over and over again?
OB>
OB>    IP address (but it is required to test that every nibble
OB> is < 256)

This can be achieved in regex, try
'.'.join(['(1?\d\d|2[0-4]\d|25[0-5])']*4)

OB>    URL
OB>    e-mail address (yes, I know very good how hard it is)
OB>    identificator (letters+digits, starting with a letter)
OB>    empty line
OB>    line consisted only with whitespaces
OB>    comment line, starting with #, may have whitespaces
OB> before #

Last three are too simple to be a pattern, IMO.





From maxm at mxm.dk  Mon Jun 24 05:48:47 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 24 Jun 2002 11:48:47 +0200
Subject: Generic Python
References: <3D16E465.F0D2509E@ifib.uni-karlsruhe.de>
Message-ID: <3D16EAFF.9010709@mxm.dk>

Uwe Mayer wrote:

> The problem is that I've got a base class and there should be many
> subclasses of it. Each subclass just overwriting an output method and
> serving as a template: 
> 
> class base1:
>     def __init__(self):
>         ...
>     def output(self):
>         ...
> 
> class template1(base1):
>     def output(self):
>         ...
> 
> class template2(base1):
>     def ouptut(self):
>         ...
> 
> I need many, many of these "template" classes, so a sorter way would be
> to accustom the base class to take another argument which then outputs
> the right thing:


This problems begs for you to create a factory function::

     def factory(objType, initValue):

         classTable = {
             'template1':template1
             'template2':template2
         }
         return classTable[objType](initValue)


This way you don't have to know the specifics anout the class that your 
create.

you would then just have to::

     import templates
     newObj = templates.factory('template2', 'Initial text')

regards Max M



From phr-n2002b at NOSPAMnightsong.com  Fri Jun 28 02:14:40 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 27 Jun 2002 23:14:40 -0700
Subject: crypt for windows?
References: <mailman.1025241336.1209.python-list@python.org>
Message-ID: <7x6604vsj3.fsf@ruckus.brouhaha.com>

Klaus Reinhardt <K.Rdt at TU-Berlin.DE> writes:
> No, I dont like cygwin. I tried pgp, but it's to
> complicated. So I decided to aks me for
> password each session. For this I can use
> python in 'the next script'.

I don't understand your question then.  When you asked about crypt and
passwords, I thought you meant something compatible with the Unix
crypt(3) function that hashes the passwords in Unix password files.  
Do you mean you just want to encrypt stuff?  There's plenty of
modules around for that.



From peter at engcorp.com  Sun Jun  2 02:08:09 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 02 Jun 2002 02:08:09 -0400
Subject: system variables
References: <WhgK8.3670$3t6.160209@ozemail.com.au>
Message-ID: <3CF9B649.39F4D6D5@engcorp.com>

Daniel Powell wrote:
> 
> hi guys,
> I'm new to python and I'm teaching myself from a website. I'm stumped on one
> thing this is the exact wording
> 
> "You will need to cause the directory containing the file named python.exe
> to be listed in your system environment variable named path."
> the website is
> http://softwaredev.earthweb.com/sdopen/article/0,,12077_626311,00.html
> it says when you do this if you type something like
> "python junk.py" it will run the "junk.py" scipt
> if anyone can help me do this i would grealty appreciate it.
> My OS is Win 98SE

Hi Daniel.  Edit the file c:\autoexec.bat to modify the PATH statement
that is there so it has the python directory listed in it.  For example,
if it reads

   SET PATH=c:\windows;c:\windows\command

then you should make it this (substituting the appropriate folder in
place of c:\python22 for your own system, depending on where you 
installed Python):

   SET PATH=c:\windows;c:\windows\command;c:\python22

Test this out manually by typing the above line at a DOS prompt,
then typing just "python" (when you are not already in the Python
folder).  If you type "SET" by itself you'll see the current PATH
setting and can adjust the above lines appropriately.

After changing AUTOEXEC.BAT you need to reboot, and then the changes
should be permanent.

-Peter


From cliechti at gmx.net  Fri Jun  7 17:43:05 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 7 Jun 2002 23:43:05 +0200
Subject: direct output to printer
References: <adr815$e6m39$2@hades.rz.uni-sb.de>
Message-ID: <Xns9226F1A931224cliechtigmxnet@62.2.16.82>

Uwe Schmitt <uwe.schmitt at procoders.net> wrote in 
news:adr815$e6m39$2 at hades.rz.uni-sb.de:
> I'd like to print barcodes with a special barcode printer.
> Therefore I have to send raw ASCII data to the printer.
> How can I do this using Python under Windows ???

there is a raw print function in win32all, but i haven't used that. for the 
barcode-printer i used, i sent a bitmap with the barcode. that way you can 
print the same barcode on paper or preview on the screen (i used wxPython).

there are some free barcode fonts around. i use this one:
http://www.squaregear.net/fonts/

chris

-- 
Chris <cliechti at gmx.net>



From m2 at plusseven.com  Fri Jun 14 23:35:06 2002
From: m2 at plusseven.com (Alex Polite)
Date: Sat, 15 Jun 2002 05:35:06 +0200
Subject: Python and SQL Server 2000
In-Reply-To: <ZCuO8.10006$9b.678135@typhoon.austin.rr.com>
References: <ZCuO8.10006$9b.678135@typhoon.austin.rr.com>
Message-ID: <20020615033506.GA8821@matijek.plusseven.com>

I tried to do this a couple of years ago. I think MS SQL Server was
version 6.x 

We used ADO connections via the win32com module. It worked but it was
kind of sluggish. Then to speed things up I wrote a wrapper around bpc
(bulk something something, look it up in the MS SQL manual). I think
we couldn't use mxODBC due to licensing issues. Then in
the end we moved to MySql.

-- 
Alex Polite
http://plusseven.com/gpg/




From jepler at unpythonic.net  Thu Jun  6 14:30:38 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Thu, 6 Jun 2002 13:30:38 -0500
Subject: Compiling Python
In-Reply-To: <838z5sgx7w.fsf@panacea.canonical.org>
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet> <3CFDF148.AA2035FF@info.unicaen.fr> <3CFE49CA.46A5FEE1@tds.net> <838z5sgx7w.fsf@panacea.canonical.org>
Message-ID: <20020606183033.GA12501@unpythonic.net>

On Thu, Jun 06, 2002 at 12:58:59PM -0400, Kragen Sitaker wrote:
> "Edward K. Ream" <edream at tds.net> writes:
> > BTW, a compiled version of a Python program would be valuable
> > commercially, as it would allow companies to protect their source code,
> > as well as (possibly) improve performance.
> 
> I'm worried about this myself; perhaps the compiler could include the
> source code in the resulting executable to prevent this.

Your sarcasm is lost on me.

Jeff




From jadestar at idiom.com  Tue Jun 11 05:07:38 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 11 Jun 2002 09:07:38 GMT
Subject: How to play a .wav sound?
References: <ad8jo7$ohl$1@wanadoo.fr> <AOWJ8.7985$_15.227658@news4.ulv.nextra.no>
Message-ID: <ae4ekq$gtg$1@news.idiom.com>

Erlend J. Leiknes <nookieNO_SP_AM at online.no> wrote:

> Think this should work... not tested.

 I doesn't.  I doesn't in principle (cat foo.wav > /dev/dsp doesn't work
 for me) and the code is wrong in at least one spot:

...
>         sdata = f.read(1024)
>         while sdata:
>             d.write(sdata)
... 

 that would only copy 1K of data

> "Bruno Bellamy" <bellamy at neverland.net> wrote in message
> news:ad8jo7$ohl$1 at wanadoo.fr...
>> I guess the question has probably been asked many times, sorry for
>> bothering...
>> But I searched here and there, and I couldn't find how to do that, simply
>> play a .wav sound from a python program using Tkinter.

>> Moreover, I need to do that in a program that would run similarly under
>> Linux and Windows.

>> I saw there's a python module called wave, but it seems it can only read
>> .wav files. I coulnd't find a way to actually play the sound. Maybe it's
>> hidden somewhere?

>> If anybody can help me, that would be cool. :)


 I was stumped for a little while last night as well.  Unfortunately
 I don't have a simple working example, yet; but it looks like
 pysolsoundserver.so and the PySol (Python Solitaire) sources might 
 be a good start.  Apparently PyGame also has some sophisticated 
 sound handling.  There doesn't seem to be a way to do it (easily)
 with the standard libraries, though.




From dalke at dalkescientific.com  Sun Jun  2 17:24:40 2002
From: dalke at dalkescientific.com (Andrew Dalke)
Date: Sun, 2 Jun 2002 15:24:40 -0600
Subject: convert string to a method of a class at runtime?
References: <5YvK8.285$Vm3.130161@news.uswest.net>
Message-ID: <ade2nc$ori$1@slb5.atl.mindspring.net>

Kevin Altis:
>I would like to be able to take a string containing the definition of a
>method, compile the method and then add it to an instance variable at
>runtime, so that the method defined in the string is part of the class just
>as if it had actually been in the class definition to begin with when the
>module was first used.

I am wary about using such tricks.  Still, try this approach instead

>>> class H:
...  def hello(self, s):
...   print s
...
>>> h = H()
>>> h.hello('world')
world
>>> s = """def world(self, s):
...  print s
...
... """
>>> d = {}
>>> exec s in d
>>> H.world = d["world"]
>>> h.world('new world')
new world
>>>

                    Andrew
                    dalke at dalkescientific.com






From marklists at mceahern.com  Thu Jun 20 08:24:59 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 20 Jun 2002 07:24:59 -0500
Subject: module string.py
In-Reply-To: <3d11bd6a.11177520@news.polito.it>
Message-ID: <NCBBLFCOHHDIKCAFGCFBAEJPLFAA.marklists@mceahern.com>

> 1)  looking at the module "string.py" I have seen that are described
> some functions like "lower", "rjust", etc. Ther problem is that it
> seems not to describe exactly how this functions are made, but there
> is only a comment about their properties. So in which file are these
> functions described?

Just so you know, those functions are also attributes of strings:

  >>> s = "foo"
  >>> s.upper()
  FOO

For documentation, rely on the __doc__ strings:

  >>> print s.upper.__doc__
  S.upper() -> string

  Return a copy of the ...

Or, use help():

  >>> help(s.upper)

(press q to close the resulting screen)

> 2) I tried to cancel the file "string.py". After that, the Python
> Interpreter doesn't work anymore (it can't even start working). Why?

What do you mean by "cancel"?  Did you delete it?

// m
-





From martin at came.sbg.ac.at  Tue Jun 11 12:02:10 2002
From: martin at came.sbg.ac.at (Martin Krallinger)
Date: 11 Jun 2002 09:02:10 -0700
Subject: tk gui for python
Message-ID: <e0b1d6f6.0206110802.7e38179c@posting.google.com>

Dear all,

I would appreciate any sample code to generate tk gui for python.
i just need to load a file, of past instead a text (protein sequence)
and run on it a c programm and return the content of it's output

any sample code is welcomed,

thanks

Martin


From Sascha.Ferley at infineon.net  Sun Jun 23 22:51:32 2002
From: Sascha.Ferley at infineon.net (Sascha Ferley)
Date: Sun, 23 Jun 2002 20:51:32 -0600 (MDT)
Subject: PIL, Python and CGI
In-Reply-To: <200206241204.16542.rjones@ekit-inc.com>
Message-ID: <Pine.GSO.4.33.0206232047450.27090-100000@infinity.infineon.net>

I actually got some nice little traceback that might be of use ..

Traceback (most recent call last):
  File "/export/server/web/CPSC461/cgi-bin/test4.py", line 10, in ?
    import _imaging
ImportError: ld.so.1: /usr/local/bin/python: fatal: libjpeg.so.6: open
failed: No such file or directory

though:

infinity:/usr/local/lib->ls -ld libjpeg*
lrwxrwxrwx   1 root     other         12 Jun 16 18:58 libjpeg.so ->
libjpeg.so.6*
lrwxrwxrwx   1 root     other         12 Jun 16 18:58 libjpeg.so.1 ->
libjpeg.so.6*
-rwxr-xr-x   1 root     bin       137924 May 21  2001 libjpeg.so.6
infinity:/usr/local/lib->

Which is interesting ... :)

Thanks
Sascha







From boud at valdyas.org  Sat Jun  1 15:04:23 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Sat, 01 Jun 2002 21:04:23 +0200
Subject: Copying a database
References: <FL6K8.771$NG1.217@newsread2.prod.itd.earthlink.net>
Message-ID: <3cf91e14$0$3864$e4fe514c@dreader4.news.xs4all.nl>

Billy Ng wrote:

> I need you guys help on mysql again.  I need to push my python program to
> production now, but I don't want to recreate the mysql database on the
> production box.  I want to copy the database from the development box to
> production box.. Would anybody please tell me how to do it, thanks!
> 

Investigate mysqldump -- it dumps your entire database to a file
that you can then install on another mysql server using the basic
mysql command line client. It's also one of the better options
for doing backups -- backing up the database files themselves is
not future-proof, as I've found to my detriment.

-- 
Boudewijn Rempt | http://www.valdyas.org


From sean.mcgrath at propylon.com  Mon Jun 24 06:08:15 2002
From: sean.mcgrath at propylon.com (Sean McGrath)
Date: Mon, 24 Jun 2002 11:08:15 +0100
Subject: a Tree data-structure class?
In-Reply-To: <20020623160005.4179.75165.Mailman@mail.python.org>
Message-ID: <5.1.0.14.0.20020624110603.00b09450@mail.digitome.com>

>On 21 Jun 2002 06:10:30 -0700, steve at ferg.org (Stephen Ferg) wrote:
>
> >I do a lot of processing of tree-structured data, so I'm thinking it
> >would be useful to develop a generic Tree class that I could adapt to
> >various applications.  Or perhaps a Tree class and a generic Node
> >class.
> >
> >But before I start trying to re-invent the wheel, I thought I'd ask if
> >anybody knows of something like this that has already been developed.

I'd suggest looking at Python's XML libraries. DOM and Pyxie in partic. are
basically tree data structures suitable for directed acyclic
graph use.

regards,
Sean







From tim.one at comcast.net  Fri Jun 28 12:22:01 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 28 Jun 2002 12:22:01 -0400
Subject: doctest for java
In-Reply-To: <afhv70$15mk$1@news.adamastor.ac.za>
Message-ID: <LNBBLJKPBEHFEDALKOLCMEMAAAAB.tim.one@comcast.net>

[Janto Dreijer]
> Are there any attempts to port the "doctest" module to Java? (All
> I can find using google is some testing class for the JavaDoc program
> calling itself "DocTest".)
>
> At a glance it looks like a worthwhile effort...extracting Jython
> "examples"
> from /**...*/ comments also seems simple to implement.

Last I heard, Python's std doctest.py worked fine under Jython.  If anything
else simple is needed, you're elected <wink>.





From shalehperry at attbi.com  Fri Jun  7 10:22:46 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 07 Jun 2002 07:22:46 -0700 (PDT)
Subject: Detecting OS and Window Manager
In-Reply-To: <3d00bfb5$1_2@hpb10302.boi.hp.com>
Message-ID: <XFMail.20020607072246.shalehperry@attbi.com>

> 
> Of course, this will probably not do what you want in this case, as kdecore
> will probably load if it is installed, without regard to whether the current
> user is running KDE.  I don't have much experience with this particular
> topic yet.
> 

correct, especially on a developers box who has GNOME, KDE, ROX, etc installed.




From mal at lemburg.com  Tue Jun 11 10:26:13 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 11 Jun 2002 16:26:13 +0200
Subject: 'for every' and 'for any'
References: <20020526135944.A32690@hishome.net> <20020526091742.A987@unpythonic.net> <mailman.1022567685.7803.python-list@python.org> <slrnaf7af3.8kn.quinn@schilling.ugcs.caltech.edu> <mailman.1022658469.27441.python-list@python.org> <ae34kt$6r3$1@news.idiom.com>
Message-ID: <3D060885.2010504@lemburg.com>

James T. Dennis wrote:
> Oren Tirosh <oren-py-l at hishome.net> wrote:
> 
> 
>>>Then drop those functions into your personal utility library and use 'em
>>>as often as you want :)
>>
> 
>>I don't have a personal utility library. It's on purpose. I don't have
>>personalized key bindings. I try to avoid customization.  Customization is
>>a big part of what makes one programmer's code difficult to maintain by
>>another programmer, what makes on programmers's workstation unusable by 
>>another and makes combining code from several sources difficult.
> 
> 
>  ...
> 
> 
>>One of the things I like about Python is that it comes with everything in
>>the box. 
> 
> 
>>	Oren
> 
> 
>  Yet, this very discussion proves that "everything" is not already in
>  the box.  I'll grant that Python includes a very large set of batteries
>  --- but I'd still say that there are many useful functions and classes
>  that can still be added.
> 
>  Those should start as personal "utility" libraries, be discussed here
>  and in PEPs, and (when their general utility is established) added to
>  appropriate (standard or optional) libraries.
> 
>  Obviously you don't re-invent the wheel for each project that you join;
>  you are re-using some code that you've used before, even if it's from 
>  "finger macros."   I can see the argument in favor of adapting your
>  functions to conform to local conventions and every programming team has
>  to collaborate to find compatible conventions and standards for their
>  projects.  Any non-trivial project (non-trivial, in this case meaning
>  "requires a *team* of programmers (> 3)) is going to find some elements
>  of their project that go beyond established coding conventions.  The
>  alternative to "building a (domain/application specific) library" is
>  to eschew code re-use.
> 
>  In this particular case the all() and every() functions that were
>  presented are elegant and seem useful.  As others have pointed out the
>  implementations are trivial; so the questions become: "does it 'pollute
>  the namespace' to add them to the core library?" and (if so) "where would 
>  we put them?"

You might want to check out mxTools which has quite a few
of these gimmicks implemented in C.

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/
Meet us at EuroPython 2002:                 http://www.europython.org/





From karczma at info.unicaen.fr  Wed Jun 19 05:21:57 2002
From: karczma at info.unicaen.fr (Jerzy Karczmarczuk)
Date: Wed, 19 Jun 2002 11:21:57 +0200
Subject: Pedagogic advice needed
Message-ID: <3D104D35.E0932297@info.unicaen.fr>

Dear Snake Charmers,

I am going to teach to some biology students (a branch called here
"neurosciences"; you may speculate yourself what kind of computation
proficiency they need...) - 

- a comp. sci. course under a buzzword title "Scientific programming".
The plan was to use Matlab as the main vehicle.
Then after a while it was clear that in order to give to *all* students
the possibility to work at home, we need a free programming package, so
the idea was to use SciLab.

Then I began to digest all that, and the obvious question was: why not
Python?

==

My questions are:

1. Does anybody here have some teaching experience in a similar context?
2. Assuming that the visualisation issues, all kind of plots, graphs 
   *and animations* are very important, how would you organize with
   Python such a work?

Of course I know Numeric Python, Scientific Python modules, and other
standard stuff permitting to do all kind of graphic exercices and
demonstrations (eg., all the wx bazar).

But I *must* avoid the low-level programming, we won't have time for that.
We will need a reasonable complete scientific 3D plotting package usable
by people without too much experience. I plan to do some non-trivial prog-
ramming with rich data structures (otherwise I would stick to Scilab), but
a decent *integrated* scientific library, I mean: a framework, not just a 
set of - powerful but rather atomic - procedures, would be quite useful.

(I checked the Obvious Suspects, the Vault of Parnassus, etc., I am veryfing
all that stuff, but perhaps some of you know something really succulent and
full of vitamines. I need *your experience*, NOT just standard Web links.)

Thank you in advance.

Jerzy Karczmarczuk
Caen, France


From hwcowan at hotmail.com  Fri Jun 14 15:47:32 2002
From: hwcowan at hotmail.com (Hugh Cowan)
Date: 14 Jun 2002 12:47:32 -0700
Subject: Creating Dynamic Web Pages (ZOPE)
References: <46ca81a0.0206062016.789d3af6@posting.google.com> <mailman.1023437832.11588.python-list@python.org> <46ca81a0.0206070654.787ac0f7@posting.google.com> <mailman.1023465260.21475.python-list@python.org>
Message-ID: <46ca81a0.0206141147.1a0d164@posting.google.com>

Hello,

> Now, I would slow down a bit. Surely Zope is a good Web tool, and the most 
> known in the Python world, but it's not the only one.
> 
> Based on your requirements it is likely the best choice for you, but it's 
> *my* opinion, so you could do worse than to give a look to a few alternative 
> options:
> 
> Webware (the most J2EE-like, and quite featureful)
> http://webware.sourceforge.net/
> 
> SkunkWeb (smaller, but well endowed)
> http://skunkweb.sourceforge.net/
> 
> Quixote (the smallest, a little jewel, PTLs are a charm)
> http://www.mems-exchange.org/software/quixote/
> 
> All of them within the reach of the Snake, of course. :^) And there's also 
> Paul Boddie's description of some of them, to save another little bit of 
> your time:
> 
> http://www.boddie.org.uk/python/web_frameworks.html

Wow -- thanks for the additional information and links.  It certainly
doesn't hurt to look at other solutions, although I am quite happy
with what I have seen so far with Zope.  I actually didn't think that
there were many *other* options out there that were available!

Is this sort / type of software what one would classify as "Web
Framework"? (as in the above link)  I am having difficulty looking for
a description or class for this type of software.  I haven't had time
to try Zope yet, but for instance I am assuming (not always a good
thing mind you) that you could create Static web pages (either coding
or WYSIWYG).  In those terms Zope is an HTML editor -- but I know that
it offers a lot more functionality that just creating web-pages.

Someone else has mentioned using programs such as Dreamweaver -- but I
didn't think that they were in the same class / category of software?

Now that you have introduced me to some other software packages (which
I will look at) I am just trying to sort out the different types of
software (for myself and so that I can better explain why we need more
than MS-FrontPage) between Zope (Web Framework), HTML Editors /
Coders, and RAD Web development tools.  I am sure that there is some
overlap between them all but I am just trying to figure out the main
differences between them.

Thanks again for the help and answers.

Hugh,


From mutsuura at mutsuura.com  Thu Jun 13 09:36:06 2002
From: mutsuura at mutsuura.com (Attila Horvath)
Date: Thu, 13 Jun 2002 09:36:06 -0400 (EDT)
Subject: Poll: Readable languages
In-Reply-To: <3D08810B.2040707@gmx.de>
Message-ID: <Pine.LNX.4.44.0206130853001.20311-100000@adams.patriot.net>

Ingo

Though I subscribe to PYTHON and am new to the language, I've programmed
for 25+ years and know various languages to varying degrees. At the risk
of stirring up everyone's ire [although not my intention], I notice
'assembler' [of any kind] is not on your list of languages. Let me 
explain...

While it sounds ludicrous to include assembler, and I'll not defend
assembler as being inherently 'readable', I want to point out that 
MANY attempts have been made in the past to specify and design
'natural-readable' computer languages - COBOL being among early ones.

Since you're polling for opinions I suggest that there is no truth that 
ANY language has succeeded in being 'readable'. My absurd suggestion that 
'assembler' is missing from your list is my attempt at offering an opinion 
that the 'best' of languages can be made horrifically unreadable while the 
'worst' of languages can be made reasonably readable. For proof, see:

	http://www.ioccc.org/

In my years of experience I have come across VERY readable assembler and
very unreadable COBOL, PASCAL, FORTRAN, ADA, C/C++, JAVA, etc. So I
suggest that a langauge does not inherently make resultant code readable.
Rather the discipline applied by the programmer/author to the work is what
inherently makes any piece of work easily understood. I'm sure that even
with PYTHON's strict indentation rules, obfuscated contests can easily be
held with very clever winning entries!

So why 'assembler'? Because, [1]with the plethera of high level languages
in existance today none can teach programmers more about the inner
workings of both the language and the machine on which it's executed -
inner workings like memory usage, CPU cycles, efficiencies, performances,
etc.; and [2]ultimately ALL languages executed on ALL machines have to be
translated (converted) minimally down to one comman denominator, assembler
[even if it's simply for debugging purposes].

Before I'm accused of missing the point, I AM NOT SUGGESTING THAT PEOPLE 
MUST PROGRAM IN ASSEMBLER. "THAT" would be ludicrous! I am suggesting 
however that assembler also qualifies for the specific questions you 
asked!

only my opinion, attila

PS: My reference to http://www.ioccc.org/ does not imply that "C" is a 
    readable language.

On Thu, 13 Jun 2002, Ingo Linkweiler wrote:

> Date: Thu, 13 Jun 2002 13:24:59 +0200
> From: Ingo Linkweiler <i.linkweiler at gmx.de>
> Reply-To: i.linkweiler at gmx.net
> To: python-list at python.org
> Subject: Re: Poll: Readable languages
> 
> Poll:
> How readable is the code of this languages?
> 
> I do not want to give an exact definition of "readable", but it should 
> include each of this questions with ~25%:
> 
> - Can someone who has never seen the language before understand what the 
> program is doing without reference material?
> - Can a developer write code that they can return to after six months 
> and still understand it?
> - Can you use this language for teaching basics of programming and for 
> showing, how an easy algorithm works?
> - What is your general optionion about his language?
> 
> 1.)
> -----
> ABC:
> ADA:
> APL:
> BETA:
> C++:
> COBOL:
> DELPHI:
> EIFFEL:
> FORTH:
> FORTRAN:
> JAVA:
> LISP:
> MODULA:
> PASCAL:
> PERL:
> PROLOG:
> PYTHON:
> SML:
> SCHEME:
> SMALLTALK:
> TCL:
> VISUAL BASIC:
> other  (  *name of language * ):
> -----
> 2.)
> My most used language is:
> -----
> 
> Please vote....
> 1 = Easy to read, like pseudocode :-)
> 2 = Good
> 3 = Average
> 4 = difficult
> 5 = nearly impossible, when not knowing the language :-(
> blank = I do not know this language or have no opinion.
> 
> Please only vote, if you know at lease 2 languages.
> Please reply to : i.linkweiler at gmx.net
> 
> This poll is done for my diploma paper at university of dortmund, germany.
> I will not use your e-mail for any purpose.
> 
> Thanks for your help.
> Ingo Linkweiler
> 
> PS: I know, that a NG is not the best place for a poll, and some people 
> do not like questions like this in a NG.
> But as I want to get opinions of many programmers, including users of 
> NGs, I hope you can accept it.
> 
> 

                                    _          
Mutsuura Associates, Inc.         /\ \         
P.O. Box 1238                    /  \ \        
Vienna, VA 22183                / /\ \ \       
                               / / /\ \ \      
E-MAIL: contact at mutsuura.com  / /_/  \ \ \     
WEB: http://www.mutsuura.com /  \ \  /  \ \    
                            / /\ \ \/ /\ \ \   
MAIN:(703)281-9722         / / /\ \/ / /\ \ \  
CELL:(703)863-1933        / / /  \  / /  \ \ \ 
FAX :(703)281-9744       / / /    \/_/    \ \_\
SBA/SDB CERTIFIED        \/_/              \/_/





From schliep at Octopussy.MI.Uni-Koeln.DE  Mon Jun 10 06:11:57 2002
From: schliep at Octopussy.MI.Uni-Koeln.DE (Alexander Schliep)
Date: 10 Jun 2002 12:11:57 +0200
Subject: Screen resolution again
References: <3D039794.9549C54@bellatlantic.net>
Message-ID: <tq9vg8rxx1u.fsf@Octopussy.MI.Uni-Koeln.DE>

The following snippet centers a splash screen dialog on the screen.

class SplashScreen(Toplevel):

    ...

    def CenterOnScreen(self):
        self.update_idletasks()
        xmax = self.winfo_screenwidth()
        ymax = self.winfo_screenheight()
        x0 = (xmax - self.winfo_reqwidth()) / 2
        y0 = (ymax - self.winfo_reqheight()) / 2
        self.geometry("+%d+%d" % (x0, y0))


This works on Windows, Linux, MacOS X ...

Cheers,
Alexander

-- 
Alexander Schliep                    schliep at zpr.uni-koeln.de


From not.this at seebelow.org  Thu Jun 13 09:42:57 2002
From: not.this at seebelow.org (Grant Griffin)
Date: Thu, 13 Jun 2002 08:42:57 -0500
Subject: why not "'in' in 'in'"?
References: <ae8dg70217g@drn.newsguy.com> <3D080482.7A7FD514@alcyone.com> <3D082A62.CF800C75@seebelow.org> <3D083298.470850D0@alcyone.com>
Message-ID: <3D08A161.A9F3136F@seebelow.org>

Erik Max Francis wrote:
> 
> Grant Griffin wrote:
> 
> > In that case, perhaps I should have thought of a better one <wink>.
> > In
> > any event, in the case of strings, I think one might stretch the
> > concept
> > of "membership" to include substrings.  In fact, note that since
> > Python
> > has no actual character type (because a character is adequately
> > represented as a string of length one) the "in" operator actually
> > _does_
> > test membership of strings in strings.  But for some reason, it's
> > limited to testing strings of length one.  Go figure!
> 
> A string is a sequence of one-character strings.  It's not a sequence of
> all the subsequences of a string.
> 
> > Honest answer: although I can't see much use for that construct, I
> > can't
> > see much harm in it either.  So although it's probably not worth
> > adding
> > to Python, I'm sure _somebody_ would find a use for it.
> 
> But that's exactly the problem; if you want [1, 2] in [1, 2, 3, 4] to
> return true (i.e., for the `in' operator to test substrings, not just
> membership), then you have an ambiguity problem.  A list can contain
> sublists as elements, after all.
> 
> > But anyway, let's "just say no" to the tuple case rather than confuse
> > ourselves; let's just make the string version of "in" a special case.
> 
> Why make a special case for it when string.find and S.find already exist
> and are included expressly for that purpose?

Again, it has to do with the fact that the illegal multi-character "in"
is intuitive and reads better.  Compare the following:

	if 'mortgage' in email_subject:
	   ...

to:

	if email_subject.find('mortgage') >= 0:
           ...

I consider the former to be highly practical, if not highly pure.

As to the latter, I frequently made this mistake at first:

	if email_subject.find('mortage'):
           ...

which is fairly intutitive, but doesn't work, of course, because "find"
returns -1 (logical true) in the failure case!  That in itself is
counter-intuitive--though it makes sense in the context of "find" being
intended as a position detector, not a substring tester.  (Maybe it
should have been called "position" instead...)  But the very fact that a
failed "find" returns true suggests that we're on the wrong path--at
least as far as intuition/readability goes.

Originally, I would find myself automatically typing the illegal "in"
form (because it's highly intuitive, though illegal), but by now I've
nearly trained myself to type the "find" form.  But even so, I find it
somewhat hard to read.

if-i-were-highly-trainable-i'd-be-using-perl-<wink>-ly y'rs,

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation            http://www.iowegian.com


From pyth at devel.trillke.net  Wed Jun  5 11:48:14 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 5 Jun 2002 17:48:14 +0200
Subject: self
In-Reply-To: <xzhL8.21696$tK.5425664@news02.optonline.net>; from NKluge@optonline.net on Wed, Jun 05, 2002 at 05:45:01AM +0000
References: <mailman.1023208220.19083.python-list@python.org> <TL7L8.79$4p7.4100181@newnews.cc.stevens-tech.edu> <mailman.1023224648.434.python-list@python.org> <xzhL8.21696$tK.5425664@news02.optonline.net>
Message-ID: <20020605174814.C30696@prim.han.de>

Nicola Kluge wrote:
> This is an interesting idea.  However, your approach doesn't handle the dot
> notation.
> For example,
> If object.x was = to 3 and g was = to 4 I need  that
> e.f='object.x*g' gives 12.
> Just having e=Evaluator({'a':'3', 'b':'a*17', 'c':
> 'b*a','object.x':'3','g':'4'})
> doesn't work.  While parsing, 'object.x' is not taken as one symbol in your
> approach.
> I need to have object capabilities in the equations.

i think i have to dig deeper into the python sources to see how 
and if your task could be done cleanly. One of the main problems seems
to be that you can't pass your own 'dictish' object to 'eval' or 'exec'.
(Actually you can, but the interception points like '__getattr__' 
don't get called. smells like a bug to me though this might be arguable).

And yes, i think that having an explicit 'self' is a very good pythonic idea (tm).
i know because i have done some pretty weird stuff like transfering 'locals()'
into the self-dict and back but it often isn't worse it. 

have fun,

    holger




From skip at pobox.com  Sun Jun 16 20:28:12 2002
From: skip at pobox.com (Skip Montanaro)
Date: Sun, 16 Jun 2002 19:28:12 -0500
Subject: xmlrpclib/python and fault codes
In-Reply-To: <20020616220103.85789.qmail@web20801.mail.yahoo.com>
References: <20020616220103.85789.qmail@web20801.mail.yahoo.com>
Message-ID: <15629.11548.588323.859686@12-248-41-177.client.attbi.com>

    Derek> Can anyone out there tell me how to check for a fault
    Derek> response from an xml-rpc server using the xmlrpclib
    Derek> python module?

Does something like this work?

    try:
        server.method(args)
    except xmlrpclib.Fault, f:
        print "code:", f.faultCode
        print "string:", f.faultString

If the faultCode isn't helpful, you can pick apart the faultString.

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html




From cliechti at gmx.net  Sat Jun 29 16:38:17 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 29 Jun 2002 22:38:17 +0200
Subject: Win32/Python serial port event character problem
References: <mailman.1025380000.22796.python-list@python.org>
Message-ID: <Xns923CE6BD96723cliechtigmxnet@62.2.16.82>

Derek Basch <dbasch at yahoo.com> wrote in 
news:mailman.1025380000.22796.python-list at python.org:

> Sorry to post this again but it didn't get any
> responses. Maybe better luck this time.

well 14hrs isn't that much time on a weekend... some people only look into 
news once per week.

> I use the following code to read data from the serial
> port when a certain event character ('\n') is
> received:
...
> Has anyone ever dealt with this? Is using the event
> character not a reliable method in windows?

no sorry, i haven't used the event mode.

....
> This is driving me batty so any help is greatly
> appreciated.

<shaemles plug>
my tip is: use http://pyserial.sf.net

and read the characters in a separate receiver thread. that thread can then 
trigger on any character or sequence (e.g. using a Queue to communicate 
with other threads)
</shaemles plug>


-- 
http://pyserial.sf.net
Chris <cliechti at gmx.net>



From maxm at mxm.dk  Tue Jun 25 17:45:39 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 25 Jun 2002 23:45:39 +0200
Subject: ASP only runs once
References: <3842735a.0206251112.25609e37@posting.google.com>
Message-ID: <3D18E483.3040808@mxm.dk>

Michael Anderson wrote:

> I've been using python in my asp pages for over a year now with no
> problems, and just ran into a problem I can't figure out. I installed
> it on a new machine and my pages only run the first time you hit them.
> After that they do nothing. Pages in VBScript and JavaScript run fine.


I am short of time so short answer.

It's a known problem with a workaround. Please search Google for answer.

regards Max M



From peter at engcorp.com  Tue Jun 25 08:15:39 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 25 Jun 2002 08:15:39 -0400
Subject: ntohl/htonl socket transfer between c and python
References: <ff4066da.0206242241.f614118@posting.google.com>
Message-ID: <3D185EEB.107547E@engcorp.com>

"Eivind B. Nilssen" wrote:
> 
> I have a bit of a problem I cant seem to solve. I am new at
> programming python,
> and my task is to make a client in python for fetching data from a
> remote server written in c.
> 
> The buffer sendt to me from the server is generated as follows:
> 
> *(int*)&bfr[bix] = htonl(*(int*)&f); #where f is a float value
> 
> The buffer received in my python program is a lot of strange asciis
> I am not able to cast this buffer to an int or a long and this again
> makes
> me unable to apply the ntohl command.
> 
> Does anybody have an idea how to extract the data from this buffer?
> Do I have to write this part in C?

No, you don't have to write it in C.  But please, always, post
code snippets from your own code showing the problem.  If possible,
use the interactive console to execute a few of your methods and
show the resulting data, or add some print statements to the code
to make it clear what's happening.

Without code to look at, we're going to waste a lot of time 
providing answers "shotgun" fashion, only some of which will hit the 
mark...

-Peter


From feldspar at ix.netcom.com  Sun Jun 30 18:37:55 2002
From: feldspar at ix.netcom.com (Antaeus Feldspar)
Date: Sun, 30 Jun 2002 19:37:55 -0300
Subject: Python needs to explain parsing? was Re: Python needs better error reporting
References: <mailman.1025315859.18149.python-list@python.org>
Message-ID: <3D1F8843.8080602@ix.netcom.com>

After reading the thread spun out of all of this, this brings up an idea 
that I've had for quite some time:  can the Python interpreter be made 
to report more information about *how* it is parsing the syntax it finds?

For instance, given the valid line "if s == ' ':" it might return 
something like the following (or at least this might be what is 
displayed to the user based on information the parser displays):

		if <condition>: ---- [next 8 lines of code]
			|
			==
		       / \
		      /   \
		     s    ' '

thus confirming that what was meant to be a condition is being parsed as 
a condition.  Clearly if the user sees that the parser is putting the 
block of code to be conditionally executed in where the condition should 
be, that's a very visual sign of trouble...

It's not quite the solution David seems to be looking for, but I think 
it would address the same problem; is it possible?

David LeBlanc wrote:

> "Syntax Error: invalid syntax" isn't very informative, yet it's tossed out
> frequently. HOW is the syntax invalid? For example:
> if s == ' '
> 
> David LeBlanc
> Seattle, WA USA
> 
> 
> 
> 



From r_proietti at yahoo.com  Thu Jun 20 11:17:15 2002
From: r_proietti at yahoo.com (NOPollution)
Date: Thu, 20 Jun 2002 15:17:15 GMT
Subject: module string.py
References: <3d11bd6a.11177520@news.polito.it> <mailman.1024576767.22581.python-list@python.org>
Message-ID: <3d11f18b.24524431@news.polito.it>

On Thu, 20 Jun 2002 14:15:06 +0200, Gerhard =?iso-8859-15?Q?H=E4ring?=
<gerhard.haering at gmx.de> wrote:

>Hi "Nop",
>
>* NOPollution <r_proietti at yahoo.com> [2002-06-20 11:38 +0000]:
>> I have two trouble:
>> 
>> 1)  looking at the module "string.py" I have seen that are described
>> some functions like "lower", "rjust", etc. Ther problem is that it
>> seems not to describe exactly how this functions are made, but there
>> is only a comment about their properties. So in which file are these
>> functions described?
>
>The behaviour of these functions is described in the documentation, for
>example in http://www.python.org/doc/current/lib/module-string.html:
>
>lower(s)
>Return a copy of s, but with upper case letters converted to lower case. 
>
>In this particular case, the docstring of the function has just as much
>info:
>
>>>> string.lower.__doc__
>'lower(s) -> string\n\n    Return a copy of the string s converted to
>lowercase.\n\n    '
>
>If you're looking at additional info for _how_ exactactly these
>functions are implemented, then the source code is the only info there
>is.

Where could I find the source code of the functions implemented in
"string.py"?


> Btw. in current Python versions the string module just delegetages
>to the string methods, which are the preferred option. So you could just
>as well use.
>
>> 2) I tried to cancel the file "string.py". After that, the Python
>> Interpreter doesn't work anymore (it can't even start working). Why?
>
>You're not supposed to delete modules from the standard library.  If you
>don't want your Python to fail in obscure ways, you'd better get that
>file back into place. Btw.  my Python 2.1.3 and 2.2.1 starts up fine
>even without the string module.
>
>Gerhard
>-- 
>mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
>web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
>public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
>reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))
>
>



From mgerrans at mindspring.com  Fri Jun  7 05:08:38 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Fri, 7 Jun 2002 02:08:38 -0700
Subject: How do you read unicode files?
References: <adpd7h$6sv$1@slb3.atl.mindspring.net> <m3vg8vaa9u.fsf@mira.informatik.hu-berlin.de>
Message-ID: <adqpph$hoi$1@slb0.atl.mindspring.net>

Thanks Mike, Francis and Martin for the quick replies!

My mistake was using readlines() to read the text file, instead of read().
(I should have posted a snippet of code in the first place and probably you
all would have noticed that immediately).

- Matt




From marklists at mceahern.com  Wed Jun 12 17:19:19 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 12 Jun 2002 16:19:19 -0500
Subject: get a list of classes in this module
In-Reply-To: <NfON8.146292$4i.13334503@bin2.nnrp.aus1.giganews.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBCEFBLEAA.marklists@mceahern.com>

[Terry Reedy]
> I believe the name of a modulue is the same inside as out.  I do not
> know inspect or whether it can work on a module under construction.
> However, if the subclasses are too numerous to build the dict by hand,
> or the set so dynamic that you do not trust yourself to update it
> properly...

Yeah, it's not so much that I don't trust myself--there's that.  ;-)  But
it's also the simple DRY principle at work.  Why should I have to list the
classes in that factory-like dictionary when Python can do that for me?

Thankfully, I didn't have to resort to globals()--but took advantage of
Python 2.2's __subclasses__ (which requires subclassing the base class from
object).  That worked smoov.

Thanks,

// mark

-





From jiri at baum.com.au  Sat Jun  1 06:35:07 2002
From: jiri at baum.com.au (Jiri Baum)
Date: Sat, 1 Jun 2002 20:35:07 +1000
Subject: Python and HTML frames
References: <ada6hf$vico3$1@ID-49758.news.dfncis.de>
Message-ID: <r08ada.b7b.ln@10.0.0.1>

Ken:
> How do I output HTML onto frames? This code below doesn't work. Can
> someone help me?

This isn't really a python question, it's a HTML question (or possibly
HTTP; are you ouputting the right headers?). However, I should note that
setting scrolling to "no" is a really, really bad idea unless the frame
contains just an image, because not all people have the same-sized fonts.

I trust you've read the guidelines for using frames at www.useit.com ?
Briefly: don't, and if you have to, make all links TARGET="_top".

> ==
> def mainpagelayout():
>        print """
>        <head><title>Languages Dictionary</title></head>

>        <frameset rows="70,40,*" scroll=no frameborder="NO" border="0">
>        <frame src="title.html" name=Title scrolling=no noresize>
>        <frame src="topmenu2.html" name=TopMenu scrolling=no noresize>

>        <frameset cols="*, 200" frameborder="NO" border="0">
>        <frame src="main2.html" name=Main scrolling=auto>
>        <frame src="sidemenu2.html" name=SideMenu scrolling=auto>
>        </frameset> </frameset>

You're missing a <noframes> here, and your <body> tags are in the wrong
place, but in any case this message is so useless that it beggars belief.
The basic point of the web is to communicate: to get information from one
person to another. Design of web pages should be guided by this goal. If it
isn't, then it becomes an impediment and the whole exercise pointless.

In any case, anyone who is shown this image will go away in disgust. You
might as well not have it. If your page really, really needs frames (and I
doubt it), at least apologise here and explain why.

>        <body>This page doesn't work without a frames-enabled
> browser.</body>
>        </noframes></body></html>
>        """


Jiri
-- 
Jiri Baum <jiri at baum.com.au>           http://www.csse.monash.edu.au/~jirib
  MAT LinuxPLC project --- http://mat.sf.net --- Machine Automation Tools


From logiplexsoftware at earthlink.net  Wed Jun 19 13:01:36 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Wed, 19 Jun 2002 10:01:36 -0700
Subject: python version?
In-Reply-To: <NCBBLFCOHHDIKCAFGCFBOEGILFAA.marklists@mceahern.com>
References: <CT%P8.33113$uk2.13710586@twister.nyroc.rr.com>
	<NCBBLFCOHHDIKCAFGCFBOEGILFAA.marklists@mceahern.com>
Message-ID: <20020619100136.27e40780.logiplexsoftware@earthlink.net>

On Wed, 19 Jun 2002 09:20:44 -0500
Mark McEahern wrote:

> Instead, you've probably got people wondering whether we will yet
> Rue the day.  ;-)

You realize that if Rue's name is mentioned one more time, he'll reappear in a
puff of autocoding to berate us for preventing him from writing software.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From gerhard at bigfoot.de  Fri Jun 28 08:10:21 2002
From: gerhard at bigfoot.de (Gerhard Haering)
Date: 28 Jun 2002 12:10:21 GMT
Subject: GUI-Control for HTML
References: <3D1C4B40.6090803@thomas-guettler.de>
Message-ID: <slrnahok93.js.gerhard@lilith.my-fqdn.de>

In article <3D1C4B40.6090803 at thomas-guettler.de>, Thomas Guettler wrote:
> Hi!
> 
> Is there are GUI-control which can display HTML?
> 
> The HTML that will be used will be simple, no javascript or
> form support is needed.
> 
> Is there are cross-plattform solution? (Unix + Win32)

Yes, wxPython is available on these platforms, and has a HTML component. There
might be a HTML viewer add-on for Tkinter, too.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From phlip_cpp at yahoo.com  Mon Jun 17 22:26:55 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 18 Jun 2002 02:26:55 GMT
Subject: ANN: PyUnitTestBrowser 006
Message-ID: <aem5pf$6ik@dispatch.concentric.net>

<posted & mailed>

PyUnit Enthusiasts:

The latest feature is integrated grep. This means if your editor does not 
have multi-file search, we do.

        http://flea.sourceforge.net/BrowseMe.html
        http://flea.sourceforge.net/browser006.zip
        http://www.c2.com/cgi/wiki?PyUnitTestBrowser

PyUnitTestBrowser is a GUI TestRunner for PyUnit that integrates, 
round-trip, with Kate, Idle or Vim.

The ChangeLog:

        005 bonds with PyChecker
        004 arbitrarily complex command lines
        003 does not require test suites
        002 does Windows
        001 exists

-- 
  Phlip
          http://www.greencheese.org/ParodyMode
  --  Appears that VIM is internationalized...
      May I ask what's the point?               --



From kragen at pobox.com  Sat Jun 15 14:59:28 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 15 Jun 2002 14:59:28 -0400
Subject: Args not set in extension class
References: <aefcav$8ef$1@reader05.wxs.nl> <aefs7n$4cc$1@reader09.wxs.nl>
Message-ID: <83sn3otlkf.fsf@panacea.canonical.org>

"Gorny" <gorny0 at zonnet.nl> writes:
> I've already dealt with the problem... Just create a function like:
> 
> static PyObject *
> wrap_pkt_object(PyObject *self, PyObject *args)
> {
>         PyObject * return_object;
>         return_object = new_pkt_object(args);
>         return return_object;
> }
> 
> But it would be very nice if someone can explain to me, why this
> is happening and it isn't possible to *directly* call new_pkt_object even
> if you prototype the latter function with two arguments like
> wrap_pkt_object!

If you prototype new_pkt_object with two args, make sure you're using
the second one.


From eager at eagercon.com  Mon Jun 10 11:22:57 2002
From: eager at eagercon.com (Michael Eager)
Date: Mon, 10 Jun 2002 08:22:57 -0700
Subject: import problem
References: <2se7guc3kkmm3bgooj776j55igpkmgpst4@4ax.com> <gaQM8.1178$g65.19246@news4.ulv.nextra.no>
Message-ID: <usg9gu4dh8iq2v7l7h8g0laop65m4kcrkr@4ax.com>

Thanks.  It was trying to open libsip.so, not looking for libsip.so.*.
Adding the link makes this problem go away (and another show up).

On Sun, 09 Jun 2002 21:56:28 GMT, "Erlend J. Leiknes"
<nookieNO_SP_AM at online.no> wrote:

>if you have the debugging application strace (its most likely you have it)
>
>write a small application who only does: from qt import *
>
>save it, run it like this: strace python your_app.py 1> log1.txt 2>log2.txt
>wait for it to finnish, and edit log2 (or log1).txt, scroll to the bottom,
>and look for io signales that tries to open libsub.so.*
>now you should find out what your filename should be.
>
>"Michael Eager" <eager at eagercon.com> wrote in message
>news:2se7guc3kkmm3bgooj776j55igpkmgpst4 at 4ax.com...
>> I'm trying to use Python 2.2 and PyQt on a RedHat-7.2 system.
>>
>> I get an error when I execute "from qt import *" -- there is a failure
>> in qt.py when it executes "import libsup".  The error is "No module
>> named libsup".
>>
>> libsup.so.9 is a link to libsup.so.9.0.3 and is in
>> /usr/lib/python2.2/site-packages.  If I enter "import libsup", I get
>> the same error message.
>>
>> Any suggestions?
>>
>> -- Mike Eager, eager at eagercon.com
>



From inkedmn at earthlink.net  Fri Jun  7 22:52:28 2002
From: inkedmn at earthlink.net (inkedmn at earthlink.net)
Date: Sat, 08 Jun 2002 02:52:28 GMT
Subject: Working with directory
References: <20020526.145248.1350573793.20446@iprimus.com.au> <83lma3udjf.fsf@panacea.canonical.org>
Message-ID: <3d016fbf.529655106@news.earthlink.net>

On 28 May 2002 18:13:08 -0400, Kragen Sitaker <kragen at pobox.com>
wrote:

>"Occean" <tritran2001 at iprimus.com.au> writes:
>> I have to write the function that read in to the existing directory and
>> display some files according to their date for example, in my PyExample
>> directory i got 
>> 
>> practice1.py    Thu Mar 16 11:54:21 2002
>> practice2.py    Thu Mar 17 10:25:22 2002
>> ..
>> practice7.py     Friday Apr 28 1:20:22 2002
>> 
>> i want to  just get 2 files from that directory which are from Mar or
>> specific time only. How can i do that, and which built in function allow
>> me to view all the file in directory with time. By looking the reference
>> os.time.stat but i can't work out the solution for this problem.
>
>os.listdir(dirname) lists the directory.
>
>os.stat(pathname)[stat.ST_MTIME] is the modification time of the file,
>as a number of seconds since 1969 GMT.
>
>time.localtime(nsecs) converts a number of seconds since 1969 GMT into
>a tuple giving year, month, day, etc.
>
>os.path.join(dirname, filename) will be useful in combining the stuff
>you get back from os.listdir into full pathnames you can pass to
>os.stat.
>
>So you want a loop something like this:
>import os, time, stat
>dirname = '.'
>for filename in os.listdir(dirname):
>    date = time.localtime(os.stat(os.path.join(dirname, filename))[stat.ST_MTIME])
>    if date[1] == 3:  # element 1 is month
>        print filename, time.asctime(date)
>
>HTH.
>


hello...

formatting the date/time is pretty easy if you use the asctime method
in the time module.

like this:
>>> import time
>>> x = time.asctime()
>>> print x
Fri Jun 07 19:46:18 2002

one of my personal favorites :)

hope this helps

ink




From pixie888 at hotmail.com  Thu Jun 20 03:11:01 2002
From: pixie888 at hotmail.com (pixie888 at hotmail.com)
Date: Thu, 20 Jun 2002 07:11:01 GMT
Subject: SafeArray problem
References: <3d105710.83643750@news.skynet.be> <3d108014$1@news.mhogaming.com> <3d108545.95472593@news.skynet.be> <3d10a3fa$1@news.mhogaming.com>
Message-ID: <3d117f5f.159498312@news.skynet.be>

On Wed, 19 Jun 2002 11:28:36 -0400, "Michael Weiss"
<FooWeissBarMike at hotmail.com> wrote:

>
>> I have not written an idl-file myself, I just used the classwizard in
>> Visual Studio to write my method on the automation object (using MFC).
>> What do you really want to know? (I looked for idl-files in my project
>> but did not find any)
>
>How about an ODL file?

not found too, I guess that when you use MFC in combination with the
Classwizard things are handled hidden for the programmer. Nonetheless:
I give you the following information for the better of it:

im my .h file

afx_msg LPDISPATCH GetRelationQuerierEx(const VARIANT FAR& vaIDs,long
lDummy);

the first argument is the safearray, the second the long

in my .cpp file

DISP_PROPERTY_PARAM(CDispCore, "RelationQuerierEx",
GetRelationQuerierEx, SetNotSupported, VT_DISPATCH, VTS_VARIANT
VTS_I4)

>
>I was just wondering if the in, in/out, retval, pointer vs. non-pointer
>symantics of the parameter were wrong. I was guessing by adding a new
>parameter (the long) that the safearray's properties were changed for the
>better. But I guess with a pure Dispatch interface (dispinterface) that this
>kind of stuff isn't done, or at least its done differently.
>
>Still, I wonder if this still isn't the problem of the function not being
>declared correctly before you made the change and then when you added that
>long the declaration of the safearray was changed.
>
>
>
>
>



From plugin at supercable.es  Mon Jun 24 11:38:46 2002
From: plugin at supercable.es (Ugo García)
Date: Mon, 24 Jun 2002 17:38:46 +0200
Subject: Redirecting StdOut
References: <1024856420.145085@seux119> <Xns9236D1FF016AAcliechtigmxnet@62.2.16.82> <1024860416.580309@seux119> <3D163550.E525D840@engcorp.com> <1024873013.652528@seux119> <3D1678F3.AD9B059B@engcorp.com>
Message-ID: <1024933181.665516@seux119>

> You'll doubtless get other replies on this, including pointers to PyGame.
> What makes you think a game implemented in Python will not have
> enough speed?  You certainly would not be writing the low level
> graphics routines in Python, but then you aren't about to do that
> in C, either (I sincerely hope).  Python can call on libraries
> which have already been written *in C* (usually) to do the grunt
> work, so it is quite fast enough.
    I've seen PyGame but It still doesn't like. The idea of making a game in
a interpreted language doesn't like very much. (Is like making a game in
Visual Basic). Perhaps I'm used to make games with C (and of course, asm)
or, perhaps this is the right way to do it. Do u know many games that are
made with an interpreted languaged? I don't. (I'm talking about GAMES, not
little demos). Well, I known one: BLADE. It use Python to define the world,
the behavour of the characters, and other things like the save games... If u
have test it u could agree that it's extremely sloowly when loading (don't
know if it's because it use Python or because the coders as Spanish, like
me.... :-) ).

> As for reliable, I have to most vigorously object to anyone
> characterizing C as "reliable".  C code is notoriously difficult
> to make reliable, generally being strewn with pointer bugs,
> memory allocation problems, and so forth.  C was really written
> only as a somewhat more human-readable Assembly language substitute
> and has been supplanted by many more suitable alternatives for
> something as sophisticated as the engine of a game.
(ufff. very difficult to understand by my poor English... Let see...) I
agree that C is more complicated than other languages; more complicated to
code and more complicated to debug. But... What engines do u know are
written in another language instead of C? (Perhaps we're are naming 'engine'
to different thing)

> Or to put it another way: don't knock it till you've tried it!
> Python is quite possibly a far more suitable solution for
> the entire application, not just for the scripting language.
I agree again. I've to try other alternatives before deciding.... but I have
already try another.... and I'm still see Python as a script language for a
engine, not for the whole engine (it depend, of course, in the kind of the
engine).

Don't know, perhaps most people who write engines are wrong (i think they're
written in C), or perhaps I'm wrong. This is the goal of the forums... to
get the right way!!! :-)

Thanks,
--ugo





From dmitri.gouliaev at telkel.net  Thu Jun  6 20:18:27 2002
From: dmitri.gouliaev at telkel.net (Dmitri I GOULIAEV)
Date: Thu, 6 Jun 2002 19:18:27 -0500
Subject: urllib, urllib2, httplib -- Begging for consolidation?
In-Reply-To: <Pine.SOL.4.30.0206062242070.21937-100000@mimosa.csv.warwick.ac.uk>; from "John J. Lee" on Thu, Jun 06, 2002 at 10:55:07PM
References: <mailman.1023295507.16177.python-list@python.org> <Pine.SOL.4.30.0206062242070.21937-100000@mimosa.csv.warwick.ac.uk>
Message-ID: <20020606191826.A17584@lifebook>

Hi, John J. Lee !


On Thu, Jun 06, 2002 at 10:55:07PM +0100, John J. Lee wrote:
> On Wed, 5 Jun 2002 brueckd at tbye.com wrote:

> > Problem #1 is what makes me throw my hands up in frustration when we talk
> > about, e.g., expanding the urllib APIs to have some way to do a HEAD
> > request: it doesn't belong on that level of functionality.

> I think you mean "it doesn't belong in a module named 'urllib2'"... or at
> least you should mean that ;)

In my opinion, he means that.
And, by the way, "it doesn't belong in a module named 'urllib2'" because I would assume, that module named urllib* works with URLs. Am I wrong ?

[...]

> > That generic
> > interface is for making it trivially easy to fetch the contents associated
> > with a URL, independent as much as possible from protocol.

> Only part of urllib2 (OpenerDirector &c) is a generic any-old-url-scheme
> API.  The rest (eg. AbstractHTTPHandler, HTTPRedirectHandler) is
> HTTP-specific, and can be used on its own, without going through the
> generic stuff.

What does HTTP-specific stuff inside urllib-module ? Why is it there ? Why not in httplib ? 


Best regards,

-- 
DIG (Dmitri I GOULIAEV)
 
All below this line is added by my e-mail provider.




From ark at research.att.com  Fri Jun  7 13:28:48 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 7 Jun 2002 17:28:48 GMT
Subject: Python 2.2.1 vs. gcc 3.1?
References: <yu99vg8v6orb.fsf@europa.research.att.com>
Message-ID: <yu99n0u76lrj.fsf@europa.research.att.com>

Incidentally, there is something suspicious here:

     gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -fPIC -I. -I/tmp/build-gnu913/Python-2.2.1/./Include -I/usr/local/include -IInclude/ -c /tmp/build-gnu913/Python-2.2.1/Modules/structmodule.c -o build/temp.solaris-2.8-sun4u-2.2/structmodule.o
     creating build/lib.solaris-2.8-sun4u-2.2
     gcc -shared build/temp.solaris-2.8-sun4u-2.2/structmodule.o -L/usr/local/lib -o build/lib.solaris-2.8-sun4u-2.2/struct.so
     WARNING: removing "struct" since importing it failed

That -L/usr/local/lib does not look right, because /usr/local/lib does not
exist on my machine.

In particular, I began the whole process by saying

        configure --prefix=/usr/gnu

The /tmp/build-gnu913 directory is where I'm actually building everything,
so that's as it should be.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From spammers.do.not.bother at void.bogus  Sat Jun  1 14:38:37 2002
From: spammers.do.not.bother at void.bogus (Magnus)
Date: Sat, 01 Jun 2002 18:38:37 GMT
Subject: Copying a database
References: <FL6K8.771$NG1.217@newsread2.prod.itd.earthlink.net>
Message-ID: <Nw8K8.41248$n4.9229522@newsc.telia.net>

Billy Ng wrote:

> I need you guys help on mysql again.  I need to push my python program to
> production now, but I don't want to recreate the mysql database on the
> production box.  I want to copy the database from the development box to
> production box.. Would anybody please tell me how to do it, thanks!
> 
> Billy Ng


Maybe this is what you are looking after:

http://www.mysql.org/documentation/mysql/bychapter/manual_MySQL_Database_Administration.html#mysqldump
http://www.mysql.org/documentation/mysql/bychapter/manual_MySQL_Database_Administration.html#mysqlhotcopy
http://www.mysql.org/documentation/mysql/bychapter/manual_MySQL_Database_Administration.html#mysqlimport

/Magnus



From martin.oberhuber at windriver.com  Thu Jun 20 07:58:47 2002
From: martin.oberhuber at windriver.com (Martin Oberhuber)
Date: Thu, 20 Jun 2002 13:58:47 +0200
Subject: Java to Python converter?
Message-ID: <YrjQ8.34$g32.995@nreader1.kpnqwest.net>

Hi all,

I've got some Java source code that I would like to convert into Python.
Does anybody know of a converter that can do this?

Thanks,
Martin




From trentm at ActiveState.com  Wed Jun 26 13:55:17 2002
From: trentm at ActiveState.com (Trent Mick)
Date: Wed, 26 Jun 2002 10:55:17 -0700
Subject: question od default args
In-Reply-To: <XFMail.20020626104856.shalehperry@attbi.com>; from shalehperry@attbi.com on Wed, Jun 26, 2002 at 10:48:56AM -0700
References: <o3vjhuguen6gfm8vnq7rbk5volmi80shl5@4ax.com> <XFMail.20020626104856.shalehperry@attbi.com>
Message-ID: <20020626105517.B28810@ActiveState.com>

On 26-Jun-2002 Gon?alo Rodrigues wrote:
> Hi,
> 
> When I want default args I usually do
> 
> def (arg = None):
>     ...
> 
> But what if I want to make None a reasonable argument too? That is, I
> want to know if the user has in fact passed an argument, and if not to
> do something special, with None (or whatever object) being a reasonable
> arg to be passed.


You could try something like this:


>>> class _NoArgSpecified:
...     pass
...
>>> def foo(arg=_NoArgSpecified):
...     if arg is _NoArgSpecified:
...         print "no arg specified"
...     else:
...         print "arg:", repr(arg)
...
>>> foo()
no arg specified
>>> foo(None)
arg: None
>>> foo(1)
arg: 1


Mind you, the user *can* still cheat, but you shouldn't have to worry
about that.

>>> foo(_NoArgSpecified)
no arg specified
>>>


Trent


-- 
Trent Mick
TrentM at ActiveState.com




From dyoo at hkn.eecs.berkeley.edu  Fri Jun 28 17:02:17 2002
From: dyoo at hkn.eecs.berkeley.edu (Daniel Yoo)
Date: Fri, 28 Jun 2002 21:02:17 +0000 (UTC)
Subject: get word base
References: <m24rfn88jp.fsf@mother.paradise.lost>
Message-ID: <afiisp$1m0q$1@agate.berkeley.edu>

John Hunter <jdhunter at nitace.bsd.uchicago.edu> wrote:

: I would like to be able to get the root/base of a word by stripping
: off plurals, gerund endings, participle endings etc...

Hi John,


PyWordNet has a function in it called wntools.morphy() which tries to
take the morphological root of a word.  For example:

###
>>> import wntools
>>> words = ['hello', 'taxes', 'thoughts', 'walked', 'rakes']
>>> for w in words: print wntools.morphy(w)
... 
hello
tax
thought
None
rake
###

morphy() doesn't quite work out of the box, but just because it
doesn't guess the part of speech --- morphy() assumes that we want the
noun root by default.  But then, we can do something like this:

###
>>> def root(w):
...     return wntools.morphy(w, "noun") or wntools.morphy(w, "verb")
... 
>>> root("walked")
'walk'
>>> root("sent")
'send'
>>> root("woke")
'wake'
###

So PyWordNet is a really useful tool if you're doing this sort of
stuff.  PyWordNet and Wordnet can be found here:

    http://pywordnet.sourceforge.net/
    http://www.cogsci.princeton.edu/~wn/

Best of wishes to you!


From garrett at bgb.cc  Thu Jun  6 16:31:58 2002
From: garrett at bgb.cc (Don Garrett)
Date: Thu, 06 Jun 2002 20:31:58 GMT
Subject: Using AT on XP to run python script
References: <mchqfu448cnq5q8toh4nq7n1sura1ug72g@4ax.com>
Message-ID: <3CFFC694.9000501@bgb.cc>

    Change "c:\python21\python.exe c:\scripts\script.py" to
"c:\python21\python.exe" "c:\scripts\script.py" otherwise it's trying to turn
that whole thing into a single file name. However, I would have expected the
last thing you tried to work.

    Try using /interactive just for testing. It may be trying to open a command
window or something. Maybe name the script .pyw

Chris Stromberger wrote:

 > I'm having trouble getting AT on XP to run a python script for me.
 > Have tried:
 >
 > at xx:xx "c:\python21\python.exe c:\scripts\script.py"
 > at xx:xx "start c:\scripts\script.py"
 > at xx:xx "c:\scripts\script.py"
 >
 > and get "the system cannot find the file specified" for the first two.
 > The last version does nothing apparently--no error indication or
 > anything.
 >
 > I imagine it is related to user accounts and such.  Anyone have any
 > pointers?
 >




From peter at engcorp.com  Mon Jun 17 20:45:02 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 20:45:02 -0400
Subject: Python 2.2.1
References: <II7O8.622$e27.275217555@twister1.starband.net>
Message-ID: <3D0E828D.B930932A@engcorp.com>

> Dave wrote:
> 
> Hi, I am 12 years old and would like to begin learning Python 
> over the summer. 

Hi Dave!  I sure wish I'd discovered Python when I was
twelve... of course, it didn't exist yet. :-(

Other replies have already covered the core question, so I'll
add two off-topic comments.

1. Please try to turn off "HTML" mode when posting messages here.
The messages are mirrored to the ASCII-based Usenet newsgroup 
comp.lang.python (in case you're using the mailing list) and 
non-ASCII messages are a problem for many readers.  In Outlook you 
can do this under the menu Tools -> Options -> Mail Format.

2. Please reply to let people know you've succeeded in following
their advice.  The payback for all the folks who spend their
time replying to questions comes when they know their efforts
paid off for the original poster!

Good luck, :-)

-Peter


From aahz at pythoncraft.com  Wed Jun 26 12:57:12 2002
From: aahz at pythoncraft.com (Aahz)
Date: 26 Jun 2002 12:57:12 -0400
Subject: Python 2.2 __slots__ Bug?
References: <mailman.1025089474.17780.python-list@python.org>
Message-ID: <afcrp8$c64$1@panix1.panix.com>

In article <mailman.1025089474.17780.python-list at python.org>,
Glyph Lefkowitz  <glyph at twistedmatrix.com> wrote:
>
>I get "I'm fine" and then a segmentation fault.  This only happens when
>__slots__ is defined on a new-style class.  Seriously, i'm not making
>those numbers up -- reliably, 43551 objects will crash the interpreter;
>fewer will not.

Interesting.  Using Python 2.2.1 on Mandrake 8.1, I get a segfault
without printing "I'm fine".  Go ahead and file a bug.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From Norman_Shelley-RRDN60 at email.sps.mot.com  Thu Jun 20 12:32:34 2002
From: Norman_Shelley-RRDN60 at email.sps.mot.com (Norman Shelley)
Date: Thu, 20 Jun 2002 09:32:34 -0700
Subject: Useful RE patterns (was: Variable Interpolation - status of PEP 215)
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <4PfQ8.44245$n4.10307683@newsc.telia.net> <yciQ8.44268$n4.10312169@newsc.telia.net>
Message-ID: <3D1203A2.B2A3AD9@email.sps.mot.com>

Fredrik Lundh wrote:

> ...
> If I were to add a dozen (or so) patterns to the (S)RE module,
> what should I pick?  What patterns do you find yourself using
> over and over again?

All kinds of numerics, e.g. scientific (1e-6, 2e6, ...) and engineering (1u,
2M and/or 2MEG, ...) notation.

Python identifiers as previously mentioned.

>
>
> </F>



From steve.bond at btinternet.com  Wed Jun 19 11:11:22 2002
From: steve.bond at btinternet.com (Steve)
Date: Wed, 19 Jun 2002 15:11:22 +0000 (UTC)
Subject: newbie: help please!
References: <aepltq$gl3$1@knossos.btinternet.com> <mailman.1024490396.5209.python-list@python.org>
Message-ID: <aeq6uq$g2h$1@knossos.btinternet.com>

Cheers Chris it was exactly as you said i'm so happy now! i just installed 
the two RPM's X11R6-contrib & XFree86-devel and updated my setup file and 
it worked!, thanks again

Steve


From boud at valdyas.org  Tue Jun  4 01:12:10 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Tue, 04 Jun 2002 07:12:10 +0200
Subject: different gui-toolkits pros/cons...
References: <adfsoo$dmik4$1@hades.rz.uni-sb.de> <_rLK8.12$fd2.979@news.ecrc.de> <adg4e0$dq6mb$3@hades.rz.uni-sb.de> <3cfbc67b$0$52024$e4fe514c@dreader3.news.xs4all.nl> <adgj1n$11edg$1@hades.rz.uni-sb.de>
Message-ID: <3cfc4c59$0$52032$e4fe514c@dreader3.news.xs4all.nl>

Uwe Schmitt wrote:

> Boudewijn Rempt <boud at valdyas.org> wrote:
> | Uwe Schmitt wrote:
> 
> |> Markus Jais <mjais at web.de> wrote:
> |> | with pyQT there might be problems with the license
> |> | if you want to make your programm freely available
> |> | (but I am not a lawyer, so I might be wrong)
> |> 
> |> That's a very important point: what do I have to pay if I used pyQt
> |> for development ??? I did not find any explicit prices at the
> |> trolltec-page...
> 
> | Trolltech doesn't sell PyQt: you need to go to http://www.thekompany.com
> | for BlackAdder.
> 
> O.K. But what if I only want pyQt ??? I suppose I have to pay Qt...
> 

Buying Qt for Windows and then compiling your own PyQt from source is
more expensive than buying BlackAdder, but it can be done -- order the
Qt from TrollTech, download your own PyQt, and make sure you've got
a supported C++ compiler.

-- 
Boudewijn Rempt | http://www.valdyas.org


From bobx at linuxmail.org  Thu Jun 13 12:33:45 2002
From: bobx at linuxmail.org (Bob X)
Date: Thu, 13 Jun 2002 16:33:45 GMT
Subject: Wanted:  Potential Python Authors
References: <ae7gke$mhb$1@ccis12.baylor.edu>
Message-ID: <3D08C96B.1020509@linuxmail.org>

Sheesh...and all he wanted was some potential authors and we debate it 
to death!  : )

Jeff Donahoo wrote:
> I'm looking for somebody to author a short book on Python in my programming
> technology series.  Who do you think would make a great author that hasn't
> already written a book on Python?  Candidates include:
> 
> - Competent and frequent posters to this newsgroup
> - Author of articles in the various trade magazines
> - Expert programmer
> 
> 



From ken at hotmail.com  Sat Jun  1 02:43:01 2002
From: ken at hotmail.com (Ken)
Date: Sat, 1 Jun 2002 16:43:01 +1000
Subject: How to open a HTML file when the python cgi program is executing?
Message-ID: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de>

How do I open a HTML file when the cgi program is executing?

The only method I can think of, which doesn't work is this:
<meta HTTP-EQUIV="REFRESH" CONTENT="0; URL=main.html">






From garrett at bgb.cc  Mon Jun 10 12:12:10 2002
From: garrett at bgb.cc (Don Garrett)
Date: Mon, 10 Jun 2002 16:12:10 GMT
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org> <ufk3bdnh675sc1@corp.supernews.com>
Message-ID: <3D04CF8D.4050407@bgb.cc>

RPM1 wrote:
> "Andrei Kulakov" <ak at silmarill.org> wrote
> 
>>Hello,
>>
>>    "hello world" type program from wxPython tutorial takes 6
>>seconds to start on my cel 366 128mb system (running Debian). Is
>>this typical? Why is it so slow? I'm looking for a graphic
>>toolkit that'd be fairly quick. I don't want to use Tk because of
>>the looks.

   This whole thread is somewhat strange.... one of the things I liked about 
wxWindows was that it seemed like one of the snappier cross plateform 
libraries I've worked with. I'm on a relatively fast machine (1 Ghz with 512M) 
with Windows (which has been mentioned here as having a faster start).

   The first start up has touch of delay (less than most commercial apps), but 
after that it's instant.

   I guess it's hard to really remember just how much faster computers have 
gotten over time.

-- 
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc



From kragen at pobox.com  Thu Jun 13 17:21:06 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 13 Jun 2002 17:21:06 -0400
Subject: M2Crypto: select() behaves weird on SSL sockets
References: <3D04ED17.5070606@NOSPAMREMOVETHISxs4all.nl> <ae2str$b0q$1@nntp6.u.washington.edu> <3D066E77.4070901@NOSPAMREMOVETHISxs4all.nl> <ae6bmb$frc$1@dahlia.singnet.com.sg> <3D0790E1.9080904@NOSPAMREMOVETHISxs4all.nl>
Message-ID: <831ybavprx.fsf@panacea.canonical.org>

Irmen de Jong <irmen at NOSPAMREMOVETHISxs4all.nl> writes:
> Ng Pheng Siong wrote:
> Well, I would appreciate a wrapper around select.select, one that
> acts on SSL sockets the way I expected (from the man page of select).
> That means: a select() wrapper that
> 1) works on regular files and sockets
> 2) works on SSL sockets
> 3) in the case of SSL sockets, honors the pending bytes that might still
>     be available after a recv() (or send): don't block but return the
>     socket in the 'read' or 'write' list if bytes are pending...
> 
> This would greatly simplify my socket reading/writing functions...

Irmen, why aren't you using asyncore?

For those of us who use asyncore, an asyncore.dispatcher-compatible
wrapper around an M2Crypto socket that you could hand to asyncore and
hang handlers on would probably be sufficient.



From bokr at oz.net  Mon Jun 17 22:58:38 2002
From: bokr at oz.net (Bengt Richter)
Date: 18 Jun 2002 02:58:38 GMT
Subject: Divorce property objects from attribute access dependency?
References: <mailman.1024336455.22417.python-list@python.org>
Message-ID: <aem7ku$td8$0@216.39.172.122>

On Mon, 17 Jun 2002 12:49:06 -0500, Gustavo Cordova <gcordova at hebmex.com> wrote:

>> 
>>  >>> class C(object):
>>  ...     def __init__(self,v):
>>  ...         self._v = v
>>  ...     def get_v(self): return self._v
>>  ...     def set_v(self,v): self._v = v
>>  ...     v=property(get_v,set_v)
>>  ...     vlist = [v,v]
>>  ...
>>  >>> c=C('something')
>>  >>> c.v
>>  'something'
>>  >>> c.v='other'
>>  >>> c.v
>>  'other'
>>  >>> c.vlist
>>  [<property object at 0x0082D4A0>, <property object at 0x0082D4A0>]
>>  >>> c.vlist[0]
>>  <property object at 0x0082D4A0>
>> 
>> Why not trigger get_v() in these accesses too, and have 
>> single mechanism to suppress it instead of a single
>> mechanism to effect it?
>>
>
>You mean that you don't recognize the difference between
>evaluating "c.v" and "c.vlist"?
>
No, I don't mean that ;-)
Also notice that I accessed c.vlist[0] to make a more direct analogue to c.v:
(BTW, disev below is just a handy thing to compile and disassemble an expression)

 >>> from miscutil import disev
 >>> c.v
 'something'
 >>> c.vlist[0]
 <property object at 0x00797370>
 >>> disev('c.v')
           0 SET_LINENO               0
           3 LOAD_NAME                0 (c)
           6 LOAD_ATTR                1 (v)
           9 RETURN_VALUE
 >>> disev('cvlist[0]')
           0 SET_LINENO               0
           3 LOAD_NAME                0 (cvlist)
           6 LOAD_CONST               0 (0)
           9 BINARY_SUBSCR
          10 RETURN_VALUE

The byte code itself is apparently not what makes the difference between property or
other attribute access. Apparently LOAD_ATTR checks to see if the attribute-referenced
object is a property or not (and I suspect as far as LOAD_ATTR is concerned, any class
attribute that dereferences to an instance of a class that has a __GET__ method might do).
What I'm saying is that (e.g.) BINARY_SUBSCR etc. could potentially also check on whether
what it was retrieving has a __GET__ method, if there were a different mechanism tying
a property to an object than via the attribute mechanism, so that object's "self" could
be passed to the __GET__ method of the property object.

>Well, for one:
>
>"c.v" is evaluated at runtime, and it's recognized as
>an instance's access to a class' property attribute,
>so all the background magic happens (which is what
>you expected). Just a note: get_v and set_v are
>methods, which take "self", and all that. Ok.
>
>On the other hand, "vlist" is created at compile
>time, and it's a list, not a property attribute;
>it's only a list with two objects. When you're
>evaluating "c.vlist", the items in the list are not
>in the correct context to be evaluated as property
>attributes, so the background magic does not happen.
IMO "correct context" refers to the current state fo affairs, not
a technical limitation. I am trying to suggest "divorcing"
properties from that "correct context" -- for the sake of discussion;-)

The reason vlist items are printed is that they *are* accessed. UIAM, first
as objects, and then normally the '__repr__' attribute is accessed to
get a method to convert the object to a string representation (or __STR__
if exlicit str() or '%s'% is involved).

However, in the case of a property-jiggered class attribute access,
something apparently discovers that the object is a property object,
and instead of looking for a __repr__ attribute immediately, attribute
access is coded to evaluate the property's __GET__ first (with the
attribute-search's associated object 'self' as argument), and *then*
with the result of that - not the property object any more -- proceed to look
for __repr__ to get the string to print. Note that it's a matter of
attribute access per se looking at an object, not looking for a particular
name. If we bind an alternate attribute name to C.v's property object
as well as using a list, the new name will trigger the property action:

 >>> class C(object):
 ...     def __init__(self,v):
 ...         self._v = v
 ...     def get_v(self): return self._v
 ...     def set_v(self,v): self._v = v
 ...     v=property(get_v,set_v)
 ...     vlist = [v,v]
 ...     a=v
 ...
 >>> c=C('something')
 >>> c.v
 'something'
 >>> c.a
 'something'
 >>> c.vlist[0]
 <property object at 0x007A9EC0>

>For one, there is no "self" to pass to the methods
There are two "selves" in the list. They would work fine
if those property selves had closure-style bindings to
the object they were a property of. Note that we can transfer
the "property" to another class and have its instances see it:

 >>> class X(object):
 ...     def __init__(self,v):
 ...         self._v = v
 ...
 >>> x=X('initial x')
 >>> x._v
 'initial x'
 >>> X.z = c.vlist[1]
 >>> x.z
 'initial x'

because the jiggered access passes the 'self' (as you point out)
of the instance involved in the attribute access.

>get_v and set_v; another, they're "property attributes",
>not "magical list items". :-)
No, they are not "property attributes," they are property
_objects_ in a list. That is my point. They don't _have_ to have anything
to do with getting accessed via a class attribute name. That's just
one conventional and handy use that provides an opportunity to pass
the associated self, which could as well be part of a property closure.

 >>> c.vlist
 [<property object at 0x00797370>, <property object at 0x00797370>]

Property object, not property attribute.
> 
>> ISTM kind of a hack to rig access via instance attribute as a 
>> mechanism for controlling referenced-object behavior (this is
>> my impression, I haven't checked code ;-). IMO default behavior
>> of an object in lhs or rhs context should not depend on how a
>> reference to it was acquired and dereferenced.
>> 
>> (Obviously, if property objects always acted as they now do 
>> through instance attribute access, "vlist=[v,v]" in this
>> example would have to have some kind of access override like
>> "vlist = [obj(v),obj(v)]", or else it would generate 
>> ['something','something'] right there. Using obj(p) a property
>> object could be passed around without evaluation and the
>> potential side effects thereof).
Sorry, that's not quite right. The vlist assignment would not find
a _v attribute from the get_v function, if the property was bound
to the class itself. I was/am thinking of the property
assignment happening in the __init__ function like
self.v=property(get_v,set_v), followed by self.vlist=[v,v].
Currently you can do that, but not with live property effect.

My generalized property factory function would have to be passed an object
to attach to, e.g., self.v=attach_property(self, self.get_v, self.set_v)
inside an __init__.

Doing it for class variables would also be valid, but of course
if you used the class itself as the attachee of the property,
you would have to have get/set methods that operated validly
in that environment.
>
>A property object *can* be passed around without side effects,
>it's only when accesed as an attribute of an instance of a
>class which defines them, that the expected behaviour
>triggers.
I know that. My post suggests a divorce, alternate ways of getting
hitched, and an alternate implementation of triggering, and
I'm talking about object-property realtionships ;-)

>
>For example, try this:
>
>>>> class klass(object): pass
>...
>>>> ob = klass()
>>>> ob._blah = 10
>>>> def _get_blah(self): return self._blah
>>>> def _set_blah(self, b): self._blah = b
>>>> ob.blah = property(_get_blah, _set_blah)
>
>What behaviour do you expect from this? Give it a try. :-)
I expect it to work as currently implemented. I am trying to
suggest an alternative implementation of property objects.

>
>> BTW, it would be nice to be able to define a property at 
>> global module scope also, and perhaps this would tie in nicely.
>> 
>> Just musing ;-) Thoughts? Am I missing a big gotcha?
>> One thing is that it will effectively make possible a ()-less 
>> function call, which will upset some people ;-)
>
>You can do that without properties:
I know interactive use triggers __repr__ and __str__ in various useful
ways. I don't want to (ab)use that. I'd like a generalized property object
that you can attach to any object and store anywhere. The normal use would
be to attach to class instances, and do that by storing as attributes of
the same instances during __init__. (Not as class attributes, note.
The class would be the normal place to define the get/set/del methods).

>
>>>> class Caller:
>...    def __init__(self, func, *args, **kw):
>...       self.func, self.args, self.kw = func, args, kw
>...    def __str__(self): return self.func(*args, **kw)
>...    __repr__ = __str__
>...
That depends on access via __str__ or __repr__, which is not
what I want. That's a worse context dependency than depending
on access via class attribute name.

>>>> def PrintTen(): print 10
>...
>>>> ten = Caller(PrintTen)
>>>> ten
>10
>
>Another thing, you *can* create "magical list items"
>with this class; since __repr__ is also defined,
>then when the list which contains them is rendered,
>the __repr__ method of the items is called. Neat. :-)
>
Neat for what it's neat for, but that's the kind of "magic"
I'm trying totalk about ;-) All those __repr__/__str__-based
ways are beside the point. I'm not looking for help with that,
thanks ;-)

>I Know, it's a simple-minded example, but you know
>what I mean. You can use this Caller class to execute
>a function when you type the name of the object that's
>related to the function.
>
>Good luck :-)
>
Thanks. I was hoping for comments on the idea.
I probably didn't make it sufficiently clear, but
maybe the above helps?

I just wanted to explore the idea of property objects
as related to what they function as properties of, potentially
associated by means other than being stored as attributes
of the instances' classes, and triggered by any access (except a
special one designed to inhibit it), not just class attribute access.

Regards,
Bengt Richter


From fredrik at pythonware.com  Thu Jun 20 06:39:38 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 20 Jun 2002 10:39:38 GMT
Subject: RE strings (was: Variable Interpolation - status of PEP 215)
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <4PfQ8.44245$n4.10307683@newsc.telia.net>
Message-ID: <KhiQ8.44269$n4.10312432@newsc.telia.net>

and while I'm at it:

> # match $$ and $var and ${var}
> _dosub = sre.compile(r'\$(?:\$|(\w+)|\{([^}]*)\})').sub

how about adding "re" strings (where the "e" tells Python
to run the string through the RE compiler):

    re"[-+]\d+"
    re"(\d+)\.(\d+)\.(\d+)\.(\d+)"
    _dosub = re'\$(?:\$|(\w+)|\{([^}]*)\})'.sub
    ishex = re'0[xX][\dA-Fa-f]'.match
    # etc

(or maybe a ".re" method that does the same thing, but
with an optional flag argument ?)

</F>






From emile at fenx.com  Tue Jun 25 10:44:58 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 25 Jun 2002 14:44:58 GMT
Subject: GOTO w/ Python?
References: <mailman.1024964672.10880.python-list@python.org>
Message-ID: <Kl%R8.317333$cQ3.17671@sccrnsc01>

Delaney, Timothy
> I hope you're not serious ... :)
>
> COME FROM is a joke. Literally. It is a parody of GOTO whose only
redeeming
> feature is that it is not GOTO.
>

Having programmed extensively using GOTOs, I'll say the redeeming
feature of _COME FROM_ is that it makes GOTO look good!   ;-)

So-much-spaghetti-so-little-space-ly y'rs,

--

Emile van Sebille
emile at fenx.com

---------



From peter at engcorp.com  Sun Jun 23 12:09:15 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 23 Jun 2002 12:09:15 -0400
Subject: (Newbie)  Counting Instances ("Hits") with Regular Expressions
References: <3d15e215.43448155@news.texas.net>
Message-ID: <3D15F2AB.95E01A13@engcorp.com>

Ben Fairbank wrote:
> 
> I am new both to Python and to regular expressions, which may account
> for my difficulty.  I must count the frequenies of certain words in
> files of moderate length (about150 k bytes).  I have been reading
> files and then using count(s,sub), which is fast and easy.  I now have
> to allow for punctuation and eliminate words within words, etc, and so
> am trying to use regular expressions instead of simple words as
> targets.  I do not, however, find a similarly easy to use count
> function in the re module.  Yet this is such common operation it must
> be there, or easy to implement.  What is the usual way of simply
> counting "hits" in the re module?  (And what have I missed in the
> documentation; where is this to be found?  I have looked through Lutz
> and Ascher)

I don't think this sort of thing is really so common as you believe.
I'm also not certain using the re module is the typical way to handle
something like this.  I suspect writing some kind of parser/tokenizer
is more common, when you talk about having to take punctuation and
such into account.

In any case, if you find re's are suitable for you, you can easily
count the number found by using "findall" from the re module and
len() to find the count.

-Peter


From sholden at holdenweb.com  Mon Jun 17 16:17:42 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 17 Jun 2002 16:17:42 -0400
Subject: Creating Dynamic Web Pages (ZOPE)
References: <46ca81a0.0206062016.789d3af6@posting.google.com> <J29M8.358696$%u2.156726@atlpnn01.usenetserver.com> <46ca81a0.0206141201.23db4a8c@posting.google.com>
Message-ID: <2srP8.127337$Fp1.45878@atlpnn01.usenetserver.com>

"Hugh Cowan" <hwcowan at hotmail.com> wrote ...
> Steve,
>
> Thanks for the suggestion and sample web-site, I will definetly
> check-it out.  Does DreamWaver support ASP then, or rather provide
> support?  I mean, I know that you could just add the ASP code mixed in
> with the HTML manually, but does it offer any RAD type tools to help
> make using ASP easier?
>
DW does indeed support ASP, though it will also generate code for Cold
Fusion or Java Server Pages. The new DW MX product apprently adds PHP and
some others.

The reason I suggested DW is its ability to quickly build reasonably exact
web page layouts, coupled with the ability to define templates whose layout
characteristics can quickly be applied to some or all of the pages in a
site. It's different from FrontPage in making it easy to define your own
templates rather than mostly using pre-built and somewhat clunky "themes"
that scream "this is a FrontPage site".

> Also, I thought that Dreamweaver was more of just a HTML Editor, while
> something like Zope provided extra functionality, tools, wizards,
> etc.. to allow you to create Dynamic Web-Sites quicker and easier?  I
> am just trying to figure out where to concentrate my search for
> software and tools.
>
Zope's acquisition features can indeed be very helpful in propagating useful
behaviours to all necessary portions of a site. I'm not aware of wizards,
but I wouldn't be the first person to ask about Zope, since although I've
run it a couple of times I'm planning to wait until v3 to really try to get
to grips with it.

> I originally looked through the various *Freeware* web-sites for HTML
> editors -- but there are so many of them out there (hard to pick one
> from the other -- almost makes you want to go back to a plain old text
> editor).  While some offer advantages over others, they all seem to
> basically assist you in generating the HTML code for web-pages (either
> by coding or using a WYSIWYG type interface)?
>
DW UltraDev comes wtih Fireworks, which is extremely capable at building
spiffy graphics like rollovers. DreamWeaver appears to generate reasonably
lightweight HTML (though the JavaScript tends to burgeon a bit when you
start to polish up your interface with dynamic graphics). It's also  fairly
liberal at letting you edit the resulting HTML without stopping you from
using WYSIWYG mode.

I think probably anything further should be done by email, as we are
drifting well away from this NG's topic. I would still like to understand
what would be involved in adding another back-end to DreamWeaver, as it
would make a great addition to something like WebWare (one of the more
interesting Python web frameworks).

regards
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------









From calves at coelce.com.br  Thu Jun 20 08:20:08 2002
From: calves at coelce.com.br (Alves, Carlos Alberto - Coelce)
Date: Thu, 20 Jun 2002 09:20:08 -0300
Subject: RES: latex-tex-packages
Message-ID: <29A97D00F387D411AC7900902770E148058F93D5@lcoeexc01.coelce.net>

Thanks,
Pardon me, but I forgot to say it's to install in my linux box at home. So I
appreciate packages for linux box based on debian distribuition (I use
kernel series 2.4).

> -----Mensagem original-----
> De: Michael Hudson [mailto:mwh at python.net]
> Enviada em: quinta-feira, 20 de junho de 2002 08:55
> Para: Alves, Carlos Alberto - Coelce
> Assunto: Re: latex-tex-packages
> 
> 
> "Alves, Carlos Alberto - Coelce" <calves at coelce.com.br> writes:
> 
> > Hi all,
> > You forgive me, c'os I know this issue it's not to this 
> list; But, anybody
> > can tell me where can I get tex/latex packages to install. 
> I just browse
> > some url but can't find'em.
> 
> Depends on OS.  I'm making a tentative guess at windows from headers,
> so miktex is a reasonable choice:
> 
>     http://www.miktex.org
> 
> Cheers,
> M.
> 
> -- 
>   Those who have deviant punctuation desires should take care of their
>   own perverted needs.                  -- Erik Naggum, comp.lang.lisp
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020620/034f59e9/attachment.html>

From bt98 at doc.ic.ac.uk  Sun Jun  2 06:03:50 2002
From: bt98 at doc.ic.ac.uk (Benjamin Tai)
Date: Sun, 02 Jun 2002 11:03:50 +0100
Subject: passing large arrays from C to Python as arguments
References: <c3835e5f.0206010620.5a277bb2@posting.google.com>
Message-ID: <3CF9ED86.F9EF898A@doc.ic.ac.uk>

Chris Fonnesbeck wrote:

> I am looking for a way of passing relatively large arrays to python as
> an argument to a method call, without hard-coding the length of the
> array using Py_BuildValue. Is there an obvious solution to this, like
> passing it in a dictionary, or something? At any rate, it is very
> awkward to have to pass each element of each array into the argument
> individually.
>
> Thanks,
> cjf

Hi,

Not too sure if the suggestion would work out for you: implementing an
extension function which return a list to Python. If so, please read on.

Assuming you have an array "a" of length "size":



PyObject* p_list_ret;

 p_list_ret = PyList_New(size); assert(p_list_ret!=NULL);

 for (ix=0; ix<size; ix++)
     PyList_SetItem(p_list_ret, ix, PyLong_FromLong(a[ix]));

 return p_list_ret;



Be careful with the borrow reference issue in the PyList_SetItem(). You
might find it helpful to look at the API document first.


Thanks

Ben



From pyth at devel.trillke.net  Thu Jun 13 12:27:51 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 13 Jun 2002 18:27:51 +0200
Subject: filter2
In-Reply-To: <20020613160557.GC25416@unpythonic.net>; from jepler@unpythonic.net on Thu, Jun 13, 2002 at 11:06:02AM -0500
References: <mailman.1023975618.6474.python-list@python.org> <lkptyvmdfz.fsf@pc150.maths.bris.ac.uk> <20020613160557.GC25416@unpythonic.net>
Message-ID: <20020613182751.D6609@prim.han.de>

Jeff Epler wrote:
> If you're benchmarking, try a solution like this in the mix.  It only
> has one map/filter call, not map+2*filter. (uses nested scopes and the
> new bool() builtin, but this can be avoided if necessary)
> 
> def filter2(test, list):
>     l1 = []
>     l2 = []
>     funcs = [l2.append, l1.append]
>     map(lambda x: funcs[bool(test(x))](x), list)
>     return l1, l2

nice!

    holger




From rapskat at hotmail.com  Sun Jun  9 21:43:58 2002
From: rapskat at hotmail.com (rapskat)
Date: Sun, 09 Jun 2002 21:43:58 -0400
Subject: Help with Black Adder...
References: <adpc36$179jq$1@ID-139866.news.dfncis.de> <3d025e96$0$3887$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <ae108u$30ha6$1@ID-139866.news.dfncis.de>

Error Log for Sat, 08 Jun 2002 15:29:00 -0400: segfault in module
"Boudewijn Rempt" - dump details are as follows...

> rapskat wrote:
> 
>> Greetings All.  I'm just getting started with python and I chose Black
>> Adder as a RAD IDE.  Since my background is (ugh!) VB, I'm used to the
>> whole Top Down IDE thing.
> 
> Well, before I started using Python a few years ago, I was a VB
> programmer, too. (And Oracle before that -- I've come a long way, I
> think). Currently, I use XEmacs for everything, so the transition can be
> made, but you need a strong incentive. For me that incentive was to have
> to write a book on PyQt and BlackAdder in docbook sgml. And from using
> XEmacs for docbook to using it for everything else (except mail and
> news) was an insidiuous process. I've now redefined KDE to accept ctrl-s
> for search, and ctrl-x-s for save :-).
> 
> 
>> My questions are:
>> 
>> 1) I am having problems with BA not finding the PyQT libs.  I installed
>> both from the rpm's (I'm using RH 7.3).  But when I start BA it gives
>> me an error that it couldn't find the PyQT modules and that they won't
>> be available for the session.  I've tried removing them and
>> reinstalling to no avail.  I've run ldconfig as well.  Has anyone ele
>> seen anything like this and any ideas on how to resolve it?
> 
> I'm not sure that the current version of BlackAdder includes the PyQt
> library -- the commercial version does, I think, but the demo version
> doesn't. Or something like that. Check whether all libs are really
> there; then check whether LD_LIBRARY_PATH includes the location of the
> relevant .so's, whether BLACKADDERDIR is correct, and whether
> BLACKADDERPYTHON is set correctly.
> 
> (There's a layouto in my book on page 39: the second export ought to be
> on the next line.)
 
Thank you for all of your responses.

I will look for your publication, Boudewijn, I think it may provide the
answers to both my current and future issues.

I will also check those variables and see if perhaps they aren't properly
set.

Thanks again!

-- 
rapskat  -  9:35pm  up 6 days, 35 min,  6 users,  load average: 10.00, 10.03, 10.06
drop the hot to mail me

Friends should be preferred to kings. -- Voltaire


From kragen at pobox.com  Sun Jun  2 22:53:46 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 02 Jun 2002 22:53:46 -0400
Subject: HTML links in a GUI tkinter
References: <mailman.1022986323.16196.python-list@python.org>
Message-ID: <83elfprs1x.fsf@panacea.canonical.org>

aurel-le-gnou <aurelienviolet at noos.fr> writes:
> How can I put in a frame or in any widget (?) an html link so that the
> browser opens at the url indicated when the user clicks ?
> Just simply as if it were an HTML link in an HTML document but in a
> tkinter GUI ?

Sorry about the empty reply --- in Python 2.x, the webbrowser module
has a webbrowser.open() method that does what you want.



From vjovanov at stevens-tech.edu  Thu Jun  6 10:40:39 2002
From: vjovanov at stevens-tech.edu (Vojin Jovanovic)
Date: Thu, 6 Jun 2002 10:40:39 -0400
Subject: self
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu> <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com> <mailman.1023294728.2322.python-list@python.org> <sWsL8.82$v_7.4260962@newnews.cc.stevens-tech.edu> <adls9d$7ji$1@peabody.colorado.edu>
Message-ID: <pvKL8.84$v_7.4260962@newnews.cc.stevens-tech.edu>

> So ':=' variables would first try to evaluate their expressions in their
> normal namespace (that of the class, the problem you're dealing with now)
and
> would then look for 'normal' variable names to resolve. Only if a name
> wasnt't found in either would a NameError be raised.

Your idea is good.  It would be exactly what I need.  It would bring a lot
to python; a kind of symbolic manipulation capability.  Unfortunately, I
don't think that my current knowledge of Python is sufficient to implement
this in a general way you are suggesting.

V.J.




From jubafre at zipmail.com.br  Fri Jun 21 13:37:59 2002
From: jubafre at zipmail.com.br (jubafre at zipmail.com.br)
Date: Fri, 21 Jun 2002 14:37:59 -0300
Subject: =?iso-8859-1?Q?text=20search=20again=3F=3F=3F=3F=3F=3F=3F=3F=3F=3F?=
Message-ID: <3D135BCE00000186@www.zipmail.com.br>

self.text.search(st, 0.0, regexp='true')
is it correct???
i?m not a exact string, i?m searchin for winzip, but in the text
appears WinZip and the search doesn?t work, how  use regexp flag?????



Juliano Freitas
www.gebrasil.hpg.com.br



------------------------------------------
Use o melhor sistema de busca da Internet
Radar UOL - http://www.radaruol.com.br







From edream at tds.net  Sun Jun 23 06:05:17 2002
From: edream at tds.net (Edward K. Ream)
Date: Sun, 23 Jun 2002 10:05:17 GMT
Subject: ConfigParser & converting strings to lists
References: <3D152A28.2177B054@tds.net> <af3kv701tsj@enews3.newsguy.com>
Message-ID: <3D159D5C.2972C263@tds.net>

JohnJacob wrote:

> You just want to get your list back from it's string representaion, right?
> Try eval:
> 
> files = eval(config.get("recent files", "recentFiles"))

Brilliant!  As Paul Erdos would say, this is the way it is written in
"The Book".  Thank you!

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


From lmbrtmarc at aol.comlaissermoi  Tue Jun 25 23:38:42 2002
From: lmbrtmarc at aol.comlaissermoi (Marc)
Date: 26 Jun 2002 03:38:42 GMT
Subject: Python and Windows XP
Message-ID: <20020625233842.29966.00000386@mb-ma.aol.com>

Hello all,
   I apologise for this newbie question, but I use Windows XP and wonder how I
can open a shell session to use Python?  There used to be the DOS command
option, but now...
-- 
Marc
if we can't benefit from an error by making jokes, what's the  point of being
here?


From catunda at pobox.com  Wed Jun 26 18:48:58 2002
From: catunda at pobox.com (Marco Catunda)
Date: Wed, 26 Jun 2002 19:48:58 -0300
Subject: geting python files for use on web
References: <k6oS8.191997$6m5.159601@rwcrnsc51.ops.asp.att.net> <Xns9239DCB2C57A0cliechtigmxnet@62.2.16.82>
Message-ID: <pan.2002.06.26.22.48.56.934699.22508@pobox.com>

To use ASP and Python I suggest you to download
activepython from www.activestate.com

Activepython is an distribution of python with
all modules for windows and much other things.

--
Marco Catunda


On Wed, 26 Jun 2002 16:39:06 -0300, Chris Liechti wrote:

> "Ali K" <akakakhel at attbi.com> wrote in
> news:k6oS8.191997$6m5.159601 at rwcrnsc51.ops.asp.att.net:
>> I would like to know where I can download the nessecary files for usin
>> python in ASP and/or CGI.
> 
> the standard python distro from python.org should do for CGI.
> 
>> Please reply.
> 
> what do you mean? by email?
> 
> chris


From tjreedy at udel.edu  Wed Jun  5 17:02:04 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 05 Jun 2002 21:02:04 GMT
Subject: writing a text in hyphenless justification
References: <32d0f89.0206042221.1910798c@posting.google.com>
Message-ID: <g%uL8.142095$Gs.13245198@bin5.nnrp.aus1.giganews.com>

"Backflip" <backflip at gmx.de> wrote in message
news:32d0f89.0206042221.1910798c at posting.google.com...
> Hi everybody,
> i?ve got a little problem again......
> i want to take a raw_input and makes it to a hyphenless
justification output.....
> how is this possible?????
> The most beautiful solution would be the one, that i can define the
width....
> has got smb a idee??? it would be very nice, if smb can help me with
my problem.....

Go to
http://mail.python.org/pipermail/python-dev/2002-June/thread.html
and check the thread 'Where to put wrap_text' (Ie, where in the
standard library, for 2.3).

Terry J. Reedy





From skip at pobox.com  Wed Jun 26 13:25:21 2002
From: skip at pobox.com (Skip Montanaro)
Date: Wed, 26 Jun 2002 12:25:21 -0500
Subject: Python 2.2 __slots__ Bug?
In-Reply-To: <20020626.060323.78699939.glyph@twistedmatrix.com>
References: <20020626.060323.78699939.glyph@twistedmatrix.com>
Message-ID: <15641.63745.509965.708519@12-248-8-148.client.attbi.com>


    Glyph> I think I have discovered a weird bug in the Python interpreter.
    Glyph> Under Python2.2, when I run this script:

    ...

Glyph,

This looks like a new bug to me.  A quick scan for __slots__ in the SF bug
tracker didn't turn up anything.  If you are unable to get ahold of the SF
site, I'll file a bug report for you, just let me know.

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html




From blomquist_scott at hotmail.com  Mon Jun 10 16:34:47 2002
From: blomquist_scott at hotmail.com (Scott Blomquist)
Date: 10 Jun 2002 13:34:47 -0700
Subject: False Queue.Empty exception ?
Message-ID: <4a53cb75.0206101234.179ba0fd@posting.google.com>

I use the Queue class to maintain a threadpool of threads which
currently have no task, rather than creating/reaping threads as they
are needed. When I have a task, I grab one out of the queue and assign
the task to it. To add complexity :) I have made the threadpool able
to have a min and a max number of threads created, and if the queue of
free threads is Empty, then I (potentially) create more threads (up to
the max), and drop them into the free queue.

The problem is that the Empty exception appears to be raised even when
the queue is in fact not empty - causing me to create threads when I
don't really need to - some sort of race condition in the following
code from the Queue class?

    def get(self, block=1):
        """Remove and return an item from the queue.

        If optional arg 'block' is 1 (the default), block if
        necessary until an item is available.  Otherwise (block is 0),
        return an item if one is immediately available, else raise the
        Empty exception.
        """
        if block:
            self.esema.acquire()
        elif not self.esema.acquire(0):
            raise Empty
        self.mutex.acquire()
        was_full = self._full()
        item = self._get()
        if was_full:
            self.fsema.release()
        if not self._empty():
            self.esema.release()
        self.mutex.release()
        return item

Keeping in mind that I am doing a non-blocking get, it seems as though
if there is a thread between the self.esema.acquire(0) and the
self.esema.release(), and it gets switched out and another thread is
switched in and comes through the get call, it will fail on the
self.esema.acquire(0), even though the queue is not neccessarily
emtpy.

Can someone help me out here? Am I just misusing/misunderstanding the
capabilities of Queue class?

Thanks,
--ScottB


From glyph at twistedmatrix.com  Wed Jun 26 11:23:45 2002
From: glyph at twistedmatrix.com (Glyph Lefkowitz)
Date: Wed, 26 Jun 2002 10:23:45 -0500 (CDT)
Subject: Python 2.2 __slots__ Bug?
In-Reply-To: <B93F7990.CECB%jonathan@onegoodidea.com>
References: <mailman.1025089474.17780.python-list@python.org>
	<B93F7990.CECB%jonathan@onegoodidea.com>
Message-ID: <20020626.102345.59647403.glyph@twistedmatrix.com>

From: Jonathan Hogg <jonathan at onegoodidea.com>
Subject: Re: Python 2.2 __slots__ Bug?
Date: Wed, 26 Jun 2002 14:00:48 +0100

> On 26/6/2002 12:03, in article
> mailman.1025089474.17780.python-list at python.org, "Glyph Lefkowitz"
> <glyph at twistedmatrix.com> wrote:
> 
> > I get "I'm fine" and then a segmentation fault.  This only happens when
> > __slots__ is defined on a new-style class.  Seriously, i'm not making those
> > numbers up -- reliably, 43551 objects will crash the interpreter; fewer will
> > not.
> 
> Interestingly I get the construction working fine, but the 'del' crashes the
> interpreter with (on Mac OS X):

Hmm.  Sprinkling print statements about shows it dying on the 'del' (this is
how I encountered the bug originally, I was trying to time the [de]allocation
of large numbers of objects under 2.1 vs. 2.2), but the stacktrace is still the
same; is there some difference between the way allocation works on OS X
vs. Linux? (I assume if you can run OS X we're using the same hardware.)

> This seems to fail for me at 5430 objects. I couldn't get the allocation to
> crash at all (or at least up to a million objects, after which I got bored).

Yep, same here.  Allocation is no problem.

-- 
 |    <`'>    |  Glyph Lefkowitz: Traveling Sorcerer   |
 |   < _/ >   |  Lead Developer,  the Twisted project  |
 |  < ___/ >  |      http://www.twistedmatrix.com      |
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020626/4b92d813/attachment.sig>

From whisper at oz.net  Fri Jun 21 02:50:58 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 20 Jun 2002 23:50:58 -0700
Subject: Dispatching a Win32 COM in the background?
In-Reply-To: <9OpQ8.90552$ks6.8015@nwrddc02.gnilink.net>
Message-ID: <GCEDKONBLEFPPADDJCOEGEKDDKAA.whisper@oz.net>

Hmmm... if this is working for you, that's great - it just looks strange to
me.

If Background_TTS is the background speacher thread, shouldn't it have the
queue.get() in it and the foreground have the queue.put()?

It's also worth noting that self isn't used or needed at the global level,
so self.speech is kind of different ;-)

Actually... tell me how this works :) :

import threading

# Private class
class _speech_processor(threading.Thread):
	def __init__(self, speech_queue, talker):
		threading.Thread.__init__(self)
		self._talk_queue = speech_queue
		self._talker = talker

	def run(self):
		tq = self._talk_queue # tiny speed up?
		talker = self._talker
		while 1: # the main thread will kill us.
			talker.Speak(tq.get())

# end of _speech_processor

class BackgroundSpeaker():
	def __init__(self, debug_mode):
		self._debug_mode = debug_mode
		self._bg = None
		self._speech_queue = None
		self._talker = None
		self._statusOK = 1 # Must be before try/excepts. Success unless failure!
		try:
			from win32com.client import constants # I think this is redundant
       		import win32com.client # since you'll get the constants here too
		except ImportError:
			if (self._debug_mode):
				print "Unable to import Win32COM extensions - are they installed?"
			self._statusOK = 0

		try:
       		self._talker = win32com.client.Dispatch("SAPI.SpVoice")
		except:
			if (self._debug_mode):
				print "Error Initializing Microsoft Speech API."
			self._statusOK = 0

	# leading '_' means private method (by agreement only - not enforced)
	def _start_bg(self):
		import Queue
		self._speech_queue = Queue.Queue()
		self._bg = _speech_processor(self._speech_queue, self._talker)
		self._bg.start()

	def say(self, speech):
		if not self._statusOK:
			if self._debug_mode:
				print "Unable to say: %s." % speech, "Components not installed."
			return
		if self._bg is None:
			self._start_bg()
		self._speech_queue.put(speech)

# end of BackgroundSpeaker

if __name__ == '__main__'

import time
import sys

debug_mode = 1

Speaker = BackgroundSpeaker(debug_mode) # It always creates, even if it
can't talk.

while 1:
	speech = raw_input("Enter a phrase to speak, then hit Enter")
	if speech == "":
		Speaker.say("All done")
		time.sleep(20) # give some time for finishing up.
		sys.exit()
	Speaker.say(speech)

I would be interested in hearing if this works! It may have typos since I
did it all manually. Don't have the speech kit to test with.

HTH,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Steven M. Castellotti
> Sent: Thursday, June 20, 2002 12:12
> To: python-list at python.org
> Subject: RE: Dispatching a Win32 COM in the background?
>
>
> Thanks a lot guys, your suggestions have worked perfectly!
>
> 	Just for the record, here's what the full implementation looks like.
>
> This code allows text to be piped to Microsoft's Text-To-Speech
> sythesizer, spoken in the background, in a thread-safe manner:
>
> #####################################################################
>
> import threading
>
> class Background_TTS(threading.Thread):
>
>    def __init__(self, debug_mode):
>       threading.Thread.__init__(self)
>
>       self.debug_mode = debug_mode
>       self.loop = 1
>
>       import Queue
>       self.text_queue = Queue.Queue()
>
>
>    #####################################################################
>
>    def Speak(self, text):
>       self.text_queue.put(text)
>
>
>    #####################################################################
>
>    def stop(self):
>
>       self.loop = 0
>
>
>    #####################################################################
>
>    def run(self):
>       from win32com.client import constants
>       import win32com.client
>       import time
>
>       speech = win32com.client.Dispatch("SAPI.SpVoice")
>
>       while self.loop:
>
>       if not(self.text_queue.empty()):
>          text = self.text_queue.get()
>          speech.Speak(text)
>          time.sleep(.001)
>
> #####################################################################
>
> initialization looks like this:
>
> try:
>    self.speech = Background_TTS(self.gameInformation.debug_mode)
>    self.speech.start()
> except:
>    if (self.gameInformation.debug_mode):
>    print "Error Initializing Microsoft Speech API."
> else:
>    self.text_to_speech_enabled = 1
>
> #####################################################################
>
> calling the synthesizer:
>
> self.speech.Speak(text)
>
> #####################################################################
>
> and finally, stopping the process:
>
> self.speech.stop()
>
>
>
> 	Once again guys, thanks!
>
> -Steve Castellotti
>  SteveC (at) innocent.com
>  http://cogengine.sourceforge.net/
>
>
>
> On Wed, 19 Jun 2002 18:56:05 -0500, David LeBlanc wrote:
>
> > Hmmm... and the run method of Background could be sitting on one end of
> > a Queue (Queue is thread-safe: see your python doc for details) that's
> > buffering the asyncronous calls on the speacher. The main task then just
> > stuffs a talk request into the queue and the Background processes it
> > when it gets to it.
> >
> > David LeBlanc
> > Seattle, WA USA
> >
> >> -----Original Message-----
> >> From: python-list-admin at python.org
> >> [mailto:python-list-admin at python.org]On Behalf Of Peter Hansen Sent:
> >> Wednesday, June 19, 2002 16:47
> >> To: python-list at python.org
> >> Subject: Re: Dispatching a Win32 COM in the background?
> >>
> >>
> >> "Steven M. Castellotti" wrote:
> >> >
> >> > 2) Make the call inside a thread which produces the same effect.
> >>
> >> This might work (I don't know anything about COM though).  Try putting
> >> your own code inside the run method of a thread:
> >>
> >> import threading
> >> class Background(threading.Thread):
> >>     def __init__(self):
> >>         threading.Thread.__init__(self)
> >>
> >>     def run(self):
> >>         print 'Starting'
> >>         pass   # your own code goes here
> >>         print 'Done'
> >>
> >> >>> b = Background()
> >> >>> b.start()
> >> Starting
> >> <possibly a pause here?>
> >> Done
> >>
> >> If at this point you get a >>> prompt back right away, but the speech
> >> generation is carrying on "in the background" then you're all set. Make
> >> sure you do this with code you've already tested outside of the thread,
> >> so you'll know that any failure is not the fault of your own code...
> >>
> >> -Peter
> >> --
> >> http://mail.python.org/mailman/listinfo/python-list
> --
> http://mail.python.org/mailman/listinfo/python-list





From geoff at gerrietts.net  Wed Jun  5 16:04:31 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Wed, 5 Jun 2002 13:04:31 -0700
Subject: htmllib examples?
In-Reply-To: <ut1nfuo5skj0d5g0ff6dngo1icelpdfmg1@4ax.com>
References: <ut1nfuo5skj0d5g0ff6dngo1icelpdfmg1@4ax.com>
Message-ID: <20020605200431.GB1455@isis.gerrietts.net>

Quoting Giulio Cespuglio (giulio.agostini.remove.this at libero.it):
> Hi there,
> 
> I'm trying to use htmllib, but I can't find any examples in the
> documentation
> (http://www.python.org/doc/current/lib/module-htmllib.html).
> Is there a (very well hidden) example codebase in www.python.org?
> Where else can I look?

I sent this out the other day -- it's probably not exactly what you're
looking for, but it's a step in the direction:

http://groups.google.com/groups?hl=en&lr=&selm=mailman.1022875851.20681.python-list%40python.org

-- 
Geoff Gerrietts             "Don't get suckered in by the comments-- 
<geoff at gerrietts net>     they can be terribly misleading.
www.gerrietts.net/geoff/     Debug only code." --Dave Storer




From pinard at iro.umontreal.ca  Sat Jun 15 14:53:03 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 15 Jun 2002 14:53:03 -0400
Subject: Pronouncing '__init__'
In-Reply-To: <20020615165915.A17174@hishome.net>
References: <20020615132519.A14452@hishome.net>
	<20020615134410.A15079@prim.han.de>
	<20020615165915.A17174@hishome.net>
Message-ID: <oqu1o44bn4.fsf@titan.progiciels-bpi.ca>

[Oren Tirosh]

> I used __init__ just as an example.  I meant how do you pronounce names
> with leading and trailing double underscore in general.  The magic name
> __init__ may be by far the one most commonly encountered in everyday
> programming but __there__ __are__ __many__ __others__.

I sometimes said "special" as a prefix (suffix in French!).  So, "special
init", "special mul", etc.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From osuchw at ecn.ab.ca  Fri Jun 14 17:36:01 2002
From: osuchw at ecn.ab.ca (waldekO)
Date: 14 Jun 2002 14:36:01 -0700
Subject: Help with ADO and dynamic insert statement...
References: <3d0826ec.240888125@netnews.attbi.com> <c5b3dbee.0206132203.793dc5a3@posting.google.com> <3d09c924.347952062@netnews.attbi.com>
Message-ID: <c5b3dbee.0206141336.35f6b34@posting.google.com>

candiazoo at attbi.com wrote in message 

> I am still having problems using dynamic sql statements to perform lookups and
> inserts into a MS Access database and an Oracle database.  
--DELETED--

I can not to answer you question directly but I may show you mine
approach to querying using ADO.

To retrive information from database

from win32com.client import Dispatch
conn = Dispatch('ADODB.Connection')
conn.Open(db_source, db_user, db_passwd)
sql = """
      select id, fname, lname, address
      from contacts
      where contact_id = %s
      """ %(contactid)
flag, rs = conn.Execute(sql)
# function name is GetRows but really it returns columns
# so one has to swing the results around uzing zip()
cols = rs.GetRows()
if rows:
    rows = zip(*cols)


# To insert I would similarily first construct sql statemnt using
string
# formatting and then call conn.Execute(sql)
# For example:
sql = """
      insert into other_contacts
      (id, fname, lname, address)
      values
      (%s, '%s', '%s', '%s')
      """

for row in rows:
    conn.Execute(sql %row)

Of course this approach is very simple and does not take into account
escaping
single quotes or exotic data types but it is sufficient for my needs
in most cases.

waldekO


From jon+usenet at unequivocal.co.uk  Tue Jun 25 11:36:03 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Tue, 25 Jun 2002 15:36:03 -0000
Subject: Python hits the spot
References: <mailman.1024928911.24460.python-list@python.org>
Message-ID: <slrnahh3f3.sar.jon+usenet@snowy.squish.net>

In article <mailman.1024928911.24460.python-list at python.org>, Tim Peters wrote:
> [Andrew Koenig]
>> Don't some implementations of malloc offer a way to ask the operating
>> system to commit to the memory being available when you allocate it?
>> If so, I can imagine testing for that facility at config time and
>> using it if available.
> 
> SourceForge remains open 24 hours a day for patches.  My bet is that nobody
> cares enough to bother -- the barrier for complaining on Usenet is one an
> ant can hurdle with ease <wink>.

It wouldn't be a good idea anyway. It is not Python's job to second-guess
the operating system's memory management policies.


From irmen at NOSPAMREMOVETHISxs4all.nl  Sat Jun 15 17:37:33 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Sat, 15 Jun 2002 23:37:33 +0200
Subject: type(socket) != SocketType
References: <3D0BAD4D.5050105@NOSPAMREMOVETHISxs4all.nl> <Xns922EEDAF2FE9Ecliechtigmxnet@62.2.16.82>
Message-ID: <3D0BB39D.1010802@NOSPAMREMOVETHISxs4all.nl>

Chris Liechti wrote in a prompt response:

> what happens to people with timoutsocket? that module repaces the 
> socket.socket class with a custom one and your check will fail....

Heh, you're probably right. I hereby declare socket.SocketType useless.

> maybe there is a cleverer solution that typechecking. the SSL socket has 
> other/additional attributes etc right?. check for them with has_attr(). or 
> use the __name__ attribute of the class.

Thanks for these suggestions. I will probably go check for the 'pending'
attribute, because that is what the SSL sockets have and regular sockets
don't. This will likely fix things....

Irmen.




From phd at phd.pp.ru  Tue Jun  4 04:45:40 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Tue, 4 Jun 2002 12:45:40 +0400
Subject: libpython as a shared library?
In-Reply-To: <Pine.BSF.4.32.0206031620070.82931-100000@localhost>; from grisha@ispol.com on Mon, Jun 03, 2002 at 08:24:28PM +0000
References: <Pine.BSF.4.32.0206031620070.82931-100000@localhost>
Message-ID: <20020604124540.E9162@phd.pp.ru>

On Mon, Jun 03, 2002 at 08:24:28PM +0000, Gregory (Grisha) Trubetskoy wrote:
> Has there been any progress in making an option of building libpython as a
> shared lib (.so)? I saw some feature requests, but couldn't quite figure
> out the ultimate status of this.

   Implemeneted in Python 2.3. Download CVS version or nightly tarball.

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From rayed at saudi.net.sa  Sun Jun  2 00:42:44 2002
From: rayed at saudi.net.sa (Rayed Al-Rashed)
Date: Sun, 02 Jun 2002 07:42:44 +0300
Subject: Interactive socket connection
References: <3CF88F9B.1080308@saudi.net.sa> <3CF8BA2E.A8C42539@engcorp.com>
Message-ID: <3CF9A244.6040505@saudi.net.sa>

Thanks for the response, flush() helped a little, but I faced another 
problem, waiting for input. I built another version it work great using 
"select" module to poll both the socket and stdin.

The finished program is a POP proxy that accept the connection from the 
user and depending on the user name it forward the connection to the proper 
POP3 server (configured by getuserserver function):
for example user "rayed" go to server "r.pop.mydomain.com"

Thanks a lot for your advice, any ideas on how to improve it? :)

------------------------------------
#!/usr/local/bin/python

import sys
import string
import socket
import select


def getuserserver(user):
     server = "%s.pop.mydomain.com" % (user[0], )
     return (user, server)


print "+OK POP3 server ready"
sys.stdout.flush()

# GET USER NAME
while 1:
         s = sys.stdin.readline()
         s = string.strip(s)
         s = string.lower(s)
         try:
                 (command, username) = string.split(s, " ", 2)
         except:
                 command = s
         if command == "user":
                 break
         elif command == "quit":
                 print "+OK Sayonara"
                 sys.exit()
         else:
                 print "-ERR Unknown AUTHORIZATION state command"
                 sys.stdout.flush()

(id,server) = getuserserver(username)


s = socket.socket(socket.AF_INET,socket.SOCK_STREAM)
s.connect((server, 110))
buf = s.recv(8000)
s.send("USER %s\r\n" % (id,))

while 1:
         r,w,e = select.select ([sys.stdin,s], [], [sys.stdin,s])

         if s in r:
                 buf = s.recv(8000)
                 if not buf:
                         # Connection closed
                         break
                 sys.stdout.write(buf)
                 sys.stdout.flush()

         if sys.stdin in r:
                 buf = sys.stdin.readline()
                 s.send(buf)
------------------------------------




Peter Hansen wrote:
> Rayed Al-Rashed wrote:
> 
>>Hello,
>>
>>I am trying to build POP3 proxy, I want it to run from "inetd" so it should
>>communicate with the user using regular stdin, stdout.
>>I built a small code that should run as general proxy for any protocol, but
>>I faced the following problem, it doesn't show any output from the server
>>until I pass it a command [...]
> 
> 
> Are you maybe in need of a sys.stdout.flush() to get the output to
> be sent after the sys.stdout.write()?
> 
> -Peter




From sholden at holdenweb.com  Mon Jun  3 08:11:25 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 3 Jun 2002 08:11:25 -0400
Subject: win32com under Cygwin
References: <3CF9846B.579606ED@wiencko.com> <Xns9221CB7B32FDcliechtigmxnet@62.2.16.82>
Message-ID: <%5JK8.180011$%u2.88943@atlpnn01.usenetserver.com>

"Chris Liechti" <cliechti at gmx.net> wrote in message
news:Xns9221CB7B32FDcliechtigmxnet at 62.2.16.82...
> Tom Wiencko <tew at wiencko.com> wrote in news:3CF9846B.579606ED at wiencko.com:
>
> > There may be an easy answer for this, but I have not yet found it...
> >
> > I noticed that the Python distribution with Cygwin does not include the
> > win32 extensions.  The only distribution I can find for them is a
> > Windows executible, which does not seem to know about Cygwin.
>
> thats the official page of win32all:
> http://starship.python.net/crew/mhammond/
>
> the binary version is compiled for the official python.org interpreter.
>
> > Can these libraries be loaded under Cygwin?  Do they even work under
> > Cygwin?  Can somebody point me toward how to load them if it is
> > possible?
>
> the sources are available from the page above. maybe you can manage to
> compile it with cygwin for cygwin python. don't know if that is easy...
>

I'm guessing that it would be easier to compile the win32all extensions
under mingw32 than under cygwin, because the former (IIRC) uses Microsoft
libraries, whereas the latter uses special libraries to give a modicum of
Posix compatibility.

It probably won't be impossible. I have compiled simpler extensions like
mxODBC and MySQLdb under cygwin with minor problems (that should go away, as
results were fed back to package authors). But don't expect a cygwin port to
be a simple recompile...

regards
 Steve
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From gbien at intekom.co.za  Tue Jun 11 11:14:33 2002
From: gbien at intekom.co.za (Geoff Bien)
Date: Tue, 11 Jun 2002 17:14:33 +0200
Subject: unknown opcode Exception then AV
Message-ID: <ae546s$bg5$1@ctb-nnrp2.saix.net>

I have one script out of many (python 2.1 activestate on win32)
which gives an "invalid opcode" exception and then also
sometimes causes an AV.

This script  uses intebase with the gvib or kinterbasdb
extension modules. since using gvib or kinterbasedb
makes no difference to the problem they are probably
not the cause

this is the only script I have problems with. it is too long
to post here.

Traceback (most recent call last):
  File "<string>", line 452, in ?
  File "<string>", line 446, in go
  File "<string>", line 425, in doPeriod
  File "<string>", line 341, in doMonth
SystemError: unknown opcode

the interpreter seems quite confused with this script,
in other places simple statements like x=x+1  are giving
the wrong results







From emile at fenx.com  Sun Jun 23 12:06:50 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 23 Jun 2002 16:06:50 GMT
Subject: Newbie question (anyone programming with curses?)
References: <af4ovd$1bt3$1@pingwin.acn.pl>
Message-ID: <ummR8.290538$352.27434@sccrnsc02>

hellboy
> I'm quite new to Python, but I'm positive that version 2.2 (for
Windows)
> comes without curses module, though documentation states it is
included.
> import curses gives me "Import Error: no module named curses" message.
> Anyone knows where to download curses module (for Windows)?
>

from http://py-howto.sourceforge.net/curses/node2.html

No one has made a Windows port of the curses module. On a Windows
platform, try the Console module written by Fredrik Lundh. The Console
module provides cursor-addressable text output, plus full support for
mouse and keyboard input, and is available from
http://effbot.org/efflib/console.

Hope this is current and helps,

--

Emile van Sebille
emile at fenx.com

---------



From ark at research.att.com  Thu Jun  6 13:14:24 2002
From: ark at research.att.com (Andrew Koenig)
Date: Thu, 6 Jun 2002 17:14:24 GMT
Subject: for , ordering
References: <mailman.1023377649.29116.python-list@python.org>
Message-ID: <yu99u1og1g9b.fsf@europa.research.att.com>

gabor> hi,
gabor> i have a list and i want to create a copy of it with a bit different
gabor> objects... by now i have something like this:

gabor> objects1 = []
gabor> objects2 = []
gabor> .
gabor> .
gabor> .

gabor> for obj1 in objects1:
gabor> 	objects2.append(obj2(obj1.x))


gabor> 1. isn't there a more elegant way to do this?

Isn't this the same as saying

        objects2 = objects1[:]

?  Am I missing something obvious?

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From mcfletch at rogers.com  Fri Jun 28 15:33:24 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Fri, 28 Jun 2002 15:33:24 -0400
Subject: wxpython threads help needed
References: <6bd9f01b.0206281055.2b01a2a6@posting.google.com>
Message-ID: <3D1CBA04.2080201@rogers.com>

http://wiki.wxpython.org/index.cgi/LongRunningTasks

HTH,
Mike

ps: better place to ask for wxPython specific advice is the wxPython 
users list...
http://lists.wxwindows.org/mailman/listinfo/wxpython-users


sameer wrote:
> Does anyone have an example of wxpython threads in action?  I am
> trying to create  an application with several (tabbed) panels, each of
> which is being updated on its on thread.  Basically, I don't want the
> application to freeze up, or to have long startup times as all the
> panels are being updated.  I want the startup to happen as soon as the
> main panel is built.  I would appreciate it if someone can give an
> example either from their own app or code, or a simple example cooked
> up on the fly.
> 
> Thanks!


-- 
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From huaiyu at gauss.almadan.ibm.com  Mon Jun  3 22:01:04 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Tue, 4 Jun 2002 02:01:04 +0000 (UTC)
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org>
Message-ID: <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com>

Gustavo Cordova <gcordova at hebmex.com> wrote:
>So, it's not actually a bug, but an artifact of the hoops
>Python needs to jump through in order to maintain total
>dynamism in data types.
>
>Instead of this:
>
>   a = ([1],[2],[3])
>   a[0] += [3]
>
>do this:
>
>   a = ([1],[2],[3])
>   t = a[0]
>   t += [3]
>
>there won't be any exception, because 't' *can* be rebinded
>to the value emmited by the += operation; *and* 'a' will
>also contain the same reference.
>
>It's not *totally* equivalent, because it's missing a
>single, last step:
>
>   a[0] = t
>
>*that* is what's raising the exception, which is what
>the += operator is trying to do in you're original version.
>

Great explanation!  The problem stems from the fact that the current +=
operator implements two semantically distinct operations, chosen according
to some properties of the operands.  This is not unlike the situation of the
division operator with distinct behaviors on floats and ints.

I'd imagine that at some future time there would be a desire to split += to
two distinct operators, one for true in-place operator and one for the fake
one currently implemented for immutables.  That would remove the spurious
exceptions for mutable members of immutable structures.

A main hindrance to the change at that time would be the difficulty to
disambiguate the true intention of existing code with the += operator.
That's not unlike the difficulty involved with the division operator change.
Maybe it's better to change earlier?


Huaiyu


From skip at pobox.com  Mon Jun 10 11:17:09 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 10 Jun 2002 10:17:09 -0500
Subject: Newbie OOP Question.(diff between function, module and class)
In-Reply-To: <B9279B52.6AEB%sarmstrong13@mac.com>
References: <B9279B52.6AEB%sarmstrong13@mac.com>
Message-ID: <15620.49909.679355.726113@12-248-41-177.client.attbi.com>

    SA> I'm still trying to grasp this OOP concept in Python. Can someone
    SA> tell me the difference between a function, a module, and a class?
    SA> Since Python defines them all as objects, how do they differ outside
    SA> of being called from inside the script or outside the script?

You posted this midday Saturday and here it's Monday morning but you still
have no replies to this question?  I would have thought people would jump
all over the opportunity to respond to a question like this.

Here's my take on things:

 * A function is an action to perform using a set of input arguments and
   returning a set of output arguments.  For example:

   >>> print sin(47)
   0.123573122745

 * A class associates a chunk of data with a set of functions which operate
   on that data.  The first argument to each function in a class (typically
   called a "method" in Python) is a reference to the data stored in that
   instance of the class.  For example:

    import math
    class Point:
        def __init__(self, x, y):
            self.x = x
            self.y = y

        def polar(self):
            return math.sqrt(self.x**2+self.y**2), math.atan2(self.y, self.x)

 * A module is an object that partitions the namespace in a hierarchy.  For
   example, you can have a function named "sin" which is distinct from the
   sin function in the math module ("math.sin").  Modules used in a
   straightforward way only add a single extra level of hierarchy to the
   namespace.  Packages generalize this concept to multiple levels (modules
   inside modules).

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html





From dysfate at gmx.de  Fri Jun 14 09:35:42 2002
From: dysfate at gmx.de (Johannes Stiehler)
Date: 14 Jun 2002 15:35:42 +0200
Subject: Writing a simple linux daemon type thing...
References: <K9eO8.393$LN6.261@newsread1.prod.itd.earthlink.net> <rn3cea.rt4.ln@127.0.0.1> <yylO8.916$TP4.88596@newsread2.prod.itd.earthlink.net>
Message-ID: <aecrfe$a4l$1@minotaurus.cip.informatik.uni-muenchen.de>

I suppose you would have to redirect all output to stdout and log directly
to a specified log file. That's usually in the code for all daemonizing
processes.
The point is to completely unconnect the process from the console where it
was started.

Johannes
-- 
Johannes Stiehler	
CIS der LMU M?nchen 


From dyoo at hkn.eecs.berkeley.edu  Sat Jun  1 03:38:02 2002
From: dyoo at hkn.eecs.berkeley.edu (Daniel Yoo)
Date: Sat, 1 Jun 2002 07:38:02 +0000 (UTC)
Subject: Internat
References: <69QJ8.68711$ux5.82517@rwcrnsc51.ops.asp.att.net>
Message-ID: <ad9tkq$g2n$1@agate.berkeley.edu>

Ali K <akakakhel at attbi.com> wrote:
: Can one use Python as a language for ASP or CGI or something?

Hi Ali,

Sure!  There's some good resources on introductory CGI stuff with
Python on the "Web Programming" topic guide:

    http://www.python.org/topics/web/


Devshed.com has a bunch of articles on doing CGI programming with
Python a while back, and they seemed prety useful:

    http://zope1.devshed.com/zope.devshed.com/Server_Side/Python/CGI/page1.html

What programming background are you coming from?  We might be able to
point you toward more relevant resources if you tell us more about
what you'd like to do.


I have to admit complete ignorance about ASP, but there are resources
out there that talk about how to drive ASP with Python.  Chris Tismer
has a site about this:

    http://starship.python.net/crew/pirx/asp/py_asp.html


Good luck to you!


From Chris.Barker at noaa.gov  Wed Jun  5 15:49:25 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Wed, 05 Jun 2002 12:49:25 -0700
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023234674.21472.python-list@python.org> <ufrupva3edm6fd@news.supernews.com>
Message-ID: <3CFE6B13.79CA4A8C@noaa.gov>

John Roth wrote:
> I don't see a compatability issue here, because
> the current behavior raises an exception.

Unless someone did the perverse thing I proposed in a earlier post, and
took advantage of the fact that the operation did, in fact, work, even
though it raised an exception!

While I completely agree with Huaiyu's argument (though I'm not sure the
"increment with re-binding" is really all that useful for mutables), and
I think the current semantics is a minor wart, I am convinced that the
original behavior that started this thread is a MAJOR WART (I'd call it
a bug). I think I have a pretty good understanding now of why it
happens, and why it isn't trivial to change, but I think it is a very
bad idea for an operation to raise an exception, but also be successful!

Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From david.abrahams at rcn.com  Mon Jun 17 21:13:22 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Mon, 17 Jun 2002 21:13:22 -0400
Subject: ImageMagick for Python incl. Source
References: <aelmhc$35m$03$1@news.t-online.com>
Message-ID: <aem1ga$e2p$1@bob.news.rcn.net>

"Achim Domma" <achim.domma at syynx.de> wrote in message
news:aelmhc$35m$03$1 at news.t-online.com...
> Hi,
>
> I have just updated http://www.procoders.net/imagemagick with some
> documentation and small examples. There is also source code to compile
> ImageMagick for Python on non-windows plattforms, but this might be a bit
> tricky. Feel free to ask, if you have problems.
>
> greetings
> Achim

Heh, you might want to fix the link to point at www.boost.org, or possibly
http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/*checkout*/boost/boost/libs/p
ython/doc/v2/reference.html, if you know what I mean ;-)

-Dave




From msoulier at nortelnetworks.com_.nospam  Mon Jun 24 13:12:17 2002
From: msoulier at nortelnetworks.com_.nospam (Michael P. Soulier)
Date: 24 Jun 2002 17:12:17 GMT
Subject: I'd give up Perl tomorrow if only...
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au> <3D12F6FC.3000608@wedgetail.com>
Message-ID: <slrnahekqq.12v.msoulier@pmerd071.ca.nortel.com>

On Fri, 21 Jun 2002 09:50:54 GMT, Derek Thomson <derek at wedgetail.com> wrote:
> 
> Among other things, it's one of the reasons I'm tinkering with a Perl ORB.

    There's also XMLRPC, and SOAP, something already supported by both
languages. 

    Mike

-- 
Michael P. Soulier, QX41, SKY  Tel: 613-765-4699 (ESN: 39-54699)
Optical Networks, Nortel Networks, SDE Pegasus
"...the word HACK is used as a verb to indicate a massive amount
of nerd-like effort."  -Harley Hahn, A Student's Guide to Unix


From rhymes at myself.com  Sat Jun 29 11:39:10 2002
From: rhymes at myself.com (Rhymes)
Date: Sat, 29 Jun 2002 17:39:10 +0200
Subject: private
Message-ID: <pan.2002.06.29.17.38.51.789527.440@myself.com>

Is there any possibility to build up a _real_ private attribute?
Such as C++ private data members...
I know that with double-underscore prefix (and the name mangling related)
I can " hide " data members but I think it's not a so crafty trick...
Is there any PEP about it? Will Guido insert a private keyword
in a future release of Python ?

-- 
Rhymes
rhymes at myself.com
" ride or die "


From opengeometry at NOSPAM.yahoo.ca  Thu Jun 27 14:24:05 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 27 Jun 2002 18:24:05 GMT
Subject: Help with an algorithm wanted
References: <rowen-CE3C0F.13020226062002@nntp1.u.washington.edu> <mailman.1025125654.6891.python-list@python.org> <rowen-3B9D8A.16590026062002@nntp3.u.washington.edu>
Message-ID: <affl84$e94us$1@ID-99293.news.dfncis.de>

Russell E. Owen <rowen at cesmail.net> wrote:
> In article <mailman.1025125654.6891.python-list at python.org>, William Park
> <opengeometry at yahoo.ca> wrote:
>>If pairing (ie. mapping between 2 objects) occurs randomly, then it's
>>virtually impossible, because you don't know which direction to go from any
>>one object.  It would be travelling from city to city, without knowing
>>which state or county.
> 
> Fortunately the pairings (conversion functions) are completely fixed. So  
> the graph never varies, and it even has a fairly simple shape (basically 
> a fork with 3 tines and a long handle, if that makes any sense).
> 
> The problem is that the graph has 10 nodes. Rather than hand code 10 
> different objects, I'm trying to create a factory that will generate the 
> required object based on which node (data representation) is a given 
> from which the others are derived.

In your previous example, C is the central "hub" for A, B, and D.  If you
have
    hub = {'A': 'C', 'B': 'C', 'D': 'C'}
then, you can find the node to which A, B, and D are attached, ie.
    >>> hub['D']
    'C'

Can you organize other objects around hubs?  After that, organize the hubs
into central hub of their own?  If so, then the traversal will be much
easier.

It's difficult to get our teeth into without knowing what the big picture
looks like... :-)

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin


From dom at edgereport.put_a_c_o_m_here  Thu Jun  6 15:36:03 2002
From: dom at edgereport.put_a_c_o_m_here (Domenic R. Merenda)
Date: Thu, 06 Jun 2002 19:36:03 GMT
Subject: Medium to Large Scale Python Deployments
References: <KHkL8.21383$eD2.2399923271@newssvr10.news.prodigy.com> <833cw0gvg3.fsf@panacea.canonical.org> <FbOL8.42027$n4.9527047@newsc.telia.net>
Message-ID: <DQOL8.21621$km1.2488527968@newssvr10.news.prodigy.com>

   Exactly.  We plan to upgrade as soon as we test the new version.

--
Domenic R. Merenda
Editor, The Edge Report
http://www.edgereport.com


"Fredrik Lundh" <fredrik at pythonware.com> wrote in message
news:FbOL8.42027$n4.9527047 at newsc.telia.net...
> Kragen Sitaker wrote:
>
> > Are you running Python natively on the AS/400?  How hard was that?
>
> chances are he's using Per Gummedals port:
>
>     http://home.no.net/pgummeda/





From Christopher.Saunter at durham.ac.uk  Wed Jun 26 06:29:52 2002
From: Christopher.Saunter at durham.ac.uk (Christopher Saunter)
Date: Wed, 26 Jun 2002 10:29:52 +0000 (UTC)
Subject: Pythoniac: Thoughts on a hardware Python processor
Message-ID: <afc52t$2pj$1@sirius.dur.ac.uk>

Dear All,

I have seen the occasional post about creating a specific hardware 'Python
Processor', that could then be built and used in programmable logic or the
like (Valves, relays, core memory - Pythoniac ;-) - in other words, create
a processor that executes Python bytecode directly.  There would be
various different reasons for doing this (the 'neat' factor etc...)  

Replies to these questions have been a bit scant, so I though I would
chuck something in.

Due to Lazy Typing Syndrome I've used a couple of abbreviations:
PVM - Python Virtual Machine / Python Interpreter 
PBC - Python Byte Code

This has been done for Java, for example
http://www.xilinx.com/prs_rls/0134lavacore.htm 
and I have seen a partially complete open source core out there somewhere,
but I can't find it right now.

I have recently been looking at doing this for Python, and it seems like a
rather odd task.  From playing with the 'dis' module and reading some
docs, it seems that Python Interpreter (PVM), which would be translated
into a processor, has these properties:


o	Operates on sequential bytecode with branching / jump
instructions.
o	Entirely stack based  (no concept of registers / memory)

So far, so good - a processor could be built to do these functions of the
interpreter.  However, we go on....

o	Rather than operating on 8/16/xxx bit words, as most hardware
processors do, the PVM processes objects (via pointers on the stack),
which are treated in a 'black box' fashion - i.e. they have add, subtract,
compare etc. methods that are called by the PVM, and that return a result.
These methods of objects may not actually be coded for in Python, but in
the machine language (x86, alpha etc, derived from C etc.) of the
underlying processor.

Here, the PVM is very different from a 'typical' hardware processor, as
the later has a strictly limited number of data types, which it is able to
manipulate, whereas Python has many, built out of the basic types, and
although some may be coded in Python, they themselves are built on top of
C code (or Fortran etc.) that are all translated to native (x86 etc)
bytecode outside of Python.  The upshot of this is that the PVM is
unaware, in the majority of cases, of what actually needs to be done to
manipulate objects, it just makes calls to non Python code.

So, how to go about dealing with this in a Python processor?  

a) Create a stack based Python bytecode processor to replace the PVM, and
have this interact with another processor (x86 etc.) which stores the
objects and executes their methods. Ick.

b) Create a stack based bytecode processor that interfaces to custom
hardware processors for each Python object.  Ick.

Neither of these are particularly nice (or for b, workable) solutions.
The one I prefer is the following:

c) Create a processor that understands Python bytecode, using a stack for
the object pointers.  However, the processor would execute a sort of
superset of PBC, call it P+.  This would be fully compatible with Python,
but would include several built in types (32bit integers, 8bit integers
for strings etc,) and would have access to a memory space.  The basic
Python types(1) would then be implemented using P+, and stored in the
memory available to P+.  

c Should present a processor that executes Python bytecode, manipulating
the Python stack.  Then, when a method of an object is called, the
appropriate piece of P+ is called, executed by the same processor, and it
modifies the data structures of the objects, and the Python stack as
necessary, before returning seamlessly to the PBC execution.

Python extensions would then be directly executable on this processor, if
coded in Python.  If they are coded in C or some other language, a C->P+
compiler would be needed...

The level of abstraction caused by having to chase down function and
object pointers will be higher than for the execution of code on a more
conventional hardware processor, but this is also the case for executing
PBC with an interpreter.  

One advanced idea (not really thought about it yet!) is to store the
Python stack in main memory, but to store the top x (e.g. 32) items of the
stack in a register file to allow bytecode parallelism.  I gather that
IA-64 may do something sort of similar.

---

So, Pythoniac: a hardware (well, FPGA(2)...) processor that directly
executes Python bytecode.  What do people think about the idea?  Is it
really feasible?  I have doubts it would be particularly quick (in the
same way I have my doubts about anti-gravity generators... ;-), but it
should be fun!  

I would like to try and implement this, initially in a simulation
environment, then in a real FPGA, with hooks back to a PC allowing the
loading, execution, single step execution and monitoring etc of PBC.
Ideally I would like to reach a state where an interactive interpreter
runs over a serial port... 

So, anyone else interested in this?  Comments, suggestions, gaping holes
in my understanding of things as outlined above etc are welcome!

Regards,
	Chris Saunter



(1) Now we have increasing type<->object integration, should we refer to
numbers, strings, lists etc as types or objects?

(2) FPGA - Field Programmable Gate Array - a chip that consists of a
'logic fabric' than can be configured to produce arbitrary circuit
designs.  If this sounds new to you, try www.fpgacpu.org  




From jjl at pobox.com  Thu Jun  6 17:55:07 2002
From: jjl at pobox.com (John J. Lee)
Date: Thu, 6 Jun 2002 22:55:07 +0100
Subject: urllib, urllib2, httplib -- Begging for consolidation?
In-Reply-To: <mailman.1023295507.16177.python-list@python.org>
References: <mailman.1023295507.16177.python-list@python.org>
Message-ID: <Pine.SOL.4.30.0206062242070.21937-100000@mimosa.csv.warwick.ac.uk>

On Wed, 5 Jun 2002 brueckd at tbye.com wrote:

> On Wed, 5 Jun 2002, John J. Lee wrote:
[...]
> Before I jump in I do want to emphasize that I'm not trying to be too
> critical of httplib/urllib/urllib2 - the OP asked if they should be
> consolidated, and my short response is "yes, and reorganized too". :-)

OK -- but no reason for consolidation of the implementation, yes.  I don't
really think there is any reason for consolidation of the interface either
(which I suppose is what you mean by reorganisation), but the
documentation for httplib might include a pointer to urllib{2,}, since as
you point out below it's probably not obvious that that's where the HTTP
redirection stuff is.

[...]
> So, depending on your specific needs, you'd "plug in" to this hierarchy at
> the appropriate level. If all you need is to go fetch some object, you'd
[...]
> The key though is that each level you'd have to reinvent as little as
> possible - you'd build on related work from lower levels.

Almost goes without saying.  But again, I think that's what we've got
already.

[...]
> functionality instead of enforced layering). The approach we have today is
> *sort of* like this, except that richer and smarter functionality is being
> added at the *top* of the hierarchy.

The redirection stuff was added between httplib and the OpenerDirector /
urlopen stuff.  That's not at the top.

> Problem #1 is what makes me throw my hands up in frustration when we talk
> about, e.g., expanding the urllib APIs to have some way to do a HEAD
> request: it doesn't belong on that level of functionality.

I think you mean "it doesn't belong in a module named 'urllib2'"... or at
least you should mean that ;)

> That generic
> interface is for making it trivially easy to fetch the contents associated
> with a URL, independent as much as possible from protocol.

Only part of urllib2 (OpenerDirector &c) is a generic any-old-url-scheme
API.  The rest (eg. AbstractHTTPHandler, HTTPRedirectHandler) is
HTTP-specific, and can be used on its own, without going through the
generic stuff.

[...]
> Probably 90% of the problem is naming and/or module organization - urllib,
> a module to retrieve the contents given a generic URL, should be just that
> and little else, and protocol-specific knowledge should be accessible from
> protocol-specific modules.

Given how it is now and need for backwards-compat., what do you suggest to
do?  I think putting a pointer in httplib docs (if there's not one there
already) is the best that can be done now.


John



From twister at raplounge.de  Tue Jun 11 15:16:46 2002
From: twister at raplounge.de (Ben Thursfield)
Date: Tue, 11 Jun 2002 21:16:46 +0200
Subject: Newbie question communicating between 2 progs
References: <3D046BF0.DC36542C@raplounge.de> <3D04A938.627EE705@engcorp.com> <3D04E8ED.C0708C2B@raplounge.de> <3D04F4EF.9040800@bgb.cc>
Message-ID: <3D064C9E.FF9B4DAE@raplounge.de>

Don Garrett schrieb:

>    The two programs are only running while the frame they handle is being
> generated.

The one programm is running as long as i want it to, because i am using a while
loop, so that is no problem. So only the one that prints needs to be refreshed. I
want it to behave in a chat window manner, where it frequently prints out the last
25 lines.

> I'm assuming you aren't using some form of server push, which has it's own
> problems.

I actually don't know what a server push is, so I don't think I am using one ;-).

>    However, the best answer to your question is to use cookies. The contents
> should be visible to all pages on the local site (you can control this) so
> they will allow you scripts to communicate with themselves and each other. You
> can store a fair amount in one (I'm not sure what the size limit is), and it
> can be binary.

I thougth that there must be a way of declaring data structers or whatever as extern
in a C manner. This does not seem to be the case, i suppose....


>    A few warnings.
>
>    Some people disable cookies. This can totally break your site.
>
>    All cookies are always uploaded as part of every page request everywhere
> they are visible. This means that you shouldn't make them too large or you
> will slow your site. If they are big, try to restrict cookie visibility to
> just the pages that really need them. You can specify where the cookies are
> visible when you create them. You can even store the state info in a database
> and use the cookie as a key to retreive it.

>
>    When it comes to security, you cannot trust the contents of the cookie. The
> contents need to be validated the same way the values from a form would be. A
> tricky black hat can force the cookie contents to be anything they want.
>

I was thinking about sth. like that actually. Only that I wanted to write the output
into an html file and then let the frame refresh that one. I then thought that there
must be a way to access the data some other way, as i didn't want to make a
directore one can write into, because of security reasons... The cookies way seems
better then. I don't know, but how do normal internet chat programms handle this? Do
they use cookies, too?

>    Two last bits of advice from an old web developer (well as old as a web
> developer can be).
>
>    Try to avoid frames. They cause a lot of problems for both you and for your
> users. For example, they often break cross-browser, they can break bookmarks,
> and they can break printing.
>

I know, on the other hand I didn't know how to do the program in another way :-( .

>    Try to avoid refreshes, especially frequent ones. Someone with a poor link
> will spend more time looking at an empty page that hasn't loaded than they
> will looking at your content.
>

Good point, too.
Thanks for all your help!

> --
> Don Garrett                             http://www.bgb.cc/garrett/
> BGB Consulting                                      garrett at bgb.cc



From chris.gonnerman at newcenturycomputers.net  Wed Jun 19 23:55:47 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Wed, 19 Jun 2002 22:55:47 -0500
Subject: python version?
References: <mailman.1024534466.22643.python-list@python.org> <3D1138EE.FDDCE5DF@engcorp.com>
Message-ID: <00e201c2180e$61910d20$0101010a@local>

----- Original Message ----- 
From: "Peter Hansen" <peter at engcorp.com>


> "Delaney, Timothy" wrote:
> > 
> > > From: George Hester [mailto:hesterloli at hotmail.com]
> > >
> > > All fixed
> > >
> > [snip questionnaire]
> > 
> > Now I'm confused. What does that mean???
> 
> I interpreted it as "I got the point, please leave it alone now."

Huh.  I got "I don't want to see your point, go away."

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net 





From stephen at xemacs.org  Mon Jun  3 02:35:45 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 03 Jun 2002 15:35:45 +0900
Subject: 'for every' and 'for any'
References: <20020526135944.A32690@hishome.net> <20020526091742.A987@unpythonic.net> <mailman.1022567685.7803.python-list@python.org> <slrnaf7af3.8kn.quinn@schilling.ugcs.caltech.edu> <mailman.1022658469.27441.python-list@python.org> <slrnafafkp.eom.quinn@regurgitate.ugcs.caltech.edu> <mailman.1022764258.26310.python-list@python.org> <slrnafd4hv.i49.quinn@regurgitate.ugcs.caltech.edu>
Message-ID: <87wutglvi6.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Quinn" == Quinn Dunkan <quinn at regurgitate.ugcs.caltech.edu> writes:

    Quinn> To me, that's "abstraction".  Logging debugging information
    Quinn> is a very common task, but python doesn't have it built in.
    Quinn> I don't write 'if debug >= whatever: print blah' all the
    Quinn> time and blame python for forcing me to jump through hoops,
    Quinn> I write util.dprint().  I don't have to complain about its
    Quinn> not being built in or wait for 2.3 or 2.2.2 or whatever,
    Quinn> since I can have exactly what I want right now.

Picking an example that suits the point: this is exactly what
PEP-something-or-other (282, in this case) is about.  Understood,
_you_ don't have to wait, and you maybe don't want to.

As I understand it, Oren's point is precisely that you rarely have to
wait very long (at the current stage of Python development) for the
appropriate PEP-something-or-other to show up -- and it's usually
pretty well thought out and will work for most people.

Furthermore, PEPs are strongly encouraged (maybe required?) to provide
an implementation early.  This means that you can use them now, with a
pretty good shot at forward compatibility.


-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
 My nostalgia for Icon makes me forget about any of the bad things.  I don't
have much nostalgia for Perl, so its faults I remember.  Scott Gilbert c.l.py


From jwilhelm at outsourcefinancial.com  Fri Jun 14 13:07:32 2002
From: jwilhelm at outsourcefinancial.com (Joseph Wilhelm)
Date: 14 Jun 2002 10:07:32 -0700
Subject: I feel stoopid ... this code is to verbose
In-Reply-To: <3D0A1791.3060307@mxm.dk>
References: <3D0A1791.3060307@mxm.dk>
Message-ID: <1024074452.1015.12.camel@jwilhelm.ofsloans.com>

On Fri, 2002-06-14 at 09:19, Max M wrote:
<snip> 
> def stringSlicer(string, chunkSize=3):
>      chunkList = []
>      reverseString = list(string)
>      reverseString.reverse()
>      for i in range(0, len(string), chunkSize):
>          chunk = reverseString[i:i+chunkSize]
>          chunk.reverse()
>          chunk = ''.join(chunk)
>          chunkList.append(chunk)
>      chunkList.reverse()
>      return '.'.join(chunkList)
> 
> 
> print stringSlicer('42000000')
> 
>  >>> 40.000.000
> 
> I just find that it's a lot of code for such a little function an it 
> annoys my sense of aestetics. I have tried a lot of different approaches 
> including using zip on a list like ['','','.'], and other weird stuff :-)
> 
> I just cannot seem to find the nice 3-liner I expected from the 
> beginning. Has anybody got a better solution ? I thought somebody might 
> find it a fun exercise. Well I have...
<snip>

How about this:

import string

def stringSlicer( str, chunkSize=3 ):
	parts = list( str )
	for x in range( ( len( parts ) - chunkSize ), 0, -chunkSize ):
		parts.insert( x, "." )
	return string.join( parts, "" )

print stringSlicer( '42000000' )

>>> 42.000.000

--Joseph Wilhelm






From marklists at mceahern.com  Sat Jun 15 10:00:03 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sat, 15 Jun 2002 09:00:03 -0500
Subject: getgrouplist() ?
In-Reply-To: <20020615001750.94759.qmail@web10304.mail.yahoo.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBOENELEAA.marklists@mceahern.com>

> I need a way to get a back a list of supplemental group ids for a
> specified uid.  Something like getgrouplist() in my C standard lib.  I
> don't see a convenient method for this, have I missed it?
> 
> The os module contains getgroups(), but that doesn't allow me to
> specify a uid; it uses only the uid of the current process.

For what it's worth, the grp module is getting some improvements for 2.3:

  http://www.python.org/dev/doc/devel/whatsnew/node8.html

// m
-





From cbbrowne at acm.org  Sat Jun  1 02:32:48 2002
From: cbbrowne at acm.org (Christopher Browne)
Date: 1 Jun 2002 06:32:48 GMT
Subject: Generating unique numbers?
References: <3CF7EBC8.4080802@verio.net> <79b2b0bb.0205312136.5365df9c@posting.google.com>
Message-ID: <ad9pqf$vard4$1@ID-125932.news.dfncis.de>

A long time ago, in a galaxy far, far away, xscottgjunk at yahoo.com (Scott Gilbert) wrote:
> VanL <vlindberg at verio.net> wrote in message news:<3CF7EBC8.4080802 at verio.net>...
>> Is there a good algorithm for generating unique numbers 
>> (ints, specifically) that can be used as object identifiers?
>> 
>> I am discounding the obvious newid += 1 because I will need 
>> to merge different sets, and while I may have some number 
>> overlap, using a simple count for each object would 
>> guarantee that I have the maximum number of clashes.
>> 
>> Specifically, I will be using these numbers as dictionary keys.
>> 
>
> If you're on windows, you could use the pythoncom module CreateGuid
> function.
>
> Basically GUIDs are 128 bit numbers that are pretty much going to be
> unique no matter where or when you call the function.

And GUIDs are basically DCE UUIDs.

<http://packages.debian.org/stable/devel/uuid-dev.html>
<http://www.gnuenterprise.org/doc/uuid-dev/>

I think this is all C stuff; it oughtn't be _too_ difficult to wrap it
into a Python library...
-- 
(concatenate 'string "cbbrowne" "@ntlug.org")
http://www.ntlug.org/~cbbrowne/oses.html
All things are possible except skiing through a revolving door.  
-- DeMara Cabrera 


From whisper at oz.net  Sat Jun 15 03:00:50 2002
From: whisper at oz.net (David LeBlanc)
Date: Sat, 15 Jun 2002 00:00:50 -0700
Subject: MS Word -- finding text
In-Reply-To: <ae58a20e.0206142249.54dfa746@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEOEFADJAA.whisper@oz.net>

Please post some snippets. That's always more likely to get useful
assistance then stating a general problem.

Where are you encountering the unicode decoding problem?

Regards,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Prema
> Sent: Friday, June 14, 2002 23:50
> To: python-list at python.org
> Subject: MS Word -- finding text
>
>
> Hi All !
> Just wondering if anyone has some sample code around how to open a
> document, find a string and return the paragraph(s) containing the
> string.
>
> I have got a good part of the way but running up against Unicode
> decoding
>
> Can post some snippets if needed
> Thanks very much in advance
>
> Kind regards
> Mike
> --
> http://mail.python.org/mailman/listinfo/python-list





From john at jrmstudios.com  Sat Jun 15 05:50:53 2002
From: john at jrmstudios.com (John R. Marshall)
Date: Sat, 15 Jun 2002 04:50:53 -0500
Subject: IDE recommendations??
References: <b9veea.l85.ln@majestix.gallien.de> <3d0b06bd$0$52043$e4fe514c@dreader3.news.xs4all.nl>
Message-ID: <aef2lb$6h5bg$1@ID-71985.news.dfncis.de>

On Saturday 15 June 2002 04:18 am, Boudewijn Rempt wrote:

> At one
> point there was a version of KDevelop that supported Python development,
> but that has disappeared, it seems.

It's (Gideon) still there :-). Instructions for getting it from CVS can be 
found here 
http://www.kdevelop.org/index.html?filename=branches_compiling.html


-- 
John R. Marshall - Web Developer

JRM Studios - http://www.jrmstudios.com
The Hotrodding Network - http://www.hotrodding.net


From gcordova at hebmex.com  Mon Jun  3 16:48:46 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Mon, 3 Jun 2002 15:48:46 -0500 
Subject: Which is fastest dict or list.index() ?
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233010A5777@mail.hebmex.com>

> 
> I need a structure to hold a correspondence between toons and id like
> 
> {'Bugs' : 0, 'Mickey' : 1, 'Babs' : 2, 'Jerry' : 3, 'Tom' : 
> 4, 'Babs' : 5}
> 
> I must be able to alpha sort these toons (without loosing 
> their id), and access a toon from an id or an id from a toon.
> 
> By now a use a dict and a list : the dict gives index from a 
> toon, and the list gives me a toon from an index.
> 
> {'Bugs' : 0, 'Mickey' : 1, 'Babs' : 2, 'Jerry' : 3,
   'Tom : 4, 'Babs' : 5}
> ['Bugs', 'Mickey', 'Babs', 'Jerry', 'Tom', 'Babs']
> 
> (the id is used somewhere else a key to a collection of elements)
> 
> My question is which is the best way (best = quickest and 
> memory safe) to handle this ?
> Should i subclasse from dict or list ?
> Should i consider using only one list with toons (and use 
> their index as id)
> ?
> Should i use 2 dicts ?
> 
> Many thanks to you for helping
> 
> s13.
> 

Hmmm... OK, first, let's analyze your data domain:

"Toon names", which are character strings;
"Toon Name Indexes", which are integer numbers.

That gives us an insight: You'll never have a clash
between a "ToonName" and a "ToonNameIndex".

In such a case (I've used this technique before,
with good results). I'd really recommend you use
a single object, a dictionary, subclassed if you
like, to store all the information.

So, you can say "toondict[5]" and obtain "Babs",
and say "toondict['Babs']" and obtain 5.

Since, now, keys and values are transposed, you need
special methods to obtain all toons and indexes,
so it's trivial to separate all contained data
to find what you need.

This will give you the piece of mind (hah!) that
you're not duplicating functionality.

I'd subclass 'dict' like this:

class ToonDict(dict):
  # All stays the same, except:
  def __setitem__(self, key, value):
    # Special-case item insertion.
    dict.__setitem__(self, key, value)
    dict.__setitem__(self, value, key)

  # Obtain list of data of a certain type.
  def __typed_data(self, *types):
    items = self.values()
    items = filter(lambda i:type(i) in types, items)
    items.sort()
    return items

  # Retrieve a sorted list of toon names.
  def toons(self): return self.__typed_data(str)

  # Retrieve a sorted list of toon name indexes.
  def toon_name_indexes(self): return self.__typed_data(int)


Good luck :-)

-gus




From jepler at unpythonic.net  Sat Jun 29 22:20:57 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Sat, 29 Jun 2002 21:20:57 -0500
Subject: possibly a dumb question
In-Reply-To: <3d1e6251$1_3@nopics.sjc>
References: <3d1e5eff$1_7@nopics.sjc> <3D1E6415.8128F3F8@engcorp.com> <3d1e6251$1_3@nopics.sjc>
Message-ID: <20020629212050.A2822@unpythonic.net>

class foo(object):
    def __new__(*args): return 0

x = foo(0)
print x, type(x)

But if you want the constructor for "foo" to return a value that is not
an instance of "foo", why not let "foo" be a function?  Since both
functions and classes are callables, it's no problem allowing both anywhere
a callable object can be used...

Jeff




From olc at ninti.com  Sat Jun  8 20:38:44 2002
From: olc at ninti.com (Michael Hall)
Date: Sun, 9 Jun 2002 10:08:44 +0930 (CST)
Subject: Removing ^M
Message-ID: <Pine.LNX.4.21.0206091005170.7346-100000@localhost.localdomain>

I am trying remove ^M characters (some kind of newline character) from an
HTML file. I've tried all sorts of string.replace and sed possibilities
but the
things just won't go away. Does anyone have a way of removing such
characters?

TIA 

Mick






From peter at engcorp.com  Sat Jun 29 16:25:52 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 29 Jun 2002 16:25:52 -0400
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com> <mailman.1025336262.12245.python-list@python.org> <3d1dc547.468619140@netnews.attbi.com> <afkv73$f2274$1@ID-139865.news.dfncis.de> <afl1ei$ffa93$1@ID-71831.news.dfncis.de>
Message-ID: <3D1E17D0.5581791E@engcorp.com>

James Kew wrote:
> 
> "Dave Kuhlman" <dkuhlman at rexx.com> wrote in message
> news:afkv73$f2274$1 at ID-139865.news.dfncis.de...
> 
> > candiazoo at attbi.com wrote:
> > > 4) I localized constants I was using in boolean
> > > tests (there were declared at the beginning of the module, I moved
> > > them to the function call just before the loop).
> >
> > Is finding things in globals slower than finding them in locals?
> 
> Yes: you have to look, and fail to find, in the locals before you look in
> the globals.

I'm no expert in this, but that doesn't sound right.  Locals are
indexed, not searched in a dictionary.  The compiler decides if
something is local, not the runtime.  I think you're confusing 
the two-level lookup with the fact that for non-locals, first it
looks in globals, then in the builtins.  (Hope I got that right.)

-Peter


From gustafl at algonet.se  Mon Jun 10 17:25:06 2002
From: gustafl at algonet.se (Gustaf Liljegren)
Date: Mon, 10 Jun 2002 21:25:06 +0000 (UTC)
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
Message-ID: <Xns9229EE722467Fgustafl@195.100.94.182>

I'm making a function that takes one integer and returns a boolean that 
says if the integer matches a number in a certain group of numbers. This 
group consists of single numbers aswell as ranges spanning thousands of 
numbers. I'm looking for an elegant way to write this function.

Thanks in advance!

Gustaf


From aahz at pythoncraft.com  Tue Jun 11 16:50:47 2002
From: aahz at pythoncraft.com (Aahz)
Date: 11 Jun 2002 16:50:47 -0400
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Jun 10)
References: <EE3CDB256E0A231C.577F9FCCC77B96EC.B451C214671092A5@lp.airnews.net> <ae5cce$hpd$1@panix1.panix.com> <mailman.1023824716.32130.python-list@python.org>
Message-ID: <ae5nr7$t6q$1@panix1.panix.com>

In article <mailman.1023824716.32130.python-list at python.org>,
Oleg Broytmann  <phd at phd.pp.ru> wrote:
>On Tue, Jun 11, 2002 at 01:35:10PM -0400, Aahz wrote:
>>>
>>>    Oleg BroytMann posted glue that automatically moves vim users to
>> 
>> Broytman
>
>   It is arguable :) This is transliteration from Russian, so there are too
>many way to do it. When my sister moved to Israel, they wrote in her
>passport "Brojtman"!
>   When I was considering my way of the transliteration I decided to use
>German-like "mann" instead of English-like "man"; after all these are
>German-like surnames. And in programs I am almost always do
>DoubleCapitalization and write "BroytMann"! :)

Oops.  I was pretty sure StudlyCaps was wrong, and I misread what I
searched for on Gooja.  My bad.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From sholden at holdenweb.com  Tue Jun  4 12:01:56 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 4 Jun 2002 12:01:56 -0400
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com>
Message-ID: <9A5L8.232032$%u2.84870@atlpnn01.usenetserver.com>

"John Roth" <johnroth at ameritech.net> wrote in message
news:ufpb3sn6332ub4 at news.supernews.com...
>
> "Huaiyu Zhu" <huaiyu at gauss.almadan.ibm.com> wrote in message
> news:slrnafo7qv.37c.huaiyu at gauss.almadan.ibm.com...
> > Gustavo Cordova <gcordova at hebmex.com> wrote:
>
> >
> > Great explanation!  The problem stems from the fact that the current
> +=
> > operator implements two semantically distinct operations, chosen
> according
> > to some properties of the operands.  This is not unlike the situation
> of the
> > division operator with distinct behaviors on floats and ints.
>
> This is not an issue with '+='. It's an issue with assignment,
> regardless
> of whether it is standalone or combined with some other operation.
>

> And I disagree that assignment implements two distinct operations.
> Assignment always rebinds. Always. In the examples cited, it
> attempted to rebind into a tuple, because that's what was asked
> for when the left hand side had a subscript.
>

>>> a = b = [1,2,3]
>>> a += [4]
>>> a, id(a), b, id(b)
([1, 2, 3, 4], 269472088, [1, 2, 3, 4], 269472088)
>>>

Where did the rebinding take place? ISTM that "a" is still bound to the same
list, which has been modified in place.

> The reason that the examples given at the top of the thread fail
> in such strange ways is simply that '+=' does not check whether
> the assignment is possible before doing the addition (or concatination,
> in the specific cases cited.)
>
That's because a type's implementation of augmented operations is allowed to
decide whether to rebind or update in place. Seems to me the exception is
caused because the tuple's hash would be changed, though I'm not familiar
with the implementation detail. That seemed reasonable to me.

> To see this, you could try the first example, using explicit '+' and
> '='. I suspect it would fail in exactly the same way, with exactly
> the same side effect.
>
Clearly an explicit assignment to a tuple element is always going to fail,
so I don't really see what this would tell you.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------






From fredrik at pythonware.com  Thu Jun 20 11:49:05 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 20 Jun 2002 15:49:05 GMT
Subject: Access to XML dataBase
References: <mailman.1024579946.6526.python-list@python.org>
Message-ID: <RPmQ8.44297$n4.10329719@newsc.telia.net>

Marta Gutierre wrote:
> I tried sothing like:
>
> import xmlrpclib
> client=xmlrpclib.Server("http://localhost:4080")
> try:
>      client.listCollections("db")
> except xmlrpclib.Fault,faultobj:
>     print "ERROR", faultobj
> ERROR <Fault 0: 'java.lang.Exception: RPC handler object not
> found for "listCollections": no default handler registered.'>
>
> Any idea of what does it mean ???

the Java server is trying to tell you that there is no
"listCollection" handler.

> I can't find good documentation of XML-RPC in Python

it's not a Python problem, really...

</F>




From kseehof at neuralintegrator.com  Sat Jun  8 20:18:25 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Sat, 08 Jun 2002 17:18:25 -0700
Subject: How could i get execution times ?
In-Reply-To: <adkjro$12183k$1@ID-146704.news.dfncis.de>
Message-ID: <EFEFIHAINIINNELLHFAIIEKKCIAA.kseehof@neuralintegrator.com>

> I'm looking for a way to get execution time at different points 
> of a program, by now i use the
> code below, but i've seen in some post that time.clock() could be 
> a better way to measure this,
> 
> So my question, is how you pythoners do it ??? and to understand _why_ ?

>>> for x in range(10):
...    print time.time()
    
1023578958.14
1023578958.14
1023578958.14
1023578958.14
1023578958.14
1023578958.14
1023578958.14
1023578958.14
1023578958.14
1023578958.14

>>> for x in range(10):
...    print time.clock()
    
88.9835171558
88.9849477866
88.9862024171
88.9876648955
88.9889245546
88.9901624231
88.9913986155
88.9926339697
88.9938701621
88.9951130592

>>> for x in range(10):
...    time.sleep(.01)
...    print time.time()
    
1023579557.1
1023579557.1
1023579557.16
1023579557.16
1023579557.16
1023579557.16
1023579557.16
1023579557.21
1023579557.21
1023579557.21

>>> pprint.pprint([time.clock() for i in range(10)])
[12.151166630349151,
 12.151180877989908,
 12.151191773244607,
 12.151202668499304,
 12.151213563754002,
 12.151224459008699,
 12.151235354263397,
 12.151246249518094,
 12.151258820965822,
 12.15126971622052]

In other words, time.clock() is far more precise for elapsed
time (the above example demonstrates precision on the order
of microseconds -- more than adequate for profiling python
code).  But time.time() apparently has a granularity on the
order of a tenth of a second (on windows), which is fine for
profiling slower applications like Microsoft Outlook :-)

I believe that the difference between the corresponding C
library functions is due to there being two different clocks
in the hardware.  One clock is very precise, and drives the CPU
but only knows about elapsed time, (this clock is only on when
the CPU is on).  The other clock is essential a digital watch
plugged into your computer (it runs on a battery, just like a
watch, when the CPU is off).

I hope this answers your questions.

- Ken Seehof






From shalehperry at attbi.com  Fri Jun  7 10:41:28 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 07 Jun 2002 07:41:28 -0700 (PDT)
Subject: question on pattern
In-Reply-To: <3D007E96.3010505@rmb.co.za>
Message-ID: <XFMail.20020607074128.shalehperry@attbi.com>

On 07-Jun-2002 Ulrich Pfisterer wrote:
> I want to do a pattern match on a string that spans multiple lines (c++ 
> style comments, re.match('(/\*.*?\*/)', filestring)
> If I use the re.findall(pattern, string) method it only finds those 
> matches that are on a single line. If I use re.match(pattern, string, 
> DOTALL) it spans multiple lines but only finds the first match.
> Any help on this one would be greatly appreciated.
> 

match only matches once you can either use a loop or use I believe findall() or
something similar.




From romberg at smaug.fsl.noaa.gov  Fri Jun  7 13:26:48 2002
From: romberg at smaug.fsl.noaa.gov (Mike Romberg)
Date: 07 Jun 2002 11:26:48 -0600
Subject: Medium to Large Scale Python Deployments
References: <mailman.1023460091.14067.python-list@python.org>
Message-ID: <ul3k7pbuhif.fsf@smaug.fsl.noaa.gov>

>>>>> " " == Michael Chermside <mcherm at destiny.com> writes:

     > Folks, I think this is a pretty useful thing to collect.  In
     > fact, it might be really nice to collect a list of large
     > (successful) Python projects and post it somewhere so the
     > Python community could use it for advocacy.

     > So do me a favor... if you have (or know of) a large Python
     > project which you think our community ought to be bragging
     > about, please either post it to this thread or email it to
     > me.

  I'm working on a project which is is a mix of C++ and python.
Currntly the python part of it is:

python ~/pycount.py -R "*.py"
   lines    code     doc comment   blank  file
     35       6       0      20       9 ./data/textFiles/GFECONFIG/Gweight.py
   1390     264       0     937     189 ./data/textFiles/GFECONFIG/gfeConfig.py

[snip]

  134062   88680    3160   24281   17941  total

  The code is for a weather forecast system being developed for the
national weather service:

http://www-md.fsl.noaa.gov/eft/

Mike Romberg (romberg at fsl.noaa.gov)


From thehaas at binary.net  Mon Jun 10 19:10:35 2002
From: thehaas at binary.net (thehaas at binary.net)
Date: Mon, 10 Jun 2002 23:10:35 GMT
Subject: wxPlotCanvas documentation?
References: <mailman.1023733395.5699.python-list@python.org>
Message-ID: <LlaN8.52033$ks5.2050662@news2.east.cox.net>

> 3. How do I find anything in the documentation?

> My inability to find my wxButt in the wxDark with both
> wxHands (with or without a wxFlashlight) is not
> limited to PlotCanvas, so I'm hoping that the problem
> is on my end.  The wxPython website, as far as I can
> tell, only has one documentation link, and it's a link
> to wxWindows, not the wxPython objects that wrapper
> it.  There's a sample code section, but every last one
> of the examples is marked "To Be Written When I Have
> More Time."

The wxPython documentation link really does goto the official wxWindows
docs, but it has notes where wxWindows and wxPython (and wxPerl, for that
matter) differ.  Not ideal, mind you, but it is there.

Did you try the wxWiki?  I've found it the best place to look first
(though it isn't complete)

http://wiki.wxpython.org/

Also, the wxPython users mailing list is a great place . . . Robin Dunn,
the man in charge of wxPython, is great about answering questions,
treating people with respect, and being honest about wx's short comings.

I love wxPython, but the initial learning curve is quite steep.  Once you
get past the very initial stage, though, it grows on you.

-- mikeh



From op73418 at mail.telepac.pt  Wed Jun 26 14:06:13 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 26 Jun 2002 19:06:13 +0100
Subject: question on properties
References: <p4pjhuc69jjqs0vvfkkmqd37rr52cq9qv3@4ax.com>
Message-ID: <ii0khu8v7v1c7p3p6ipe1nspgcmnpl2kqo@4ax.com>

On Wed, 26 Jun 2002 17:04:32 +0100, Gon?alo Rodrigues
<op73418 at mail.telepac.pt> wrote:

>Hi,
>
>Suppose in a class Klass I have a property prop. Suppose also that I
>have
>
>class AnotherKlass(Klass):
>    ...
>
>I can override property prop in AnotherKlass by just redefining it, but
>what if I want to acess the base class prop how do I do it?
>
>TIA and all the best,
>Gon?alo Rodrigues

Got it! Use (for getting):

Klass.prop.__get__(self)

All the best,
Gon?alo Rodrigues


From bkelley at wi.mit.edu  Mon Jun  3 11:02:50 2002
From: bkelley at wi.mit.edu (Brian Kelley)
Date: Mon, 03 Jun 2002 11:02:50 -0400
Subject: Graph distances, how can I speed it up?
References: <33803989.0206030235.4134d33e@posting.google.com> <FyKK8.39522$4i.4685714@bin2.nnrp.aus1.giganews.com>
Message-ID: <3CFB851A.80102@wi.mit.edu>

While a C or C++ implementation would be prefereable, there are more 
efficient algorithms for shortest distance between nodes. 
Floyd-Warshall comes to mind.  (I think Andrew Dalke helped me with this 
one a while ago just to give appropriate credit)

import Numeric
# Finds the all-pairs minimum distance between two points.
#
# Initial matrix must be a symmetric "weight" matrix w[i][j] where
# w[i][i] == 0 and w[i][j] is the direct distance between i and j.
# This is usually 1 but can be weighted for importance.
#
# Must have a symmetric matrix for this implementation

def floyd_warshall(w):
     assert(len(w.shape) == 2) # must be a 2D array
     N = w.shape[0]
     assert(N == w.shape[1]) # must be a square 2D array

     k = 0
     dold = w # destroys contents of w
     # (Use this if you don't want to destroy)
     #dold = Numeric.array(w)

     dnew = Numeric.zeros((N,N))

     for k in range(1, N):
         for i in range(N):
	    #  We can do this simplification because of symmetry
             for j in range(i+1, N):
		x = min(dold[i][j], dold[i][k] + dold[k][j])
                 dnew[i][j] = x
                 dnew[j][i] = x

         dnew, dold = dold, dnew
     return dold

Brian Kelley
Whitehead Institute for Biomedical Research





From jadestar at idiom.com  Mon Jun 17 04:29:30 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 17 Jun 2002 08:29:30 GMT
Subject: Which is fastest dict or list.index() ?
References: <adaih2$vtb8c$1@ID-146704.news.dfncis.de> <adcms5$vm6j0$1@ID-146704.news.dfncis.de>
Message-ID: <aek6la$n07$1@news.idiom.com>

Shagshag13 <shagshag13 at yahoo.fr> wrote:

> Thanks !!! (i use some tricks i find in seqdict)

>> If you want a skiplist version, let me know and I'll send it.
>> Raymond Hettinger

> I'm sorry, but i don't understand what you mean by a "skiplist version"

> So here is how i do it now, i think i miss to explain that i must keep some
> kind of id (which i use as a key in another table, and i want to keep 
> this other table indexed by integer instead of words to save mermory). 
> <- thoug i don't know if this is a real saving...

 I'd probably use a simple list of tuples in a class.  The class
 would be responsible for maintaining the sorted order of the list
 via the "bisect" module's insort function and it would provide
 a short binary search function to find and return any element 
 of the list.  

 Actually almost any technique is likely to be overkill for this
 particular problem domain.  There are max. a few thousand toons
 and any modern computer could could simply brute force them using
 x=dictionaries.items(); x.sort().

> i'm huge newbie but i feel python really pleasant, great, etc... (even if i
> had lots of trouble
> with memory run out... on a solaris system with  1024mo...)

> i post for comments or advices as i'm not a skilled pythoner...

> thanks in advance for reading and helping

> class IKeys:
>  def __init__(self):
>   self._size = 0
>   self._term = {}  # element -> int key - ex: {a : 0, b : 2, c : 1}
>   self._iKeys = [] # contains elements, int key -> element - ex: [a, c, b]

>  def __str__(self):
>   s = 'iKey (' + "\n"
>   s += str(self._term) + "\n"
>   s += str(self._iKeys) + "\n"
>   s += str(self._size) + " # size)\n"
>   return s

>  def __len__(self):
>   return self._size

>  def __getitem__(self, pos):
>   return self._iKeys[pos]

>  def index(self, element):
>   return self._term[element]

>  def add(self, element):
>   if self._term.has_key(element):
>    return self._term[element]
>   self._iKeys.append(element)
>   self._term[element] = self._size
>   self._size += 1
>   return (self._size - 1)

>  def keys(self):
>   return self._iKeys

>  def get_size(self):
>   return self._size

> which i use like this :

>>>> toons = IKeys()
>>>> for t in ['Bugs', 'Mickey', 'Babs', 'Jerry', 'Tom', 'Babs']:
>  toons.add(t)
> 0
> 1
> 2
> 3
> 4
> 2  # already in toons so we don't had it, we get id
>>>> print toons
> iKey (
> {'Mickey': 1, 'Babs': 2, 'Tom': 4, 'Bugs': 0, 'Jerry': 3}
> ['Babs', 'Bugs', 'Jerry', 'Mickey', 'Tom']
> 5 # size)
>>>> t[2]
> 'b'
>>>> toons[2]
> 'Babs'
>>>> toons.index('Jerry')
> 3
>>>> tmp = toons.keys()
>>>> tmp.sort()
>>>> for t in tmp: # here it's sorted
>  i = toons.index(t)
>  print "Toons [%s] has id [%s]" % (t, i)
> Toons [Babs] has id [2]
> Toons [Bugs] has id [0]
> Toons [Jerry] has id [3]
> Toons [Mickey] has id [1]
> Toons [Tom] has id [4]






From aahz at pythoncraft.com  Mon Jun 17 14:44:29 2002
From: aahz at pythoncraft.com (Aahz)
Date: 17 Jun 2002 14:44:29 -0400
Subject: If you use Python -OO is it a permanent condition??
References: <rXpP8.32471$ks6.25445@nwrddc02.gnilink.net>
Message-ID: <aelamd$fo1$1@panix1.panix.com>

In article <rXpP8.32471$ks6.25445 at nwrddc02.gnilink.net>,
Jeff Sasmor <jeff at sasmor.com> wrote:
>
>So  I tried:
>
>python -OO P1.pyo
>
>and it works just fine.  Does this mean that the interpreter must
>always be started with -OO to execute files compiled with -OO?  I can't
>imagine why this would be the case, but it seems to be -- at least as
>far as I can see from my little bit of experimentation.

Python doesn't search for .pyo files unless it is running in optimized
mode.  Try setting the PYTHONOPTIMIZE env var.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From daves_spam_dodging_account at yahoo.com  Mon Jun 10 14:22:31 2002
From: daves_spam_dodging_account at yahoo.com (David Brady)
Date: Mon, 10 Jun 2002 11:22:31 -0700 (PDT)
Subject: wxPlotCanvas documentation?
Message-ID: <20020610182231.26180.qmail@web21108.mail.yahoo.com>

I have searched the wxPython and wxWindows
documentation, peered through the wxdemo files, and
searched the web in the hopes of finding the answers
to my questions, but to no avail.

I would like to write a simple graphing calculator
program with wxPython.

1. How do I create a wxPlotCanvas in a wxFrame window?

I can see in wxPython/demo/wxPlotCanvas.py  how to
create a PlotCanvas object if I wanted, say, to run my
graphing calculator inside the wxDemo framework, but I
don't.  How do I instantiate this guy and stick him in
a wxFrame class of my own deriving?

2. How do I manipulate wxPlotCanvas?

>From the demo code I see there are PolyMarker and
PolyLine objects, as well as a PlotGraphics function. 
However, since I can find exactly zero documentation
on PlotCanvas, this is all I know about this object.

3. How do I find anything in the documentation?

My inability to find my wxButt in the wxDark with both
wxHands (with or without a wxFlashlight) is not
limited to PlotCanvas, so I'm hoping that the problem
is on my end.  The wxPython website, as far as I can
tell, only has one documentation link, and it's a link
to wxWindows, not the wxPython objects that wrapper
it.  There's a sample code section, but every last one
of the examples is marked "To Be Written When I Have
More Time."

The wxPython homepage proudly shows a long list of
people who claim that wxPython is a Java and/or
TkInter killer.  All I want to know is, how the hell
did these people figure out how to use it?

I'm not complaining about wxPython, mind you.  What
I've gotten to work makes me believe that it really is
wonderful.  But I'm at wit's end here trying to find
out how to do anything that isn't already in a
tutorial.  Like get the client rect of a wxFrame
object.  Like find out what events EVT_PAINT might
throw.  Like use a wxBitmap to double-buffer a wxDC. 
Like do anything beyond the first four tutorial pages.

HELLLLLP!

(thanks)

-dB

=====
David Brady
daves_spam_dodging_account at yahoo.com
I'm feeling very surreal today... or *AM* I?

__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com




From aoeu at go.com  Mon Jun 17 06:32:32 2002
From: aoeu at go.com (Thinkit)
Date: 17 Jun 2002 03:32:32 -0700
Subject: you guys needs a logo
References: <7fe97cc4.0206142325.4df3a174@posting.google.com>
Message-ID: <3eeda89d.0206170232.675596b8@posting.google.com>

xah at xahlee.org (Xah Lee) wrote in message news:<7fe97cc4.0206142325.4df3a174 at posting.google.com>...
> you guys need a logo.
> 
> the light-bulb is quite lame as far as logo goes.
> 
> I suggest that the O'Reilly's python book cover a good choice. It
> should not be something funny and cute.
> 
> The MacPython's some 16 Ton weight block is also a good logo.
> 
> Perhaps you guys can run a logo competition.
> 
>  Xah
>  xah at xahlee.org
>  http://xahlee.org/UnixResource_dir/complang.html#python

How about a penguin masturbating?  To show what else besides python linux males do.


From jepler at unpythonic.net  Fri Jun 14 14:15:54 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Fri, 14 Jun 2002 13:15:54 -0500
Subject: I feel stoopid ... this code is to verbose
In-Reply-To: <3D0A1791.3060307@mxm.dk>
References: <3D0A1791.3060307@mxm.dk>
Message-ID: <20020614181549.GF30070@unpythonic.net>

On Fri, Jun 14, 2002 at 06:19:29PM +0200, Max M wrote:
> Hmm ... I am working on a problem. In Danish we have a number format 
> that looks like:
> 
> 42.000.000,00 which is 42 millions

See the 'locale' module, functions 'format' and (internal) _group.
If you can call locale.setlocale() to one which requires these
separators, it should take care of the rest on calls to .format().

>>> locale.setlocale(locale.LC_ALL, "da_DK")
'da_DK'
>>> locale.format("%.2f", 42000000.0, 1)
'42.000.000,00'

Jeff




From geoff at gerrietts.net  Sat Jun  1 00:55:08 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Fri, 31 May 2002 21:55:08 -0700
Subject: Extracting data from HTML
In-Reply-To: <1022883953.6150.5802.camel@localhost>
References: <82096df4.0205311152.5891a17b@posting.google.com> <1022883953.6150.5802.camel@localhost>
Message-ID: <20020601045508.GF13794@isis.gerrietts.net>

Quoting Ian Bicking (ianb at colorstudy.com):
> On Fri, 2002-05-31 at 14:52, Hazel wrote:
> > how do i write a program that
> > will extract info from an HTML and print
> > of a list of TV programmes, its Time, and Duration
> > using urllib?
> 
> You can get the page with urllib.  You can use htmllib to parse it, but
> I often find that regular expressions (the re module) are an easier way
> -- since you aren't looking for specific markup, but specific
> expressions.  You'll get lots of false negatives (and positives), but
> when you are parsing a page that isn't meant to be parsed (like most web
> pages) no technique is perfect.

Definitely agree with this sentiment.

I'll go a step farther, and do a little compare/contrast.

Once upon a time, I wanted to grab data from the
weatherunderground.com website. I know there are lots of better ways
to go about getting this information, these days, but I was not so
well-informed back then.

So I wanted to grab this information, and I tried using regular
expressions to mangle the page. But truthfully, it was just too hard
to do. I could guess about where in the file the table with all the
info would appear, but getting a regular expression that was inclusive
enough to catch all the quirks, yet exclusionary enough to filter out
all the other embedded tables, proved a very large challenge.

That's when the idea of a parser made a lot of sense.

I could push the whole page through a parser, looking for one
particular phrase in a <TH> element, and from that point forward, map
<TH> elements to <TD> elements effectively. It became a very simple
exercise, because I knew how to find that info.

But as Ian rightly points out, htmllib and a real parser can be very
heavy if you're just looking to grab unformatted info -- or if you
can't rely on the formatting to be reliable.

Both techniques are worth knowing -- but better than either would be
finding a way to get the information you're after via XML-RPC or some
other protocol that's designed to carry data rather than rendering
instructions.

Best of luck,
--G.

-- 
Geoff Gerrietts             "If life were measured by accomplishments, 
<geoff at gerrietts net>     most of us would die in infancy." 
http://www.gerrietts.net/       --A.P. Gouthey




From phd at phd.pp.ru  Thu Jun  6 09:08:47 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 6 Jun 2002 17:08:47 +0400
Subject: vim, compile python
Message-ID: <20020606170847.A7683@phd.pp.ru>

Hello!

For those who use vim editor.

   Put this into your .vimrc:

" Set options for python files
autocmd FileType python set autoindent smartindent
   \ cinwords=class,def,elif,else,except,finally,for,if,try,while
   \ makeprg=compyle4vim.py
   \ errorformat=%E\ \ File\ \"%f\"\\,\ line\ %l\\,\ column\ %c,%C%m |
   \ execute "autocmd BufWritePost " . expand("%") . " call DoPython()"

" Compile (clearing *.cgi[co] files after compilation)
" and if it is script, make it executable
function DoPython()
   !compyle %
   if expand("%:e") != "py"
      !rm -f %[co]
   endif
   if getline(1) =~ "^#!"
      !chmod +x %
   endif
endfunction

   The compyle and compyle4vim.py programs are attached.

   Now comments. The first line (autocmd) does not need any explanation. In the
second line there are listed keywords after which <Enter> not only opens
new line but also does indent.
   The third and fourth lines give the name of makeprg (the program for the
:make command) and the format of its error output. If you have a python
program with syntax errors open the program in vim and execute the command
:make %
   vim will run makeprg (compyle4vim.py) and parse its output, then put
cursor to the line and column with the error and report the error on the
status line.
   The fifth line teaches vim to call DoPython function upon saving the
file. The function runs compyle (shell script that uses py_compile.py), and
if the file is a script (starts with #!) makes it executable.

   The compyle script is useful for all users, not only for the users of
vim, and probably those who use Emacs could use compyle4vim.py :)

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.
-------------- next part --------------
#! /usr/local/bin/python -O


import sys


filename = sys.argv[1]
infile = open(filename, 'r')
codestring = infile.read()
infile.close()

try:
   compile(codestring, filename, "exec")
except SyntaxError, detail:
   pass
else:
   sys.exit()

msg, (_fname, lineno, offset, line) = detail

sys.stderr.write("""  File "%s", line %d, column %d
SyntaxError: %s
""" % (filename, lineno, offset, msg))
-------------- next part --------------
#! /bin/sh

if [ "$1" = "-1" ]; then
   only_one=1
   shift
fi

if [ -z "$1" ]; then
   echo "Usage: compyle [-1] file.py..."
   exit 1
fi

TEMPLATE="from py_compile import compile; compile('"

if [ "$only_one" = "1" ]; then
   pgm=$TEMPLATE$1"')"
   python -c "$pgm" || exit 1
else
   for file in "$@"; do
      pgm=$TEMPLATE$file"')"
      python -c   "$pgm" || exit 1
      python -OOc "$pgm" || exit 1
   done
fi

#pgm="$HOME/lib/python/com.py"
#python $pgm "$@" && python -O $pgm "$@"

From peter at engcorp.com  Sat Jun 29 22:30:27 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 29 Jun 2002 22:30:27 -0400
Subject: possibly a dumb question
References: <3d1e5eff$1_7@nopics.sjc> <3D1E6415.8128F3F8@engcorp.com> <3d1e6251$1_3@nopics.sjc>
Message-ID: <3D1E6D43.75D2820D@engcorp.com>

Adonis wrote:
> 
> sure:
> 
> class foo:
>     def __init__(self, value):
>         return value
> x = foo(0)
> print x ;yeilds 0
> 
> i know the code provided is wrong, but its the general idea.

Ahh, you simply cannot do that.  The __init__() method is
defined to return a reference to the object which it initializes
and you can't change that.  Note the error message given:

  TypeError: __init__() should return None

Anyway, you really don't need to do this, since in your other
post you give your goal (sort of).  I'll paste the message here:

"""an addition:
   why not use a function blah blah, yes, but this is going to be 
   a modularized project, essentially i want to load some user made 
   modules which will be classes and will be loaded into python 
   dynamically. like a memory resident cgi server skipping the use of 
   pipes."""

Okay, trying again:  The above is still not quite correct, because
"modules which will be classes" is wrong.  I guess you mean the modules
will contain class definitions?  In that case, you still need to 
create instances of the classes (like the "x = foo()" part above)
and then call methods of the classes.

Can you clarify further?  What you're trying to accomplish is not
hard (I believe... once we figure out exactly what it is) but you
are starting off on the wrong foot.

-Peter


From gh_pythonlist at gmx.de  Tue Jun 18 18:10:51 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 19 Jun 2002 00:10:51 +0200
Subject: how tranform in String??
In-Reply-To: <3D0F9DAE000002DE@www.zipmail.com.br>
References: <3D0F9DAE000002DE@www.zipmail.com.br>
Message-ID: <20020618221051.GA1879@lilith.my-fqdn.de>

* jubafre at zipmail.com.br <jubafre at zipmail.com.br> [2002-06-18 18:49 -0300]:
> i have s='xavante brasil', i want to transform in
> 
> s='xavante', 'brasil' with two indexes
> 
> how i make this???????????

Check out the split method of strings in the docs. Using it works like:

    s = 'xavante brasil'.split()

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 25.0 ?C      Wind: 1.0 m/s




From sholden at holdenweb.com  Mon Jun  3 16:03:30 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 3 Jun 2002 16:03:30 -0400
Subject: thread module under Windows
References: <59d8a5f6.0206030730.1be14ffb@posting.google.com> <3CFB9030.4030802@attbi.com>
Message-ID: <V0QK8.197551$%u2.178262@atlpnn01.usenetserver.com>

"Pete Shinners" <shredwheat at attbi.com> wrote in message
news:3CFB9030.4030802 at attbi.com...
> Dmitri wrote:
> > I have been searching the conference for the simple question like "
> > where the thread module is " and got the answer that it's inside the
> > python if the threading support is on.
>
> Thread support defaults to "on" for all python compiles. all windows users
> will have thread support, unless they have specifically compiled there own
> without. an easy way to test is to just try it..
>
[ ... ]
>
Erm, isn't cygwin the exception here, due to an occasional buglet which
Jason hasn't yet finally squashed? I semm to remember I had to explicitly
rebuild my cygwin python to get threading.

regards
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From emile at fenx.com  Tue Jun  4 00:54:09 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 04 Jun 2002 04:54:09 GMT
Subject: Please help with script...
References: <adheqc$nhn$1@netnews.upenn.edu>
Message-ID: <QJXK8.103614$cQ3.2146@sccrnsc01>

"Topy" <hobiewon at hotmail.com> wrote in message
news:adheqc$nhn$1 at netnews.upenn.edu...
> Hey all,
> I'm having trouble with a problem I have with this python program.
It's
> kind of long, so if anyone has the patience, I would gladly appreciate
your
> time(the programming is fairly simple I think.)  Anyway, the problem I
need

s/problem I nees/homework assignment/

Ask again when you've addressed the problem and have a specific
implementation issue or question.

Learning to ask != learning.

--

Emile van Sebille
emile at fenx.com

---------



From pyth at devel.trillke.net  Wed Jun 19 16:20:56 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 19 Jun 2002 22:20:56 +0200
Subject: python version?
In-Reply-To: <slrnah1l0i.6o5.huaiyu@gauss.almadan.ibm.com>; from huaiyu@gauss.almadan.ibm.com.trillke.net on Wed, Jun 19, 2002 at 06:57:22PM +0000
References: <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com> <3D1010C3.C321F8B0@engcorp.com> <raXP8.47926$GY.15260255@twister.nyroc.rr.com> <3D107948.B3FEE4BB@engcorp.com> <CT%P8.33113$uk2.13710586@twister.nyroc.rr.com> <slrnah1l0i.6o5.huaiyu@gauss.almadan.ibm.com>
Message-ID: <20020619222056.K15079@prim.han.de>

Huaiyu Zhu wrote:
> Being able to communicate effectively in Usenet may be an important skill
> for your students to master.  You may want to teach them some lessons you
> have learned yourself.  Here's a mini multiple choice questionair to get you
> started:

> ...

wow! 

this monster-thread begins to be quite amusing (hopefully for everybody!)

    holger





From fperez528 at yahoo.com  Wed Jun 19 15:35:35 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 19 Jun 2002 13:35:35 -0600
Subject: Problems with mayavi, pyvtk
References: <7396d2b2.0206191025.3351a888@posting.google.com>
Message-ID: <aeqme8$avd$1@peabody.colorado.edu>

Lemniscate wrote:

> 
> I just downloaded and installed the latest versions of pyvtk and
> mayavi.??However,?when?I?try?to?follow?the?instructions?at
> http://mayavi.sourceforge.net/docs/guide/c827.html#VIZ-DATA for using
> mayavi, I get the ImportError as below.??For?the?like?of?me,?I?can't
> seem to figure out why I don't have any vtkpython modules anywhere
> (doing searches on the internet for the module hasn't been a big help
> so far

Well, google's second hit on 'vtkpython' is 
http://basic.netmeg.net/godzilla/, which sounds like what you want. At least 
it should be a good start.

Cheers,

f


From uwe at rocksport.de  Fri Jun  7 05:56:22 2002
From: uwe at rocksport.de (Uwe Schmitt)
Date: 7 Jun 2002 09:56:22 GMT
Subject: problem with distutil....
Message-ID: <adq006$e29eb$4@hades.rz.uni-sb.de>

Hi, I use the following setup.py for generating a distribution:

    from distutils.core import setup
    setup(name="graph",
	  version="1.0", 
	  author="Uwe Schmitt",
	  author_email="uwe.schmitt at procoders.net",
	  url="http://www.procoders.net",
	  py_modules=["graph"])

The module to distribute is graph.py and is located in the same
directory as setup.py.
When I do "python setup.py sdist" a dist-folder and a .tar.gz-file
are generated. But when I unpack this tar.gz, I cant find the module
graph.py...

Greetings, Uwe

-- 
Dr. rer. nat. Uwe Schmitt      ICQ# 159647634   Uwe.Schmitt at num.uni-sb.de


From jdhunter at nitace.bsd.uchicago.edu  Fri Jun 28 16:14:50 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Fri, 28 Jun 2002 15:14:50 -0500
Subject: get word base
Message-ID: <m24rfn88jp.fsf@mother.paradise.lost>

I would like to be able to get the root/base of a word by stripping
off plurals, gerund endings, participle endings etc...  Here is a
totally naive first attempt that gets it right sometimes:

import re

rgx = re.compile( '(\w+?)(?:ing|ed|es|s)')

def get_base(word):

    m = rgx.match(word)
    if m:
        return m.group(1)
    else:
        return word

words = ['hello', 'taxes', 'thoughts', 'walked', 'rakes']

for word in words:
    print word, get_base(word)

Produces the following output
> python get_baseword.py
hello hello
taxes tax
thoughts thought
walked walk
rakes rak


I can think of a few things to do to refine this, but before I forge
ahead, I wanted to solicit advice.

Thanks,
John Hunter


From candiazoo at attbi.com  Sat Jun 29 21:32:03 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Sun, 30 Jun 2002 01:32:03 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com> <mailman.1025336262.12245.python-list@python.org> <3d1dc547.468619140@netnews.attbi.com> <afkv73$f2274$1@ID-139865.news.dfncis.de> <OxnT8.45616$n4.10835602@newsc.telia.net>
Message-ID: <3d1e5f28.508011921@netnews.attbi.com>

Oops!  Mental typo.

:)

Either way, the few changes I made really made a difference.  There is no
escaping the hit I am taking with my awkward queries to the Oracle database that
are gathering additional data... but the other changes which affected the code
in the loop really paid off!

Mike

On Sat, 29 Jun 2002 19:05:18 GMT, "Fredrik Lundh" <fredrik at pythonware.com>
wrote:

>Dave Kuhlman wrote:
>
>"from time import asctime", I hope.



From kalle at lysator.liu.se  Thu Jun 27 16:31:19 2002
From: kalle at lysator.liu.se (Kalle Svensson)
Date: Thu, 27 Jun 2002 22:31:19 +0200
Subject: self destruction :)
In-Reply-To: <XFMail.20020627131351.shalehperry@attbi.com>
References: <f4KS8.1326$qq1.14832@castor.casema.net> <XFMail.20020627131351.shalehperry@attbi.com>
Message-ID: <20020627203119.GH11208@i92.ryd.student.liu.se>

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

[Guyon Mor?e]
> > i've tried: 'del self', but somehow it didn't work. afterwards it should
> > also _not_ be in the list anymore.

[Sean 'Shaleh' Perry]
> mylist = [obj1, obj2, ....]
> del obj1 # mylist still has a reference to obj1 so
>          # the del simply decrements the ref count

By the way, this program

- ----- 8< -----
import sys
allobjects = []

class C:
    def __init__(self):
        print "Before append, RC: %s." % (sys.getrefcount(self) - 2)
        allobjects.append(self)
        print "After append, RC: %s." % (sys.getrefcount(self) - 2)

    def __del__(self):
        print "In __del__, RC: %s." % (sys.getrefcount(self) - 2)

    def delete(self):
        print "Before remove, RC: %s." % (sys.getrefcount(self) - 2)
        allobjects.remove(self)
        print "After remove, RC: %s." % (sys.getrefcount(self) - 2)

print "Creating..."
C()
print
print "Deleting..."
allobjects[0].delete()
print
print "Exiting..."
- ----- >8 -----

prints

- ----- 8< -----
Creating...
Before append, RC: 3.
After append, RC: 4.

Deleting...
Before remove, RC: 2.
After remove, RC: 1.
In __del__, RC: 3.

Exiting...
- ----- >8 -----

when run.  Where do the extra references in __init__ and __del__ come
from?

Peace,
  Kalle
- -- 
Kalle Svensson, http://www.juckapan.org/~kalle/
Student, root and saint in the Church of Emacs.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)
Comment: Processed by Mailcrypt 3.5.6 <http://mailcrypt.sourceforge.net/>

iD8DBQE9G3YMdNeA1787sd0RAqHiAJ9Fmja/KUvq08+qls9OZEEHLWrEbACfXWIW
URLMCDEABgB1pkvj1eYxV6Q=
=APuL
-----END PGP SIGNATURE-----




From shalehperry at attbi.com  Fri Jun 28 12:10:24 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 28 Jun 2002 09:10:24 -0700 (PDT)
Subject: Can i write chating system using python
In-Reply-To: <3d1c8661$1_1@news.iprimus.com.au>
Message-ID: <XFMail.20020628091024.shalehperry@attbi.com>

On 28-Jun-2002 tritran wrote:
> Can anyone provide me the website or resource to create the chating system
> or instance messeger using python.
> 

just go to google and do some searching.  There are irc clients, jabber
clients, talk clients, etc in python.  Plus there is a full network
implementation so you can invent your own protocol and client/servers as well.




From christophe.delord at free.fr  Fri Jun 14 13:05:35 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Fri, 14 Jun 2002 19:05:35 +0200
Subject: I feel stoopid ... this code is to verbose
References: <3D0A1791.3060307@mxm.dk>
	<20020614183945.17cc73f7.christophe.delord@free.fr>
	<20020614184908.5a174d7a.christophe.delord@free.fr>
Message-ID: <20020614190535.62ae384c.christophe.delord@free.fr>

On Fri, 14 Jun 2002 18:49:08 +0200
Christophe Delord <christophe.delord at free.fr> wrote:

> 
> Here is a 3-liner ;-)
> 
> 
> > > I just cannot seem to find the nice 3-liner I expected from the 
> > > beginning. Has anybody got a better solution ? I thought somebody might 
> > > find it a fun exercise. Well I have...
> 
> def stringSlicer(st, chunkSize=3):
> 	def slc(st): return st and slc(st[:-chunkSize])+[st[-chunkSize:]] or []
> 	return ".".join(slc(st))
> 
> 
> 

Or maybe this 1-liner:

def stringSlicer(st, sz=3, d=''): return st and stringSlicer(st[:-sz],sz,'.')+st[-sz:]+d or ""



From gbreed at cix.compulink.co.uk  Mon Jun 24 05:56:48 2002
From: gbreed at cix.compulink.co.uk (gbreed at cix.compulink.co.uk)
Date: Mon, 24 Jun 2002 09:56:48 +0000 (UTC)
Subject: Python HTML Data Collection
References: <Pine.GSO.4.33.0206232352240.18910-100000@fissure.scar.utoronto.ca>
Message-ID: <af6qd0$98k$1@thorium.cix.co.uk>

Paul Barrett wrote:

>    I'm fairly new to Python but I do have a little bit of experience 
> from
> a class I took which taught some Python. I'm working at a job where I
> query a CGI script, collect the data and then post it to another CGI
> script. My main question is whether I could write a python script which
> would automate this process. Both of these scripts are on external
> websites and I only need a select amount of the data. I thought that if
> any language would do this job, then it would be Python, but I haven't
> been able to find any information for something like this. Do you folks
> have any suggestions?

See urllib[2].urlopen


                     Graham


From cliechti at gmx.net  Fri Jun 21 17:46:32 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 21 Jun 2002 23:46:32 +0200
Subject: TimeoutQueue.py
Message-ID: <Xns9234F249D7D85cliechtigmxnet@62.2.16.82>

Anyone ever needed a Queue with timeout?
I have some external hardware, talking over pyserial to my app. The app 
uses threads, i.e one is listening on the serial port, any thread can
send data but receives answers from the listening one using a Queue.
Now sometimes the hardware does not respond (e.g. switched off etc.) and 
the sending thread would be waiting forever in a Queue.get(). Therefore i 
needed a get(timeout) so that i can raise an exception in cas of a hardware 
failure.
I hacked up a version of Queue to TimeoutQueue and attached it for everyone 
who's interested. Its basicaly a merge between Queue.py and threading.py's 
_Event.wait().

have fun,
chris


----- file: TimeoutQueue.py -------------
"""A multi-producer, multi-consumer queue.

now with timeout value for get(). it waits up to 'timeout' seconds
and return the item if one is available within that time or it raises
an Empty exception otherwise.

cliechti at gmx.net
"""

>from time import time as _time, sleep as _sleep

class Empty(Exception):
    "Exception raised by Queue.get(block=0)/get_nowait()."
    pass

class Full(Exception):
    "Exception raised by Queue.put(block=0)/put_nowait()."
    pass

class Queue:
    def __init__(self, maxsize=0):
        """Initialize a queue object with a given maximum size.

        If maxsize is <= 0, the queue size is infinite.
        """
        import thread
        self._init(maxsize)
        self.mutex = thread.allocate_lock()
        self.esema = thread.allocate_lock()
        self.esema.acquire()
        self.fsema = thread.allocate_lock()

    def qsize(self):
        """Return the approximate size of the queue (not reliable!)."""
        self.mutex.acquire()
        n = self._qsize()
        self.mutex.release()
        return n

    def empty(self):
        """Return 1 if the queue is empty, 0 otherwise (not reliable!)."""
        self.mutex.acquire()
        n = self._empty()
        self.mutex.release()
        return n

    def full(self):
        """Return 1 if the queue is full, 0 otherwise (not reliable!)."""
        self.mutex.acquire()
        n = self._full()
        self.mutex.release()
        return n

    def put(self, item, block=1):
        """Put an item into the queue.

        If optional arg 'block' is 1 (the default), block if
        necessary until a free slot is available.  Otherwise (block
        is 0), put an item on the queue if a free slot is immediately
        available, else raise the Full exception.
        """
        if block:
            self.fsema.acquire()
        elif not self.fsema.acquire(0):
            raise Full
        self.mutex.acquire()
        was_empty = self._empty()
        self._put(item)
        if was_empty:
            self.esema.release()
        if not self._full():
            self.fsema.release()
        self.mutex.release()

    def put_nowait(self, item):
        """Put an item into the queue without blocking.

        Only enqueue the item if a free slot is immediately available.
        Otherwise raise the Full exception.
        """
        return self.put(item, 0)

    def get(self, timeout=None):
        """Remove and return an item from the queue.

        If optional arg 'timout' is None (the default), block if
        necessary until an item is available.  Otherwise (timout is a
        number), return an item if one is available within the specified
        time, else raise the Empty exception.
        """
        if timeout is None:
            self.esema.acquire()
        else:
            # code from threading.py: _Event.wait()
            # Balancing act:  We can't afford a pure busy loop, so we
            # have to sleep; but if we sleep the whole timeout time,
            # we'll be unresponsive.  The scheme here sleeps very
            # little at first, longer as time goes on, but never longer
            # than 20 times per second (or the timeout time remaining).
            delay = 0.0005 # 500 us -> initial delay of 1 ms
            endtime = _time() + timeout
            while 1:
                if self.esema.acquire(0):
                    break
                remaining = endtime - _time()
                if remaining <= 0:  #time is over and no element arrived
                    raise Empty
                delay = min(delay * 2, remaining, .05)
                _sleep(delay)       #reduce CPU usage by using a sleep
        self.mutex.acquire()
        was_full = self._full()
        item = self._get()
        if was_full:
            self.fsema.release()
        if not self._empty():
            self.esema.release()
        self.mutex.release()
        return item

    def get_nowait(self):
        """Remove and return an item from the queue without blocking.

        Only get an item if one is immediately available.  Otherwise
        raise the Empty exception.
        """
        return self.get(0)

    # Override these methods to implement other queue organizations
    # (e.g. stack or priority queue).
    # These will only be called with appropriate locks held

    # Initialize the queue representation
    def _init(self, maxsize):
        self.maxsize = maxsize
        self.queue = []

    def _qsize(self):
        return len(self.queue)

    # Check whether the queue is empty
    def _empty(self):
        return not self.queue

    # Check whether the queue is full
    def _full(self):
        return self.maxsize > 0 and len(self.queue) == self.maxsize

    # Put a new item in the queue
    def _put(self, item):
        self.queue.append(item)

    # Get an item from the queue
    def _get(self):
        item = self.queue[0]
        del self.queue[0]
        return item

-------------------
-- 
Chris <cliechti at gmx.net>



From fperez528 at yahoo.com  Tue Jun 25 15:24:41 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 25 Jun 2002 13:24:41 -0600
Subject: Creating an execution environment
References: <mailman.1024992873.4084.python-list@python.org>
Message-ID: <afag1q$cun$2@peabody.colorado.edu>

Juan M. Casillas wrote:

> It's?posible?to?have?multiple?environments?with?a?single?python
> interpreter???its?posible?to?have?multiple?interpreters?in?a
> C?program???Also,?I?want?to?load?chunks?of?code?for?multiple
> files,?and?then?add?the?symbols?to?the?right?environment...
> 

Hola Juan,

You can use IPython if you want (http://www-hep.colorado.edu/~fperez/ipython) 
which is set up for exactly that. A single call (documented in the manual and 
supplied examples) will start an environment which looks similar to the 
normal interpreter (actually with many enhancements) but with a completely 
independent namespace. Multiple such environments can coexist either in 
parallel or embedded within one another. Since they are all python programs, 
I imagine you can just make the call from within your C code. So far I've 
only tested the embedding facilities by calling them from within other python 
programs, but I don't see any reason why making the calls from within C would 
be a problem.

Suerte!

f.


From tdelaney at avaya.com  Thu Jun 13 22:55:52 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 14 Jun 2002 12:55:52 +1000
Subject: filter2
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A38C@natasha.auslabs.avaya.com>

> From: Delaney, Timothy [mailto:tdelaney at avaya.com]
>
> Python 1.5.2 ...
> 
> filter2:   5.0598633013
> filter3t:  4.76191299505
> filter3l:  4.37944378728
> 
> Python 2.2.1 ...
> 
> filter2:   5.40253108615
> filter2t:  4.38322776899
> filter2l:  4.40173851257
> filter3t:  4.32153526376
> filter3l:  4.26489291133

Jython 2.1.2

C:\python\modules\test10.py:6:[SyntaxWarning]: local name 'funcs' in
'filter2t' shadows use as global in nested scopes
C:\python\modules\test10.py:6:[SyntaxWarning]: local name 'test' in
'filter2t' shadows use as global in nested scopes
C:\python\modules\test10.py:13:[SyntaxWarning]: local name 'funcs' in
'filter2l' shadows use as global in nested scopes
C:\python\modules\test10.py:13:[SyntaxWarning]: local name 'test' in
'filter2l' shadows use as global in nested scopes
filter2:   5.827999949455261
filter3t:  5.078000068664551
filter3l:  4.983999967575073

Tim Delaney




From eric.brunel at pragmadev.com  Tue Jun 11 06:15:28 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 11 Jun 2002 10:15:28 +0000
Subject: python, shell, environment variable
References: <a7c2c93c.0206102219.28e2d2f7@posting.google.com>
Message-ID: <ae4b7s$jqc$1@wanadoo.fr>

oliver wrote:

> hi, folks,
> 
> I am trying to get a set of python and bash scripts to work, and I
> constantly run into trouble with environment variables:
> 
> First, my python script need to modify environment variable, say,
> PATH, and want to keep the modification even after the script is done.
> os.environ["PATH"]= ... doesn't seem to work, any idea?

Modifying environment variables this way is simply impossible: a process 
run by another process cannot change the environment of its parent, so your 
Python script will not be able to change environment variables in a 
persistent way.

> Second, it would be better if my python script can call bash shell and
> still keep the environment variable modification done by bash. I tried
> os.popen("source some_shell"), it doesn't work. ?

Same answer: this is impossible. os.popen runs the command in a subshell, 
and the environment variable changes made by the "source" will only be 
visible in that sub-shell.

One way to achieve what you want to do is to describe the environment 
variables you want to change in a file created by the sub-process. That 
file will then be used by the parent to do the actual variable settings. 
Here is an example:

---------foo.sh---------------------
cat > /tmp/envvars <<.
MYVAR1=value1
MYVAR2=value2
.
------------------------------------

---------foo.py---------------------
import os
os.system('foo.sh')
for l in open('/tmp/envvars').readlines():
  var, val = l.strip().split('=')
  os.environ[var] = val
# ...
print os.environ['MYVAR1']
------------------------------------

This works from a bash child to a Python parent. The reverse is also 
possible using the same mechanism: the Python script may simply create a 
file containing "setenv" commands that will be source'd by the parent bash 
script.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From michael at damaru.com  Mon Jun 10 19:39:55 2002
From: michael at damaru.com (Michael Davis)
Date: Mon, 10 Jun 2002 19:39:55 -0400
Subject: newbie question - python blocking on int()?
Message-ID: <4LaN8.2488$Vr2.586019@news20.bellglobal.com>

Hi,

I'm writing a specialized ftp client. I'm parsing the output of ftp.dir, 
which gives me a string like this (call it str):

-rw-r--r--   1 fred       527 Jun  4 22:58 report.php

and I'm getting the various parts like this:

details = string.split( str )
permissions = details[0]
size = details[3]
name = details[7]
debug( "added remote file size %5d: %s" % (int(size), name) )

This works. But when I replace the 3rd line with this:

size = int( details[3] )

python hangs on that line. Why?

Thanks a lot!

-- 
Michael Davis
Damaru
Custom Programming - Web Development - Database Design
http://www.damaru.com
416-540-1284


From stefnin.nospam at yahoo.fr  Wed Jun 19 12:23:34 2002
From: stefnin.nospam at yahoo.fr (Stephane Ninin)
Date: 19 Jun 2002 16:23:34 GMT
Subject: tkinter menu (newbie question)
References: <Xns9232778FCD335stefninnospamyahoofr@213.228.0.138> <aepps0$al$1@wanadoo.fr> <Xns9232967929C7Fstefninnospamyahoofr@213.228.0.138> <aeq0cv$p5g$1@wanadoo.fr>
Message-ID: <Xns9232B9BE55D67stefninnospamyahoofr@213.228.0.138>

Eric Brunel <eric.brunel at pragmadev.com> a tapote dans
news:aeq0cv$p5g$1 at wanadoo.fr: 

> 
>> If I want to use several of these windows,
>> can still use for base class Tk, instead of Frame ?
> 
> Nope: the instance of Tk is your main window. There should only be
> one, and closing it will end your application. If you want to create
> other windows, you should create instances of Toplevel.
> 

So, what would be the easiest way to attach a menu to a window which is 
not the main window ? (that is which doesnot inherit from Tk class)

It seems that Frame class has no menu key, 
so I cannot use config(menu=...) 
for a class that would inherit for Frame...

 

-- 
  Stephane Ninin
stefnin.nospam at yahoo.fr
Les cons, ca ose tout... C'est meme a ca qu'on les reconnait.


From imbosol at vt.edu  Tue Jun 25 15:49:04 2002
From: imbosol at vt.edu (Carl Banks)
Date: Tue, 25 Jun 2002 15:49:04 -0400
Subject: Suggestions for good programming practices?
References: <mailman.1025027914.18364.python-list@python.org>
Message-ID: <gfhafa.hb.ln@grey.aerojockey.localdomain>

brueckd at tbye.com wrote:
> On 25 Jun 2002, Donn Cave wrote:
> 
>> | So... rather than teaching "avoid W!", let's say "be careful with W 
>> | because of X, Y, and Z". I still wouldn't use eval/exec on code posted 
>> | through a web form, for example, but there are times when they are very 
>> | useful and I can use them in good confidence because I understand their 
>> | risks.
>> 
>> But you weren't going to be deterred by that pronouncement anyway.
>> 
>> "Avoid exec, execfile, eval, and input" is good advice.  Taken at face
>> value, it doesn't necessarily absolutely prohibit their use - if I said
>> "avoid walking in the road", you could reasonably reasonably assume I'm
>> saying something like "walk on the sidewalk when possible".
> 
> If you mean the latter, say the latter. Neither newbies nor veterans have 
> to do any reading between the lines. Even better, say "walk on the 
> sidewalk when possible because ___".

Fair enough.

* Avoid using exec, execfile, eval, and input unless you are
  deliberately asking your user to supply Python code, and take
  serious precautions if the program is running at a higher privledge
  level than the user.  All other uses of them should be thoroughly
  thought out, and the implications understood.


>> Someone whose software engineering skills have been honed by years
>> of wrestling with ugly code monsters will apply a different perspective
>> to that advice.  A 1-week newbie could do worse than to follow that
>> advice religiously.
> 
> I understand your point, but the OP is new to Python, not programming.  
> This isn't a big deal; I was simply pointing out that just saying "don't
> use those!" (Why? Are they broken? Deprecated? What?) isn't as helpful as
> explaining the risks.
> 
> In the specific case of exec/eval, people are quick to strike them down,
> often citing the untrusted input example (what about os.system, .fork,
> etc.?), but that's a pretty narrow and uncommon usage scenario.

I completely disagree.  I've seen at least two potentially
security-compromizing uses of eval posted here in the last few days as
"good" ways to do something.

I imagine most people see eval and think, "what a cool, easy way to
turn a string into a list!" without stopping to consider that it could
also turn a string into a syntax error, or an object that erases your
whole filesystem when its __getitem__ method is called.


> When I
> first learned Python I used them in some of the "forbidden" ways because I
> hadn't learned getattr/setattr yet. So what? My programs worked fine (a
> little slow) and later I learned better ways and now use those instead.

You were lucky that your forbidden uses of eval and exec weren't
dangerous in that case.  (Or maybe you weren't lucky but were already
aware that eval could be dangerous, but that's not the case for all
newbies.)


> If a person is new to programming altogether then what kind of a teacher
> will draw attention to exec/eval in the first place anyway?

A good one.  Realizing that one does not need to call attention to
eval for a newbie to find it, a good teacher explains that it is
dangerous before the newbie has a chance to hurt himself with it.


> Besides, the
> people who understand the potential for problems with something tend to be
> the ones that don't mishandle it (whether it be streets, guns, or Python
> functions).

Which is why it is important to tell people who don't understand the
potential for problems to not use eval until they do.

You seem to have this idea, based on this last paragraph, that a good
way to keep newbies from misusing eval and exec is to keep them in
ignorance.  I think that's absurd.  I don't know if you're just
grasping for arguments, because you previously seemed to think it's
best to explain WHY, as opposed to simply saying, "Don't do it."


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From shalehperry at attbi.com  Tue Jun 18 20:04:37 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 18 Jun 2002 17:04:37 -0700 (PDT)
Subject: procmail replacement in Python
In-Reply-To: <20020618235901.GA2197@lilith.my-fqdn.de>
Message-ID: <XFMail.20020618170437.shalehperry@attbi.com>

On 18-Jun-2002 Gerhard H?ring wrote:
> Does such a thing exist? I've recently looked into the procmail sources
> and run away screaming. I'd very much prefer a Python replacement to
> such badly written C code.
> 
> Does anything exist that I can start from? I've even contemplated not
> inventing yet another filter language, but just using plain Python
> functions.
>  

there is also a nice filter language from CMU called Sieve (I think).  You
could implement a python based Sieve interpreter.





From info at mjais.de  Tue Jun  4 17:07:16 2002
From: info at mjais.de (Markus Jais)
Date: Tue, 04 Jun 2002 23:07:16 +0200
Subject: property in Py2.2
References: <3CFC9814.4761F3CB@gol.ge>
Message-ID: <adja44$11k99k$1@ID-75083.news.dfncis.de>

hi
this is an example I am using:
############
class Address(object):

        __slots__ = ('firstname', 'lastname', 'street', 'zipcode', '
city', 'country', '_email', 'phone_private',
                                 'phone_work', 'phone_mobile', 'birthday', 'comment')
                                 
        def __init__(self):
                self.firstname     = ""
                self.lastname      = ""
                self.street        = ""
                self.zipcode       = ""
                self.city          = ""
                self.country       = ""
                self._email        = ""
                self.phone_private = ""
                self.phone_work    = ""
                self.phone_mobile  = ""
                self.birthday      = ""
                self.comment       = ""


        def set_email(self, email):
                if re.match(".*@.*\..*", email):  # must contain '@' and a '.'
                        self._email = email
                else:
                        print "wrong"  # TODO raise Exceptionm

        def get_email(self):
                return self._email
        email = property(get_email, set_email, None, 'Setting the email adress')

############

markus



Giorgi Lekishvili wrote:

> Hi all!
> 
> I wonder if someone finds time to send me a real code of how to use
> properties in Py2.2
> 
> I would be much obliged, if you also send me a version of the code for
> Py2.1, with __setattr__  and __getattr__.
> 
> Thx,
> Giorgi
> 
> PS. Prior to post here, I have read the materials on www.python.org. I
> find them very helpful.



From fcolaco at demnet.ubi.pt  Sat Jun 22 05:44:17 2002
From: fcolaco at demnet.ubi.pt (=?iso-8859-1?q?Francisco_Miguel_Cola=E7o?=)
Date: Sat, 22 Jun 2002 10:44:17 +0100
Subject: =?iso-8859-1?b?RWxlaef1ZXM=?= Brasileiras/Brasilian Election
References: <mailman.1024707421.31952.python-list@python.org>
Message-ID: <pan.2002.06.22.09.44.12.42287.8944@demnet.ubi.pt>

On Sat, 22 Jun 2002 01:40:12 +0100, KEVIN MITNICK wrote:

> 		PORTUGU?S
> O que um governo corrupto ? capaz de fazer pra manter-se no poder?!
> TUDO:manipula??o da imprensa, da economia, das pesquisas eleitorais,
> mentiras e falsas acusa??es s?o apenas alguns dos expedientes usados
> por FHC e sua equipe (de corruptos) pra manter-se no poder.
> Acordem brasileiros. N?o podemos deixar que esse terrorista social
> volte a eleger os seus. Basta!Vamos dar uma resposta nas urnas! Vamos
> dar uma chance a esquerda. Vamos eleger LULA pra presidente!!! 
> 
> 		ENGLISH
> What a corrupt goverment is capable to make to keep the power itself?!
> EVERYTHING:manipulation of press, of economy, of electoral research,
> lies and false accusation are just some  resources used
> by FHC and his team (of corrupts) to keep the power themselves.
> Brasilians wake up. We can't let this social terrorist elect his
> friends again. That's enough!Let's give an answer in the elections.
> Let's give a chance to Left Hand. Let's elect LULA for president of Brazil.

I think this is the function you want:

if government_is corrupt:
  manipulate (press, economy, poles)
  say (lies, false_accusations)

  def rotate_corrupts (scoundrel, color):
    elect (scroundel, color)

  rotate_corrupts ('Lula da Silva', 'red')
  
Thanks for recurring to the python help desk.

  FHC (Francisco Hon?rio Cola?o, actually)



From m2 at plusseven.com  Wed Jun 12 17:34:37 2002
From: m2 at plusseven.com (Alex Polite)
Date: Wed, 12 Jun 2002 23:34:37 +0200
Subject: Something faster then sgmllib for sucking out URLs
In-Reply-To: <Pine.LNX.4.44.0206121503560.24931-100000@ns0.tbye.com>
References: <20020612202828.GB16934@matijek.plusseven.com> <Pine.LNX.4.44.0206121503560.24931-100000@ns0.tbye.com>
Message-ID: <20020612213437.GC16934@matijek.plusseven.com>

On Wed, Jun 12, 2002 at 03:15:45PM -0700, brueckd at tbye.com wrote:
> How fast do you need it to be? Using regular expressions + other junk (see
> below) I get about 1 MB/sec on a 900 MHz Pentium III - that's a lot

I think this will be fast enough. Thanks a lot.

-- 
Alex Polite
http://plusseven.com/gpg/




From mhammond at skippinet.com.au  Sun Jun  2 19:12:46 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 02 Jun 2002 23:12:46 GMT
Subject: windows install put short path names to win register
References: <61f6bfa8.0206021304.2abedc3c@posting.google.com>
Message-ID: <3CFAA6E0.5010806@skippinet.com.au>

Toni Gaya wrote:
> Windows python 2.2.1 installation program puts short file names to
> some path values in windows register.
> 
> This makes that programs that read python path or some other
> information as python path from windows register crash
> -i'm using NTFS (maybe with FAT32, due that FAT system works with
> short file names and long file names, works, but as NTFS only works
> with long file names, some other installation programs just crash)-.
> Examples: try to install mysqldb (mysql connection) or py2exe on
> windows with a NTFS disk, and these programs crash.

NTFS supports both long and short file names.  The short file name can 
be opened in exactly the same way as the long file name - ie, the 2 
names are exactly equivilent.

I can't see a good reason for the Python installer to convert to short 
names (my guess is code generated by the Wise "Wizard"), but I can also 
see no reason it would not work, and I can assure you that the simple 
fact that Python writes short names is not enough to prevent mysql from 
working :)

Mark.



From tebeka at cs.bgu.ac.il  Wed Jun 19 04:41:22 2002
From: tebeka at cs.bgu.ac.il (Miki Tebeka)
Date: 19 Jun 2002 01:41:22 -0700
Subject: Help: win32gui & threads
References: <33803989.0206170318.9d3a225@posting.google.com> <3D0E6849.3050909@skippinet.com.au>
Message-ID: <33803989.0206190041.43b788ac@posting.google.com>

Hello Mark,

> * Is the DestroyWindow() call being made?
Yes and OnDestroy is also called.
> * Does the main thread ever appear to end - ie, if you put a print 
> statement after the PumpMessages(), will it appear.
Yes.

The OS is winnt4

BTW: I'm sure you recognize the code copy from win32gui_taskbar demo, 10x :-)


From eric.brunel at pragmadev.com  Mon Jun  3 06:33:10 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 3 Jun 2002 10:33:10 +0000
Subject: How to suspend execution for at least x seconds in python ? (ie does "sleep" exist in Python ?)
References: <adf7t3$ki71@news.rd.francetelecom.fr>
Message-ID: <adf996$h97$1@wanadoo.fr>

Nicolas Torzec wrote:
> Dear all,
> I was wondering if an utility like "sleep x" exist in python ?
> Under Unix-like operating systems, the sleep utility suspends execution
> for at least x seconds under...

Of course it does:

import time
time.sleep(x)

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From bokr at oz.net  Sun Jun 16 14:39:44 2002
From: bokr at oz.net (Bengt Richter)
Date: 16 Jun 2002 18:39:44 GMT
Subject: Divorce property objects from attribute access dependency?
Message-ID: <aeim1g$gs2$0@216.39.172.122>

 >>> class C(object):
 ...     def __init__(self,v):
 ...         self._v = v
 ...     def get_v(self): return self._v
 ...     def set_v(self,v): self._v = v
 ...     v=property(get_v,set_v)
 ...     vlist = [v,v]
 ...
 >>> c=C('something')
 >>> c.v
 'something'
 >>> c.v='other'
 >>> c.v
 'other'
 >>> c.vlist
 [<property object at 0x0082D4A0>, <property object at 0x0082D4A0>]
 >>> c.vlist[0]
 <property object at 0x0082D4A0>

Why not trigger get_v() in these accesses too, and have single mechanism to
suppress it instead of a single mechanism to effect it?

ISTM kind of a hack to rig access via instance attribute as a mechanism for controlling
referenced-object behavior (this is my impression, I haven't checked code ;-). IMO default
behavior of an object in lhs or rhs context should not depend on how a reference to it was
acquired and dereferenced.

(Obviously, if property objects always acted as they now do through instance attribute
access, "vlist=[v,v]" in this example would have to have some kind of access override like
"vlist = [obj(v),obj(v)]", or else it would generate ['something','something'] right there.
Using obj(p) a property object could be passed around without evaluation and the potential
side effects thereof).

BTW, it would be nice to be able to define a property at global module scope also,
and perhaps this would tie in nicely.

Just musing ;-) Thoughts? Am I missing a big gotcha?
One thing is that it will effectively make possible a ()-less function call, which will
upset some people ;-)

Regards,
Bengt Richter


From jepler at unpythonic.net  Tue Jun 11 12:20:40 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 11 Jun 2002 11:20:40 -0500
Subject: Defining a method final
In-Reply-To: <6glN8.35376$86.884025@twister1.libero.it>
References: <6glN8.35376$86.884025@twister1.libero.it>
Message-ID: <20020611162039.GA2348@unpythonic.net>

The following code uses a metaclass to attempt to enforce "final"
methods.  Warning: Use of metaclasses has exploded brains.  Accessing
directly through __dict__ or object.__setattr__ can bypass these
protections.

def finalmethod(f):
    """Marks a method as a final method.  Use like classmethod(), etc.  Only
       works in subclasses of FinalMethods:
       class C(FinalMethods):
           def f(x): ...
           f = finalmethod(f)"""
    f.__final__ = 1
    return f

class FinalMethodsMeta(type):
    """Metatype for FinalMethods"""
    def __init__(cls, name, bases, dict):
	super(FinalMethodsMeta, cls).__init__(name, bases, dict)
	for base in cls.__mro__:
	    if base is cls: continue
	    for attrname in base.__dict__:
		attr = base.__dict__[attrname]
		if hasattr(attr, "__final__") and attr.__final__:
		    func = getattr(cls, attrname).im_func
		    print attrname, attr, func
		    if func is not attr:
			raise ValueError, \
			    "Subclasses of %s may not override the method %s" \
			    % (base.__name__, attrname)

    def __setattr__(cls, name, attr):
	oldval = getattr(cls, name, None)
	if hasattr(oldval, "__final__") and oldval.__final__:
	    raise ValueError, "final methods may not be overridden by assignment to class"
	super(FinalMethodsMeta, cls).__setattr__(name, attr)

class FinalMethods:
    """Subclasses of FinalMethods may define a method to be a finalmethod().
    This probably only works with instancemethods, not classmethods or 
    staticmethods.  A subclass may not override a finalmethod"""


    __metaclass__ = FinalMethodsMeta

    def __setattr__(self, name, attr):
	oldval = getattr(self, name, None)
	if hasattr(oldval, "__final__") and oldval.__final__:
	    raise ValueError, "final methods may not be overridden by assignment to instance"
	super(FinalMethods, self).__setattr__(name, attr)

def _test():
    class c1(FinalMethods):
	def f(x): pass
	f = finalmethod(f)

	def g(x): pass

    class c2(FinalMethods):
	def g(x): pass
	g = finalmethod(g)

    # Test that FinalMethods classes can be subclassed when they don't
    # interfere with each others' final methods
    class c3(c2, c1): pass
    assert c3.f.im_func is c1.f.im_func

    # Test that a FinalMethods class can be subclassed when the added
    # methods do not interfere with the super's final methods
    class c4(c1):
	def g(x): pass

    # Test that when a final method is concealed by a non-final method,
    # it blows up
    try:
	class c5(c1, c2): pass
    except ValueError, detail:
	print "Caught expected ValueError:", detail
    else:
	print "Did not catch expected error"
	print c5.g, c1.g, c2.g
	raise RuntimeError

    # Test that when a final method is overridden in the subclass, it
    # blows up
    try:
	class c5(c1):
	    def f(x): pass
    except ValueError, detail:
	print "Caught expected ValueError:", detail
    else:
	print c5.f, c1.f
	print "Did not catch expected error"
	raise RuntimeError

    # Test that when a final method is overridden in the sub-subclass,
    # it blows up
    try:
	class c6(c3):
	    def f(x): pass
    except ValueError, detail:
	print "Caught expected ValueError:", detail
    else:
	print c6.f, c3.f
	print "Did not catch expected error"
	raise RuntimeError

    # Test that assigning to non-final attributes works (class)
    class c7(c1): pass
    c7.g = lambda: None

    # Test that assigning to final attributes fails (class)
    try:
	class c7(c3): pass
	c7.f = None
    except ValueError, detail:
	print "Caught expected ValueError:", detail
    else:
	print "Did not catch expected error"
	raise RuntimeError

    # Test that assigning to non-final attributes works (instance)
    i = c1()
    i.g = None

    # Test that assigning to final attributes fails (instance)
    try:
	i = c1();
	i.f = None
    except ValueError, detail:
	print "Caught expected ValueError:", detail
    else:
	print "Did not catch expected error"
	raise RuntimeError

    print "seems ta work"

if __name__ == '__main__': _test()




From Sascha.Ferley at infineon.net  Sun Jun 23 23:43:56 2002
From: Sascha.Ferley at infineon.net (Sascha Ferley)
Date: Sun, 23 Jun 2002 21:43:56 -0600 (MDT)
Subject: Resolution PIL, Python and CGI
In-Reply-To: <200206241308.33270.rjones@ekit-inc.com>
Message-ID: <Pine.GSO.4.33.0206232142030.27303-100000@infinity.infineon.net>


On Mon, 24 Jun 2002, Richard Jones wrote:

> On Mon, 24 Jun 2002 13:02, Sascha Ferley wrote:
> > On Mon, 24 Jun 2002, Richard Jones wrote:
> > > Your web server environment most likely doesn't have /usr/local/lib in
> > > its LD_LIBRARY_PATH.
> >
> > You probly mean Apache... I'll take a look for at the conf files for it..
> > Any ideas where to statically define the lib paths for apache.. i doubt in
> > the httpd.conf file..
>
> You've basically got three choices as I see it:
>
> 1. recompile PIL to statically link libjpeg etc.
> 2. recompile PIL to use -R (or whatever the flag is) so it remembers where the
> dynamic libs are
> 3. modify your apache start script to set the environment var appropriately.
>
> 1 or 2 are going to be the easiest, and safest options.

Thanks to Richard, this problem is resolved.
The issue:
	Apache didn't have all the library path's that it needed.
	to fix. set in httpd.conf  the SetEnv LD_LIBRARY_PATH
	env-variable.

Thanks to everyone..


Sascha






From d_blade8 at hotmail.com  Wed Jun 19 14:25:25 2002
From: d_blade8 at hotmail.com (Lemniscate)
Date: 19 Jun 2002 11:25:25 -0700
Subject: Problems with mayavi, pyvtk
Message-ID: <7396d2b2.0206191025.3351a888@posting.google.com>

Hi everybody,

I just downloaded and installed the latest versions of pyvtk and
mayavi.  However, when I try to follow the instructions at
http://mayavi.sourceforge.net/docs/guide/c827.html#VIZ-DATA for using
mayavi, I get the ImportError as below.  For the like of me, I can't
seem to figure out why I don't have any vtkpython modules anywhere
(doing searches on the internet for the module hasn't been a big help
so far, and it isn't ANYWHERE on my harddrive).  Am I missing
something simple (the fact that other people don't seem to be having
this problem makes me think I am).  Please let me know.  Thanks a
bunch.

Lem


>>> import mayavi
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
  File "C:\Python22\Lib\site-packages\mayavi\__init__.py", line 5, in
?
    import Common, Base.Objects
  File "C:\Python22\Lib\site-packages\mayavi\Base\__init__.py", line
19, in ?
    import Objects
  File "C:\Python22\Lib\site-packages\mayavi\Base\Objects.py", line
20, in ?
    import Tkinter, vtkpython, math, tkColorChooser
ImportError: No module named vtkpython
>>>


From jeff at sasmor.com  Tue Jun 18 21:40:08 2002
From: jeff at sasmor.com (Jeff Sasmor)
Date: Wed, 19 Jun 2002 01:40:08 GMT
Subject: If you use Python -OO is it a permanent condition??
References: <mailman.1024445663.26965.python-list@python.org>
Message-ID: <YhRP8.63558$uX3.23256@nwrddc01.gnilink.net>

Actually, once you use Python it's a permanent condition :-)


--
#--------------------------------
Jeff Sasmor
jeff at sasmor.com
"Delaney, Timothy" <tdelaney at avaya.com> wrote in message
news:mailman.1024445663.26965.python-list at python.org...
> For some reason, I parse the subject line as "If you use Python in an
> object-oriented way, is it a permanent condition?"
>
> I'm not sure if that's a philosophical question, but it confused the hell
> out of me ;)
>
> Tim Delaney
>
>




From jtt at hnc.com  Tue Jun  4 15:39:18 2002
From: jtt at hnc.com (Jim)
Date: 4 Jun 2002 12:39:18 -0700
Subject: DCOracle2 - Can't get cursor.execute to return data when SQL query involves datetimes.
Message-ID: <be0d725a.0206041139.1f8f6647@posting.google.com>

Hello,

I am trying to write a simple program to retrieve 2 colums from a
table.
I am running on Solaris 2.8, Python 2.1.1, DCOracle2 and Oracle
8.1.7.3.


The columns are:

acct_number   ( one char long)
transaction_datestamp (oracle datetimestamp)

Here is my program:

import DCOracle2
trandt = '19970110121212'

db = DCOracle2.connect("uid/pwd")
c = db.cursor()

dt = DCOracle2.Timestamp(int(tran_date[0:4]), int(tran_date[4:6]),
int(tran_date[6:8]), int(tran_date[8:10]), int(tran_date[10:12]),
int(tran_date[12:14]))

results = c.execute,('select acct_nbr, trn_dt from testtable where
ACCT_NBR = :1 and TRN_DT = :2',"1",dt)

aaa = c.fetchone()

print aaa


The execute never return any thing from query when the date timestamp
is part of the query.  If I only query on the acct_nbr it returns the
expected row.

What am I missing ???

Thanks,
jim.


From joe+usenet at sunstarsys.com  Sat Jun  1 10:14:39 2002
From: joe+usenet at sunstarsys.com (Joe Schaefer)
Date: 01 Jun 2002 10:14:39 -0400
Subject: Migrating from PHP to Python...
References: <xWyk7.4824$v%4.492814@news20.bellglobal.com> <l4ihfu0nhp1a9p53vgch3n3kcc25s8c00p@4ax.com> <slrnafhj0d.dp4.gerhard@lilith.my-fqdn.de>
Message-ID: <m3znyf3x2o.fsf@mumonkan.sunstarsys.com>

Gerhard =?iso-8859-15?Q?H=E4ring?= <gerhard at bigfoot.de> writes:

[f'ups disobeyed]

[...]

> X-Post and F'up to appropriate group. Please continue trolling over
> there.

Please don't spam clp.misc with garbage like this.
IMO you've done a disservice to both communities.

-- 
Joe Schaefer   "What a waste it is to lose one's mind. Or not to have a mind is
                           being very wasteful. How true that is."
                                               -- Dan Quayle



From kragen at pobox.com  Sat Jun  1 16:12:02 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 01 Jun 2002 16:12:02 -0400
Subject: How to open a HTML file when the python cgi program is executing?
References: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de> <ad95l1.i0.1@kserver.org> <ad9vsm$vhjnu$1@ID-49758.news.dfncis.de> <9m6ada.31b.ln@10.0.0.1> <adaeh8$uum7t$1@ID-49758.news.dfncis.de> <ada3k7.lo.1@kserver.org>
Message-ID: <83n0uevjvx.fsf@panacea.canonical.org>

Sheila King <usenet at thinkspot.net> writes:
> print "Location: http://replacement-url\n"
> 
> You should have nothing before this, and there's no point in having
> anything after it, since the page is being redirected.

Some browsers (in particular old versions of Opera) have problems with
redirects every once in a while; that's why redirects traditionally
contain HTML, in case the browser is broken.



From chris_mk at hotmail.com  Tue Jun  4 01:28:14 2002
From: chris_mk at hotmail.com (Christopher)
Date: 3 Jun 2002 22:28:14 -0700
Subject: Problem with urllib.urlopen()
References: <cc14131d.0206031445.5154de31@posting.google.com>
Message-ID: <cc14131d.0206032128.5b2efcf6@posting.google.com>

There were some problem with posting the whole link so I'll break each
one up...

1st Link: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?
cmd=PureSearch&db=PubMed&details_term=%28%28%28%22organization%20and%
20administration%22%5BMeSH%20Subheading%5D%20OR%20organ%5BText%20Word%5D%29%
20AND%20%28%22diabetes%20mellitus%22%5BMeSH%20Terms%5D%20OR%20diabetes%5BText%
20Word%5D%29%29%20AND%20%28%28%22glucagon%22%5BMeSH%20Terms%5D%20OR%20glucagon%
5BText%20Word%5D%29%20AND%20%28%22insulin%22%5BMeSH%20Terms%5D%20OR%20insulin%
5BText%20Word%5D%29%29%29


2nd Link: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?
cmd=Display&db=PubMed&details_term=%28%28%28%22organization%20and%
20administration%22%5BMeSH%20Subheading%5D%20OR%20organ%5BText%20Word%5D%29%
20AND%20%28%22diabetes%20mellitus%22%5BMeSH%20Terms%5D%20OR%20diabetes%5BText%
20Word%5D%29%29%20AND%20%28%28%22glucagon%22%5BMeSH%20Terms%5D%20OR%20glucagon%
5BText%20Word%5D%29%20AND%20%28%22insulin%22%5BMeSH%20Terms%5D%20OR%20insulin%
5BText%20Word%5D%29%29%29&dopt=XML&query_key=1

Sorry about that.

Chris

chris_mk at hotmail.com (Christopher) wrote in message news:<cc14131d.0206031445.5154de31 at posting.google.com>...
> Hi all,
> 
> I'm having a bit of a problem and I'm not sure what I should do.  I've
> tried a couple of things and I'm running out of ideas.
> 
> Basically, I want to be able to access, directly, the XML versions of
> the NCBI results of a search.  There is a default display of 'Summary'
> and then there are numberous other displays, one of which is 'XML' You
> can view the following link in XML (sort of) merely by changing the
> display option (that was just some backgroud info, nothing spectacular
> that really matters).
>  Here is a link to an example page (do a search, then click details,
> then the URL button and PubMed gives the actual url it uses (otherwise
> it is just a cgi url; got that from O'Reilly's "Developing
> Bioinformatics Computer Skills": 
> http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=PureSearch&db=PubMed&details_term=%28%28%28%22organization%20and%20administration%22%5BMeSH%20Subheading%5D%20OR%20organ%5BText%20Word%5D%29%20AND%20%28%22diabetes%20mellitus%22%5BMeSH%20Terms%5D%20OR%20diabetes%5BText%20Word%5D%29%29%20AND%20%28%28%22glucagon%22%5BMeSH%20Terms%5D%20OR%20glucagon%5BText%20Word%5D%29%20AND%20%28%22insulin%22%5BMeS
%
> 0Terms%5D%20OR%20insulin%5BText%20Word%5D%29%29%29
> 
> 
> Now, try this url in your webbrowser and see what you get (you should
> see the actual xml version of the page above):
> http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Display&db=PubMed&details_term=%28%28%28%22organization%20and%20administration%22%5BMeSH%20Subheading%5D%20OR%20organ%5BText%20Word%5D%29%20AND%20%28%22diabetes%20mellitus%22%5BMeSH%20Terms%5D%20OR%20diabetes%5BText%20Word%5D%29%29%20AND%20%28%28%22glucagon%22%5BMeSH%20Terms%5D%20OR%20glucagon%5BText%20Word%5D%29%20AND%20%28%22insulin%22%5BMeSH%2
T
> rms%5D%20OR%20insulin%5BText%20Word%5D%29%29%29&dopt=XML&query_key=1
> 
> You should, like I said, get the XML version of the original links
> (with some extra information, which is why I want to interface with
> the XML site, btw).  Now, if you do an interface with Python instead
> of your browser you get a response of an error page instead.  Why is
> that and what can I do to change it?  If you have any information, I
> would appreciate it ver much.  Thanks.
> 
> Chris


From garth at deadlybloodyserious.com  Mon Jun 17 01:49:32 2002
From: garth at deadlybloodyserious.com (Garth T Kidd)
Date: 16 Jun 2002 22:49:32 -0700
Subject: a possibly foolish question about slices
References: <yu99sn3prbqm.fsf@europa.research.att.com> <eppstein-FCC2A4.12082514062002@news.service.uci.edu>
Message-ID: <4c877253.0206162149.bff3ae3@posting.google.com>

> x[p:][:n]

Does the parser take the appropriate short-cut, here, or do we end up
with a temporary list?

Regards,
Garth.


From cliechti at gmx.net  Mon Jun  3 16:17:46 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 3 Jun 2002 22:17:46 +0200
Subject: Created a "file" like command in python (see man 4 magic) and face a showstoper bug
References: <359e75aa.0206031113.6a8bfed1@posting.google.com>
Message-ID: <Xns9222E32F2F52Fcliechtigmxnet@62.2.16.82>

thomas.mangin at free.fr (Thomas Mangin) wrote in 
news:359e75aa.0206031113.6a8bfed1 at posting.google.com:
> I wrote some code perform "file" like operation and I am facing a
> nasty bug.
> 
> the files can be found at : ftp://www.slhan.org/source/circle/
> 
> It produce the same output that "file":
> 
> # ./magic.py core
> core: ELF 32-bit LSB core file of 'vi' (signal 11), Intel 80386,
> version 1 (SYSV)
> 
> # ./magic.py jpg_good.jpg
> jpg_good.jpg: JPEG image data, JFIF standard 1.01, resolution (DPI),
> 50 x 50
> 
> # ./magic.py jpg_bad.jpg
> jpg_bad.jpg: JPEG image data, EXIF standard %d.73, 10752 x 2048
> 
> Note the %d !
> 
> The code causing problem is :
> 

1. initialize replace with None

> if replace:

use this instead:
  if replace is not None:

and have fun progamming python!
chris

>   try:
>     mime = mime % replace
>   except:
>     pass
> 
> In both case mime is containing the string "%d." and replace is in
> both case of type int and valued 1 and 0.
> 
> Any idea ?
> 
> Thomas Mangin
> 



-- 
Chris <cliechti at gmx.net>



From neal at metaslash.com  Thu Jun 13 22:36:25 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Fri, 14 Jun 2002 02:36:25 GMT
Subject: Shrinking Python
References: <mailman.1023999624.27508.python-list@python.org>
Message-ID: <pan.2002.06.14.02.36.14.452776.7540@metaslash.com>

On Thu, 13 Jun 2002 16:15:39 -0400, Mahrt, Dallas wrote:

> I am looking into using Python in a small device with limited space for
> binaries. Due to this, I am both investigating creating a shared library
> with the interpreter and shrinking the footprint of this library. I have
> found many threads talking bout shared libraries, so I am not too
> concerned about that aspect.
> 
> I am concerned about shrinking the library's binary footprint. I did
> find several links to a "Deeply Embedded Python" based on 1.5.1 which
> has since vanished. I have stripped the library to improve size some (to
> ~1.13 MB) but would like to reduce it further. Has anyone tackled this
> or even attempted? Are there any links to resources that may help me
> that I haven't found?

How small do you want it to go?  My stripped sizes on Linux are:

-rw-r--r--    1 neal     cvs        702902 Jun 13 22:10 libpython2.3.a
-rwxr-xr-x    1 neal     cvs        682428 Jun 13 22:13 python*

This is with the latest CVS version and passing --without-doc-strings
to ./configure.  You should be able to easily drop the library
down by 90k, by removing:  unicode, regular expressions, and posix.

Without too much work, ie. by defining -DWITHOUT_COMPLEX, and 
./configure --without-doc-strings --without-unicode \
            --without-signal-module --disable-shared
I get this:

[neal at epoch src]$ ls -l libpython2.3.a python 
-rw-r--r--    1 neal     cvs        686126 Jun 13 22:29 libpython2.3.a
-rwxr-xr-x    1 neal     cvs        664828 Jun 13 22:29 python*

Neal


From jbublitzNO at SPAMnwinternet.com  Tue Jun 18 14:48:36 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Tue, 18 Jun 2002 18:48:36 GMT
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de> <aeftmj$m29$1@bob.news.rcn.net> <3d0c6d67@nntp.server.uni-frankfurt.de> <3D0CE698.1050806@SPAMnwinternet.com> <aeiu3a$4ff$1@bob.news.rcn.net> <3D0D3223.5070500@SPAMnwinternet.com> <aekrd1$4ao$1@bob.news.rcn.net> <3D0E2606.5050001@SPAMnwinternet.com> <aen6ch$d5u$1@bob.news.rcn.net>
Message-ID: <3D0F807A.40203@SPAMnwinternet.com>

David Abrahams wrote:
> "Jim" <jbublitzNO at SPAMnwinternet.com> wrote in message

>>Template-based code does require special treatment with sip.
>>In most cases where I've wrapped template-based types or
>>classes, I manually write C++ code for conversion C++ <-> Py
>>(sip has some functions that make that easier) and implement
>>it as a sip 'mapped type', which lets me refer to the same
>>conversion code from any return value or arg.

> That's a given with Boost.Python; with v2 conversions are automatically
> registered in a global repository so they can be re-used automatically
> across extension modules.

For C++ code with a lot of templates Boost.Python has an
advantage, particularly if you're early on the learning curve.

>>I generally
>>assign a new name (eg QList<int> -> IntList) but wouldn't
>>have to (sip will parse QList<int> and look for a mapped
>>type definition corresponding).

>>sip won't handle all possible C++ syntax, and it sometimes
>>takes some creativity to shoehorn something in, but something
>>well over 80% of h file code translates directly (after
>>removing variable names), and only 1 or 2% really takes
>>major effort.

>     .../your/ h file code...

Sure - you can extend sip in almost any way you want also
(the bindings generated are C++ code), but it may be a
little trickier fitting things in the binding's framework.
It's also possible to embed custom Python or C++ code in
sip description files and I do that when appropriate (rarely).
It kind of goes against the design philosophy of sip IMHO,
so I try to avoid it personally.

>>The real problem wouldn't seem to be C++ types as such, but
>>finding a useful (to the Python programmer) representation
>>via PyObject, and still maintaining the range of abilities
>>built into the C++ code (esp things like subclassing).

> That's built-in to Boost.Python. V1 had to implement a complicated Python
> class look-alike itself; V2 uses Python's new-style classes so subclassing
> "just works". Making virtual functions overridable in Python requires a
> small amount of boilerplate on the part of the programmer. I often wish I
> had a good C++ parser front-end which I could use to automate that part.

New-style classes lock you into Python 2.2+ though - I've
got users still at 1.5.2 and place a lot of emphasis on long
term version support. sip does the virtual code boilerplate'
automatically, unless the method requires handwritten code
for some other reason (rare). On the stuff I do, I see a lot
more virtual code than templates, but that obviously varies
a lot.

>>>>The other features that are helpful on large projects are the
>>>>built-in support for versioning

>>>How so? Can you say more about this?

>>sip uses directives: %If (A - B) ... %End to wrap a block
>>of code for versioning over a range. It could also be
>>%If (A - ) (>= A) or %If ( - B) (< B). Acceptable values
>>for A and B are defined in a %Timeline declaration; actual
>>values for A and B are passed to sip via the command line,
>>usually  derived from a version #define in the C++ source
>>code via configure (at least on Linux). Evaluation is by
>>lexical ordering of the A and B values. A and B are
>>strings like 'KDE_3_0_0'.

>>Not very different from #ifdefs, but maybe a little
>>easier.

> OK. The C++ code you're wrapping still has to get by with #ifdef, I
> suppose...

Depends on the code I suppose. Qt and KDE just issue all
new files - no versioning in the C++ code.

>>>Do you mean docstring generation?
>>
>>No - similar, but sip provides %ExportDoc ... %End directives
>>inside which you could place just about anything you want I
>>suppose. For PyQt/PyKDE the directives wrap SGML marked-up text.
>>You can construct an input file which along with the proper sip
>>command line switch, will collect and concatenate all of the
>>doc blocks distributed over the description files. The
>>concatenated file can be processed (in the case of SGML) to
>>provide HTML docs, or whatever you require.

> Oh, there are plenty of automated tools for extracting doc from C++. Lots of
> people use Doxygen with Boost.Python. No need to invent a new way to do it!

It's basically no different than Python - just %ExportDoc instead
of """, and SGML and tools are hardly new. You could use doxygen
if you wanted to (sip supports comments - I'm guessing sip files
are sufficiently parseable for doxygen) and you don't have to
use %ExportDoc. You have to generate markup either way if you
want HTML or other formats.

>>My guess is that Boost does at least some of this a little more
>>naturally. I'm not sure the extent that Boost can modify things
>>like return values or args without resorting to C++

> There is no "resorting to C++" with Boost.Python -- Boost.Python *embraces*
> C++! Since you're coding the wrappers in C++ anyway, it's trivial to make a
> "thin wrapper function" which calls the function being wrapped if you want
> to change the interface somehow. Then you just wrap the thin wrapper.

I finally went back to your website - I took your 'like an
IDL' comments a little too literally. IMHO, the fact that
Boost.Python *is* C++ is a big advantage for people who
want/need that. Personally, I don't want to write C++,
which is how I got into this in the first place.

Thin wrappers are easily implementable with sip too. In
contrast to the very large packages like PyQt and PyKDE,
they tend to be very lightweight and quick to build. I
can see ways though that I might use Boost.Python to avoid
having to choose between writing my own thin wrapper and
having to bind an entire C++ fileset. OTOH, I've written
C++ wrappers where the only difference betwen the h file
and the sip file is the addition of a %HeaderCode block
for sip, and I could write a simple py script to do that.
Doesn't happen often, but normally isn't a lot more work
than that.

>>Yes, but I don't write it - I just edit it mostly to correct
>>mistakes my description file generator creates and to handwrite
>>code where the generator has flagged that. Crude but effective (TM)
>>Even without automation, the process is easiest if you start by
>>saving the h file with a .sip extension and then (mostly) delete
>>the things that sip doesn't want/need, so it's still mostly
>>editing (sometimes A LOT of editing).

> That seems to be the key difference. It's not clear wether it's really an
> advantage, but I suppose that time will tell.

I doubt that it's ever clearly an advantage. It works very
well for PyQt and PyKDE because of the C++ style those projects
use generally - very consistent, relatively small amount of
template usage, lots of enums and of course sip has features to
support Qt. sip has been used outside of the Qt/KDE area, but
perhaps not on anything available publicly yet.

There's certainly plenty of opportunities for something like
"benchmarketing". Either one of us could contrive examples to
make one or the other look clearly better, but like most things
it's probably more rational to look at what the user actually
needs to accomplish. I've avoided actual code comparisons,
because the simple cases tend to be one-sided one way or the
other.

>>I should probably go back and review Boost. I had thought that
>>with Boost it was possible to wrap as much or as little of a
>>C++ library as you want to.

> That's true.

Big selling point IMHO.

>>I had the impression SWIG was supposed to do most of that in a single 
 >>step. Just wondering if I could implement something like that with 
Boost,
 >>or need to code a line manually for each method/class/whatever.

> If you have something which can mostly parse your headers, you can generate
> much of the wrapping code automatically.

Yes - it looks like that should be possible. Virtual code
might require a little more work in generating Boost.Python
code automatically, but the tradeoff is probably less things
flagged as requiring handwritten code.

> Yep. BTW, Boost is much more than just my Python binding library, so please,
> for the sake of the other authors: "Boost.Python" when discussing my lib.

Noted - I've tried to do better in this post.

> Not sure that you'll find much to steal; they're so different in approach...

I can be very creative when stealing other people's code :)

Actually "under the hood" sip and Boost.Python have similarities.


Jim



From aahz at pythoncraft.com  Thu Jun 13 09:43:31 2002
From: aahz at pythoncraft.com (Aahz)
Date: 13 Jun 2002 09:43:31 -0400
Subject: Threading tutorial wanted.
References: <mailman.1023954137.22915.python-list@python.org>
Message-ID: <aea7i3$71t$1@panix1.panix.com>

In article <mailman.1023954137.22915.python-list at python.org>,
Alex Polite  <m2 at plusseven.com> wrote:
>
>I few minutes of googling on python + threads doesn't yield a
>lot. Maybe someone here can point me to a good primer. Maybe someone
>here can write a good primer.

Take a look at my web page, but look at the OSCON 2001 tutorial, not the
one from 2000 suggested by Anoop.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From phlip_cpp at yahoo.com  Thu Jun 13 01:07:31 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 13 Jun 2002 05:07:31 GMT
Subject: Python GUI
References: <9ddd154e.0206122021.1b06f89a@posting.google.com>
Message-ID: <ae99aj$jp0@dispatch.concentric.net>

H. Safarzadeh wrote:

> -Tkinter: A nice simple lib, but really simple! It dosn't have enough
> widgets, and I do not like its look in addition!

Look closer.

The other ones you listed are fine, and you might end up productively going 
with them, but do not judge Tkinter based only on its lone /tkinter page on 
the python.org site.

Firstly, you probably overlooked Python MegaWidgets. Tkinter is not simple, 
it's primitive. That means you must assemble it to get things like a list 
box WITH a scrollbar. Pmw puts together all the pieces into a fat framework 
with splitter bars, canned dialog boxes, combo boxes, spinners, sliders, 
formatted entry fields; the works. And because it comes with a framework 
for this assembly, new assemblages are just an extension away.

Secondly, whenever anyone thinks of a new kind of canvas, such as 
Visualization Toolkit, the first thing they do is make sure it bonds with 
Tcl/Tk. Hence the second thing it bonds with is Tkinter. Search the web for 
these Widgets; they are all over the place. I have found several for flat 
charting (including Blt), and for R3 and R4 presentations like OpenGL.

Thirdly, Tkinter's Canvas widget is essentially a mini-UI kit within a kit. 
You can float a wide number of kinds of graphic objects (shapes, images, 
lines, text, etc.), all represented in memory as objects, not pixels. You 
can bind UI events directly to these objects instead of perform manual hit 
detection outside the black box. They come with a full set of properties so 
you don't need to manage their state and representation for them. And you 
can float primitive windows >and< other Tkinter widgets inside the canvas, 
down among the graphic primitives. Hence you can roll elaborate 
super-widgets yourself with very few lines of code. Just last week, over a 
couple of hours, I implemented visual feedback (a drop shadow) under 
graphics that were clickable, then granted the user the ability to click 
and drag these graphics around, with a bungee line connecting dynamically 
back to their origin. This canvas gave me all that in a very few lines of 
code, and prevented me from wasting many days performing primitive graphic 
manipulations within a hand-made framework. Just because other libraries 
make rolling your own widgets hard don't look at raw Tkinter's low widget 
count as any sign of weakness; it is in fact a strength.

A very nice example of the Canvas in action is TreeBox.py by Arpad Kiss, 
featured here:

        http://c2.com/cgi/wiki?PyUnitTestBrowser

Fourth, Tkinter has been around a very long time, and any UI question you 
could think to ask (such as idle timer functions, multi-level callbacks, 
elaborate key chord detection, mousewheel support) has been added to the 
system.

Put them all together, and you have a >lot< of widgets, a lot of abilities 
under the hood, and a lot of potential for very facile coding.

End of zealotry. The next time someone floats this question I'l 
passionately recommend Fltk or PyAmulet or something.

-- 
  Phlip
       http://www.greencheese.org/LucidScheming
  --  In business always remember:
      The customer is always funny!  --



From peter at engcorp.com  Fri Jun 28 18:45:56 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 28 Jun 2002 18:45:56 -0400
Subject: Dialing out on MODEM
References: <mailman.1025270142.10532.python-list@python.org> <Xns923BE7FFF29D3cliechtigmxnet@62.2.16.82>
Message-ID: <3D1CE724.1B204E88@engcorp.com>

Chris Liechti wrote:
> 
> "David McInnis" <david at dataovation.com> wrote in
> news:mailman.1025270142.10532.python-list at python.org:
> 
> > This is my object
> >
> >>>> import serial
> >>>> modem = serial.Serial
> >>>> modem = serial.Serial(2)
> >>>> modem.write("atdt3120992\n")
> 
> somtimes it it's a '\r\n' and maybe you need to send the escape sequence
> first, to enter the command mode. if i remember correctly thats 3 times ESC
> and then wait 100ms or so. either google or your modem manual should help.
> i had to do somthing like that to dial with pyserial, but unfortunately the
> code is on an other machine...

Most often, it's three plusses "+++" preceded and followed by a one-second
delay (if I recall correctly... I'm amazed I know longer know this with
certainty, given the number of times I've had to do it).

In addition, it's always a wise thing to write an empty line (just
the terminate, whether it's \n or \r\n or \r) before doing anything
after you open the serial port a modem, since a previous application
could have sent a few characters without a terminator and they're still
sitting in the modem's buffer.  It might look to your modem as though
you had actually sent "ATDT31ATDT3120992\n" for example, which wouldn't
help much.

By the way, the need for \r\n has nothing to do with Windows.  It's
entirely up to your modem what termination strings it wants.  Some
can handle just \n I think, and I'm pretty sure others will automatically
adapt to \r\n as well, even if they don't normally take it.  Check
the manual for the modem.

-Peter


From whisper at oz.net  Wed Jun 19 13:40:25 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 19 Jun 2002 10:40:25 -0700
Subject: newbie: help please!
In-Reply-To: <002b01c2178e$3e6693e0$0101010a@local>
Message-ID: <GCEDKONBLEFPPADDJCOECEDGDKAA.whisper@oz.net>

Errr... since that's tk.h doing the include, I am pretty sure that it's
trying to include tcl/include/x11/xlib.h and not one from a X11 distro. You
just need to add tcl/include to your include path.

These files do have some minor differences to add support for Tcl.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Chris Gonnerman
> Sent: Wednesday, June 19, 2002 5:38
> To: Steve
> Cc: python-list at python.org
> Subject: Re: newbie: help please!
>
>
> ----- Original Message -----
> From: "Steve" <steve.bond at btinternet.com>
> Newsgroups: comp.lang.python
> To: <python-list at python.org>
> Sent: Wednesday, June 19, 2002 5:20 AM
> Subject: newbie: help please!
>
>
> > hello,
> >
> > i have python 2.2.1 installed on my mandrake 8.1 system, which seems to
> > work fine.
> >
> > The problem is im trying to install Tk exstensions, i have
> downloaded both
> > tcl8.3.4 and tk8.3.4 files and installed using the ./configure, make and
> > then make install commands. I then when into the setup file in the
> > Python-2.2.1/Modules directory where i understand i have to enable the
> > tkinter extension module, by uncommenting the approritae lines.
> >
> > Lastly i run 'make' at the top of my Python-2.2.1 directory but
> i get this
> > error:
> >
> > gcc -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
> > -DHAVE_CONFIG_H   -DWITH_APPINIT -I/usr/local/include
> -I/usr/X11R6/bin -c
> > ./Modules/_tkinter.c -o Modules/_tkinter.o
> > In file included from ./Modules/_tkinter.c:49:
> > /usr/include/tk.h:83:29: X11/Xlib.h: No such file or directory
> >
> > i have the Xlib.h in the Python-2.2.1/Include directory
> >
> > the tkinter part of my setup file looks like this and everything was
> > installed in the default places
>
> The standard Python 2.2 configuration system should automatically add tk
> support if all components are there.  Your error is because you don't have
> the X11 development RPM(s) installed.  Get out your Mandrake install set
> and add any RPM's which are specifically for X11 development (I run SuSE
> so I don't know what RPM's you need for Mandrake).
>
> The Xlib.h in the Python-2.2.1/Include directory isn't the right one.  You
> need the X11 version.
>
> >
> > # *** Always uncomment this (leave the leading underscore in!):
> >  _tkinter _tkinter.c tkappinit.c -DWITH_APPINIT \
> > # *** Uncomment and edit to reflect where your Tcl/Tk libraries are:
> >         -L/usr/local/lib \
> > # *** Uncomment and edit to reflect where your Tcl/Tk headers are:
> >         -I/usr/local/include \
> > # *** Uncomment and edit to reflect where your X11 header files are:
> >         -I/usr/X11R6/bin \
>
> X11 header files are conventionally in /usr/X11R6/include or in
> /usr/include/X11, and on many distro's both exist, one being a link to the
> other.  /usr/X11R6/bin is almost certainly NOT the right location.
>
> > # *** Or uncomment this for Solaris:
> > #       -I/usr/openwin/include \
> > # *** Uncomment and edit for Tix extension only:
> > #       -DWITH_TIX -ltix8.1.8.2 \
> > # *** Uncomment and edit for BLT extension only:
> > #       -DWITH_BLT -I/usr/local/blt/blt8.0-unoff/include -lBLT8.0 \
> > # *** Uncomment and edit for PIL (TkImaging) extension only:
> > #     (See http://www.pythonware.com/products/pil/ for more info)
> > #       -DWITH_PIL -I../Extensions/Imaging/libImaging  tkImaging.c \
> > # *** Uncomment and edit for TOGL extension only:
> > #       -DWITH_TOGL togl.c \
> > # *** Uncomment and edit to reflect your Tcl/Tk versions:
> >         -ltk8.3 -ltcl8.3 \
> > # *** Uncomment and edit to reflect where your X11 libraries are:
> >         -L/usr/X11R6/lib \
> > # *** Or uncomment this for Solaris:
> > #       -L/usr/openwin/lib \
> > # *** Uncomment these for TOGL extension only:
> > #       -lGL -lGLU -lXext -lXmu \
> > # *** Uncomment for AIX:
> > #       -lld \
> > # *** Always uncomment this; X11 libraries to link with:
> >         -lX11
> >
> > Any help would be much appreciated!
>
> Install the missing RPM's and correct your setup file as noted, and
> all should build fine.
>
> > Thanks
>
> You're welcome.  Hope this helps.
>
> Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
> http://newcenturycomputers.net
>
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From kragen at pobox.com  Sat Jun 15 15:05:34 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 15 Jun 2002 15:05:34 -0400
Subject: IDE recommendations??
References: <b9veea.l85.ln@majestix.gallien.de> <3d0b06bd$0$52043$e4fe514c@dreader3.news.xs4all.nl>
Message-ID: <83lm9gtla9.fsf@panacea.canonical.org>

Boudewijn Rempt <boud at valdyas.org> writes:
> I've tried to use most IDE's out there -- from Pythonwin to Pythonworks. 
> None was really comfortable, except for Wing IDE, which is really powerful. 
> It doesn't include integration of a GUI designer, though, which makes it 
> little more than a glorified editor/debugger (but _very_ glorified). At one
> point there was a version of KDevelop that supported Python development, but
> that has disappeared, it seems.
> 
> I use bash & XEmacs & Designer nowadays, and that gives me a fair 
> productivity.

I'd like to improve XEmacs to be a better Python IDE.  What kind of
integration does it need with Designer?  I haven't used GUI-designer
IDEs much, so I don't really know what features I'm missing.


From jubafre at zipmail.com.br  Wed Jun 26 15:41:41 2002
From: jubafre at zipmail.com.br (jubafre at zipmail.com.br)
Date: Wed, 26 Jun 2002 16:41:41 -0300
Subject: =?iso-8859-1?Q?py2exe=20erro=20message=3F=3F=3F?=
Message-ID: <3D19FB0A000002A8@www.zipmail.com.br>

C:\Python22\Lib\SITE-P~1\py2exe>c:\python22\python setup.py py2exe
usage: setup.py [global_opts] cmd1 [cmd1_opts] [cmd2 [cmd2_opts] ...]
   or: setup.py --help [cmd1 cmd2 ...]
   or: setup.py --help-commands
   or: setup.py cmd --help

error: invalid command 'py2exe' (no module named 'distutils.command.py2exe')

what does mean this error??

Juliano Freitas
www.gebrasil.hpg.com.br



------------------------------------------
Use o melhor sistema de busca da Internet
Radar UOL - http://www.radaruol.com.br







From gabor at realtime.sk  Thu Jun  6 17:36:05 2002
From: gabor at realtime.sk (gabor)
Date: 06 Jun 2002 17:36:05 -0400
Subject: for , ordering
Message-ID: <1023399365.7322.50.camel@wintermute.atriaky.sk>

hi,

i have a list and i want to create a copy of it with a bit different
objects... by now i have something like this:

objects1 = []
objects2 = []
.
.
.

for obj1 in objects1:
	objects2.append(obj2(obj1.x))


1. isn't there a more elegant way to do this?
2. is that guaranteed that the for cycle iterates over the elemenst of
the list in the correct order ( 0,1,2,3,4,....)?

thanks,
gabor







From candiazoo at attbi.com  Thu Jun 27 13:36:13 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Thu, 27 Jun 2002 17:36:13 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com>
Message-ID: <3d1b4d09.306764359@netnews.attbi.com>

         Completed at Thu Jun 27 10:50:13 2002

         11623 function calls in 152.079 CPU seconds

   Ordered by: standard name

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
        1    0.001    0.001  151.962  151.962 <string>:1(?)
      429    0.140    0.000    0.279    0.001 catalog_model.py:109(__init__)
      429    0.148    0.000    0.148    0.000 catalog_model.py:114(set_base_fee)

      429    0.014    0.000    0.014    0.000 catalog_model.py:118(set_page_fee)

      429    0.015    0.000    0.015    0.000 catalog_model.py:121(to_string)
     1703    0.104    0.000    0.104    0.000 catalog_model.py:509(fix_date_rang
e)
      848    0.157    0.000    0.157    0.000 catalog_model.py:57(__init__)
        1  150.396  150.396  151.961  151.961 catalog_model.py:573(generate_cata
log)
      848    0.027    0.000    0.027    0.000 catalog_model.py:66(set_wrk_inst)
      848    0.020    0.000    0.020    0.000 catalog_model.py:69(set_title)
      848    0.017    0.000    0.017    0.000 catalog_model.py:72(set_pub_name)
      848    0.017    0.000    0.017    0.000 catalog_model.py:75(set_std_num)
      848    0.016    0.000    0.016    0.000 catalog_model.py:78(set_date_range
)
      848    0.137    0.000    0.137    0.000 catalog_model.py:81(set_source_id)

       33    0.000    0.000    0.000    0.000 catalog_model.py:820(test_refresh)

       36    0.353    0.010    0.353    0.010 catalog_model.py:823(test_echo)
        1    0.000    0.000    0.000    0.000 catalog_model.py:826(test_cancel)
      419    0.015    0.000    0.033    0.000 catalog_model.py:89(__init__)
      419    0.136    0.000    0.136    0.000 catalog_model.py:92(to_string)
        1    0.000    0.000    0.000    0.000 connections.py:85(cursor)
        1    0.011    0.011    0.189    0.189 cursors.py:108(__do_query)
        1    0.022    0.022    0.022    0.022 cursors.py:135(_fetch_row)
        1    0.000    0.000    0.000    0.000 cursors.py:145(_check_for_warnings
)
        1    0.178    0.178    0.178    0.178 cursors.py:160(_get_result)
        2    0.002    0.001    0.002    0.001 cursors.py:162(close)
        1    0.000    0.000    0.212    0.212 cursors.py:167(_query)
      671    0.027    0.000    0.036    0.000 cursors.py:174(fetchone)
        1    0.000    0.000    0.000    0.000 cursors.py:21(__init__)
        1    0.000    0.000    0.000    0.000 cursors.py:28(__del__)
        2    0.000    0.000    0.000    0.000 cursors.py:31(close)
      671    0.009    0.000    0.009    0.000 cursors.py:36(_check_executed)
        2    0.000    0.000    0.000    0.000 cursors.py:46(_get_db)
        1    0.000    0.000    0.212    0.212 cursors.py:51(execute)
        0    0.000             0.000          profile:0(profiler)
        1    0.117    0.117  152.079  152.079 profile:0(test.generate_catalog())


From Sascha.Ferley at infineon.net  Sun Jun 23 23:02:51 2002
From: Sascha.Ferley at infineon.net (Sascha Ferley)
Date: Sun, 23 Jun 2002 21:02:51 -0600 (MDT)
Subject: PIL, Python and CGI
In-Reply-To: <200206241258.04096.rjones@ekit-inc.com>
Message-ID: <Pine.GSO.4.33.0206232101300.27090-100000@infinity.infineon.net>


On Mon, 24 Jun 2002, Richard Jones wrote:

> On Mon, 24 Jun 2002 12:51, Sascha Ferley wrote:
> > I actually got some nice little traceback that might be of use ..
> >
> > Traceback (most recent call last):
> >   File "/export/server/web/CPSC461/cgi-bin/test4.py", line 10, in ?
> >     import _imaging
> > ImportError: ld.so.1: /usr/local/bin/python: fatal: libjpeg.so.6: open
> > failed: No such file or directory
> >
> > though:
> >
> > infinity:/usr/local/lib->ls -ld libjpeg*
> > lrwxrwxrwx   1 root     other         12 Jun 16 18:58 libjpeg.so ->
> > libjpeg.so.6*
> > lrwxrwxrwx   1 root     other         12 Jun 16 18:58 libjpeg.so.1 ->
> > libjpeg.so.6*
> > -rwxr-xr-x   1 root     bin       137924 May 21  2001 libjpeg.so.6
> > infinity:/usr/local/lib->
> >
> > Which is interesting ... :)
>
> Your web server environment most likely doesn't have /usr/local/lib in its
> LD_LIBRARY_PATH.

You probly mean Apache... I'll take a look for at the conf files for it..
Any ideas where to statically define the lib paths for apache.. i doubt in
the httpd.conf file..
Sascha






From jjl at pobox.com  Tue Jun  4 17:38:27 2002
From: jjl at pobox.com (John J. Lee)
Date: Tue, 4 Jun 2002 22:38:27 +0100
Subject: Login Authentication with cookies...
In-Reply-To: <e271e93c.0206031905.73796d31@posting.google.com>
References: <e271e93c.0206031905.73796d31@posting.google.com>
Message-ID: <Pine.SOL.4.30.0206042233550.10073-100000@mimosa.csv.warwick.ac.uk>

On 3 Jun 2002, chandan wrote:
[...]
> import ClientCookie
[...usage snipped...]
> Error message:
> Traceback (innermost last)
>   File "c:\python22\programs\demos\trial", line 7, in ?
>     result = ClientCookie.urlopen(request)
> NameError: global name 'types' is not defined
[...]

This was fixed a few days ago -- please either download 0.0.7b, or just
add the line 'import types' at the top of ClientCookie/_ClientCookie.py.

Sorry about that -- this will (/should) teach me not to make a 'small
change' without retesting.

BTW, please send bug reports / questions to me rather than the Python list
in the 1st instance --  I may not always notice them here immediately.


John



From whisper at oz.net  Wed Jun 26 03:27:12 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 26 Jun 2002 00:27:12 -0700
Subject: autocoder redux (was Re: Reminder of a python Project.)
In-Reply-To: <3D1941BA.7AFF5C92@engcorp.com>
Message-ID: <GCEDKONBLEFPPADDJCOEEEDPDMAA.whisper@oz.net>

<snip>
> Timothy Rue wrote:
> >
> > On 25-Jun-02 22:38:16 Peter Hansen <peter at engcorp.com> wrote:
> > >I'm afraid there is no longer any need for an autocoder.
> >
> > Geee, maybe you should tell that to NIST
>
> No need.  They called me the other day and I let them know, but
> it was too late to get it into the Computerworld article.
>
> Seriously, there's really no issue with software development any
> more.  The problems are solved.  I guess if you've invented the
> world's first autocoder, it kinda sux to be you then, eh?
>
> -Peter
> --

Yes, software development techniques have gotten so good, they're now going
to have Windows XP running atomic Reactors, Air Traffic Control and the
Space Shuttle. I expect they where encouraged by US Navy tests of a warship
being run on Windows NT where the ship (conventional fueled carrier) ended
up laying dead in the water and tugs had to be summoned to bring her into
port (source: Procedings of the United States Naval Institute). Gives "blue
screen of death" a whole new meaning.

Dave LeBlanc
Seattle, WA USA





From skip at pobox.com  Fri Jun  7 14:00:10 2002
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 7 Jun 2002 13:00:10 -0500
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <200206071747.g57HlbH01057@europa.research.att.com>
References: <yu99vg8v6orb.fsf@europa.research.att.com>
        <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
        <200206071747.g57HlbH01057@europa.research.att.com>
Message-ID: <15616.62634.635544.905055@12-248-41-177.client.attbi.com>

    Andrew> PPS: I tried building one of the .so files by hand, essentially
    Andrew> by snarfing the output from the Makefile log and executing it
    Andrew> manually:

    ...

    Andrew> So apparently it is building a .so file with references
    Andrew> correctly resolved.

That's good to hear

    Andrew> How do I go about finding out the cause of the failure?

Try running the just built python executable and then manually execute
"import struct".  What error message(s) do you see?

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html





From max at alcyone.com  Wed Jun 12 00:43:30 2002
From: max at alcyone.com (Erik Max Francis)
Date: Tue, 11 Jun 2002 21:43:30 -0700
Subject: newbie: copying dictionaries
References: <gdAN8.8432$nn1.3195949@news1.news.adelphia.net>
Message-ID: <3D06D172.FE5F7A23@alcyone.com>

"Jon J. Morin" wrote:

> Hi.  I'm a newbie to python and have a question about copying
> dictionary
> objects.  I'm working out of Learning Python by Mark Lutz & David
> Ascher
> (O'Reilly).  The following bit of code was an exercise in the book and
> it
> works, but what I'm asking is why?  I see where the dictionary keys
> get
> copied from one object to the other, but how do the values get copied
> over
> as well?

With the assignment statement:

	newDict[key] = aDict[key]

This assigns the key in the new dictionary to the value of the old.

Note you could shorten the function using the .update method:

	def copyDict(aDict):
	    newDict = {}
	    newDict.update(aDict)
	    return newDict

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Who'd ever think it / Such a squalid little ending
\__/ The American and Florence, _Chess_
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From wurmy at earthlink.net  Sat Jun  8 11:35:32 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Sat, 08 Jun 2002 15:35:32 GMT
Subject: RE Module How to escape $
References: <ae58a20e.0206080509.44da4bcb@posting.google.com>
Message-ID: <3D0223D5.4050900@earthlink.net>

Prema wrote:
> Hi People !
> I'm trying to use the re.sub function to replace several instances of a
> PHP Variable eg "$Amount". Could someone please help with an example
> Thanks very much
> Mike

Sure...

 >>> import re
 >>> s = "I have $amount dollars"
 >>> re.sub("\$\S+", "foo", s)
'I have foo dollars'
 >>>

HTH,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/



From marc.spoorendonk at pretection.com  Wed Jun 19 07:05:08 2002
From: marc.spoorendonk at pretection.com (Marc Spoorendonk)
Date: Wed, 19 Jun 2002 13:05:08 +0200
Subject: connect to running internet explorer (IE) via shdocvw.dll
Message-ID: <3D106564.4040602@pretection.com>

This is how my problem can be handled in VB:

Dim SWs As New SHDocVw.ShellWindows
Dim IE As SHDocVw.InternetExplorer

For Each IE In SWs
       Set Doc = IE.document
       If TypeOf Doc Is HTMLDocument Then
          List3.AddItem Doc.url
       End If
Next

The question is simple: How can I accomplish this same thing in Python?

Mark Hammond answered me this by email:
"I think you will find SHDocVw is a vtable interface, and therefore not 
supported by Python."

I've read up about vtable, COM etc. but I am fairly new to this. (I was 
born as a Unix programmer.)  I've played with the interfaces and 
makepy.py. Should I write a python-plugin in C++ or can it be done by 
native pytoncode?

Regards,

Marc



From cliechti at gmx.net  Thu Jun 13 17:24:09 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 13 Jun 2002 23:24:09 +0200
Subject: Type subclassing: bug or feature
References: <aeau48$111$1@panix1.panix.com>
Message-ID: <Xns922CEE783C10Dcliechtigmxnet@62.2.16.82>

aahz at pythoncraft.com (Aahz) wrote in news:aeau48$111$1 at panix1.panix.com:

> Consider the following code:
> 
> class MyStr(str):
>     def contains(self, value):
>         return self.find(value) >= 0
> 
> s = MyStr("hello, world!")
> s = s.capitalize()
> if s.contains('Hello'):
>     print "Found it!"
> 
> It fails with an AttributeError when it calls s.contains(), because
> s.capitalize() returned a str instead of a MyStr.  Anyone want to take a
> whack at defending this as the correct behavior?

solving this at the C level would require patching many stingobject factory 
functions... e.g. PyString_FromStringAndSize for capitalize(). but it would 
indeed be desireable.

i think it would like to have a even more genral mechanism. i would like to 
have a possibility to replace the builtin type. e.g. i could replace the 
builtin "str" and all strings from now on would be instaces of my class. 
but that's clearly impossible in the current implementation.

<hacker warning>
like Bengt already sugested just automated ;-) ...

>>> class mstr(str):
... 	for name,m in str.__dict__.items():
... 		if type(m) == type(str.find):
... 			exec """def %s(*args, **kwargs):
... 					q = str.%s(*args, **kwargs)
... 					if type(q) is str: return mstr(q)
... 					else: return q\n""" % (name, name)
... 	def __contains__(self, other):
... 		return str.find(self, other) >= 0
... 	
>>> s = mstr("hello")
>>> type(s.capitalize())
<class '__main__.mstr'>
>>> 'll' in s.capitalize()
1
</hacker warning>

chris

-- 
Chris <cliechti at gmx.net>



From emile at fenx.com  Thu Jun 27 10:14:35 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 27 Jun 2002 14:14:35 GMT
Subject: Difference between HTMLParser and htmllib.HTMLParser
References: <mailman.1025178934.31606.python-list@python.org>
Message-ID: <f5FS8.325922$352.35979@sccrnsc02>

"Dave Swegen" <dswegen at software.plasmon.com> wrote in message
news:mailman.1025178934.31606.python-list at python.org...
> I've just started playing with these, and I'm curious which is the
> better of the two (I noticed the HTMLParser module only came in in
2.2).
> And is there any reason why the names are so confusingly similiar? I
> tried searching in google for info on this, but the daft naming made
it
> rather hard ;)
>

A quick look at the code shows that HTMLParser is part of a family that
subclasses markupbase.ParserBase and doesn't need to specify each of the
individual tags it understands, allowing it to adapt more easily.

>From the initial cvs checkin just over a year ago:
(ref
http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/dist/src/Li
b/HTMLParser.py )

"""
A much improved HTML parser -- a replacement for sgmllib.  The API is
derived from but not quite compatible with that of sgmllib, so it's a
new file.  I suppose it needs documentation, and htmllib needs to be
changed to use this instead of sgmllib, and sgmllib needs to be
declared obsolete.  But that can all be done later.

This code was first published as part of TAL (part of Zope Page
Templates), but that was strongly based on sgmllib anyway.  Authors
are Fred drake and Guido van Rossum.
"""

HTH,

--

Emile van Sebille
emile at fenx.com

---------



From look at sig.invalid  Mon Jun 24 04:57:47 2002
From: look at sig.invalid (Simo Salminen)
Date: Mon, 24 Jun 2002 08:57:47 +0000 (UTC)
Subject: HTTP authentication
References: <TWtR8.86654$Am4.23833590@twister.columbus.rr.com> <3D166ED4.FBE9E2C8@alcyone.com> <3D1679BF.7F2D3A1@engcorp.com>
Message-ID: <slrnahdnoc.iop.look@pikkukorppi.cs.tut.fi>

* Peter Hansen [Sun, 23 Jun 2002 21:45:35 -0400]
> I think another solution is just to specify the username and 
> password in the URL like so:
> 
>  ref = urllib.urlopen("http://username:password at host.domain/path/file")
> 

and if this does not work, the site might be using digest-authentication
instead of basic. then one can use urllib2 and HTTPDigestAuthHandler (look
urllib2.py __doc__ for mor info).

-- 
simo <dot> salminen <at> iki <dot> fi


From goodger at users.sourceforge.net  Thu Jun 27 21:53:33 2002
From: goodger at users.sourceforge.net (David Goodger)
Date: Fri, 28 Jun 2002 01:53:33 GMT
Subject: Python Documentation generation
In-Reply-To: <mailman.1024599266.494.python-list@python.org>
References: <mailman.1024599266.494.python-list@python.org>
Message-ID: <B9413A18.2506B%goodger@users.sourceforge.net>

Mahrt, Dallas wrote:
> I have found several systems for generating documentation of Python modules
> (HappyDoc, pythondoc, pydoc) but I don't believe any of them fulfill all of
> these desires. Any suggestions from the group?

I'd say that HappyDoc is the best there is right now.  If you're
looking for something better, and have ideas as to what could make
it better, how about helping out with Docutils
(http://docutils.sourceforge.net/)?  We're working on generating
docs from Python modules, but we've a long way to go yet.  Input
and participation is welcome!

-- 
David Goodger  <goodger at users.sourceforge.net>  Open-source projects:
  - Python Docutils: http://docutils.sourceforge.net/
    (includes reStructuredText: http://docutils.sf.net/rst.html)
  - The Go Tools Project: http://gotools.sourceforge.net/



From opengeometry at NOSPAM.yahoo.ca  Thu Jun 27 14:45:16 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 27 Jun 2002 18:45:16 GMT
Subject: Pythoniac: Thoughts on a hardware Python processor
References: <afc52t$2pj$1@sirius.dur.ac.uk>
Message-ID: <affmfs$e5snt$1@ID-99293.news.dfncis.de>

Christopher Saunter <Christopher.Saunter at durham.ac.uk> wrote:
> Dear All,
> 
> I have seen the occasional post about creating a specific hardware 'Python
> Processor', that could then be built and used in programmable logic or the
> like (Valves, relays, core memory - Pythoniac ;-) - in other words, create
> a processor that executes Python bytecode directly.  There would be
> various different reasons for doing this (the 'neat' factor etc...)  

The only route that has any chance of success is 
    Python -> Forth -> stack engine
which leverage works that has already gone into Forth area.  But, my days
of hand assembling 68000 or programming HP calculators are long gone... :-)

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin


From K.Rdt at TU-Berlin.DE  Thu Jun 27 03:46:38 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Thu, 27 Jun 2002 09:46:38 +0200
Subject: system-output to wxpython-window
In-Reply-To: <7647A9F4B298E74DB6793865DA67285004AD94@exchange.adrembi.com>
Message-ID: <963VF032VR76ARQMHDC93ZX61WT32VS.3d1ac2de@FRITZweb>

Am 26.06.02 15:27:51, schrieb "Roman Yakovenko" <romany at actimize.com>:

>Also I think after a few minutes of running this code you will want to write multi threading. 
>It is also simple. I have working example. I think exactly what you need. 
---------------------------------------------------------------------
Hi

Can you give me an example. I read the
doc and tried this ..

import sys,os,thread
# y=thread.start_new_thread("c:\\windows\\netstat.exe","-a -n tcp 20")
y=thread.start_new_thread("dir","-a -n tcp 20")
sys.exit()

but it naturally -) doesn't work.

		K at Rdt
---------------------------------------------------------------------







From pyth at devel.trillke.net  Sun Jun 30 03:58:30 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 30 Jun 2002 09:58:30 +0200
Subject: Python needs better error reporting
In-Reply-To: <3D1E5376.32716.DF5C94@localhost>; from opus@value.net on Sun, Jun 30, 2002 at 12:40:22AM -0700
References: <3D1EAF12.84CE5914@cascade-sys.com> <3D1E5376.32716.DF5C94@localhost>
Message-ID: <20020630095830.C20310@prim.han.de>

Opus wrote:

> At the most, I might suggest that there be an option to run just the parser, 
> with more verbose output.  Maybe even a different parser would be in order.  
> Idea for anyone????

i'd suggest a heuristic approach.  Compute statistics for which token
usually comes after a group of others. Just look at AST-structures not
values.  This way you can suggest a hint to the user what he might have
done wrong.  I think it's not feasible and too verbose to tell the user
what is actually correct.  A simple "Maybe you forgot ..." would cover
90% of the cases, i guess.

    holger




From hesterloli at hotmail.com  Mon Jun 17 04:21:31 2002
From: hesterloli at hotmail.com (George Hester)
Date: Mon, 17 Jun 2002 08:21:31 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org>
Message-ID: <f_gP8.42357$GY.12362923@twister.nyroc.rr.com>

Probably not.  Where do I import that from and how do it?  I am trying to put a
script in ASP using Python as the script language to give me the version of
Python that I am running when I call the ASP.  The more the script works out of
the box the better I'll be able to catch on to this language.  Thanks.

--
George Hester
_________________________________
"Gerhard H?ring" <gerhard at bigfoot.de> wrote in message
news:mailman.1024182500.12788.python-list at python.org...
> * George Hester <hesterloli at hotmail.com> [2002-06-15 22:58 +0000]:
> > Python ActiveX Scripting Engine error '80020009'
> >
> > Traceback (innermost last): File "<Script Block >", line 5, in ?
> > Response.Write(sys.version) NameError: name 'sys' is not defined
> >
> > /python.asp, line 14
> >
> > Response.Write(sys.version)
> >
> > I think I need a script to do this.
>
> You did not import the sys module.
>
> Gerhard
> --
> mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
> web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
> public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
> reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))
>
>




From fperez528 at yahoo.com  Wed Jun 19 19:01:03 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 19 Jun 2002 17:01:03 -0600
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu>
Message-ID: <aer2fn$frc$1@peabody.colorado.edu>

Tom Loredo wrote:

> I've been using Python heavily on a Solaris box and a Mac for a
> few years now.  For reasons I'll spare you from, I now am venturing
> into Windows territory (a Dell Inspiron 7500 laptop, if that matters).

If you're an old unix hand and want to preserve your sanity, you have two 
options:

1- put linux on that laptop and be happy.
2- if (1) is not an option for some reason, put cygwin on it. At least you'll 
have a decent working environment. If you have to work in the normal, 
crippled windows environment, that laptop is going to suffer a painful death 
after being hurled through the nearest window very soon :)

Cheers,

f.


From gerhard at bigfoot.de  Tue Jun 18 21:54:56 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 19 Jun 2002 03:54:56 +0200
Subject: I have .py, I want .exe, how??
In-Reply-To: <3D0F6623000012D8@www.zipmail.com.br>
References: <3D0F6623000012D8@www.zipmail.com.br>
Message-ID: <20020619015456.GC2739@lilith.my-fqdn.de>

* jubafre at zipmail.com.br <jubafre at zipmail.com.br> [2002-06-18 22:01 -0300]:
> i have a file.py how i make e executable file .exe for run in
> windows???

It's a FAQ and answered here:

http://www.python.org/cgi-bin/faqw.py?req=show&file=faq04.028.htp

Gerhard, who considers making an editor macro for this question
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 22.4 ?C      Wind: 1.0 m/s




From fperez528 at yahoo.com  Thu Jun  6 05:00:19 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 06 Jun 2002 09:00:19 +0000
Subject: self
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu> <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com> <mailman.1023294728.2322.python-list@python.org> <sWsL8.82$v_7.4260962@newnews.cc.stevens-tech.edu> <adls9d$7ji$1@peabody.colorado.edu> <pvKL8.84$v_7.4260962@newnews.cc.stevens-tech.edu>
Message-ID: <adntop$bnl$1@peabody.colorado.edu>

Vojin Jovanovic wrote:

>> So ':=' variables would first try to evaluate their expressions in their
>> normal namespace (that of the class, the problem you're dealing with now)
> and
>> would then look for 'normal' variable names to resolve. Only if a name
>> wasnt't found in either would a NameError be raised.
> 
> Your idea is good.  It would be exactly what I need.  It would bring a lot
> to python; a kind of symbolic manipulation capability.  Unfortunately, I
> don't think that my current knowledge of Python is sufficient to implement
> this in a general way you are suggesting.

Well, I've given it a bit of thought on the side but unfortunately right now 
I'm swamped with other things. I'll put it in the idea-bucket for IPython, 
hopefully later I'll have a chance to look at it. By the way, the ':=' part 
is absolutely trivial to implement for me, the tricky side is the problem 
you've been struggling with. So if you ever can build a class which evaluates 
the way you've been trying, I could easily incorporate it into the ipython 
framework with a special syntax pre-processor (I already have that 
infrastructure in place).

But the problem is not trivial, esp. if you want to be able to mix and match 
'normal' python variables with these 'delayed evaluation' ones. Good luck :)

f


From tismer at tismer.com  Wed Jun 19 09:38:46 2002
From: tismer at tismer.com (Christian Tismer)
Date: Wed, 19 Jun 2002 15:38:46 +0200
Subject: Tcl adept wanted for Stackless problem
Message-ID: <3D108966.9050402@tismer.com>

Dear Lists,

there is still some problem with Tcl and Stackless hiding.

Is there anybody around who knows the Tcl/Tk sources
about as I know Python's?

My big question is:
When does Tcl use C stack entries as globals, which
are passed as function arguments to interpreter calls?
This is of special interest when Python callbacks
are invoked.
The problem is, that I use stack slicing, which moves
parts of the C stack away at some times, and I need
to know when I am forbidden to do that. It is solved
to some extent, but not all.

That's unfortunately not all.
A friend is running the Tcl/Tk mainloop in one real thread,
and stackless tasklets are running in another one (of course
calling into Tcl). When tasklet switches are performed
(again, this is moving stack contents around), there appear
to be crashes, too.
This kind of slicing should be allowed IMHO, since these
are different contexts, which shouldn't interact at all.
Are there any structures which are shared between threads
that use Tcl/Tk? Something that may not disappear during
some operations?

Which part of the documentation / the Tcl/Tk source should
I read to find this out wihout learning everything?

This is really an urgent problem which is blocking me.
If somebody has good knowledge, please show up! :-)

thanks & cheers - chris

-- 
Christian Tismer             :^)   <mailto:tismer at tismer.com>
Mission Impossible 5oftware  :     Have a break! Take a ride on Python's
Johannes-Niemeyer-Weg 9a     :    *Starship* http://starship.python.net/
14109 Berlin                 :     PGP key -> http://wwwkeys.pgp.net/
work +49 30 89 09 53 34  home +49 30 802 86 56  pager +49 173 24 18 776
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
      whom do you want to sponsor today?   http://www.stackless.com/






From roger at efn.org  Fri Jun 28 19:58:12 2002
From: roger at efn.org (Roger)
Date: Fri, 28 Jun 2002 16:58:12 -0700
Subject: Why the colon?
In-Reply-To: <4.3.2.7.2.20020628190712.02450328@mail.gte.net>; from res04ft9@gte.net on Fri, Jun 28, 2002 at 07:33:31PM -0400
References: <XFMail.20020628153929.pedrobados@eresmas.net> <4.3.2.7.2.20020628190712.02450328@mail.gte.net>
Message-ID: <20020628165812.A2224@efn.org>

Around Fri,Jun 28 2002, at 07:33,  Erv Young, wrote:
> 
> 1. def ask_ok(prompt, retries=4, complaint='Yes or no, please!'):
> 2.     while 1:
> 3.         ok = raw_input(prompt)
> 4.         if ok in ('y', 'ye', 'yes'): return 1
> 5.         if ok in ('n', 'no', 'nop', 'nope'): return 0
> 6.         retries = retries - 1
> 7.         if retries < 0: raise IOError, 'refusenik user'
> 8.         print complaint
> 
> 
> Then what am I to make of the colons at the end of lines 1 and 
> 2?  "Yoo-hoo, interpreter!  I'm about to start indenting now!"  Hmmm....
> 

http://python.org/doc/current/tut/node5.html
section 3.2

essentially,

while  1:   # : indicates beginning of a block of code
   line a
   line b
   line c
line d

there is no 'end' statement.  Quit indenting when you want the block to
end.  lines a,b and c will be executed for the 'while' loop.


-- 
Roger
---------------------------------------------------------------------------




From huaiyu at gauss.almadan.ibm.com  Wed Jun 19 15:23:40 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 19 Jun 2002 19:23:40 +0000 (UTC)
Subject: What If.....    Strong Types
References: <3D0C0F0D.4030900@bgb.cc> <slrnagv0ue.46u.huaiyu@gauss.almadan.ibm.com> <3D100C9A.10502@bgb.cc>
Message-ID: <slrnah1mhs.6o5.huaiyu@gauss.almadan.ibm.com>

Don Garrett <garrett at bgb.cc> wrote:
>
>   Have you, by chance, ever used Sather? It was (I think development is 
>stopped) an Eiffel inspired language that seperated interface and 
>implementation inheritance (amoung several other cool ideas). Unfotunatly, I 
>never got to use it for any large projects, so never felt that I learned it well.
>
>   http://www.icsi.berkeley.edu/~sather/

Not yet.  It's on my todo list.  Thanks for pointing out this feature, which
moves it much higher on my list. 

Huaiyu


From pinard at iro.umontreal.ca  Tue Jun 25 11:07:03 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 25 Jun 2002 11:07:03 -0400
Subject: Parsing strings (\n and \\)
In-Reply-To: <3D188006.7050202@thomas-guettler.de>
References: <3D188006.7050202@thomas-guettler.de>
Message-ID: <oq1yavz9bc.fsf@titan.progiciels-bpi.ca>

[Thomas Guettler]

> Is there a function for parsing c-like
> strings? I need to parse strings containing
> '\n' and '\\' '\"'.

A simple avenue is to get Python itself to evaluate the string as a constant
(you ensure this by removing evaluation context).  Something like this:

    PARSED = eval(UNPARSED, {}, {})

But this is a bit rough, there are cases you want more precise control
over the string syntax.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From mgerrans at mindspring.com  Sun Jun 23 17:02:09 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Sun, 23 Jun 2002 14:02:09 -0700
Subject: Python hits the spot
References: <mailman.1024772248.9399.python-list@python.org> <yppfsn3dlpco.fsf@hotlips.cs.chalmers.se> <3D1634ED.B9ACF9EC@engcorp.com>
Message-ID: <af5d0r$nlt$1@slb4.atl.mindspring.net>

> > for i in range(50*1000*1000):
> >    pass

You call that a bug?   Did you consider trying xrange()?




From sandeep182 at hotmail.com  Wed Jun 26 15:22:42 2002
From: sandeep182 at hotmail.com (Sandeep Gupta)
Date: Wed, 26 Jun 2002 19:22:42 GMT
Subject: Pychecker and /var/tmp/python2-2.2-root/usr/lib/python2.2/.....
References: <d71cc7f6.0206241327.17ebabfa@posting.google.com> <pan.2002.06.25.01.46.31.138959.7184@metaslash.com> <d71cc7f6.0206251220.4ae17aa0@posting.google.com>
Message-ID: <6woS8.9386$4a2.453054@typhoon.austin.rr.com>

I found the solution:
cd /
sudo python2 /usr/lib/python2.2/compileall.py -f /usr/lib/python2.2


Thanks for the lead, Neal!

"Sandeep Gupta" <sandeep182 at hotmail.com> wrote in message
news:d71cc7f6.0206251220.4ae17aa0 at posting.google.com...
> Neal Norwitz <neal at metaslash.com> wrote in message
news:<pan.2002.06.25.01.46.31.138959.7184 at metaslash.com>...
> > On Mon, 24 Jun 2002 17:27:30 -0400, Sandeep Gupta wrote:
> >
> > > I'm using pychecker on my application, which uses BaseHTTPServer.py
and
> > > unittest.py.  Both those files have "No method" warnings that are
> > > displayed as:
> > > /var/tmp/python2-2.2-root/usr/lib/python2.2/BaseHTTPServer.py and
> > > /var/tmp/python2-2.2-root/usr/lib/python2.2/unittest.py
> > >
> > > Does anyone know why /var/tmp/python2-2.2-root gets prepended to the
> > > file name?
> > > How can I remove the warnings for these files?
> >
> > I had this same problem.  From what I remember, Redhat compiles
> > the .py -> .pyc/.pyo files in the /var/tmp/.... directory.
> >
> > I didn't find a great way to fix it in pychecker, but to
> > fix your problem, you should just have to do:
> >
> > cd /usr/lib/python2.2
> > python compileall.py .
> >
> > Neal
>
> I tried the above steps replacing the second line with:
> sudo python2 compileall.py .
> Successful compile.  I verified that
> /usr/lib/python2.2/BaseHTTPServer.pyo and unittest.pyo exist.  But I'm
> still receiving the same error message from pychecker.  Any other
> ideas?
> Thanks
> Sandeep




From gnufnork at fuckspam.hetdigitalegat.nl  Sun Jun 23 14:24:15 2002
From: gnufnork at fuckspam.hetdigitalegat.nl (Wladimir)
Date: Sun, 23 Jun 2002 20:24:15 +0200
Subject: The best GUI library in python
References: <8d3f4438.0206220457.b820418@posting.google.com> <3D1481DE.C0331204@engcorp.com> <8d3f4438.0206221422.6f71287a@posting.google.com> <3D14FFB5.90808865@engcorp.com> <8d3f4438.0206230713.535169aa@posting.google.com>
Message-ID: <pan.2002.06.23.18.24.15.215365.242@fuckspam.hetdigitalegat.nl>

On Sun, 23 Jun 2002 17:13:09 +0200, d2002xx wrote:

> Currently, I know only wxPython and PyQT, they are fast enough, but I
> want to know are there any GUIs better than the two.
 
python-gtk and python-gnome?

Wladimir


From dkuhlman at rexx.com  Thu Jun 27 13:11:11 2002
From: dkuhlman at rexx.com (Dave Kuhlman)
Date: Thu, 27 Jun 2002 10:11:11 -0700
Subject: [Q] How Can I call python function from C source?
References: <afe3ji$gvk$1@news.kreonet.re.kr>
Message-ID: <affguf$e3cn5$1@ID-139865.news.dfncis.de>

Kim Chul Min wrote:

> Hello there,
> 
> I want to call python function from c source.
> 
> for example,
> 
> in file "main.c"
> --------------------------------------------------------------
> int main(void)
> {
> 
>     /** want to call python function add(a,b) that return a+b **/
> 
> }
> --------------------------------------------------------------
> 
> in the file "add.py"
> --------------------------------------------------------------
> def add(a,b):
>     return(a+b)
> --------------------------------------------------------------
> 
> Can I call any function that written by python from C source ?
> 
> If possible, can anyone show me simple example like above??
> 
> Plz,,, Help me....Thankx in advance...
> 
> Ciao,
> Farangan Xavio
> cmkim at shinbiro.com

Here is a simple example.

You will find the following especially helpful:

    http://www.python.org/doc/current/api/importing.html
    http://www.python.org/doc/current/api/object.html

/* =========================================== */

#include "Python.h"

int main(int argc, char ** argv)
{
    PyObject * module;
    PyObject * function;
    PyObject * result;
    long cresult;
    int arg1;
    int arg2;

    if (argc != 3)
    {
        printf("Usage: test1 <arg1> <arg2>\n");
        exit(-1);
    } /* if */

    arg1 = atoi(argv[1]);
    arg2 = atoi(argv[2]);

    Py_Initialize();
    module = PyImport_ImportModule("add");
    if (PyObject_HasAttrString(module, "add"))
    {
        function = PyObject_GetAttrString(module, "add");
        if (PyCallable_Check(function))
        {
            result = PyObject_CallFunction(function, "ii", arg1, arg2);
            cresult = PyLong_AsLong(result);
            printf("arg1: %d  arg2: %d  result: %d\n", arg1, arg2, cresult);
        }
        else
        {
            printf("Error -- attribute is not a function/method\n");
        } /* if */
    }
    else
    {
        printf("Error -- Module does not contain \"add\" function.\n");
    } /* if */
    Py_Exit(0);
} /* main */

/* =========================================== */



-- 
Dave Kuhlman
dkuhlman at rexx.com
http://www.rexx.com/~dkuhlman



From dgallion1 at yahoo.com  Tue Jun 25 09:31:33 2002
From: dgallion1 at yahoo.com (Darrell)
Date: 25 Jun 2002 06:31:33 -0700
Subject: PYREX
References: <3d1749d3.19107439@news1.news.adelphia.net>
Message-ID: <b0c5da3a.0206250531.1fe72a82@posting.google.com>

I've spent very little time with it. Just made it work and that's why
I love it. I use MSC 6 and the following setup.py, which I got from
Alex M. at the activestate python cookbook.

from distutils.core import setup, Extension
setup (name = "",
       version = "1.0",
       maintainer = "",
       maintainer_email = "",
       description = "Sample Python module",
       ext_modules = [Extension('tcb',sources=['tcb.c','tcbBase.c'])]
)

#=======================
tcbBase.c has the detailed 'c' code and tcb.pyx has the interface.

cdef extern int openTcbC(int readTimeOut)
cdef extern int writeTcbC(char *cmdData, unsigned int cmdLen)
cdef extern char *readTcbC(unsigned int *bytesRead)
cdef extern PyString_FromStringAndSize(char *v,  int len)
cdef extern int closeTcb()

def open(int readTimeOut):
    return openTcbC(readTimeOut)

def write(data):
    return writeTcbC(data, len(data))

def read():
    cdef unsigned int bytesRead
    cdef char *s
    s=readTcbC(&bytesRead)
    p=PyString_FromStringAndSize(s, bytesRead)
    return p

def close():
    closeTcb()

#==============================
This batch builds it all

python \python22\pyrexc.py tcb.pyx
setup.py build
copy C:\Python22\Lib\site-packages\Pyrex\Demos\build\lib.win32-2.2\*.pyd
.

#Yes I built in the demo directory :(

--Darrell

edlsoft at mindspring.com (Burt Leavenworth) wrote in message news:<3d1749d3.19107439 at news1.news.adelphia.net>...
> Has anyone installed Pyrex on Windows? If so, can you share the
> details?
> 
> Burt


From edlsoft at mindspring.com  Mon Jun 24 12:34:57 2002
From: edlsoft at mindspring.com (Burt Leavenworth)
Date: Mon, 24 Jun 2002 16:34:57 GMT
Subject: PYREX
Message-ID: <3d1749d3.19107439@news1.news.adelphia.net>

Has anyone installed Pyrex on Windows? If so, can you share the
details?

Burt


From romany at actimize.com  Sun Jun 16 10:00:20 2002
From: romany at actimize.com (Roman Yakovenko)
Date: Sun, 16 Jun 2002 16:00:20 +0200
Subject: ConfigParser
Message-ID: <7647A9F4B298E74DB6793865DA67285004B354@exchange.adrembi.com>

Hi all. I need some help with this class(ConfigParser). 
(It seems that help it is not enough for me).

I have one ini file. I have some dialog to configure this file. 
Also I have some program that will run using this ini file. 
I'd like to use interpolation( %(base_path)temp\run.exe).

The problems is: when I use this file from GUI I don't want variables to be expanded,
but when I use it from "run" program I do want it. How I do it?

One answer is to use 2 files. Is there are others?

Thanks. Roman.






From DLNXPEGFQVEB at spammotel.com  Thu Jun 20 10:34:36 2002
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Thu, 20 Jun 2002 17:34:36 +0300
Subject: How to find all emails (on the same line)
References: <mailman.1024560806.17188.python-list@python.org>
Message-ID: <omp3huc333dcpvur2g4ja29h8f1gqumdic@4ax.com>

On Thu, 20 Jun 2002 10:12:20 +0200, rumours say that "A"
<printers at sendme.cz> might have written:

>Hi,
>I need to find ALL emails in a file. I can use re module but I do not 
>know how to find all emails if there are more emails on the same 
>line because re module ( as far as I know )finds only one/first 
>occurrence on that line.
>Thanks for help.
>Ladislav

At first, as /F noted first, you are very vague.
Presumably you want to find all e-mail *addresses* in a text file.
If that is the case, you can start by opening a Python interpreter, and
give:

help(file.read)
import re
help(re.findall)

terse-but-hopefully-not-as-vague-ly y'rs
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')


From shalehperry at attbi.com  Wed Jun  5 16:41:21 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Wed, 05 Jun 2002 13:41:21 -0700 (PDT)
Subject: Detecting OS and Window Manager
In-Reply-To: <slrnafstki.3ne.tim@vegeta.ath.cx>
Message-ID: <XFMail.20020605134121.shalehperry@attbi.com>

On 05-Jun-2002 Tim Hammerquist wrote:
> Andrew Wilkinson graced us by uttering:
>> I'm writing a program that integrates with a number of different OSes
>> and Window Managers (under Linux), but I need to check what the user
>> is running.  I know that...
>> 
>> import os
>> if os.name == 'nt':
>> 
>> will tell me they're running NT. But under Linux I'd like to know if
>> they are running KDE or Gnome, does anyone know a (preferably) simple
>> test for this?
> 
> Well, windowmanagers aren't as static as operating systems are.  It's
> actually possible to run more than one wm at a time.
> 
> In any case, I went and experimented with the various wms on my system
> and in all cases the $WINDOWMANAGER shell var was set to the full path
> of the running wm executable.  This appears to be the only consistent
> and reliable var in my environment; for instance, here's a dump of my
> wm-related vars:
> 

WINDOWMANAGER is not one I am familiar with -- is KDM or GDM setting it?  None
of the machines I looked at had it set.




From emile at fenx.com  Sat Jun  1 13:20:42 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 01 Jun 2002 17:20:42 GMT
Subject: Generating unique numbers?
References: <3CF7EBC8.4080802@verio.net> <3CF8FC1C.7060903@lindbergs.org>
Message-ID: <Kn7K8.81436$352.4720@sccrnsc02>

VanL
> Based on what people have said, I was thinking of using some
> combination of DateTime.gmticks and the id of the obj,
> making it (I think) extremely unlikely that there would be a
> collision.


Last time GUIDs got my attention I found the link in this post helpful:

http://groups.google.com/groups?selm=3051e77a.0201130853.557edffc at postin
g.google.com

--

Emile van Sebille
emile at fenx.com

---------



From dom at edgereport.put_a_c_o_m_here  Thu Jun  6 15:34:30 2002
From: dom at edgereport.put_a_c_o_m_here (Domenic R. Merenda)
Date: Thu, 06 Jun 2002 19:34:30 GMT
Subject: Medium to Large Scale Python Deployments
References: <KHkL8.21383$eD2.2399923271@newssvr10.news.prodigy.com> <b10fcc9a.0206060655.443d06c2@posting.google.com>
Message-ID: <aPOL8.21620$tl1.2488471578@newssvr10.news.prodigy.com>

Benji,

   At some point, I think it is likely.  The company is in the process of
being acquired, so it's really up to the new owners.  I am a 20%
stockholder, and I've brought up the issue already.  The only major
complication is giving up the competitive advantage.  And, believe me, we'd
love the community participation.  We only have 7 programmers working on the
project now.


--
Domenic R. Merenda
Editor, The Edge Report
http://www.edgereport.com


"Benji York" <benji_york at cknainc.com> wrote in message
news:b10fcc9a.0206060655.443d06c2 at posting.google.com...
> "Domenic R. Merenda" <dom at edgereport.put_a_c_o_m_here> wrote in message
news:<KHkL8.21383$eD2.2399923271 at newssvr10.news.prodigy.com>...
> >    I am running a home-grown Enterprise Resource Planning system,
written
> > entirely in Python...
>
> Do you have any plans to open source your system. I, and others, would
> be most interested in using, and contributing to such a project.
>
> --
> Benji York
> benji_york at cknainc.com




From wesc at deirdre.org  Sat Jun  1 13:25:44 2002
From: wesc at deirdre.org (Wesley J. Chun)
Date: 1 Jun 2002 10:25:44 -0700
Subject: Core Python programming
References: <337619fa.0205121618.26e557e0@posting.google.com> <3CE30D2A.2090502@woeltje.org>
Message-ID: <fcd9edc1.0206010925.6b2ea2d0@posting.google.com>

> Raphael Ribeiro wrote:
> > I wanna buy this book , for i heard it is PERFECT for a beginner , but
> > i saw it , and it was released in 2000, will i get obsolete if buy it
> > today, and will i have problems with the new features included after
> > 2000?


would it be a conflict-of-interest to post to this thread?  :-)

the book was sent to the printers in Dec 2000 and released
late that month and Jan 2001, depending on the distributor.
the writing phase of the book was quite challenging as i
started with 1.5.2 as the most current release.  as the
final deadlines approached, 1.6 and 2.0 became reality
rather quickly, and i had to lobby the publisher to delay
the release of the book until i had been able to bring the
book up-to-date as much as possible.

the book was designed for readers familiar with programming
to pick up Python as quickly as possible.  yes, it is par-
tially a reference book, but certainly not to the extent of
Beazley's Python Essential Reference -- it's a portable
version of the Python Library Reference!

in fact, once you "graduate" from Core Python Programming,
PER is a great book to carry around with you -- I know I do!
then if you have to go back to look up a topic for further
elucidation, then you can pick up CPP again.

some people have asked whether a 2e of CPP is forthcoming,
and i have to say that this is not the right timeframe yet.
i have to integrate all the changes from 2.1 and 2.2 into
the meat of the book as well as add new sections which i
felt were missing when the book went to press.  such chap-
ters may include topics such as:  Internet Client Program-
ming (FTP, NNTP, POP3, etc.), Databases/SQL, Zope, Win32
Programming, XML, Java/Jython/Swing, etc.

in the meantime, i will be coming out with some supplemental
reading, i.e., the new features found in 2.1, 2.2, etc., and
posting the PDFs to the book's website, keeping CPP as cur-
rent as possible.  look for those in the next few months.

i've also written a few articles for Linux Journal.  Jun 2002
features an interview with Guido regarding 2.2, and July 2002
will have an expose on some of the new features in 2.2 and
how to use them.

you can also read others' reviews about CPP on the book's
website here:

http://starship.python.net/crew/wesc/cpp/reviews.htm

David Mertz has finally reviewed CPP, adding one more
article to his bank of Python book reviews:

http://www-106.ibm.com/developerworks/linux/library/l-pbook3.html

if you have any questions, feel free to write!  i'm
open to any and all suggestions for the 2nd edition
as well as other topics you may be interested in.

cheers!

-wesley

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

"Core Python Programming", Prentice Hall PTR, ? 2001
    http://starship.python.net/crew/wesc/cpp/

Silicon Valley-San Francisco Bay Area Python Users Group (BayPIGgies)
    http://deirdre.org/baypiggies

wesley.j.chun :: wesc at deirdre.org
cyberweb.consulting : henderson, nv : cyberweb_consulting at yahoo.com
http://roadkill.com/~wesc/cyberweb/


From imbosol at vt.edu  Mon Jun 24 13:44:20 2002
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 24 Jun 2002 13:44:20 -0400
Subject: Generic Python
References: <3D16E465.F0D2509E@ifib.uni-karlsruhe.de>
Message-ID: <kpl7fa.kb.ln@grey.aerojockey.localdomain>

Uwe Mayer wrote:
> Is it somehow possible to have the "base2" class return a class instead
> of an instance?
> Is there another way to solve this problem?


In Python 2.1 and above, you can take advantage of nested scopes to
generate different classes from a template.  The idea is to put the
template class definition inside a function, and have the class refer
to the function's arguments.

def template_factory(baseclass,text):
    class template(baseclass):
        def output(self):
            print text
    return template

Remember to use from __future__ import nested_scopes if you're on
Python 2.1.  If you're using 2.0 or lower, you can still fake the
nested scopes thing with a default argument:

def template_factory(baseclass,text):
    class template(baseclass):
        def output(self,text=text):
            print text
    return template


Here's an example session:

Python 2.1.3 (#1, Apr 20 2002, 10:14:34)
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> from __future__ import nested_scopes
>>> def template_factory(text):
...     class template:
...         def output(self):
...             print text
...     return template
...
>>> a = template_factory("hello, world")
>>> b = a()
>>> b.output()
hello, world
>>> c = template_factory("hello, my name is inigo montoya")
>>> d = c()
>>> d.output()
hello, my name is inigo montoya


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From ian at nasoftware.co.uk  Thu Jun 20 05:46:32 2002
From: ian at nasoftware.co.uk (Ian McConnell)
Date: 20 Jun 2002 10:46:32 +0100
Subject: Inter Process Communication with Numeric arrays
Message-ID: <wx1yb2uvt3.fsf@velocita.nasoftware>

I want to send the results of some calculation on a Numeric array to a
second process. These arrays will be fairly large (about 1Mb) so can anyone
recommend a package for simple IPC?

I'm writing both processes (though one might be in C) so ORB/CORBA seems to
be overkill and I can't get through to the memory mapped files at
        http://pobox.com/~kragen/sw/arrayfrombuffer/
(Internal server error)

Suggestions?


From kragen at pobox.com  Sat Jun  8 22:43:36 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 08 Jun 2002 22:43:36 -0400
Subject: Using popen interactively
References: <slrnag0fhu.6u3.joe@zork.plover.net> <adqm9k$22u6$1@nntp6.u.washington.edu>
Message-ID: <83sn3x41ev.fsf@panacea.canonical.org>

[note Joe is talking about a game by a guy named Don about a cave, or
a derivative thereof]

"Donn Cave" <donn at u.washington.edu> writes:
> Quoth Joe Mason <joe at notcharles.ca>:
> ...
> | So I'm trying to use popen2.Popen3("executable-name") to get handles to
> | stdin and stdout.  Works fine, except that I'm not sure how to detect
> | when the app has stopped printing and is waiting for input.  readline()
> 
> Right.  I expect you'd do better to throw away the file objects
> popen2 makes for you, and use the pipe file descriptors directly.
> Instead of
>     infile = app.fromchild
>     ...
>         line = infile.readline()
> say
>     infile = app.fromchild.fileno()
>     ...
>         data = posix.read(infile, 8192)

But not too soon --- if the program does multiple writes, this might
return before the last of them.

>         #  split into lines if you care
> 
> The stdio buffering in a file object is only trouble for you.
> Part of this trouble is on the other end, if your external
> process buffers output - in that case you'll have to figure
> out how to use ptys.
> 
> 	Donn Cave, donn at u.washington.edu



From larooy at xtar.co.nz  Sat Jun  1 03:25:11 2002
From: larooy at xtar.co.nz (John La Rooy)
Date: Sat, 1 Jun 2002 19:25:11 +1200
Subject: Why no open(f, "w").write()?
References: <mailman.1022702940.7544.python-list@python.org>
	<slrnafagek.64g.huaiyu@gauss.almadan.ibm.com>
	<slrnafcsor.12v.msoulier@pmerd071.ca.nortel.com>
	<20020531225504.248c92ce.larooy@xtar.co.nz>
	<slrnafg8jn.12v.msoulier@pmerd071.ca.nortel.com>
Message-ID: <20020601192511.7ec4a585.larooy@xtar.co.nz>

On 1 Jun 2002 01:23:42 GMT
"Michael P. Soulier" <msoulier at nortelnetworks.com_.nospam> wrote:

> On Fri, 31 May 2002 22:55:04 +1200, John La Rooy <larooy at xtar.co.nz> wrote:
> > 
> > Do you explicitly del all your objects too, or wait for them to disappear
> > when they go out of scope?
> 
>     I only delete them in C++. ;-) Your point is taken, but we must draw the
> line somewhere. If you don't care when the file is closed, as long as it is,
> then implicit should be fine. If however you do care when it is closed, then
> explicit should be better, no?
> 
>     Mike

I've often used the 'open(f, "w").write()' right at the end of a program where
I know the file is about to get closed anyway. From what I've seen here about
Jython, it seems that I can't even count on that behaviour. eeek!

Perhaps the specification will some day change to make the objects returned by
open() to be *smarter* so we can count on them knowing when they should flush
and when they should close the toilet seat.

A few people seem to be getting stuck on the explicit instead of implicit idea
but I think that is mostly because we are used to be told to close files
and we've all had a few late nights because we forgot to. Just as lots of us
learned we should always free() our mallocs() and now we think nothing of
letting python do that housekeeping for us.

cloudcuckoo-land-ly y'rs - John


From kopecjc at att.net  Fri Jun 28 22:34:14 2002
From: kopecjc at att.net (Joseph C. Kopec)
Date: Sat, 29 Jun 2002 02:34:14 GMT
Subject: Bitmaps in PIL
Message-ID: <3D1D1B56.5070803@att.net>

Can anyone point me to examples of algorithmically generating bitmps, 
storing them in a list and/or file and displaying them via PIL?  Thanks 
in advance.



From peteris_martinsons at swh-t.lv  Mon Jun  3 08:20:38 2002
From: peteris_martinsons at swh-t.lv (Peteris Martinsons)
Date: 3 Jun 2002 05:20:38 -0700
Subject: calldll with error handling
Message-ID: <44da3708.0206030420.be3dae7@posting.google.com>

I have modified the calldll code to handle GPF messages when foreign
function is called with incorrect parameters. Instead of throwing GPF
it raises Python SystemException which can be catched and dealed with.

Question: can it cause delayed problems due to stack corruption etc?
As far as I have tested it, repeatedly calling WinAPI functions with
incorrect parameter types, it does not crash Python nor cause delayed
GPFs.

Source code of the modified module provided below (MSVC 6). Change are
in lines 106-132 and 161-163.


// -*- Mode: C++; tab-width: 4 -*-
//	Author: Sam Rushing <rushing at nightmare.com>

#include <windows.h>

/* Define this if you want to keep track of <membuf> allocation */
#undef DEBUG_MEMBUF

#include "Python.h"

static PyObject *calldll_module_error;

/* It'd be nice to use __declspec (thread), but it doesn't work with
 * DLL's */

DWORD thread_state_tls_index;

#define FETCH_THREAD_STATE() ( (PyThreadState *) TlsGetValue
(thread_state_tls_index) )
#define STORE_THREAD_STATE(value) ( TlsSetValue
(thread_state_tls_index, ((LPVOID)value)) )

/* This is used by trampolines created by the gencb module, in place
 * of a normal call to PyEval_CallObject */

PyObject *
thread_safe_call_object (PyObject * fun, PyObject * args)
{
  PyObject * result;

  PyEval_RestoreThread (FETCH_THREAD_STATE());

  result = PyEval_CallObject (fun, args);

  STORE_THREAD_STATE (PyEval_SaveThread());

  return result;
}

//
// adapated from MIT Scheme, 7.3, microcode/ntgui.c.
//

static long
call_ff_really (long function_address,
				char * arg_format_string,
				PyObject * arg_tuple,
				long * result_addr)
{
  {
	/*	use a struct for locals that live across the foreign function call
		so that their position in the stack is the right end of the stack
		frame with respect to the stacked C arguments */
	struct {
	  long c_args[30];
	  long old_esp;
	} local;
	
	long result;

	/*	We save the stack pointer and restore it because the called
function
		may pop the arguments (pascal/__stdcall) or expect us to (__cdecl).
*/

	/*	The stack pointer is saved in a static variable so that we can
find
		it if the compiler does SP-relative addressing with a broken SP */
	
	/*	The implication is that things will break if this gets
overwritten.
		This will happen if the foreign function directly or indirectly 
		allows a Scheme interrupt to be processed (eg by calling as scheme
		function with interrupts enabled and that function gets rescheduled
		in the threads package. */
	
	static long saved_esp;
	
	{
	  long *arg_sp = &local.c_args[10];
	  long	funadd = function_address;

	  if (!PyArg_ParseTuple (arg_tuple,
							 arg_format_string,
							 &local.c_args[10],
							 &local.c_args[11],
							 &local.c_args[12],
							 &local.c_args[13],
							 &local.c_args[14],
							 &local.c_args[15],
							 &local.c_args[16],
							 &local.c_args[17],
							 &local.c_args[18],
							 &local.c_args[19],
							 &local.c_args[20],
							 &local.c_args[21],
							 &local.c_args[22],
							 &local.c_args[23],
							 &local.c_args[24],
							 &local.c_args[25],
							 &local.c_args[26],
							 &local.c_args[27],
							 &local.c_args[28],
							 &local.c_args[29])) {
		return 0;
	  }

	  STORE_THREAD_STATE (PyEval_SaveThread());

	  arg_sp = &local.c_args[10];
	  local.old_esp = saved_esp;
	  //try-except block to deal GPFs.
	  __try{
		__asm
			{
			  // Important: The order of these instructions guards against
			  // stack pointer relative addressing.
			  mov	eax, dword ptr [funadd]
			  mov	dword ptr [saved_esp], esp
			  mov	esp, dword ptr [arg_sp]
			  call	eax
			  mov	esp, dword ptr [saved_esp]
			  mov	dword ptr [result], eax
			}
			saved_esp = local.old_esp;
			PyEval_RestoreThread (FETCH_THREAD_STATE());
			
			*result_addr = result;
			return 1;
	  }

	  __except(EXCEPTION_EXECUTE_HANDLER ){
		result = NULL;
	  
		PyEval_RestoreThread (FETCH_THREAD_STATE());
		*result_addr = result;
		return 0;
	  }


	}
  }
}


static
PyObject *
call_foreign_function (PyObject * self, PyObject * args)
{
  long function_address;
  char * arg_format_string;
  char * result_format_string;
  PyObject * arg_tuple;
  long result;

  if (!PyArg_ParseTuple (args, "lssO",
						 &function_address,
						 &arg_format_string,
						 &result_format_string,
						 &arg_tuple)) {
	return NULL;
  } else {
	long status;

	status = call_ff_really (function_address, arg_format_string,
arg_tuple, &result);

	if (!status) {
	  PyErr_SetString (PyExc_SystemError, "system exception in called
function, check parameters");
	  return NULL;

	} else {
	  return Py_BuildValue (result_format_string, result);
	}
  }
}

static PyObject *
py_load_library (PyObject * self, PyObject * args)
{
  LPCTSTR name;
  if (!PyArg_ParseTuple (args, "s", &name)) {
	return NULL;
  } else {
	return Py_BuildValue ("l", ::LoadLibrary (name));
  }
}

static PyObject *
py_get_module_handle (PyObject * self, PyObject * args)
{
  LPCTSTR name;
  if (!PyArg_ParseTuple (args, "s", &name)) {
	return NULL;
  } else {
	return Py_BuildValue ("l", ::GetModuleHandle (name));
  }
}

static PyObject *
py_free_library (PyObject * self, PyObject * args)
{
  HMODULE dll_handle;
  if (!PyArg_ParseTuple (args, "l", &dll_handle)) {
	return NULL;
  } else {
	return Py_BuildValue ("i", (int) ::FreeLibrary (dll_handle));
  }
}

static PyObject *
py_get_proc_address (PyObject * self, PyObject * args)
{
  HMODULE dll_handle;
  LPCSTR proc_name;
  
  if (!PyArg_ParseTuple (args, "ls", &dll_handle, &proc_name)) {
	return NULL;
  } else {
	return Py_BuildValue ("l", (long) ::GetProcAddress (dll_handle,
proc_name));
  }
}

static PyObject *
py_read_long (PyObject * self, PyObject * args)
{
  void * address;

  if (!PyArg_ParseTuple (args, "l", &address)) {
	return NULL;
  } else {
	return Py_BuildValue ("l", *((long *)address));
  }
}

static PyObject *
py_read_byte (PyObject * self, PyObject * args)
{
  void * address;

  if (!PyArg_ParseTuple (args, "l", &address)) {
	return NULL;
  } else {
	return Py_BuildValue ("b", *((char *)address));
  }
}

static PyObject *
py_read_string (PyObject * self, PyObject * args)
{
  void * address;
  int length = 0;

  if (!PyArg_ParseTuple (args, "l|i", &address, &length)) {
	return NULL;
  } else if (length == 0) {
	return Py_BuildValue ("s", (char *)address);
  } else {
	return Py_BuildValue ("s#", (char *)address, length);
  }
}

// ===========================================================================
// memory buffer objects
// ===========================================================================

typedef struct {
	PyObject_HEAD
	void *	buffer;
	size_t	size;
    int allocated;
} membuf_object;

#ifdef DEBUG_MEMBUF
unsigned int membuf_allocated;
unsigned int membuf_freed;
#include <stdio.h>
#endif

static void
membuf_object_dealloc(membuf_object * m_obj)
{
  if (m_obj->allocated) {
#ifdef DEBUG_MEMBUF
	membuf_freed += m_obj->size;
#endif  
	free (m_obj->buffer);
  }
  PyMem_DEL(m_obj);
}

static PyObject *
membuf_read_method (membuf_object * self,
					PyObject * args)
{
  unsigned int offset = 0;
  unsigned int size = 0;
  
  if (!PyArg_ParseTuple (args, "|ii", &offset, &size)) {
    return(NULL);
  }

  if ((offset + size) > self->size) {
	PyErr_SetString (PyExc_ValueError, "offset+size is larger than
buffer");
	return NULL;
  } else {
	return Py_BuildValue ("s#", ((char *)(self->buffer)) + offset, size ?
size : self->size);
  }
}

static PyObject *
membuf_write_method (membuf_object * self,
					 PyObject * args)
{
  int offset=0;
  unsigned int length=0;
  int string_length;
  char * data;

  if (!PyArg_ParseTuple (args, "s#|ii", &data, &string_length,
&offset, &length)) {
    return(NULL);
  } else if (length == 0) {
	length = string_length;
  }

  if (length > self->size) {
	PyErr_SetString (PyExc_ValueError, "input string size is larger than
buffer");
	return NULL;
  } else {
	memcpy (((char *)(self->buffer))+offset, data, length);
	Py_INCREF (Py_None);
	return (Py_None);
  }
}

static PyObject *
membuf_size_method (membuf_object * self,
					PyObject * args)
{
  if (!PyArg_ParseTuple (args, "")) {
	return NULL;
  } else {
	return Py_BuildValue ("i", self->size);
  }
}

static PyObject *
membuf_address_method (membuf_object * self,
					   PyObject * args)
{
  if (!PyArg_ParseTuple (args, "")) {
	return NULL;
  } else {
	return Py_BuildValue ("l", (long)(self->buffer));
  }
}

static struct PyMethodDef membuf_object_methods[] = {
  {"read",		(PyCFunction)	membuf_read_method,		1},
  {"write",		(PyCFunction)	membuf_write_method,	1},
  {"size",		(PyCFunction)	membuf_size_method,		1},
  {"address",	(PyCFunction)	membuf_address_method,	1},
  {"__int__",	(PyCFunction)	membuf_address_method,	1},
  {NULL,		NULL}		/* sentinel */
};

static PyObject *
membuf_object_getattr(membuf_object * self, char * name)
{
  return Py_FindMethod (membuf_object_methods, (PyObject *)self,
name);
}

static PyTypeObject membuf_object_type = {
	PyObject_HEAD_INIT(&PyType_Type)
	0,							// ob_size
	"membuf",					// tp_name
	sizeof(membuf_object), 		// tp_size
	0,							// tp_itemsize
	// methods
	(destructor) membuf_object_dealloc,// tp_dealloc
	0,							// tp_print
	(getattrfunc) membuf_object_getattr, // tp_getattr
	0,							// tp_setattr
	0,							// tp_compare
	0,							// tp_repr
	0,							// tp_as_number
};

static PyObject *
new_membuf_object (PyObject * self, PyObject * args)
{
  membuf_object * m_obj;
  void * address = NULL;
  int size;

  if (!PyArg_ParseTuple (args, "i|l", &size, &address)) {
	return NULL;
  } else {
	m_obj = PyObject_NEW (membuf_object, &membuf_object_type);

	if (!m_obj) {
	  PyErr_SetString (PyExc_MemoryError, "couldn't allocate memory for
buffer object");
	  return NULL;
	}
	
	if (address) {
	  /* we want a pointer into existing memory */
	  m_obj->allocated = 0;
	  m_obj->buffer = address;
	  m_obj->size = size;
	  return (PyObject *) m_obj;
	} else {
	  /* we want a new memory buffer */
	  void * buffer = (void *) malloc (size);

	  if (!buffer) {
		PyErr_SetString (PyExc_MemoryError, "couldn't allocate memory for
buffer");
		PyMem_DEL (m_obj);
		return NULL;
	  } else {
#ifdef DEBUG_MEMBUF
		membuf_allocated += size;
#endif
		m_obj->allocated = 1;
		m_obj->size = size;
		m_obj->buffer = buffer;
		return (PyObject *) m_obj;
	  }
	}
  }
}

#ifdef DEBUG_MEMBUF
static PyObject *
membuf_allocation (PyObject * self, PyObject * args)
{
  if (!PyArg_ParseTuple (args, "")) {
	return NULL;
  } else {
	return Py_BuildValue ("ll", membuf_allocated, membuf_freed);
  }
}
#endif

// ===========================================================================
//
// ===========================================================================

/* List of functions exported by this module */
static struct PyMethodDef calldll_functions[] = {
  {"call_foreign_function",	(PyCFunction) call_foreign_function,1},
  {"load_library",			(PyCFunction) py_load_library,		1},
  {"free_library",			(PyCFunction) py_free_library,		1},
  {"get_module_handle",		(PyCFunction) py_get_module_handle, 1},  
  {"get_proc_address",		(PyCFunction) py_get_proc_address,	1},
  {"read_long",				(PyCFunction) py_read_long,			1},
  {"read_byte",				(PyCFunction) py_read_byte,			1},
  {"read_string",			(PyCFunction) py_read_string,		1},
  {"membuf",				(PyCFunction) new_membuf_object,	1},
#ifdef DEBUG_MEMBUF
  {"membuf_allocation",		(PyCFunction) membuf_allocation,	1},
#endif
  {NULL,			NULL}		 /* Sentinel */
};

extern "C"

__declspec(dllexport) void initcalldll(void)
{
	PyObject *dict, *module;
	module = Py_InitModule ("calldll", calldll_functions);
	dict = PyModule_GetDict (module);
	calldll_module_error = PyString_FromString ("calldll error");
	PyDict_SetItemString (dict, "error", calldll_module_error);

	/* initialize our thread local storage index */
	thread_state_tls_index = TlsAlloc();

	/* This is not actually needed, we could just use
	   calldll.get_proc_address to get them! */

	PyDict_SetItemString (
	  dict,
	  "addrs",
	  Py_BuildValue (
		"(lll)",
		&Py_BuildValue,
		&PyInt_AsLong,
		/* &PyEval_CallObject */ /* OR */ &thread_safe_call_object
		)
	);
}


From siegfried.gonzi at kfunigraz.ac.at  Sat Jun 22 10:49:21 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Sat, 22 Jun 2002 16:49:21 +0200
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <m21yazxtgs.fsf@mother.paradise.lost>
Message-ID: <3D148E71.2070201@kfunigraz.ac.at>

John Hunter wrote:

> You are jumping in to the deep water pretty fast, wrapping external
> code with SWIG and F2PY.  When I first started wrapping my C++
> classes, I encountered some system freezes as well.

They work on Windows and they work on Linux under Python (normally). I 
have tested the functions
(I can even plot the phase with Dislin of the Mie results).

Note: the first 20 minutes everything went on right (including the two 
external functions). On Windows after the first wavelength the system 
sucks down too, but it keeps up.

> If you post some code, you might get some real help, as well as advice
> on how to debug where the freezes are occurring and why.  There is
> support for debugging mixed language code.

I can send you the code (without) data, but that won't cure the 
situation. I do not believe that you are happy with maybe 1000 lines of 
bad commented code.


S. Gonzi



From wesc at alpha.ece.ucsb.edu  Mon Jun 10 19:03:34 2002
From: wesc at alpha.ece.ucsb.edu (Wesley J. Chun)
Date: Mon, 10 Jun 2002 16:03:34 -0700 (PDT)
Subject: BayPIGgies meeting Wed 6/12 7:30pm
Message-ID: <200206102303.QAA28411@alpha.ece.ucsb.edu>

What:	Python for (Perl) Programmers
Who:	Silicon Valley-San Francisco Bay Area Python Users Group (BayPIGgies)
When:	Wednesday, June 12, 2002
Where:	Stanford University, Palo Alto, CA
Time:	7:30pm - 9pm

This is a talk aimed at introducing Python to programmers
familar with Perl and other languages, and serves as a warm-up
to the full tutorial that will be presented at the O'Reilly
Open Source Convention in San Diego near the end of July.
Although the non-Python examples use Perl, it's aimed at
experienced programmers of all sorts. 

Our guest speaker is Aahz, long-time active member of the
Python community.  He has been kicking around the computer
industry for more than two decades, doing tech support,
programming, consulting, tech writing, and training. Aahz
recently signed a book contract for an intermediate-level
Python book, which will be published in early 2003. 

More information + directions:  http://deirdre.org/baypiggies/





From mjais at web.de  Wed Jun  5 08:37:53 2002
From: mjais at web.de (Markus Jais)
Date: Wed, 05 Jun 2002 14:37:53 +0200
Subject: python cookbook at oreilly's web site
Message-ID: <dCnL8.5$566.1250@news.ecrc.de>

hello

just found

http://www.ora.com/catalog/pythoncook/

I am really looking forward to reading this.
hope it is as good as the Perl Cookbook or "The Ruby Way" (a cookbook
for Ruby Fans by Hal Fulton, Samspublishing)


regards
markus


From fredrik at pythonware.com  Mon Jun 17 16:28:37 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Mon, 17 Jun 2002 20:28:37 GMT
Subject: '__builtin__' vs. __builtins__
References: <Pine.GSO.4.33.0206171543060.8570-100000@techst02.technion.ac.il>
Message-ID: <VDrP8.43848$n4.10189424@newsc.telia.net>

Beni Cherniavksy wrote:

> Why is the '__builtin__' module available under the __builtins__ name?
> By what's the logic one is supposed to remember it?  (If I do remember
> that it's different I would still confuse which is what...)

forget about __builtins__; it's an implementation detail.

if you need to explicitly refer to stuff in the built-in namespace,
*always* import the __builtin__ module.  never use __builtins__
in your code.

</F>




From gerhard at bigfoot.de  Mon Jun 17 14:36:53 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 17 Jun 2002 18:36:53 GMT
Subject: C or gcc lists
References: <mailman.1024335262.13419.python-list@python.org>
Message-ID: <slrnagsb26.1a2.gerhard@lilith.my-fqdn.de>

Alves, Carlos Alberto - Coelce wrote in comp.lang.python:
> This message is in MIME format.

Please don't do that.
 
> Does anybody can indicate me lists of C or gcc discusions?!

The gcc project has several mailing lists:
http://www.gnu.org/software/gcc/lists.html gcc-help, for example.

There is a comp.lang.c newsgroup. Be sure to read the FAQ before
posting: http://www.eskimo.com/~scs/C-faq.top.html

Please also disable multipart/alternative in your mail software.
Usenet is a text-only medium, as are mailing lists.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From jones at jones.jones  Fri Jun 14 00:19:54 2002
From: jones at jones.jones (Jones)
Date: Fri, 14 Jun 2002 04:19:54 GMT
Subject: Writing a simple linux daemon type thing...
Message-ID: <K9eO8.393$LN6.261@newsread1.prod.itd.earthlink.net>

Hi everyone,
I hope this is a simple question for you guys.  I'm new to Python and Linux.
I'm wondering if Python (and this group) can help me.
I want a program to always be running in the background that, connects to
PostgreSQL every few seconds, queries a table, and for each record it finds,
shells out to the system to execute program.

My research indicates that "nohup" might do the trick?  So would it be as
simple as writing the program in Python and then puting something like
"nohup python myprog" in the Linux startup sequence (wherever that may be)?

That's my primary goal.  Now... assuming I get past that part, I may want to
take it a few steps further later on.  I'd be happy if you guys could take a
stab at telling me if these are possible and any tips too...
a) Make it multi-threaded...   As I said before, the number of times it
executes the external program equals the number of records it finds in the
table.  It would be nice if I could execute the program multiple times
concurrently in a *controlled* manner (depending on my system resources)
instead of having to wait for each one to be done before starting another
one.
b) Make it fail-safe.  I would like to figure out how I could re-launch the
program if it was to crash.

P.S.  You're probably wondering why I don't just use a trigger.  One reason
is that I don't think I can execute an external program from PostgreSQL and
I don't know C to write a wrapper function.  Second reason is that because I
am running an external program (out of my "control"), I'm afraid it might
crash, etc and that record will never get processed.

Thank You!!!




From erno-news at erno.iki.fi  Wed Jun  5 12:04:49 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 05 Jun 2002 19:04:49 +0300
Subject: os.popen and place in file
References: <3CFE2762.519E4344@motorola.com>
Message-ID: <kud6v5ogny.fsf@lasipalatsi.fi>

In article <3CFE2762.519E4344 at motorola.com>, Stephen Boulet
<stephen.boulet at motorola.com> writes:

| I'm running windows NT.
| The command: 
|   f = os.popen('dir')
| returns a file object 'f'. I can't however use the seek command on it [...]
| Anyone know why?

it is because the text output by the "dir" program is relayed directly
to your program. the operating system does not store it away anywhere
in case you want to revisit parts of it later. sockets and many
unix device files work the same way. otherwise there would
be problems when transferring large amounts of data through
pipes.

if you want to access the data like a regular file, you could redirect
the output from the command to a disk file and open that. or, read
all of the data from the pipe and put it in a StringIO object.

  -- erno


From wtri at wtresearch.de  Thu Jun  6 05:32:53 2002
From: wtri at wtresearch.de (wtr)
Date: Thu, 6 Jun 2002 11:32:53 +0200
Subject: pygoogle search restricted by date
References: <adfmsk$205i$1@ariadne.rz.tu-clausthal.de>
Message-ID: <1023351766.561304@lix-049dz-008.inside.impress.com>

Hi,
maybe what mx is thinking what a Julian Date is does not
match what Google expects it to be.
What values are given for today?
The complete Julian Date is a very large number, counting days since
more than 6700 years. Sometimes, Julian Dates are given just as offset
of the current year, i.e. day number in year.
Bye,
wolfgang teschner

"Lutz Schroeer" <ifls at rz.tu-clausthal.de> wrote in message
news:adfmsk$205i$1 at ariadne.rz.tu-clausthal.de...
> Hi Pythonistas,
>
> The Goolge API doc tells me that I can restrict my search to a specific
date
> range using "daterange:<date1>-<date2>" where date1 and date2 are Julian
Day
> Numbers. I put the term in the restrict part auf the doGoogleSearch
function
> (-> code snippet). Nevertheless I'm getting the same results.
>
> Any hints?
>
> Lutz
>
> <--snip-->
>
> import google
> from mx.DateTime import *
>
> google.setLicense('<mykey>')
> today = int(now().jdn)
> yesterday = today - 1
>
> restrict_query = 'daterange:%d-%d' % (yesterday, today)
>
> data = google.doGoogleSearch('python', restrict = restrict_query)
> for x in data.results:
>   print x.title, x.URL
>
> --
>
>




From jdhunter at nitace.bsd.uchicago.edu  Thu Jun 13 15:02:10 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Thu, 13 Jun 2002 14:02:10 -0500
Subject: dictionary: keys() and values()
Message-ID: <m2elfbuhn1.fsf@mother.paradise.lost>

For some dictionary m, if we call sequentially

keys = m.keys()
values = m.values()

Do we know if the values come out in the same order, that is does
m[keys[i]] necessarily equal values[i] ?

Thanks,
John Hunter


From fredrik at pythonware.com  Sat Jun 29 08:51:02 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 29 Jun 2002 12:51:02 GMT
Subject: Python needs better error reporting
References: <LNBBLJKPBEHFEDALKOLCEEOKAAAB.tim.one@comcast.net> <mailman.1025351378.7466.python-list@python.org>
Message-ID: <W2iT8.45588$n4.10825262@newsc.telia.net>

Walter D?rwald wrote:

> This is easy: Use the _G_uido _C_loning _C_ommand! ;)

the one with a -Wall option?  scary...

</F>




From shalehperry at attbi.com  Thu Jun 20 12:53:23 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 20 Jun 2002 09:53:23 -0700 (PDT)
Subject: Function to add commas in numbers?
In-Reply-To: <1024589333.15471.82.camel@naked>
Message-ID: <XFMail.20020620095323.shalehperry@attbi.com>

> i have some problems with locale module on debian sid:
> 
> 

Debian no longer ships ALL of the locales enabled.  It is up to the sysadmin to
enable the locales you need.  This makes the package smaller, easier to build
and also takes up less room on your system.

Edit /etc/locale.gen and follow the directions in the comments.





From mfranklin1 at gatwick.westerngeco.slb.com  Tue Jun 25 05:22:21 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Tue, 25 Jun 2002 09:22:21 +0000
Subject: problem recursively removing directories on an NFS mount
In-Reply-To: <3D17579F.9030905@bioeng.ucsd.edu>
References: <3D17579F.9030905@bioeng.ucsd.edu>
Message-ID: <200206250825.g5P8PxM13488@helios.gatwick.geco-prakla.slb.com>

On Monday 24 Jun 2002 5:32 pm, you wrote:
>
> Using python1.5.2 on an IRIX 6.5 system, this function usually works
> fine.  However, sometimes nfs files of the form ".nfs07D7761" get

Don't think this helps but I see these files all the time (without involving 
python) I assume they are a 'feature' of NFS.....  maybe google can teach us 
somthing....... <time passing>..... ahh yes.

http://www.cse.unsw.edu.au/faq/questions/help-nfs.html

Cheers
Martin




From maxm at mxm.dk  Tue Jun 11 08:58:37 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 11 Jun 2002 14:58:37 +0200
Subject: Sorting list (maybe stupid)
References: <mailman.1023789674.28761.python-list@python.org>
Message-ID: <3D05F3FD.7020202@mxm.dk>

Harvey Thomas wrote:

> import re
> 
> r = re.compile('\d+')
> afile = open('myfile.txt')
> tuplelist = []
> for aline in afile:
>     m = r.match(aline)
>     if m:
>         tuplelist.append((int(m.group(0)), aline))
>     else:
>         tuplelist.append((0, aline))
> tuplelist.sort()
> for x in tuplelist:
>     print x[1]


Old Perl programmer ? ;-)

afile = open('myfile.txt')
tuplelist = []
for aline in afile:
     nummericValue = int(aline[:aline.find('.')])
     tuplelist.append((nummericValue, aline))
tuplelist.sort()
sortedLines = [tup[1] for tup in tuplelist]
for line in sortedLines:
     print line


regards Max M



From whisper at oz.net  Fri Jun  7 16:14:59 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 7 Jun 2002 13:14:59 -0700
Subject: Where is Berkeley DB on your system? Can you live without DB 1.85?
In-Reply-To: <15617.3834.292130.446065@12-248-41-177.client.attbi.com>
Message-ID: <GCEDKONBLEFPPADDJCOEAEGJDHAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Skip Montanaro
> Sent: Friday, June 07, 2002 12:52
> To: python-list at python.org
> Subject: Where is Berkeley DB on your system? Can you live without DB
> 1.85?
>
> 1.85.
>
> Finally, is compilation with an older version (say, version 2)
> important to
> you in the presence of a newer installed version (version 3 or 4)?  Let me
> know that as well.
>
> Thx,
>
> --
> Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
> Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html

If Sleepycat will allow it, i'm all in favor of updating to the latest
version! However, they want money for commercial use of versions > 1.85
AFAIK.

(FWIW, bsddb lives in python22/dlls on Windows :-). )

David LeBlanc
Seattle, WA USA





From djc at object-craft.com.au  Mon Jun 24 00:01:09 2002
From: djc at object-craft.com.au (Dave Cole)
Date: 24 Jun 2002 14:01:09 +1000
Subject: PIL, Python and CGI
References: <Pine.GSO.4.33.0206232101300.27090-100000@infinity.infineon.net> <mailman.1024888231.4213.python-list@python.org>
Message-ID: <m34rftfhq2.fsf@ferret.object-craft.com.au>

>>>>> "Richard" == Richard Jones <rjones at ekit-inc.com> writes:

Richard> On Mon, 24 Jun 2002 13:02, Sascha Ferley wrote:
>> On Mon, 24 Jun 2002, Richard Jones wrote: > Your web server
>> environment most likely doesn't have /usr/local/lib in > its
>> LD_LIBRARY_PATH.
>> 
>> You probly mean Apache... I'll take a look for at the conf files
>> for it..  Any ideas where to statically define the lib paths for
>> apache.. i doubt in the httpd.conf file..

Richard> You've basically got three choices as I see it:

Richard> 1. recompile PIL to statically link libjpeg etc.
Richard> 2. recompile PIL to use -R (or whatever the flag is) so it
Richard> remembers where the dynamic libs are 3. modify your apache
Richard> start script to set the environment var appropriately.

Richard> 1 or 2 are going to be the easiest, and safest options.

Number 3 is pretty easy.  Use the SetEnv directive in your Apache
config.

        http://httpd.apache.org/docs/mod/mod_env.html#setenv

- Dave

-- 
http://www.object-craft.com.au


From K.Rdt at TU-Berlin.DE  Sun Jun 23 17:01:56 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Sun, 23 Jun 2002 23:01:56 +0200
Subject: mail-problem over t-online
In-Reply-To: <3D163622.E7733360@engcorp.com>
Message-ID: <NH2WBAGDXS8D0EQOYV41WB88VRT.3d163744@lap>

Am 23.06.02 22:57:06, schrieb Peter Hansen <peter at engcorp.com>:

>Klaus Reinhardt wrote:
>>         host=smtplib.SMTP('mailszrz.zrz.tu-berlin.de')
>>         host.sendmail(zwi_eck(From),zwi_eck(Tp),al)
>>         host.quit
>> authentification required .. Providing my log and pass
>> produces errors, too.

>Try connecting manually with TELNET first, to ensure you
>have a clear channel to that server and port.  Many ISPs
>will block port 25 except to their own mail server, to 
>prevent you using them to send spam through other systems
>and having them suffer the consequences...
>
>-Peter
---------------------------------------------------------------------
Hi
Thanks for the hint (port 25); the telnet is working
and on anis.tu I have a python, so I can launch a
script; but this is the long way. My goal is to fiddle all
together to one script. For now I couldn't establish
a working telnet a.s.o. - But these are my python-
beginnings. -) 
		K at Rdt
---------------------------------------------------------------------







From roelsch at acm.org  Tue Jun  4 01:53:24 2002
From: roelsch at acm.org (Robert Oelschlaeger)
Date: Tue, 04 Jun 2002 05:53:24 GMT
Subject: Please help with script...
References: <adheqc$nhn$1@netnews.upenn.edu>
Message-ID: <oBYK8.13175$sJ4.1366450361@newssvr12.news.prodigy.com>

"Topy" <hobiewon at hotmail.com> wrote in message
news:adheqc$nhn$1 at netnews.upenn.edu...
> Hey all,
> I'm having trouble with a problem I have with this python program.  It's
> kind of long, so if anyone has the patience, I would gladly appreciate
your
> time(the programming is fairly simple I think.)  Anyway, the problem I
need
> to do is this :
<SNIP>

> It's not much, but I'm not sure where to go from here...can anyone help?
>
> Thanks so much!
> Topy

This might get you started; the worksheet is left open for you to examine
and save via the File menu -- Bob

#!/cygdrive/c/Python22/./python
# vim:ts=4:sw=4:et

def main():
    ProductID = int(raw_input("Please enter the ProductID: "))
    OrderQty = int(raw_input("Please enter the cutoff order quantity: "))
#   CSV = open('a:/sales.txt', 'r')
    CSV = open('sales.txt', 'r')

    import win32com.client
    xl = win32com.client.Dispatch('Excel.Application')
    xl.Visible=1
    xl.Workbooks.Add()

    # start at the top of the worksheet
    xl.Range("A1").Select()

    # report selection parameters and skip a line
    xl.Range("A1:B1").Value = ("ProductID:", ProductID)
    xl.Range("A2:B2").Value = ("OrderQty:", OrderQty)
    xl.ActiveCell.Offset(4).Activate()

    # fill in the column headers
    xl.ActiveCell.Range("A1:E1").Value = ("OrderID", "OrderDate",
"RegionID", "ProductID", "OrderQty")
    xl.ActiveCell.Offset(2).Activate()

    # process each line in the CSV file
    for line in CSV.readlines():
        # remove trailing \n
        line = line.rstrip()

        # get the raw data as strings
        inOrderID, inOrderDate, inRegionId, inProductId, inOrderQty =
line.split(",")

        # is this item one that is interesting?
        if int(inProductId) == ProductID:
            if int(inOrderQty) >= OrderQty:
                # write the input data to the output file
                xl.ActiveCell.Range("A1:E1").Value = (inOrderID,
inOrderDate, inRegionId, inProductId, inOrderQty)
                # and activate the next line
                xl.ActiveCell.Offset(2).Activate()


if __name__ == "__main__":
    main()






From huaiyu at gauss.almadan.ibm.com  Tue Jun 18 21:42:34 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 19 Jun 2002 01:42:34 +0000 (UTC)
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com> <gYgP8.42356$GY.12361719@twister.nyroc.rr.com> <aekcug$7js8n$1@ID-114079.news.dfncis.de> <9YBP8.44418$GY.13865843@twister.nyroc.rr.com> <mailman.1024423525.24359.python-list@python.org> <20020619091900.70ae2b7f.larooy@xtar.co.nz>
Message-ID: <slrnagvoca.46u.huaiyu@gauss.almadan.ibm.com>

John La Rooy <larooy at xtar.co.nz> wrote:
>I've been utterly astonished reading this thread. I was sure George was flame-baiting
>but after his last couple of posts I'm not so sure. Well said Cliff

Some people are simply unable to admit their ignorance in front of a big
audience.  They try to imply that others are ignorant, in an attempt to
stimulate some response or explanation.  When they realize the errors of
their arrogance they start picking on various things others have said/done,
not knowing that this makes them sounding sillier than ever.  What makes the
matter worse is when others behave unexpectedly nice, robbing them the
chance to fight an exit in a confusing battle.  We have probably just
witnessed one example of such people.

Huaiyu



From see at below  Sat Jun  1 03:22:32 2002
From: see at below (Paul Foley)
Date: 01 Jun 2002 19:22:32 +1200
Subject: Why no open(f, "w").write()?
References: <mailman.1022702940.7544.python-list@python.org> <slrnafagek.64g.huaiyu@gauss.almadan.ibm.com> <slrnafcsor.12v.msoulier@pmerd071.ca.nortel.com> <20020531225504.248c92ce.larooy@xtar.co.nz>
Message-ID: <m2lm9zh39j.fsf@mycroft.actrix.gen.nz>

On Fri, 31 May 2002 22:55:04 +1200, John La Rooy wrote:

>> Well, as one of the credos of Python is that explicit is better than
>> implicit, why are we relying on the implicit nature of file objects to close
>> themselves? It's not nearly as easy to read.

> Do you explicitly del all your objects too, or wait for them to disappear when they go out of scope?

Neither.  You can't delete objects in Python ("del" doesn't do
anything to objects; it does something to variables), and in theory
they persist forever; GC is just an optimization to allow reuse of
memory used by objects that the program will never access again,
because you don't really have infinite memory.

Correct optimizations don't change the behaviour of correct code.

[Therefore, if the behaviour of your code does change depending on
whether or not the optimization is done, either the optimization is
in error or your code is.  Thus, relying on GC to close files (or do
anything else) is Wrong.  Python really ought to print a warning when
it GCs an open file]

-- 
You don't have to agree with me; you can be wrong if you want.

(setq reply-to
  (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))


From jepler at unpythonic.net  Tue Jun  4 12:51:15 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 4 Jun 2002 11:51:15 -0500
Subject: Looking for a list subclassing example...
In-Reply-To: <b0c5da3a.0206040833.456d184e@posting.google.com>
References: <adi4dv$118tko$1@ID-146704.news.dfncis.de> <b0c5da3a.0206040833.456d184e@posting.google.com>
Message-ID: <20020604115115.D24006@unpythonic.net>

On Tue, Jun 04, 2002 at 09:33:08AM -0700, Darrell wrote:
> class LimitList(list):
>     """ list limited in size """
>     def __init__(self, limit):
>         self.limit=limit
>     def append(self, item):
>         list.append(self, item)
>         diff =  len(self) - self.limit
>         if diff == 1:
>             del self[0]
>         else:
>             self = self[-self.limit:]

I don't think the line
>             self = self[-self.limit:]
can do what you want (it only changes something about the locals in
the method).  You almost certainly need to use slice deletion
	      del self[:-self.limit]
or
	      if diff > 0: del self[:diff]
(these seem to work in all cases, even when diff == 1). You also haven't
overridden the slice assignment method, extend(), __add__, or __mult__
to make sure the length invariant is preserved in cases like
    x[:0] = ['ha ha'] * 1000
    x = x + ['ha ha'] * 1000
    x = x * 1000

Jeff




From drew_csillag at yahoo.com  Tue Jun 11 05:23:31 2002
From: drew_csillag at yahoo.com (Drew Csillag)
Date: Tue, 11 Jun 2002 14:23:31 +0500
Subject: [ANN] SkunkWeb 3.2.2
Message-ID: <3D05C193.3060004@yahoo.com>

SkunkWeb 3.2.2 has just been released.

The release is available from the SkunkWeb home page at
   http://skunkweb.sourceforge.net

or directly from
       http://prdownloads.sourceforge.net/skunkweb/skunkweb-3.2.2.tgz

This is mainly a bugfix release.

Changes in 3.2.2
========================================
* The <:compargs:> tag can now be used in an STML block (used to only work
  at the top level)
* Works with Python 2.1 again (broke in 3.2.1)
* new geeknews feed from python.org
* fixed lwn geeknews feed
* added support for DCOracle2

Drew





From cliechti at gmx.net  Thu Jun 13 18:54:37 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 14 Jun 2002 00:54:37 +0200
Subject: newbie: from .. import *
References: <56e1eff0.0206131437.4ca8ed9e@posting.google.com>
Message-ID: <Xns922D9AB17E08cliechtigmxnet@62.2.16.82>

jseb at cs.mcgill.ca (Jean-S?bastien Bolduc) wrote in 
news:56e1eff0.0206131437.4ca8ed9e at posting.google.com:
> I've been working with Python for a while, but never really felt
> confortable with the way namespaces are imported.
> 
> I'm working on a simple application that is to be used in the
> interactive mode. The idea is that the behavior of a function is
> determined, in part, by some global(?) flags. So I can change my
> flags, and explicitely call my function. Really simple. Except that
> for some reason, it does not work.
> 
> In "test.py", I have the following:
> 
>   flag = 1
>   def foo():
>     print flag
>   foo()
>   flag = 0
>   foo()
>   
> In the interactive mode, if I import the code above as:
>>>> from test import *
> 1
> 0
> 
> The output is as expected. However, if I continue:
>>>> flag = 1
>>>> foo()
> 0
> 
> (!) Okay... not what _I_ expected. It sounds really silly, but I've
> spent a couple of hours on that, without success. I thought at some
> point that using "global" would help me. Of course, if I just do
> "import test", everything works fine --- but in interactive mode, I
> don't want to type "test.foo()"
> 
> Some help?

the function still runs in the namespace of its module.
you can explicit use a module to store global values. e.g. you could
use __main__.

test.py:
import __main__
__main__.flag = 0
def foo():
    	print __main__.flag
    	...

of course you can use any module for that, not only __main__.

chris

-- 
Chris <cliechti at gmx.net>



From gh_pythonlist at gmx.de  Sun Jun  9 07:22:09 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sun, 9 Jun 2002 13:22:09 +0200
Subject: MySQLdb warnings problem
In-Reply-To: <3D022ACE.A545BF98@anansispaceworks.com>
References: <3D022ACE.A545BF98@anansispaceworks.com>
Message-ID: <20020609112209.GA438@lilith.my-fqdn.de>

* Terry Hancock <hancock at anansispaceworks.com> [2002-06-08 09:03 -0700]:
> I'm writing this crawler now, and it's my first
> experience with "Python Database API 2.0", which seems
> really overblown and complicated compared to the simple
> interface described in O'Reilly's MySQL/mSQL book, or
> with the ZSQL interface in Zope. I suppose this is
> for the benefit of more complicated RDBMs than MySQL.

The DB-API 2.0 is really a rather slim interface, and in no way
overblown or complicated. Good beginner-level online docs on it are
sparse, though.

> Anyway, I hope I'm not making a mistake using it --
> I'm just trying to use the most "standard" Python
> interface to MySQL, and I'm under the impression that
> this is it. Please advise if I'm wrong.

No, MySQLdb is the way to go, and it is a really fine interface.

> [snip traceback]

Sorry, I don't know what's the reason for this traceback.
 
> For reference, here's a snippet of the code where this
> broke:
> 	sqlcmd= """
> 		UPDATE user SET nposts=%d, nrecent=%d, keytopic='%s'
> 		WHERE username='%s'
> 		""" % ( n_posts, n_recent, keytopic, user[0] )
> 
> 	dbcursor.execute(sqlcmd)

Here we have the single most common misuse of the cursor's execute
statement. Consider what happens if your string contains a single
apostrophe. Yes, right, the SQL will break.

That's why the cursor's execute command has an optional second
parameter:

.execute(operation[,parameters]) 
           
    Prepare and execute a database operation (query or command).
    Parameters may be provided as sequence or mapping and will be bound
    to variables in the operation.  Variables are specified in a
    database-specific notation (see the module's paramstyle attribute
    for details).

Ok, so let's rewrite this to the appropriate form:

    sqlcmd= """UPDATE user SET nposts=%d, nrecent=%d, keytopic='%s'
               WHERE username='%s'"""

    dbcursor.execute(sqlcmd, (n_posts, n_recent, keytopic, user[0]))

This form ensures that the appropriate quoting is used for each of the
parameters (n_posts, n_recent, keytopic and user[0]).

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 16.8 ?C      Wind: 2.4 m/s




From syver-en+usenet at online.no  Thu Jun 27 17:43:53 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Thu, 27 Jun 2002 21:43:53 GMT
Subject: How to find out operating system
References: <mailman.1025210498.607.python-list@python.org>
Message-ID: <ubs9wjt3i.fsf@online.no>

"David LeBlanc" <whisper at oz.net> writes:
> C:\>python
> Python 2.2.1 (#34, Apr 9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
> 
> Type "help", "copyright", "credits" or "license" for more information.
> 
> >>> import os
> >>> os.environ.has_key('WINOS')
> 1
> >>> os.environ['WINOS']
> 'WIN2000'
> >>>
> 

Strange (see below) where is the variable defined on your machine?


Here's what I get.

Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> os.environ['WINOS']
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "d:\devtools\python22\lib\os.py", line 387, in __getitem__
    return self.data[key.upper()]
KeyError: WINOS
>>> 

-- 

Vennlig hilsen 

Syver Enstad


From rganesan at myrealbox.com  Tue Jun 18 23:36:18 2002
From: rganesan at myrealbox.com (Ganesan R)
Date: 19 Jun 2002 09:06:18 +0530
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <D7NP8.46301$GY.14487785@twister.nyroc.rr.com>
Message-ID: <ouu1o0djnx.fsf@andlx-anamika.cisco.com>

>>>>> "George" == George Hester <hesterloli at hotmail.com> writes:

> Don't worry I won't return here.  This is not a group to get help but a
> group of individuals that think the world owes them something when they
> participate here.

Oh, my! I am simply overcome with grief *sob*

Ganesan

-- 
Ganesan R (rganesan at debian.org>)     | http://www.debian.org/~rganesan/
1024D/5D8C12EA, fingerprint F361 84F1 8D82 32E7 1832  6798 15E0 02BA 5D8C 12EA


From not.this at seebelow.org  Thu Jun 13 10:59:33 2002
From: not.this at seebelow.org (Grant Griffin)
Date: 13 Jun 2002 07:59:33 -0700
Subject: why not "'in' in 'in'"?
References: <mailman.1023977529.27368.python-list@python.org>
Message-ID: <aeac0l013k@drn.newsguy.com>

In article <mailman.1023977529.27368.python-list at python.org>, Tim says...
>
>[Grant Griffin]
>> See, that's where we bump into "practicality beats purity".
>>
>> note-that-every-one-of-tim's-principles-is-subjective-ly y'rs,
>
>They only appear that way until one has developed objective good taste.
>

Yes, guilty as charged!  Here's an example of my poor taste.  Since "find" is
sort of a boolean concept ("I found/didn't find my tennis shoes"), and since the
"find" function really returns the index (position) of a substring, I probably
would have called it "position" or maybe just "pos" (in admiration for BASIC).

hindsight-need-not-be-20/20-<wink>-ly y'rs,

=g2

_________________________________________________________________________

Grant R. Griffin                                           g2 at dspguru.com
Publisher of dspGuru                               http://www.dspguru.com
Iowegian International Corporation                http://www.iowegian.com



From gashton at cmedltd.com  Wed Jun 12 05:52:48 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 12 Jun 2002 10:52:48 +0100
Subject: using m2crypto to encrypt credit card numbers
In-Reply-To: <NCBBLFCOHHDIKCAFGCFBGEBOLEAA.marklists@mceahern.com>
References: <NCBBLFCOHHDIKCAFGCFBGEBOLEAA.marklists@mceahern.com>
Message-ID: <1023875569.603.44.camel@walter>

On Tue, 2002-06-11 at 21:12, Mark McEahern wrote:

> Also, another reason I need to store the credit card number is in the case
> of chargebacks, which don't go through the payment processor--rather, they
> go through the bank.  I don't fully understand this part, but I do believe I
> need the credit card number in order to link the chargeback to a
> transaction.

I'd forgotten all about that side of it. My only experience of
chargebacks involved banks querying fraudulent transactions (i.e.
somebody had entered somebody else's credit card into our site). It
happened a hell of a lot and cost us (my last company) an enormous
amount of developer time writing tools to combat fraud. If you get too
much fraud (I think the limit was 40% fraud by value in the UK, but my
memory is hazy) going through your site you can get your merchant ID
status revoked by the banks (i.e. you get your transaction processing
account chopped).

Anyway, the only information that the banks could give us that enabled
us to uniquely identify fraudulent transactions (so that we could refund
them) was the card number and the time the transaction was made.

This makes my idea somewhat less feasible (but you could still store
card numbers in a more tightly secured database than the one that stores
the card hashes if you could find a provider that could do recurring
charges). Sadly, I don't know of any.

-- 
Graham Ashton





From donn at u.washington.edu  Tue Jun  4 13:51:15 2002
From: donn at u.washington.edu (Donn Cave)
Date: 4 Jun 2002 17:51:15 GMT
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <3CFCF3E7.AE891EBE@noaa.gov>
Message-ID: <adiumj$22g2$1@nntp6.u.washington.edu>

Quoth Chris Barker <Chris.Barker at noaa.gov>:
| Huaiyu Zhu wrote:
|> Great explanation!  The problem stems from the fact that the current +=
|> operator implements two semantically distinct operations, chosen according
|> to some properties of the operands.
|
| exactly, and it wa kind of a weird decision. Personally, I feel the only
| compelling reason to add The += (and friends) operator was to support
| in-place addition, most notably for Numeric arrays. Frankly it ought to
| just be disallowed for non-mutable types, but all the folks that want it
| for integers would scream. to be honest, I've kind of lost track of why
| it is a good idea for numbers to be immutable, but that's another topic!

That would be almost as cool as the old implementations of FORTRAN
where you could change the value of 1.

|> I'd imagine that at some future time there would be a desire to split += to
|> two distinct operators, one for true in-place operator and one for the fake
|> one currently implemented for immutables.  That would remove the spurious
|> exceptions for mutable members of immutable structures.
|
| Do we really need a new notation? is a = a + b so bad?

No, it's fine.

| ... In fact, are the
| current semantics of += s bad, or should this bug just be fixed?

Not really - the current semantics are actually better than the fixed
version, inasmuch as they are at least consistent.  As I interpret the
proposal, the fixed += would sometimes work without rebinding (if the
object supports it) and sometimes would be an assignment (falling back
to +.)  To me, to allow an object to dictate its relationship to its
computational environment like that is anathema.

...
| By the way, the += and friends operators also cause similar confusion
| with name scopes in functions, += is assumed to be a re-bind, when it
| isn't necessarily intended to:
|
| >>> l = [1,2,3]
| >>> def test():
| ...     l += [4]
| ... 
| >>> test()
| Traceback (most recent call last):
|   File "<stdin>", line 1, in ?
|   File "<stdin>", line 2, in test
| UnboundLocalError: local variable 'l' referenced before assignment
|
| but:
|
| >>> def test2():
| ...     l.extend([4])
|
| >>> test2()
| >>> l
| [1, 2, 3, 5]
|
| Works just fine. Perhaps  Huaiyu is right, we need different operators.
| However, then the operator for incrementing a non-mutable would just be
| a synonym for a = a + b, which looks like unnecessary syntactical sugar
| to me.

Yes.

	Donn Cave, donn at u.washington.edu


From irmen at NOSPAMREMOVETHISxs4all.nl  Mon Jun 10 15:20:21 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Mon, 10 Jun 2002 21:20:21 +0200
Subject: M2Crypto: select() behaves weird on SSL socket
References: <3D04ECED.2030202@NOSPAMREMOVETHISxs4all.nl> <slrnag9rie.enn.akuchlin@ute.mems-exchange.org>
Message-ID: <3D04FBF5.8090407@NOSPAMREMOVETHISxs4all.nl>

A.M. Kuchling wrote:

> Perhaps there really is no more data on the socket because OpenSSL has
> read it all, but the decrypted data is now sitting in a buffer

Sounds like a reasonable explanation. But still, I have a problem then.
I'm using select() to realize a timeout on socket reads (and writes)
(much like timeoutsocket.py does), and I am reading the data on the
socket in chunks. select() is called again before reading the next chunk.

This works nicely with the regular sockets.
It breaks when using M2Crypto SSL sockets.

I can't seem to think of a solution, other than disabling the
timeout function if using SSL...

My software heavily relies on the chunk reading because what
I'm doing is: read a fixed size block that tells how large
the next block is. Then read that next block.


Irmen


PS my reading function:

def _recv_msg(sock,size,timeout):
	msglen=0
	msglist=[]
	while msglen<size:
		# XXX this breaks when using SSL... :-(
		r,w,e=select.select([sock],[],[],timeout)
		if not r:
			raise TimeoutError()
		chunk=sock.recv(size-msglen)
		if not chunk:
			raise ConnectionClosedError()
		msglist.append(chunk)	
		msglen+=len(chunk)
	return ''.join(msglist)



From gcordova at hebmex.com  Thu Jun 27 15:58:31 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 27 Jun 2002 14:58:31 -0500
Subject: Pythoniac: Thoughts on a hardware Python processor
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233015DE6E0@mail.hebmex.com>

> 
> Which is truly a shame; it's more because of commercial
> and market pressures, and nothing to do with true
> throughput or technical merit.
> 

Argh!!

Please forgive me!!

This last paragraph sounds vaguely "Tim Rue"-ish, 
and for that I grovel forgiveness.

-gus

pd: *grovel*  *grovel*




From max at alcyone.com  Wed Jun 26 15:42:39 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 26 Jun 2002 12:42:39 -0700
Subject: Python 2.2 __slots__ Bug?
References: <mailman.1025089474.17780.python-list@python.org> <3D1A057F.A81FF84@alcyone.com> <mailman.1025116128.17820.python-list@python.org>
Message-ID: <3D1A192F.DE47ACBE@alcyone.com>

Jeff Epler wrote:

> When run in a thread, it dies on my redhat 7.3 machine.  When not run
> in
> a thread, it seems to succeed even for very large values.

I can reproduce that crash in a thread on Slackware 8.0, as well.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ See the son in your bad day / Smell the flowers in the valley
\__/ Chante Moore
    Bosskey.net: Aliens vs. Predator 2 / http://www.bosskey.net/avp2/
 A personal guide to Aliens vs. Predator 2.


From ark at research.att.com  Tue Jun  4 12:52:24 2002
From: ark at research.att.com (Andrew Koenig)
Date: Tue, 4 Jun 2002 16:52:24 GMT
Subject: self
References: <BI5L8.78$Vv1.4155449@newnews.cc.stevens-tech.edu>
Message-ID: <yu99sn43kmuv.fsf@europa.research.att.com>

Vojin> I have been working on a project trying to implement some
Vojin> advanced functionality.  Now, one thing that has been a
Vojin> constraint for me is the concept of self.  Why is it that
Vojin> within a class one has to use self. (or "any word". ) to refer
Vojin> to instance variables? Would it not be more logical to just use
Vojin> the name of the variable like it is for example in C++?

It has advantages and disadvantages.  The obvious disadvantage of
having to use "self" to refer to every instance variable is that
you have to use "self" to refer to every instance variable.

The advantage of having to use "self" to refer to every instance
variable is that there is never any doubt about whether a given name
represents an instance variable.  These doubts can cause trouble for
C++ programmers; whether the convenience of not having to utter
"this->" constantly is a matter of personal taste and opinion.

For Python, requiring the explicit use of "self" is clearly the
right choice because variables are so dynamic that it would not
generally be possible to tell during compilation whether a
particular name refers to an instance variable.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From syver-en+usenet at online.no  Wed Jun  5 07:40:06 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: 5 Jun 2002 04:40:06 -0700
Subject: manipulating child processes under Windows
References: <lcptz7zw0m.fsf@gaffa.mit.edu> <3cfd6d6d$1@news.nz.asiaonline.net>
Message-ID: <d7226884.0206050340.168d43ce@posting.google.com>

"Colin Brown" <cbrown at metservice.com> wrote in message news:<3cfd6d6d$1 at news.nz.asiaonline.net>...
> "Douglas Alan" <nessus at mit.edu> wrote in message
> news:lcptz7zw0m.fsf at gaffa.mit.edu...
> >
> > (2) What's the best way for me to solve the problem.  I.e., what's the
> >     best way to get my Python script to wait for the real program to
> >     terminate?
> >
> > |>oug
> 
> To locate a grandchild process and get its pid and handle on Win2000 you
> could examine the output of the freeware program "pslist -t"
> (www.sysinternals.com). After that use win32 routines as usual.

I thought that the program was already running and used some form of
interprocess communication to send some information to the other
program and then exit.


From bokr at oz.net  Fri Jun  7 17:23:54 2002
From: bokr at oz.net (Bengt Richter)
Date: 7 Jun 2002 21:23:54 GMT
Subject: map
References: <3D007666.2B6D3582@gol.ge> <3d00fe37@news.mhogaming.com>
Message-ID: <adr89a$vbj$0@216.39.172.122>

On Fri, 7 Jun 2002 14:33:45 -0400, "Michael Weiss" <FooWeissBarMike at hotmail.com> wrote:

>Would this work?
>
Yes, after
 >>> from math import pow
;-)

>##########
>x = [1,2,3,4]
>y = 2
>
>#this?
>print map(lambda z: pow(z,y), x)
Why not just show that it does work ;-)
 >>> print map(lambda z: pow(z,y), x)
 [1.0, 4.0, 9.0, 16.0]
>
># or this?
>print map(pow, x, [y for z in xrange(len(x))])
ditto ;-)
 >>> print map(pow, x, [y for z in xrange(len(x))])
 [1.0, 4.0, 9.0, 16.0]
or just
 >>> map(pow, x, [y]*len(x))
 [1.0, 4.0, 9.0, 16.0]

>##########
>
math.pow might not be optimal for speed (or accuracy for e.g., 3L**100),
depending on arguments. If y is integer 2 or 3 for example,
code for x**y is probably pretty fast. (I am speculating
that the BINARY_POWER byte code opeator is optimized for
some special cases).

If y is fixed for the run through x, but you don't know it
beforehand, you could generate a lambda on the fly with a constant
built into the code with one eval and pass that to map, e.g.,

 >>> x = [1,2,3,4]
 >>> y=2
 >>> map(eval('lambda x: x**%s' % y), x)
 [1, 4, 9, 16]
 >>>
 >>> import dis
 >>> dis.dis(eval('lambda x: x**%s' % y))
           0 SET_LINENO               1
           3 LOAD_FAST                0 (x)
           6 LOAD_CONST               1 (2)      <<-- note constant from y
           9 BINARY_POWER
          10 RETURN_VALUE
 >>>
 >>> x=map(float,x); x
 [1.0, 2.0, 3.0, 4.0]
 >>> y=2.0
 >>> map(eval('lambda x: x**%s' % y), x)
 [1.0, 4.0, 9.0, 16.0]
 >>> dis.dis(eval('lambda x: x**%s' % y))
           0 SET_LINENO               1
           3 LOAD_FAST                0 (x)
           6 LOAD_CONST               1 (2.0)    <<-- note constant from y
           9 BINARY_POWER
          10 RETURN_VALUE

also note that the eval can effectively fold constant-valued expressions:
 >>> x=range(1,5); x; y=2; y
 [1, 2, 3, 4]
 2
 >>> map(eval('lambda x: x**%s' % (1.0/y)), x)
 [1.0, 1.4142135623730951, 1.7320508075688772, 2.0]
 >>> dis.dis(eval('lambda x: x**%s' % (1.0/y)))
           0 SET_LINENO               1
           3 LOAD_FAST                0 (x)
           6 LOAD_CONST               1 (0.5)     <<-- note constant from 1.0/y
           9 BINARY_POWER
          10 RETURN_VALUE

I suppose you could look at the eval('lambda ... as a kind of currying.

>"Giorgi Lekishvili" <gleki at gol.ge> wrote in message
>news:3D007666.2B6D3582 at gol.ge...
>> Hi all!
>>
>> Is there a way to map a function with several arguments to a list? E.g.,
>>
>> >>>import math
>> >>>import MLab
>> >>>a=MLab.rand(3,5)
>> >>>b=map(math.pow,a)
>>
>> of course, this doesn't work.
>>
>> What to do? My task is to optimize the power (i.e., y in pow(x,y)) to
>> achieve maximal performance.

Need more info about x and y for that ;-)

Regards,
Bengt Richter


From jubafre at zipmail.com.br  Tue Jun 18 17:58:19 2002
From: jubafre at zipmail.com.br (jubafre at zipmail.com.br)
Date: Tue, 18 Jun 2002 18:58:19 -0300
Subject: =?iso-8859-1?Q?results=20of=20stat=5BST=5FATIME=5D=3F=3F=3F=3F=3F=3F=3F?=
Message-ID: <3D0F9DAE0000034E@www.zipmail.com.br>

Hello!!!

when i use st[ST_ATIME] a int number is returned, how i can transform
this number to a "time number"???????

mode=st[ST_ATIME]
print mode
2134567890

i want just the file access time, please help-me!!!

Juliano Freitas
www.gebrasil.hpg.com.br





------------------------------------------
Use o melhor sistema de busca da Internet
Radar UOL - http://www.radaruol.com.br







From SBrunning at trisystems.co.uk  Tue Jun 25 10:16:44 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Tue, 25 Jun 2002 15:16:44 +0100
Subject: Suggestions for good programming practices?
Message-ID: <31575A892FF6D1118F5800600846864DCBD435@intrepid>

> From:	Emile van Sebille [SMTP:emile at fcfw.fenx.com]
> Simon Brunning 
> > It isn't in the 2.2 series until 2.2.1, IIRC.
> > 
> 
> Am I missing something?
> 
> F:\Python22>python
> Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
> Type "help", "copyright", "credits" or "license" for more information.
> >>> import this
> The ...
 
Clearly I *didn't* recall correctly...

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.




From johannes at zellner.org  Tue Jun 18 08:17:31 2002
From: johannes at zellner.org (Johannes Zellner)
Date: Tue, 18 Jun 2002 14:17:31 +0200
Subject: xml fo->pdf processor wanted
Message-ID: <slrnagu96r.9jc.johannes@orbital.zellner.org>

Hi,

is there a fo (formatting object) to pdf converter in python?

-- 
   Johannes


From tchur at optushome.com.au  Mon Jun 24 16:01:05 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Tue, 25 Jun 2002 06:01:05 +1000
Subject: SOLVED (Re: Python hits the spot)
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at> <af4vrl$ffe$0@216.39.172.122> <3D16DD5C.1050708@kfunigraz.ac.at> <uhdume2n1mhbba@corp.supernews.com> <3D171291.2060905@kfunigraz.ac.at>
Message-ID: <3D177A81.4D960B5C@optushome.com.au>

Siegfried Gonzi wrote:
> It was really the laptop processor. I tried the calculation on a
> stationary machine (Pentium II 450 MHz and SuSE 8.0), and it works.
> 
> Die Typen k?nnen jetzt was erleben.

Could someone who understands idiomatic ?sterreichisches Deutsch
confirm that the above means "And sorry for being such a prat."?

Tim C




From christophe.delord at free.fr  Thu Jun 20 16:53:55 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Thu, 20 Jun 2002 22:53:55 +0200
Subject: How to represent the infinite ?
References: <3d11ee9b$1@news.swissonline.ch>
	<20020620193230.53d58f8f.christophe.delord@free.fr>
	<aet4p6$33kk$1@nntp6.u.washington.edu>
	<20020620202732.68e67873.christophe.delord@free.fr>
	<slrnah4dnd.8m7.huaiyu@gauss.almadan.ibm.com>
Message-ID: <20020620225355.7b8dd9db.christophe.delord@free.fr>

On Thu, 20 Jun 2002 20:11:25 +0000 (UTC)
huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu) wrote:

> Christophe Delord <christophe.delord at free.fr> wrote:
> [regarding 1e1000 giving inf]
> >With Python 2.2 and Linux it works. The example I gave is a copy-paste from IDLE. It also works with my Python 1.5.2 and 2.3 (I'm under Linux on an Intel processor)
> 
> Does 1e200**2 also work for you out of the box?  I had to get rid of the
> artificial check for overflow in the source to really use inf in
> computations.  I don't remember if 1e1000 works on my box without patch.


>>> 1e200**2
Traceback (most recent call last):
  File "<pyshell#1>", line 1, in ?
    1e200**2
OverflowError: (34, 'Numerical result out of range')
>>> 1e400
inf

When the overflow occurs, the OverflowError exception is thrown. inf can't be the result of a computation (without patch and without catching exception).

> 
> Huaiyu


-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From whisper at oz.net  Mon Jun 24 22:24:53 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 24 Jun 2002 19:24:53 -0700
Subject: Suggestions for good programming practices?
In-Reply-To: <013901c21bec$8a950910$bd5d4540@Dell2>
Message-ID: <GCEDKONBLEFPPADDJCOEAEOODLAA.whisper@oz.net>

Structs and tuples are different. They're both of the same general category
of aggregate data, but they are different as Brian notes.

It's like people: boys are people and girls are people, but if you try to
access a strange girl by position, you're apt to get your face smacked or
worse! ;-)

<tongue value='most firmly in cheek'/>

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Brian Quinlan
> Sent: Monday, June 24, 2002 19:04
> To: 'Peter Hansen'; python-list at python.org
> Subject: RE: Suggestions for good programming practices?
>
>
> Peter Hansen wrote:
> > Brian Quinlan wrote:
> > >
> > > Peter Hansen wrote:
> > > > You _are_ returning them "in a struct".  It's spelled "tuple"
> > > > but other than that what's the difference?
> > >
> > > struct members are access by name while tuple elements are accessed
> > > positionally.
> >
> > Okay.  So why is that a problem?
>
> I'm not saying it is. I'm saying that tuples and structs are different.
>
> Cheers,
> Brian
>
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From daniel.dittmar at sap.com  Fri Jun 21 04:58:39 2002
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Fri, 21 Jun 2002 10:58:39 +0200
Subject: RE strings (was: Variable Interpolation - status of PEP 215)
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <4PfQ8.44245$n4.10307683@newsc.telia.net> <KhiQ8.44269$n4.10312432@newsc.telia.net> <29e28c51.0206201156.7e61a7f5@posting.google.com> <3D127B89.EE1C17F4@engcorp.com>
Message-ID: <aeuprv$ek$1@news1.wdf.sap-ag.de>

> Cimarron Taylor wrote:
>>   m = re'^EMP:([^,]*),([^,]*),([^,]*),([^,]*)$'.match(line)
>>   if m:
>>         ...
>>
>> is much clearer and more maintainable than
Peter Hansen replied:
> Given that the only real difference is you skip an import, and
> at the cost of modifying the language definition in an unusual
> manner, is there any point to this at all?

The implementation could choose to cache the result of re.compile at module
level, thus
- speeding up the 'naive' re usage
- freeing the programmer from doing the caching in code

I'm still against it, the road to Perl is paved with syntax extensions.

But as this thread started with Variable Interpolation: maybe someone could
come up with a general way of extendable string literals? Something which
would transform re"..." to re.compile ("...") and pass this to the parser,
something which would transform v'value of x: $x' to 'value of x' % x and
pass this to the parser, ...

Daniel





From bac at OCF.Berkeley.EDU  Tue Jun  4 01:18:26 2002
From: bac at OCF.Berkeley.EDU (Brett Cannon)
Date: Mon, 3 Jun 2002 22:18:26 -0700 (PDT)
Subject: ANN: Pure Python strptime()
In-Reply-To: <15612.14121.529921.388554@12-248-41-177.client.attbi.com>
References: <f85b5ab1.0206031354.132f3148@posting.google.com>
 <15612.14121.529921.388554@12-248-41-177.client.attbi.com>
Message-ID: <Pine.SOL.4.44.0206032214350.22223-100000@death.OCF.Berkeley.EDU>

On Mon, 3 Jun 2002, Skip Montanaro wrote:
>     Brett> Do I ... change the time module into a Python module with the
>     Brett> current module renamed _time and then just have a stub module to
>     Brett> globally import the _time module?
>
> Yes, that's the usual technique used to get a mixed Python/C module into the
> standard library.
>
>     Brett> Do I just need to convince Guido to include it?
>
> I've heard that often helps. ;-)
>

=) Guess it's time to join the python-dev mailing list and starting
talking to the BDFL.

>     Brett> I have posted this as a patch at SF (patch #474274), but I am
>     Brett> wondering if I have to do anything that I just previously
>     Brett> mentioned on top of this.  Any guidance on this would be helpful.
>
> You just did it.  I assigned it to myself and will try to take a look at it.
>

Thank you very much, Skip.

> (I don't suppose you have a unit test for this beast laying about do you?)
>

There is a large PyUnit suite with the patch (which I notied you found).
I am hammering out the bugs you discovered (actually, only the month one,
the %c directive didn't get reproduced for me; can you run strftime() with
%c as the format and email it to me?).

> --
> Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
> Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html
>
>

-Brett C.





From mhuening at zedat.fu-berlin.de  Wed Jun 26 13:18:03 2002
From: mhuening at zedat.fu-berlin.de (Matthias Huening)
Date: 26 Jun 2002 17:18:03 GMT
Subject: Problems using samples
References: <337619fa.0206260913.41e08940@posting.google.com>
Message-ID: <Xns9239C44CFB16Emhueningzedatfuberli@130.133.1.4>

gusraan at terra.com.br (Raphael Ribeiro) wrote in 
news:337619fa.0206260913.41e08940 at posting.google.com:

> this is my code, it says there's an error in line 6, but i can't find
> what **** error is that, can anyone help me?

your line 6 is:
> 6 if x = y:

should be:
> 6 if x == y:

Matthias


From johnroth at ameritech.net  Fri Jun 21 07:59:02 2002
From: johnroth at ameritech.net (John Roth)
Date: Fri, 21 Jun 2002 07:59:02 -0400
Subject: CR+LF problem
References: <3d11bc33.1129927479@news.mad.ttd.net> <12q3hukse2cepcm9tpern92jc71a5vrhvi@4ax.com> <3d1309ff.1215379262@news.mad.ttd.net>
Message-ID: <uh656hci5vkh52@news.supernews.com>

"Manuel Ruiz" <mruiz at safa.es> wrote in message
news:3d1309ff.1215379262 at news.mad.ttd.net...
>
> Hi,
>
> char * txt = "print 123\r\n" is only a sample, I am storing scripts in
> a database, I put the script into the database with \n but when I get
> it, It has \r\n as end of line, I have assumed this is a database
> problem.

Thanks. That defines it better. Yes, it's a data base interface
problem. Whatever you're using isn't using the C standard for
newlines, it's using the platform standard, and returning \r\n.
It should return either \n, or nothing.

> The rare thing is that PyRun_Simple() function can't execute this kind
> of python code using \r\n as end of line, however I saved that code to
> file and ran "python2 myfile.py" (with \r\n as end of line too) and it
> ran fine, very fine.

Well, Python is built on top of C. In C (and C++) all platform
dependent line endings are translated to a simple \n.

> Where is the diferrence between the api of python and the interprete
> itself. Because first one can't execute this code and the second one
> can do it ?

There's no difference once it actually tries to execute.
What's happening to you is that the process
of writing it to the file and then reading it back is cleaning up the
input. More specifically, the process of reading it back is removing
the \r, and on input, the interpreter expects lines to end with \n, and
removes them before passing it to the execute function.

For your problem, there are two possible solutions:

1. Your data base is misconfigured. If you tell it that the program
is a C (or C++) interface, it might not give you the \r. I have no
idea if this is true, or if your data base even supports this kind of
smarts, and I'm not the one that can help you on it.

2. You need to remove the \r\n sequence yourself. Please
notice that this is a requirement when you read lines from a
file - Python always leaves the \n in place, for reasons that
are spelled out in the FAQ (I think.) In other words, you will
always have to remove the \n, if it exists.

John Roth





From jbublitzNO at SPAMnwinternet.com  Sun Jun 16 20:49:52 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Mon, 17 Jun 2002 00:49:52 GMT
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de> <aeftmj$m29$1@bob.news.rcn.net> <3d0c6d67@nntp.server.uni-frankfurt.de> <3D0CE698.1050806@SPAMnwinternet.com> <aeiu3a$4ff$1@bob.news.rcn.net>
Message-ID: <3D0D3223.5070500@SPAMnwinternet.com>

David Abrahams wrote:
> "Jim" <jbublitzNO at SPAMnwinternet.com> wrote in message
> news:3D0CE698.1050806 at SPAMnwinternet.com...

>>On the other hand, I maintain PyKDE (I should point out I'm not the sip
>>author however), and for an actively maintained project with nearly a
>>dozen libraries and several hundred files, sip seems preferable.

> I'd really like to know what you find preferable about sip -- aside your own
> familiarity with it, of course, which for me usually counts more than any
> other single factor ;-). I'm actively developing v2 of Boost.Python and
> would really like to make it the best tool possible for this job, so it
> would be useful to know more.

I inherited sip rather than choosing it after 'careful evaluation',
and certainly familiarity is a big part of it. In fact I've only
given other alternatives a cursory look, so what I like about sip
may already be available in Boost or other packages.

The biggest plus for sip is that the 'IDL' is the h file syntax
with a few subtractions (variable names, forward declarations,
inline code, but inline methods stay), so that most of the
description file creation is easy to automate and still not
that hard to do manually for a small number of files. sip
supports most C++ h file syntax.

The other features that are helpful on large projects are the
built-in support for versioning and embedded docs (both also fairly
easy to automate), as well as support for 'platforms' and
'features'. So for example PyQt (not my project) can support Qt
from 1.4.2 through 3.0.4 with different mixes of features (with
SQL or not, QTable or not) across Windows, Linux/Unix/BSD, and
(soon) Mac, with a single set of source files, and still be
manageable by one person.

sip is flexible enough that you can handwrite code to "Pythonize"
C++ code (return a Python list or dict, or for example,
'void rgb (int&, int&, int&)' returns (int, int, int) and takes
no arguments in PyKDE), or create type substitutions/conversions
('mapped types') once and re-use them across a project. You can
do C++ with anything of course, but sip has it well-integrated
(manageable) and has support code to make it easier (arg unpacking,
object conversions, etc). IIRC, Boost might not need handwritten
C++ code quite as much.

I suppose those are a series of small advantages, but the end
result seems to me that you write less code and you do it in
fewer but well-known locations. It seems to add up to a big
difference on a large project.

Boost (IIRC) also has a big advantage in being able to bind much
more selectively. sip is easiest when binding everything in a lib
and probably takes a little more effort to be selective - you
can do it, but it isn't as 'mindless' (ie automatable) as most
sip file creation is. And of course Boost is actually
documented - no small thing.

How would you see Boost performing on a very large project -
say 10 libs, 20 classes/lib, 10 exposed methods/class (2000
methods overall)? I haven't looked at Boost in a while -
could you incorporate conditional compilation (eg versions,
features) and embed docs easily? Could you automate code
generation to some extent? The automation for generating
PyKDE is pretty crude at the moment, and in all fairness
isn't generally available, so this example would be a lot
of work using sip as it comes out of the box. Is Boost
really intended for projects of this size (there probably
aren't many), or do you have different goals for it?


Jim



From buzzard at urubu.freeserve.co.uk  Mon Jun 24 18:40:02 2002
From: buzzard at urubu.freeserve.co.uk (Duncan Smith)
Date: Mon, 24 Jun 2002 23:40:02 +0100
Subject: ? and %s placeholders, help?
References: <af5oq7$sj1$1@news8.svr.pol.co.uk> <m2k7oolcrm.fsf@video.paradise.lost> <3D178542.604@mxm.dk>
Message-ID: <af87cj$n5b$1@newsg3.svr.pol.co.uk>

Thanks both,
                   But I'm still having problems with this.  I want to use
prepared statements to avoid unnecessary parsing, and I get the impression
(from what information I can find, eg Python Programming on Win32) that I
should (need to?) use the '?' placeholder to achieve this.  But I cannot
come up with anything that works.


>>> query = 'INSERT INTO %s (%s) VALUES (%s)' % (tblname, string.join(vars,
', '), string.join(['%s']*len(vars), ', '))
>>> query
'INSERT INTO tbl (var1, var2, var3) VALUES (%s, %s, %s)'
>>> curs.execute(query % ("'a'", "'b'", "'c'"))
1L

#Hurray, but does this avoid parsing the statement on each INSERT?

>>> query = 'INSERT INTO %s (%s) VALUES (%s)' % (tblname, string.join(vars,
', '), string.join(['?']*len(vars), ', '))
>>> query
'INSERT INTO tbl (var1, var2, var3) VALUES (?, ?, ?)'
>>> curs.execute(query, ("'a'", "'b'", "'c'"))
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
  File "C:\Python22\Lib\site-packages\MySQLdb\cursors.py", line 70, in
execute
    raise ProgrammingError, m.args[0]
ProgrammingError: not all arguments converted
>>>

I can always write the data to a text file and dump it to a MySQL table.
But I would like to get this tried out first.  Cheers.

Duncan




From kragen at pobox.com  Sun Jun  2 22:18:07 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 02 Jun 2002 22:18:07 -0400
Subject: convert string to a method of a class at runtime?
References: <5YvK8.285$Vm3.130161@news.uswest.net>
Message-ID: <83y9dxrtpc.fsf@panacea.canonical.org>

"Kevin Altis" <altis at semi-retired.com> writes:
> >>> s
> 'def world(self, s):\n    print s\n\n'
> >>> c = compile(s, '', 'exec')
> >>> c
> <code object ? at 0164B620, file "", line 1>

Now 'c' is not the code object that is the body of 'world'; it is a
code object which, if executed, will create a function called 'world'.

> >>> import new
> >>> f = new.function(c, globals())
> >>> h.__class__.__dict__['world'] = new.instancemethod(f, None, h.__class__)
> >>> h.world('new world')
> Traceback (most recent call last):
>   File "<input>", line 1, in ?
> TypeError: ?() takes no arguments (2 given)

And, not surprisingly, c takes no args.

> Fingers crossed, that this kind of dynamic method addition is doable.

For Q&D, you could probably get away with exec astring in globaldict,
aclass.__dict__.

This works for me in Python 2.1.1 and 1.5.2:

class MethodBroken(Exception): pass

def methodadd(astring, aclass, globaldict=globals()):
    localdict = {}
    exec astring in globaldict, localdict
    methodnames = localdict.keys()
    if len(methodnames) != 1:
        raise MethodBroken("Method code didn't define exactly one method",
                           astring,
                           methodnames)
    name = methodnames[0]
    setattr(aclass, name, localdict[name])

I was surprised to learn not long ago that setattr() on class objects
will turn ordinary functions into instance methods for you.

> But I need to be able to do the addition outside the shell, so compiling the
> string is necessary.

exec will compile for you if you give it a string instead of a code
object.  If you want, you can compile separately:

    code = compile(astring, '', 'exec')
    exec code in globaldict, localdict

That might give you better control over error handling, for example.

> In more complex examples, the method needs the namespace to be right
> so it can make use of whatever imports and other global variables
> have been setup.

Unfortunately, I don't think there's a way to extract the globals from
a class unless it has methods.



From gerhard at bigfoot.de  Wed Jun  5 14:17:57 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 5 Jun 2002 18:17:57 GMT
Subject: new SmallEiffel elj-win32 release (-0.74)
References: <8cd6cc47.0206050602.1d69a237@posting.google.com>
Message-ID: <slrnafslcu.116m.gerhard@lilith.my-fqdn.de>

Geoff Eldridge wrote in comp.lang.python:
> A new version of elj-win32 has been prepared and is now available from ..

I wonder if we should post all our Python announcements to the Eiffel
newsgroup, too. That sure would increase their traffic a lot ;-)

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From amckay at merlintechnologies.com  Fri Jun  7 12:49:17 2002
From: amckay at merlintechnologies.com (Andy McKay)
Date: Fri, 7 Jun 2002 09:49:17 -0700
Subject: Medium to Large Scale Python Deployments
In-Reply-To: <3D000B36.5C7BCEBF@engcorp.com>
References: <mailman.1023405789.30972.python-list@python.org> <aJTL8.21666$ZF5.2505752733@newssvr10.news.prodigy.com> <3D000B36.5C7BCEBF@engcorp.com>
Message-ID: <20020607164917.GA11618@mckay.merlintechnologies.com>

> Running pycount (http://starship.python.net/crew/gherman/playground/pycount/ )
> as "pycount.py -R *.py" in the top level of your source tree would tell
> you something more accurate if you wished.

According to cvs we had 90k LOC. After pycount, it became 60k of which only
30k was actually code, the rest comments etc. How to make a project smaller.
And a serious refactoring could get it down to about a third of that. 

We'll match those perl guys and get it down to a one line app soon :)
-- 
  Andy McKay
  Merlin Technologies




From eric.brunel at pragmadev.com  Thu Jun  6 06:32:10 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Thu, 6 Jun 2002 10:32:10 +0000
Subject: prototyping good OOdesign in Python?
References: <XFMail.20020601131915.shalehperry@attbi.com> <Pine.LNX.4.33.0206030725420.5660-100000@rnd.onego.ru> <mailman.1023310034.25601.python-list@python.org>
Message-ID: <adn6bb$pv2$1@wanadoo.fr>

Geoff Gerrietts wrote:
[snip]
> Since Python doesn't care about most of this stuff, modelling it is
> only useful if you expect to port the code later.

Eeeer, maybe I'll start yet another flame war here, but I strongly disagree 
with this statement! Modelling is *always* useful if your application isn't 
a "quick and dirty" one, is significantly large and/or if it may have 
several versions and/or several persons that will work on it. UML diagrams 
are not just another representation of the code: it allows the developers 
to take a step back and look at the whole picture instead of just their 
little part of the code. If you have several cooperating classes, it also 
allows to visually design these cooperations, instead of having to look 
everywhere in the code to see where a particular class is used (and also 
*if* it's still used at all...). In fact, I find UML design particularly 
useful on large/medium Python applications, precisely *because* of Python's 
dynamic features and lack of strong typing. If you want your code to be 
readable by other people than just yourself, and maybe at a time where 
you'll not even be around anymore, a good documentation of the code is 
really needed...

Whether UML is well-suited for documenting Python applications is of course 
debatable. I personaly find it quite convenient:

- on one hand, it counter-balances the "fuzzy" part of Python (no 
declarations, everything is dynamic) by allowing to actually declare 
everything. Let's be honest here: in a significantly large application, a 
big part of the code make assumptions on the objects it handles, and may 
benefit from strong typing. Not having to declare everything's type is 
quite handy, but may in the end make the code unreadable if you do not have 
something that tells you what is needed on the objects you manipulate...

- on the other hand, UML is quite open and may also allows to document 
"fuzzy" features, or things really specific to a language like Python. For 
example, UML doesn't know anything about Python dictionaries, since 
dictionaries are quite often not natively supported by other languages. But 
if you know your target language will be Python, you may just use a 
specific "custom" type like "dict<key type, value type>" in your UML spec. 
Of course, if the tool you use to design your UML diagrams is too 
restrictive, it may report an error on this kind of declaration, but 
well..., no big deal... The whole purpose of UML diagrams is documentation, 
not making correct diagrams for Rational Rose or whatever other tool...

Of course, many features of UML tools cannot be used at all with Python: 
semantics checking may be disturbed by your custom Python types, and I 
don't know a single UML tool that can generate Python code (if anybody 
knows one, please let me know!). But again: no big deal... When you'll get 
back to your code in a few months or years, you'll be really glad to have 
your UML spec at hand to understand how your few dozens classes speak to 
each other...
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From brueckd at tbye.com  Wed Jun 12 18:15:45 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 12 Jun 2002 15:15:45 -0700 (PDT)
Subject: Something faster then sgmllib for sucking out URLs
In-Reply-To: <20020612202828.GB16934@matijek.plusseven.com>
Message-ID: <Pine.LNX.4.44.0206121503560.24931-100000@ns0.tbye.com>

On Wed, 12 Jun 2002, Alex Polite wrote:

> I'm working on a webspider to fit my sick needs. The profiler
> tells me that about 95% of the time is spent in sgmllib. I use sgmllib
> solely for extracting URLs. I'm looking for a faster way of doing
> this. Regular expressions, string searches? What's the way to go? I'm
> not a python purist. Calling some fast C program with the html as
> argument and getting back a list of URLs would be fine by me.

How fast do you need it to be? Using regular expressions + other junk (see
below) I get about 1 MB/sec on a 900 MHz Pentium III - that's a lot of
HTML and probably faster than you'll be reading it off the 'net. Also, you
can probably simplify it as I needed to get nearly all URLs (not just
hrefs); also I specifically ignore images and uglified it in a few more 
places for speed reasons.

The code below returns a list of start/end URL indices but you can easily
modify it to return the actual URLs (make sure you return them from the
original buffer because for speed this function lowercases the entire web
page first and some web servers are case sensitive). Disclaimer: I've
added to this function over time as I've encountered weird cases, but that
obviously doesn't mean it now handles all weird cases. :)

-Dave

# Note that group(1) of validTagRE tells which tag we matched
validTagRE = re.compile('<(a|embed|object|param)\s.*?>', re.DOTALL)
urlStartRE = re.compile('(href|src|codebase|value)\s*=\s*[\'"]?\s*')
urlEndRE = re.compile('\s+|>')
def FindLinks(data):
    'Returns a list of (start, end) url indices'
    data = data.lower() # much faster than using re.I
    cur = 0
    links = []
    while 1:
        # Locate a tag we're interested in
        tagMatch = validTagRE.search(data, cur)
        if not tagMatch: break
        tagEnd = tagMatch.end()
        cur = tagEnd + 1

        # Now search for attributes within that tag that contain urls
        urlCur = tagMatch.end(1) + 1 # e.g. for '<embed ...', start=7
        while 1:
            urlMatch = urlStartRE.search(data, urlCur, tagEnd)
            if not urlMatch: break

            # Find the end of the url - extra work required because we
            # find HTML like href=foo, href="foo" href=" foo" href='foo', 
etc.
            urlStart = urlMatch.end()
            urlEnd = -1
            lastChar = urlMatch.group().strip()[-1]
            if lastChar == '=':
                # URL not quoted, end is whitespace or '>'
                endMatch = urlEndRE.search(data, urlStart, tagEnd)
                if not endMatch: break # Bleh! Bad HTML!

                urlEnd = endMatch.start()
                if urlEnd > tagEnd: break # More bad HTML!

                # One last check: In Javascript, might have escaped quotes
                if data[urlStart:urlStart+2] == '\\"' and \
                   data[urlEnd-2:urlEnd] == '\\"':
                    urlStart += 2
                    urlEnd -= 2
            elif lastChar == '"':
                # End will be a double quote
                if data[urlStart] != "'": # Make sure not part of 
Javascript
                    urlEnd = data.find('"', urlStart)
            elif lastChar == "'":
                # End will be a single quote
                if data[urlStart] != '"': # Make sure not part of 
Javascript
                    urlEnd = data.find("'", urlStart)

            if urlEnd != -1:
                # Skip not-really-URLs
                ok = 1
                ch = data[urlStart]
                if ch == 'j' or ch == 'm':
                    if data[urlStart:urlStart+11] == 'javascript:' or \
                       data[urlStart:urlStart+7] == 'mailto:':
                        ok = 0
                if ok:
                    links.append((urlStart, urlEnd))
            else:
                urlEnd = tagEnd

            urlCur = urlEnd + 1
            if urlCur >= tagEnd:
                break
            
    return links





From cliechti at gmx.net  Thu Jun 20 19:51:14 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 21 Jun 2002 01:51:14 +0200
Subject: Are generators in nested functions possible?
References: <mailman.1024614876.11270.python-list@python.org>
Message-ID: <Xns92341349C8797cliechtigmxnet@62.2.16.82>

"Bjorn Pettersen" <BPettersen at NAREX.com> wrote in 
news:mailman.1024614876.11270.python-list at python.org:

> I've got the following code to generate the indicies of all nearest
> neighbours (span levels) in a multi dimensional qube:
> 
> def generateIndexes(length=5, span=1):
>     spanvalues = range(-span, span+1)
>     
>     def genInd(res, length):
>         if length == 1:
>             for val in spanvalues:
>                 yield res + [val]
>         else:
>             for i in range(length):
>                 for val in spanvalues:
>                     genInd(res + [val], length-1)

here you generate generator objetcs and throw them away 'cause you dont use 
them in a loop aor anything!

if you replace that last line with:
    	    	    	    	 for x in genInd(res + [val], length-1):
        	    	    	     	yield x

it yields some data, don't know if it's what you want ;-)


chris





>     return genInd([], length)
> 
> However, when I try to use it it doesn't seem to return anything:
> 
>  >>> for i in generateIndexes(3):
>   ...   print i
>   ...
>  >>>
> 
> I tried getting the iterator (or is it generator?) directly and calling
> next() on it, and it seems like it raises a StopIteration exception
> immediately:
> 
>  >>> g = generateIndexes(3)
>  >>> g.next()
>   Traceback (most recent call last):
>     File "<stdin>", line 1, in ?
>   StopIteration
>  >>>
> 
> Can anyone help me understand what is going on here?
> 
> -- bjorn
> 
> 
> 



-- 
Chris <cliechti at gmx.net>



From gringro1 at umbc.edu  Thu Jun 20 19:09:41 2002
From: gringro1 at umbc.edu (mike)
Date: Thu, 20 Jun 2002 19:09:41 -0400
Subject: Raw String Output
Message-ID: <aetnkf$tot$1@news.umbc.edu>

If I have a string variable with the value '\n' how do I output it to a file
with the value of '\012' instead ?

Thank you in advance!




From prabhu at aero.iitm.ernet.in  Wed Jun 12 16:39:56 2002
From: prabhu at aero.iitm.ernet.in (Prabhu Ramachandran)
Date: Thu, 13 Jun 2002 02:09:56 +0530
Subject: ANN: MayaVi-1.2 released
Message-ID: <15623.45468.950445.451894@monster.linux.in>

Hi,

This is to announce the availability of the MayaVi Data Visualizer
version 1.2.

MayaVi is a free, easy to use, scientific data visualizer.  It is
written in Python and uses the Visualization Toolkit (VTK) for the
graphics.  It provides a GUI written using Tkinter.  MayaVi is free
and distributed under a BSD license.  It is also cross platform and
should run on any platform where both Python and VTK are available.

For more information, downloadable sources and binaries, screenshots,
installation instructions, documentation etc. visit the MayaVi home
page at:

	http://mayavi.sourceforge.net

Also bundled with MayaVi is a VTK pipeline browser written in Python
and a new utility module that makes using VTK easier from the
interpreter.


New in this release:

    (*) Changed license from GPL/LGPL to BSD. [1]

    (*) Added an integrated data picker.  Thanks to Jose Paulo for
    initial contributions.

    (*) Added a light control toolkit (contributed by Raymond Maple).
    This lets you configure 8 different lights using a GUI.

    (*) Added a CellData to PointData filter.

    (*) Fixed bad problem in the lookup table editor that would freeze
    MayaVi under Win32.

    (*) The scalar/vector range can be set to predetermined values.  A
    visible range for data can also be specified.  Anything outside
    this range is made transparent.

    (*) Added --display and --geometry command line options and a new
    help menu.

    (*) Geomview OOGL export option (thanks to Jeremy Gill).

    (*) RPM and Debian packages are available in addition to source
    and binary releases for Win32 and GNU/Linux.  PDF docs are also
    available at the site.


[1] Please don't flame me on the license change issue.  I posted about
    it on the MayaVi-users list and gave it considerable thought.  You
    can read the original post here:

 http://sourceforge.net/mailarchive/forum.php?thread_id=694778&forum_id=1824
 
    It seemed that moving to BSD would be the best possible option.


Have fun!
prabhu




From cliechti at gmx.net  Thu Jun 20 17:18:01 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 20 Jun 2002 23:18:01 +0200
Subject: application development - separating content from function
References: <yroQ8.260949$i6.17986369@amsnews02.chello.com>
Message-ID: <Xns9233ED737D952cliechtigmxnet@62.2.16.82>

"Henry Baumgartl" <h.baumgartl at chello.NOSPAM.nl> wrote in
news:yroQ8.260949$i6.17986369 at amsnews02.chello.com: 

> For eight months now, I am preparing for the development of an open
> source Python application to support individuals in Life Long
> Learning. It will be a mixture of PIM, extended diary, groupware,
> project management, and modules to interact with centralised knowledge
> / skill exchange servers. 
> 
> Most of the user activity is done on individual boxes. However, when
> and if needed, the application must provide intuitive and standardised
> interaction with fellow network members (e.g. project workgroups,
> mentor, and such) and centralised server modules.
> 
> What needs to be done is:
> 1. Develop GUIs and modules that are interchangeable on, at least,
> Linux, Windows, and Mac.

tkinter or wxWindows come to mind, they should run right out of he box 
without special efforts of your side on those platforms, they're portable.

something like the "observer pattern" should help separating GUI and 
problem domain and it make easier to read code.

> 2. Use the modularity of Python to allow personalised application
> adaptation, thus requiring an effective way of integrating personal
> variables into almost all scripts.

the standard module ConfigParser provides .ini style configuaration. this 
is easy to save window positions, user settings in general and much more.

if you need real plugin with dynamic loaded functions and classes:
import+reload, module imp or execfile() can be used to load plugins (.py 
files) at runtime.

in that case you migh define a simple interface like making it mandatory 
for an object to inherit from a special base class and or the file must 
provide an init() function to register itself in the main app.

> 3. And, last but not least ;-) the application needs to be
> multi-lingual. 

start with:
def _(s): return s

and write all the string you want to localize later with 
_("hello")
_("this can be translated later by replacing the _ function")

later you can use locale or a simple dictionary lookup to translate.
i saw that technique in "cplay" http://www.tf.hut.fi/~flu/cplay/

simple with a dict:
translation = {'hello':'ole'}
def _(s):
    	return translation.get(s,s)
 
> The more I have tried to solve these requirements in an elegant and
> efficient manner, the further I seem to stray from anything remotely
> resembling a viable solution. In other words, I can't see the forest
> for the trees anymore.

do one after the other, use small scripts/interactive prompt to test hacks.
maybe use CVS or some other source repository so you can try out things 
without fear as you can go back in time to get the last working version.

> Anyone working on something similar? Or, just plain brilliant in all
> things Python, and ready and willing to point me in the right
> direction and make it all worth while again :-)
> 
> Help!
> 
> Henry Baumgartl
> Dynamic Knowledge Network
> 
> *Please remove .NOSPAM from my e-mail address when abswering me
> directly.* 
> 
> 
> 



-- 
Chris <cliechti at gmx.net>



From irmen at NOSPAMREMOVETHISxs4all.nl  Thu Jun 13 18:16:15 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Fri, 14 Jun 2002 00:16:15 +0200
Subject: M2Crypto: select() behaves weird on SSL sockets
References: <3D04ED17.5070606@NOSPAMREMOVETHISxs4all.nl> <ae2str$b0q$1@nntp6.u.washington.edu> <3D066E77.4070901@NOSPAMREMOVETHISxs4all.nl> <ae6bmb$frc$1@dahlia.singnet.com.sg> <3D0790E1.9080904@NOSPAMREMOVETHISxs4all.nl> <831ybavprx.fsf@panacea.canonical.org>
Message-ID: <3D0919AF.7050903@NOSPAMREMOVETHISxs4all.nl>

Kragen Sitaker wrote:

 > Irmen, why aren't you using asyncore?
 >
 > For those of us who use asyncore, an asyncore.dispatcher-compatible
 > wrapper around an M2Crypto socket that you could hand to asyncore and
 > hang handlers on would probably be sufficient.

My system (Pyro) is built around threads all the way.
Switching to asyncore/medusa requires a big restructuring of
my code, and that wasn't part of the plan for the next release of Pyro.

I'm still not certain if asyncore/medusa fits my requirements for Pyro,
but perhaps it is worthwile to investigate this for a future release.
(I need to be able to accept new connections, and start processing
the requests, while Pyro is still busy processing other requests.
I don't see how asyncore can do this. And if we're running on a
multi-processor box, how can asyncore take advantage of the extra CPUs?)

Irmen



From thenault at nerim.net  Thu Jun 13 05:30:00 2002
From: thenault at nerim.net (Sylvain Thenault)
Date: 13 Jun 2002 11:30:00 +0200
Subject: How to check version of Python under CGI
References: <rbmggus7ln1uiutktf5libf4bbngu29ji7@4ax.com> <873cvrin1q.fsf@tux.ntw23.fr> <cgnggu40ethjhpqcmltnfb06d5qvuf078k@4ax.com> <t1ogguoua7b9ne9n3mb3n7rkp8mrqe2nee@4ax.com> <87y9djh6tg.fsf@tux.ntw23.fr> <ndpggu8uqsmi6ppumkem4c16rt7p5i6k9o@4ax.com>
Message-ID: <87ptyvh6g7.fsf@tux.ntw23.fr>

Johann <observer at NOSPAM.space.pl> writes:

> On 13 Jun 2002 11:22:03 +0200, Sylvain Thenault <thenault at nerim.net>
> wrote:
> 
> >I think the best way to know if a module is available is to try to import it:
> 
> But it is difficult to import all modules. I found much better
> sollution. Ready to run  CGI script at http://snakefarm.org/
> 
> Thanx to Simon :)

I didn't see the solution on the newsgroup, could you share it ?

-- 
Sylvain Th?nault 


From loewis at informatik.hu-berlin.de  Tue Jun 18 07:54:17 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 18 Jun 2002 13:54:17 +0200
Subject: Unicode to default encoding...
References: <aen3ej$1ko$1@norfair.nerim.net>
Message-ID: <j4elf4x0nq.fsf@informatik.hu-berlin.de>

"Gillou" <nospam at bigfoot.com> writes:

> I'm looking for a way to encode an Unicode string to the default encoding
> for the output (console, file...).
> 
> someunicodestring.encode('iso8859-1')
> 
> works for a file (windows & unix) but gives ugly characters on a windows
> console.

This is difficult to do, since you need to find out the "OEM
codepage". On Western European and Northern American Windows
installations, this is either 437 or 850, so "cp437" or "cp850" may
produce proper results.

Notice that the user can change the codepage away from the default.

I believe there is currently no API to find out what code page is used
in a certain console window. You can try parsing the output of chcp.exe.

Regards,
Martin



From hesterloli at hotmail.com  Tue Jun 18 20:46:16 2002
From: hesterloli at hotmail.com (George Hester)
Date: Wed, 19 Jun 2002 00:46:16 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <yXBP8.44417$GY.13865423@twister.nyroc.rr.com> <3D0FD592.93BFBCDB@engcorp.com>
Message-ID: <svQP8.46757$GY.14772070@twister.nyroc.rr.com>

Your apology (?) is well taken and I apologize myself.  Later I tried to think
of code in ASP using VBScript to determine the version of ASP running and I
realized that does not come right off the top of my head.  So there is no reason
why it should be any easier in Python.  My apologies again Please!  I will look
at what you wrote again without jumping down your throat.

--
George Hester
_________________________________
"Peter Hansen" <peter at engcorp.com> wrote in message
news:3D0FD592.93BFBCDB at engcorp.com...
> George Hester wrote:
> >
> > Listen you aren't my daddy.  If you do not want to help constructively then
> > please get off my boat.
>
> I thought I was being constructive in two distinct ways, although
> perhaps I was more blunt than you were able to handle.  If so,
> my apologies for appearing to patronize you.  That was not my
> intention.
>
> I did point out two helpful things (I thought).  One was to
> show you by explicit example how you can't access sys.version
> until you've imported the sys module.  You didn't seem to
> understand that was required from your response in another post,
> and I believe you might have missed the point again in
> reading my response.  If you read the post again, you'll see
> that I got the same error you did until I did "import sys"
> and then the problem went away.  As that would have solved
> your initial problem, I felt it would be fairly helpful.
>
> Secondly, and I believe I'm still correct here, I pointed out
> that you might have a lot of trouble trying to juggle two
> things at the same time: learning Python, and using it in
> an unforgiving environment as an ASP scripting language.
> I have been in your position many times in the past, and
> every time I have been burned by ignoring my own instincts.
> Never attempt to take on two difficult tasks at once when
> you can partition the problem and approach one task at a
> time.  Learn Python first (without ASP involved), _then_
> try to use it for ASP.
>
> You may choose to ignore my advice.  It was, after all,
> given freely and, as always, is worth what it costs you...
>
> Cheers,
> -Peter




From i.linkweiler at gmx.de  Sun Jun  9 08:02:50 2002
From: i.linkweiler at gmx.de (Ingo Linkweiler)
Date: Sun, 09 Jun 2002 14:02:50 +0200
Subject: Logic programing with python?
Message-ID: <3D0343EA.4070100@gmx.de>

Hello,

is it possible to use Python as "logic lanuage" ?
I have found the project "formula", a logical language coded with 
python, and Chris Meyers "Prolog interpreter".
I have heard from "holmes", but can not find any information about it.

Do you know other sources?

Ingo



From tjreedy at udel.edu  Tue Jun 11 12:44:08 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Tue, 11 Jun 2002 16:44:08 GMT
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
References: <Xns9229EE722467Fgustafl@195.100.94.182> <DafN8.125526$4i.12087144@bin2.nnrp.aus1.giganews.com>
Message-ID: <sNpN8.132081$4i.12440184@bin2.nnrp.aus1.giganews.com>

> "Gustaf Liljegren" <gustafl at algonet.se>
wants elegant function matching int to any of multiple numbers and
ranges

My very-pseudo code had parity error at end.  Real Python code:

import bisect

# noncontiguous and sorted numbers and ranges: 1, 3, 10-300, 2999,
12345-98765
# expanded to (lo, hi+1) ranges and flattened to list - exercise for
OP
rangenums = (1,2, 3,4, 10,301, 2999,3000, 12345,98766)

for i in (0,1,2, 3,4, 10,20, 300,301, 90000,98765,98766,100000):
  print bisect.bisect(rangenums, i) %2,

0 1 0 1 0 1 1 1 0 1 1 0 0
>>>





From max at alcyone.com  Sun Jun  2 20:50:28 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sun, 02 Jun 2002 17:50:28 -0700
Subject: What is a classmethod for?
References: <addgv30f4d@enews4.newsguy.com> <addlm6$g44$1@panix1.panix.com> <adebub0mdl@enews1.newsguy.com>
Message-ID: <3CFABD54.C03329DD@alcyone.com>

Dennis Peterson wrote:

> ...much like Java static methods. I see the point there, I'm just
> trying to
> figure out the purpose of the implicit class object in classmethod.

It simply generalizes a case.  When you have a function associated with
a class, you might want the class passed in as another argument.  That
would facilitate, for instance, having the same function (converted to a
static method with staticmethod) act as a static method in multiple
classes.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Who'd ever think it / Such a squalid little ending
\__/ The American and Florence, _Chess_
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From cimarron+google at taylors.org  Thu Jun 20 15:56:47 2002
From: cimarron+google at taylors.org (Cimarron Taylor)
Date: 20 Jun 2002 12:56:47 -0700
Subject: RE strings (was: Variable Interpolation - status of PEP 215)
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <4PfQ8.44245$n4.10307683@newsc.telia.net> <KhiQ8.44269$n4.10312432@newsc.telia.net>
Message-ID: <29e28c51.0206201156.7e61a7f5@posting.google.com>

I like it! 

To me

  m = re'^EMP:([^,]*),([^,]*),([^,]*),([^,]*)$'.match(line)
  if m:
        ...

is much clearer and more maintainable than

  import re
  emp_re = re.compile(r'^EMP:([^,]*),([^,]*),([^,]*),([^,]*)$')
  m = re.match(emp_re, line)
  if m:
      ...

Cim


From bokr at oz.net  Sat Jun 22 13:53:26 2002
From: bokr at oz.net (Bengt Richter)
Date: 22 Jun 2002 17:53:26 GMT
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at>
Message-ID: <af2dim$sh6$0@216.39.172.122>

On Sat, 22 Jun 2002 15:59:10 +0200, Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote:

>A few weeks ago I complaint here that Python on Windows XP sucks. I had
>a problem which would took theoretically about 1.5 hours; but in reality
>it takes 6 hours on Windows XP and Python 2.2. After that 6 hours I must
>reboot my system in order to work on.
If you have the funds, and Tim is available and willing, perhaps you could
hire him to solve the problem. If the problem is a bug in Python, it would
be great to get it fixed ;-)

If the problem is in the way your code uses Python, e.g., consuming memory
resources and not releasing them for reuse, it would be good to identify
the pattern as a Python programming pitfall. If it is a problem with your
use of SWIG, or F2PY, likewise it would be good to know what the real problem is.
Also you could get your program fixed, and the FUD from this unresolved
problem could be dispelled.

If somehow your program has exposed a problem shared by Windows and Linux,
that would be interesting too ;-)

>
>In the following I got the advice to install a real user system on my
>laptop. Okay, I bought Linux SuSE 8.0 and installed it on my laptop (256
>MB RAM, 1000 MHz Celeron). I also installed Python 2.2 (the ActivePython
>version, otherwise I couldn't install idle due to the annoying
>"free_software_garbage_2_weeks_installation_nightmare").
>
>SWIG helped me to make my external Python C function (with gcc); and
>F2PY helped me to create my external Python Fortran 90 function (with
>the Intel Fortran 95 compiler).
Did you run individual tests on these? Are you mixing support libraries
and compiled code from different environments??
>
>The directories and all the other programs and data are the same as on
>Windows.
>
>
>1. After starting the process from within idle the complete system
>freezes after the first wavelength (after about 20 minutes). I couldn't
>kill a process or open a different session (via F2) in order to kill
>python. The last resort: turn off the power supply.
What did you determine about why it "froze"??
>
>2. I thought it is better to start the process from the command line and
>in a X terminal: python forcing.py
>The same behavior as aforementioned: last resort: power supply. That is
>really good and fine because a Linux system really is greedy for "shut
>down immediately".
If there were a problem with your code, why wouldn't it move to the
new context? ISTM you have decided the problem is in the context, and
are desperately trying to find a context that will make your code work,
but what if the problem is in the code?

>
>3. I thought it has to do with Gnome and I logged in to a KDE session.
>Same behavior as in point 1. and 2.
Sounds like looking for another part of the context to be at fault.
(Do you actually need a GUI, BTW?)
>
>
>My conclusion: I will rewrite my simulation in Fortran 90 and will
>abadon Python. Python is maybe good for some small scripts but not
>appropiate for serious programming.
Rewriting in Fortran may solve your problem, because it will change
your code, not just the context.
>
>Thank you Billy Boy for a forgiving user system Windows XP which in turn
>relieves Python's "wreak havoc" behavior.
?? ISTM you are pointing fingers, but you have not yet really diagnosed
the problem (I may have missed some posts, but I would think you'd mention
in this post if you knew what the real problem cause was).

>
>I cannot believe that Linux is responsible for Python's ill behavior
>under Linux. Before you start ranting about note the following: In a
No, I doubt whether the context is responsible, especially since you
have tried two. But I suspect it's not Python per se either.

>first step it has nothing to do with my programming style (admittedly
>not very good and famous) it has to do that the same calculation cannot
>be performed on Linux and Python 2.2!
The "same calculation" is evidently not "the same." ;-)

Have you never chased a bug that only shows up in secondary effects?
In C code, the most common way for that to happen is storing something
through a bad pointer, but instead of getting a memory fault, it corrupts
some data structure or data value. The value could be used 20 minutes later
and cause GPF then, or e.g., it could be data that's part of a calculation,
but maybe it's only overwriting the least significant half of a floating
point number, so you get the effect systematic bias or random error injected
into your calculation, but nothing else. If you don't look for explanations
of last-decimal errors, this kind of thing can persist a long time. OTOH,
you could be corrupting numbers that are used in an iterative calculation
that mysteriously starts to take longer to converge, and does so in a systematically
increasing way, because the corrupting data varies systematically with time.
Or you could be corrupting data structures and sabotaging memory management
without causing immediate GPF, etc. etc.

The classic thing then is to compile a debug version, and have it run just fine,
because memory layout is different, and the bad pointer usage just throws a value
into an unused space (e.g., space at the end of an input line buffer that is never
used with max length lines). (Combine this with multi threading and multiprocessing,
and you may have a _really_ challenging debugging problem ;-)

If you move from windows to Linux, the chances are something is laid out differently ;-)
That's just an example. There are lots of ways to generate secondary-symptom bugs.

Another possibility that could lead to bogging down is a sensitive iterative algorithm.
If your problem is running in a very sensitive region, just the differences in a few
least significant bits returned by different library routines for the same function
can do it. Or differences due to different order of evaluation between different compilers.

Your result could be bouncing around chaotically and only satisfying convergence
criteria with some systematically decreasing probability, instead of deterministically
converging. Hopefully you would know if you are operating on the hairy edge, but such things
can be hidden in the way an algorithm is implemented. If C is involved, float instead of double
somewhere, in conjunction with some unhappy choice of coordinate system origins, could
easily cause problems.  BTW, if you have iterative algorithms, have you printed out iteration
counts, for starters? Anything surprising there?

Do you have some _exact_ test problems that you've used for unit tests? _ANY_ deviation from
_exactly_ correct results needs an explanation.

What happens if you systematically stub out major routines? Can the program run
to completion without slowing down? Can you identify the routine that is causing
the problem (or inducing Python to cause the problem)? How about just timing various
parts, to see which are constant and which is bogging down?

You apparently have a lot of different stuff glued together. Have you tested the
separate pieces separately? Do you have some test problems that you can use to validate
numerical results *exactly* (trailing decimal differences need an explanation. They
can be benign/acceptable or not).

I hate to see Python (or anything) bashed without proof. So far I haven't seen the proof,
nor an explanation.

If your code is open source, maybe someone will be interested in discovering why
it bogs Python down (or Python bogs it down ;-)

Regards,
Bengt Richter


From davidma at eskimo.com  Wed Jun 12 21:50:11 2002
From: davidma at eskimo.com (David Margrave)
Date: 12 Jun 2002 18:50:11 -0700
Subject: socket module htonl/ntohl bug
Message-ID: <f2f390f9.0206121750.1fa384b2@posting.google.com>

Hello,

The following statement should work, but does not, on python 1.5.2 and
python 2.2:

>>> socket.htonl(pow(2L,32)-1)
Traceback (innermost last):
  File "<stdin>", line 1, in ?
OverflowError: long int too long to convert

2^32-1 is a valid 32-bit value, equal to 32 1s in binary, and htonl
should be able to handle it.  This is not related to the argument
being a long int and htonl not accepting that type, because
socket.htonl(4294967295) raises the same exception, and 4294967295 can
be represented as a regular int in python2 (but not 1.5).

By way of comparison, the following simple C program works properly:

#include <stdio.h>
#include <sys/socket.h>

int main(void)
{
        printf ("%u\n", htonl(4294967295));
}

When run, it prints out '4294967295' as expected.


Dave


From kragen at pobox.com  Sat Jun  1 01:55:30 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 01 Jun 2002 01:55:30 -0400
Subject: matching one time through a loop
References: <B43D149A9AB2D411971300B0D03D7E8BF0A335@natasha.auslabs.avaya.com> <mailman.1022806551.14232.python-list@python.org> <3CF6D7D2.4894DF36@engcorp.com>
Message-ID: <83adqfwnjh.fsf@panacea.canonical.org>

Peter Hansen <peter at engcorp.com> writes:
> I believe "programmer productivity" is significantly improved when 
> using Python, especially because of the emphasis on readability.
> 
> Productivity needs to include not just the time to type the statement,
> which is a miniscule portion of the overall effort, but also the
> time to revise the design repeatedly (assuming code-and-fix, which 
> 95% of programmers use) or the time to write test cases, whether
> after (for 4% of programmers) or incrementally during (for those
> using Test-Driven Development).  In all cases, code maintenance is
> a far greater proportion of the work than mere typing, and I 
> really don't believe the minor time savings when typing provided by Perl

I find aline = re.sub(r'\s+', aline, ' ') less readable than
$aline =~ s/\s+/ /g;, but maybe that's because I've used Perl since
1994 and Python only since 1999 or so.  In particular, I never have to
look up the order of arguments in the Perl construct.  aline =
re.compile(r'\s+').sub(aline, ' ') is only marginally better, and its
readability (like that of recent Python-URL!s) suffers because of
verbosity.



From rnd at onego.ru  Fri Jun  7 08:25:17 2002
From: rnd at onego.ru (Roman Suzi)
Date: Fri, 7 Jun 2002 16:25:17 +0400 (MSD)
Subject: Efficient python programming...
In-Reply-To: <3D00A456.4315EDA3@engcorp.com>
Message-ID: <Pine.LNX.4.44.0206071620570.3531-100000@suzi.com.onego.ru>

On Fri, 7 Jun 2002, Peter Hansen wrote:

> Eddie Corns wrote:
> > 
> > A friend is fond of quoting this to me:
> > 
> >   Rules of Optimization:
> >     Rule 1: Don't do it.
> >     Rule 2 (for experts only): Don't do it yet.
> >   -- M.A. Jackson
> > 
> > I wasn't going to bother contributing since others have said the same thing
> > but the more people say the more you'll believe it :)
> > 
> > Especially for a beginner the important thing is to get clear and obvious
> > code.  Later, IF IT ISN'T FAST ENOUGH, then you think about ways to speed it
> > up.
> 
> You forgot the even more important first thing for a beginner: 
> get it correct!  The only good way to do this is to write unit
> tests that prove it.  If you don't have tests, you don't have 
> working code (or how can you prove it?).

To prove that code is correct is very difficult (practically impossible).
Unittests are there to prove that you have errors, not the other way
around.

Of course, they can increase your confidence in that optimized code is
equivalent to what you had before.

> -Peter

Sincerely yours, Roman A.Suzi
-- 
 - Petrozavodsk - Karelia - Russia - mailto:rnd at onego.ru -
 





From a-steinhoff at web.de  Sun Jun 23 07:56:30 2002
From: a-steinhoff at web.de (Armin Steinhoff)
Date: 23 Jun 2002 04:56:30 -0700
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at>
Message-ID: <ddc19db7.0206230356.777bf1b3@posting.google.com>

Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote in message news:<3D1482AE.8FA299D8 at kfunigraz.ac.at>...
> A few weeks ago I complaint here that Python on Windows XP sucks. I had
> a problem which would took theoretically about 1.5 hours; but in reality
> it takes 6 hours on Windows XP and Python 2.2. After that 6 hours I must
> reboot my system in order to work on.
> 
> In the following I got the advice to install a real user system on my
> laptop. Okay, I bought Linux SuSE 8.0 and installed it on my laptop (256
> MB RAM, 1000 MHz Celeron). I also installed Python 2.2 (the ActivePython
> version, otherwise I couldn't install idle due to the annoying
> "free_software_garbage_2_weeks_installation_nightmare").

It seems you have not a way with that "free_software_garbage"

> SWIG helped me to make my external Python C function (with gcc); and
> F2PY helped me to create my external Python Fortran 90 function (with
> the Intel Fortran 95 compiler).
> 
> The directories and all the other programs and data are the same as on
> Windows.
> 
> 
> 1. After starting the process from within idle the complete system
> freezes after the first wavelength (after about 20 minutes). I couldn't
> kill a process or open a different session (via F2) in order to kill
> python. The last resort: turn off the power supply.

Seems so that _YOUR_ Python version shuts down the whole machine ... 
It's simple -> BS in means BS out :) 

 
> 2. I thought it is better to start the process from the command line and
> in a X terminal: python forcing.py
> The same behavior as aforementioned: last resort: power supply.

Your (extended) Python version seems to be really CRAPPY !!

> That is
> really good and fine because a Linux system really is greedy for "shut
> down immediately".
> 
> 3. I thought it has to do with Gnome and I logged in to a KDE session.
> Same behavior as in point 1. and 2.

Why the heck should your problem related to Gnome or KDE ??
_YOU_ have implemented a problem into Python ...

(geborener Weltmeister ??)
 
> My conclusion: I will rewrite my simulation in Fortran 90 and will
> abadon Python.

Your problem has nothing to do with Python  ... your implementation is the problem.

>Python is maybe good for some small scripts but not
> appropiate for serious programming.

Just bizarr!! Are you able to do serious program development??

Have a look to the SQL database 'gadfly' ... and then tell me again 
that Python isn't appropriate for serious programming 
 
> Thank you Billy Boy for a forgiving user system Windows XP

 ... and its crazy users :)

Armin


which in turn
> relieves Python's "wreak havoc" behavior.
> 
> I cannot believe that Linux is responsible for Python's ill behavior
> under Linux. Before you start ranting about note the following: In a
> first step it has nothing to do with my programming style (admittedly
> not very good and famous) it has to do that the same calculation cannot
> be performed on Linux and Python 2.2!
> 
> 
> S. Gonzi


From mruiz at safa.es  Thu Jun 20 07:33:55 2002
From: mruiz at safa.es (Manuel Ruiz)
Date: Thu, 20 Jun 2002 11:33:55 GMT
Subject: CR+LF problem
Message-ID: <3d11bc33.1129927479@news.mad.ttd.net>

Hi,

------------------------
.
char * txt = "print 123\r\n";
result = PyRun_Simple( txt, Py_file_input, PyDict_New(), NULL );
.
-----------------------

this code fails to exec txt, I got a SyntaxError exception,
however I if save to file the statement "print 123\r\n" and try to run
it ussing python interprete (version 2.2.1) It run fine.

Can anyone tell me anything about this rare behaviour

Regards
Manuel Ruiz


From narnett at mccmedia.com  Mon Jun 17 10:44:11 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Mon, 17 Jun 2002 07:44:11 -0700
Subject: MySQL database location
In-Reply-To: <aekasd$96t$1@news.onet.pl>
Message-ID: <NDBBLAEMOMEHKALAMKKJKEADLAAA.narnett@mccmedia.com>


> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Pawel Lewicki
> Sent: Monday, June 17, 2002 2:42 AM
> To: python-list at python.org
> Subject: MySQL database location
>
>
> Hi,
> Is is possible to get the physical path to the MySQL database file using
> MySQLdb or any other pythonous way?

Yes.  Execute "SHOW VARIABLES" and you'll get a list of name-value pairs for
all of MySQL's current variables.  The data directory is "datadir".

For example, to print them:

	self.dbh.execute("SHOW VARIABLES")
	print self.dbh.fetchall()

Nick





From idbaxter at semdesigns.com  Sat Jun  1 09:40:43 2002
From: idbaxter at semdesigns.com (Ira Baxter)
Date: Sat, 1 Jun 2002 08:40:43 -0500
Subject: python tool: finding duplicate code
References: <mailman.1022676587.23318.python-list@python.org>
Message-ID: <3cf8cc82$1@giga.realtime.net>

If you are interested in strong clone detection,
you should check out
http://www.semdesigns.com/Products/Clone/index.html

--
Ira Baxter, Ph.D. CTO Semantic Designs
www.semdesigns.com  512-250-1018

"Michal Wallace" <sabren at manifestation.com> wrote in message
news:mailman.1022676587.23318.python-list at python.org...
>
> In "Refactoring: Improving the Design of Existing Code",
> Martin Fowler and Kent Beck list duplicate code as their
> number one "Code Smell"
>
> Since I'm looking to clean up some of my projects, I went
> looking for a way to find duplicate lines in my python
> project... So I wrote a little python program to do it for
> me.
>
> Features:
>
>     - lists all pairs of overlapping files
>     - shows how many lines are in common
>     - shows each duplicated line
>     - ignores indentation
>     - filters out try, pass, if __name__=="__main__", etc.
>
> It's just string-matching, so it won't find duplicate logic
> with different variable names or layout, but it *can* find
> cut and paste issues.
>
> (hmm... Come to think of it, someone could probably find
> *some* duplicate logic by running source files through the
> tokenizer first. I wonder if that would work...)
>
> Anyway, just thought I'd share:
>
> http://cvs.sabren.com/sixthdev/cvsweb.cgi/sdunit/overlaps.py?rev=1.1
>
> Cheers,
>
> - Michal   http://www.sabren.net/   sabren at manifestation.com
> ------------------------------------------------------------
> Learn to build web apps!      http://www.webAppWorkshop.com/
> ------------------------------------------------------------
>
>
>




From nico at tekNico.net  Sat Jun  8 16:55:55 2002
From: nico at tekNico.net (Nicola Larosa)
Date: Sat, 08 Jun 2002 22:55:55 +0200
Subject: Removing ^M
References: <Pine.LNX.4.21.0206091005170.7346-100000@localhost.localdomain>
Message-ID: <3D026F5B.7060401@tekNico.net>

> I am trying remove ^M characters (some kind of newline character) from an
> HTML file. I've tried all sorts of string.replace and sed possibilities
> but the things just won't go away. Does anyone have a way of removing such
> characters?

Python source distribution includes, at least since 1.5.2 , the files 
crlf.py and lfcr.py in Tools/scripts/ . Here is crlf.py :


#!/usr/bin/env python

"Replace CRLF with LF in argument files.  Print names of changed files."

import sys, re, os
for file in sys.argv[1:]:
     if os.path.isdir(file):
         print file, "Directory!"
         continue
     data = open(file, "rb").read()
     if '\0' in data:
         print file, "Binary!"
         continue
     newdata = re.sub("\r\n", "\n", data)
     if newdata != data:
         print file
         f = open(file, "wb")
         f.write(newdata)
         f.close()


The only difference in lfcr.py are the description and the re.sub line; they 
become:

"Replace LF with CRLF in argument files.  Print names of changed files."
...
     newdata = re.sub("\r?\n", "\r\n", data)


There are lots of interesting nuggets in the Demo/ and Tools/ directories of 
the Python source distribution.


-- 
"Too much cleverness in the parser can turn against you."
   Guido Van Rossum

Nicola Larosa - nico at tekNico.net






From d2002xx at myrealbox.com  Sat Jun 22 18:22:28 2002
From: d2002xx at myrealbox.com (d2002xx)
Date: 22 Jun 2002 15:22:28 -0700
Subject: The best GUI library in python
References: <8d3f4438.0206220457.b820418@posting.google.com> <3D1481DE.C0331204@engcorp.com>
Message-ID: <8d3f4438.0206221422.6f71287a@posting.google.com>

Peter Hansen <peter at engcorp.com> wrote in message news:<3D1481DE.C0331204 at engcorp.com>...
> How fast do you need your GUI to be?  Are you talking about 
> an unacceptably slow loading time, or response time within 
> the application?

Response time within the application, "idle" is. I think it's not due
to python, because boa-constructor (using wxWindows) and some examples
of PyQT work much faster than idle and others based on Tkinter.

Well, if you doubt what I said, try to use grail. Or write a listbox
containing 10,000 items, then wait......


From anoop79 at myrealbox.com  Sun Jun 30 18:09:47 2002
From: anoop79 at myrealbox.com (Anoop P B)
Date: Mon, 01 Jul 2002 03:39:47 +0530
Subject: problem with spawnle
Message-ID: <3D1F81AB.1070906@myrealbox.com>

Hi,

	I am trying to call a java application through my python script on 
windows. After consulting the documentation, i quickly learnt that the 
os.spawn*p versions are not available on windows. so, i tried the following:

os.spawnle(os.P_NOWAIT, "java.exe", "java", "testProg", "testParam", 
os.environ)

As I understand, this should work by picking up the path system variable 
  from os.environ for setting up the environment for the new process (or 
am I missing something here?). But what I get is....

Traceback (most recent call last):
   File "server.py", line 48, in ?
     ct.run()
   File "server.py", line 25, in run
     os.spawnle(os.P_NOWAIT, "java.exe", "java", "testProg", 
"testParam",  os.environ)
   File "c:\python22\lib\os.py", line 541, in spawnle
     return spawnve(mode, file, args[:-1], env)
OSError: [Errno 2] No such file or directory

Can someone please throw some light on possible solutions to this 
problem. My ultimate aim is only to run the java program through my 
python script. However, I cannot use os.system() since my python program 
gets blocked which is not acceptable. I need my python script to call 
the java program and continue executing.

Thanks in advance,
Anoop

note: the java class is in the same directory in which my python script lies





From dfackrell at DELETETHIS.linuxmail.org  Mon Jun 24 12:12:03 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Mon, 24 Jun 2002 10:12:03 -0600
Subject: How to stop a SocketServer?
References: <7xhejv9af2.fsf_-_@ruckus.brouhaha.com>
Message-ID: <3d1744d4$1_1@hpb10302.boi.hp.com>

I faced this problem with a simple chat server I wrote.  I expected that
termination of the server with <ctrl-C> should have stopped it immediately,
but it instead waits until some communication is received from each thread
that was stopped at a read().

It appears that the current threading implementation does not provide for
termination of threads inside a single python statement (such as a
filelike.read()), so I am considering a move to character I/O, checking to
verify that a character is available before each read.

It's been a while since I looked at it, so I don't have the actual methods
fresh in my mind, but hopefully this will help.

--
Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.


"Paul Rubin" <phr-n2002b at NOSPAMnightsong.com> wrote in message
news:7xhejv9af2.fsf_-_ at ruckus.brouhaha.com...
> I want to run a simple network server based on
> SocketServer.ThreadingTCPServer.  I followed the docs
> and implemented a handler class which works fine.
> Then I instantiate the server and run the serve_forever() method.
> The server listens for connections, starts threads for them
> and handles them just like it should.
>
> Question: how do I shut down the server once it's running, besides
> doing something crude like killing the process?  I'd like to be able
> to connect to the server and enter a "shutdown" command and have the
> server gracefully exit and free the port.  I haven't been able to
> figure out how to do that.
>
> Suggestions?  Thanks.




From SBrunning at trisystems.co.uk  Fri Jun 21 07:21:48 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Fri, 21 Jun 2002 12:21:48 +0100
Subject: I'd give up Perl tomorrow if only...
Message-ID: <31575A892FF6D1118F5800600846864DCBD40B@intrepid>

> From:	David Mitchell [SMTP:djmitchell at NOSPAM.optushome.com.au]
> ASCII<->EBCDIC conversion is pretty easy to implement in just about any
> language, but it's also easy to make a typo somewhere when you're typing
> in
> lots and lots of ASCII codes.  I know; I've done it ;->
 
See
<http://groups.google.com/groups?selm=mailman.1021958915.15497.python-list%4
0python.org>.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.




From ajw126 at york.ac.uk  Wed Jun  5 14:53:21 2002
From: ajw126 at york.ac.uk (Andrew Wilkinson)
Date: Wed, 5 Jun 2002 19:53:21 +0100
Subject: Detecting OS and Window Manager
Message-ID: <ufsnh8pmmde6cf@corp.supernews.com>

Hi,

I'm writing a program that integrates with a number of different OSes and
Window Managers (under Linux), but I need to check what the user is running.
I know that...

import os
if os.name == 'nt':

will tell me they're running NT. But under Linux I'd like to know if they
are running KDE or Gnome, does anyone know a (preferably) simple test for
this?

Thanks in advance
Andrew Wilkinson




From loewis at informatik.hu-berlin.de  Thu Jun 13 03:48:35 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 13 Jun 2002 09:48:35 +0200
Subject: Extension objects in Python 1.5 and 2.2
References: <mailman.1023952757.30131.python-list@python.org>
Message-ID: <j4sn3r7h64.fsf@informatik.hu-berlin.de>

Martin Sj?gren <martin at strakt.com> writes:

> How do I write extension objects so that they work in both Python 1.5 and
> the 2.x series? 

This can work only on the source level. Binary compatibility of
modules is, in general, not possible between 1.5 and 2.x (even though
it will work fine in certain cases).

> Until now, I've used PyObject_New and PyObject_Del for 2.x, are
> there an equivalent set for 1.5 and if so, can they be used in 2.x
> too, or do I have to #ifdef?

In 1.5, you need to use PyObject_NEW and PyMem_DEL. You can get the
1.5 documentation from

http://www.python.org/doc/1.5/

> What's the correct way to do this? And gee, it would be nice if the
> objects were classes in 2.2 too, but I suppose that'd be a LOT of
> #ifdefs.

You would normally use #ifdefs, indeed.

To make extension types classes, the biggest problem is that you need
to fill additional slots in the type objects, slots that were not
available in 1.5. So you #ifdef part of the type object (potentially
also the functions in the slots, although that is not strictly
necessary).

Regards,
Martin



From a.schmolck at gmx.net  Mon Jun 10 13:33:27 2002
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 10 Jun 2002 18:33:27 +0100
Subject: Could someone submit this to the sf-bug tracker, please?(Was: Re: __imul__ broken for 'objects' in 2.2.1)
References: <yfsd6v3qa86.fsf@black132.ex.ac.uk> <lk660tlnon.fsf@pc150.maths.bris.ac.uk> <yfs660r8dgl.fsf_-_@black132.ex.ac.uk> <mailman.1023722773.12051.python-list@python.org>
Message-ID: <yfsr8jf6ntk.fsf@black132.ex.ac.uk>

>     >> Hmm.  You should file a bug on sf so this doesn't get forgotten
>     >> (unless it's already been reported, but I don't think so).
> 
>     Alexander> Well yes, but then I *can't* which is why I've posted the
>     Alexander> previous message.
> 
>     Alexander> I tried again today, but logging in from the submission form
>     Alexander> has no effect, I'm back to the same form that tells me to
>     Alexander> "please log-in" *after* I have completed the log-in step (if
>     Alexander> I submit the wrong password, however I do get an error
>     Alexander> message). I've tried 3 different browsers (mozilla .94,
>     Alexander> konqueror and netscape 4.7)
> 
> Do you have cookies enabled?

I had cookies enabled for konqueror (it was the first thing I made sure, after
encountering the problem), but I must admit that I stupidly forgot to check
again later when I tried it with mozilla and netscape and it turns out that I
hadn't.  It doesn't seem possible to submit a bug report with the version of
konqueror I'm using, but it worked OK with netscape just now and the bug
report is now finally filed. My excuses for not figuring this out earlier.

> 
>     Alexander> If my problems don't reflect the weirdness of the proxy I'm
>     Alexander> forced to use, I'd really suggest that either the decision
>     Alexander> not to allow anonymous bug reports should be reviewed again.
> 
> This is really not an option.  Far too often an incomplete bug report would
> be filed.  When a python developer wanted more input there was nobody
> listening.   There was then no choice but to close the bug report
> unresolved.

Pitty, but understandable.

alex


From logiplexsoftware at earthlink.net  Tue Jun 11 20:15:53 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 11 Jun 2002 17:15:53 -0700
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
In-Reply-To: <sNpN8.132081$4i.12440184@bin2.nnrp.aus1.giganews.com>
References: <Xns9229EE722467Fgustafl@195.100.94.182>
	<DafN8.125526$4i.12087144@bin2.nnrp.aus1.giganews.com>
	<sNpN8.132081$4i.12440184@bin2.nnrp.aus1.giganews.com>
Message-ID: <20020611171553.077d5ebd.logiplexsoftware@earthlink.net>

On Tue, 11 Jun 2002 16:44:08 GMT
Terry Reedy wrote:

> 
> > "Gustaf Liljegren" <gustafl at algonet.se>
> wants elegant function matching int to any of multiple numbers and
> ranges
> 
> My very-pseudo code had parity error at end.  Real Python code:
> 
> import bisect
> 
> # noncontiguous and sorted numbers and ranges: 1, 3, 10-300, 2999,
> 12345-98765
> # expanded to (lo, hi+1) ranges and flattened to list - exercise for
> OP
> rangenums = (1,2, 3,4, 10,301, 2999,3000, 12345,98766)
> 
> for i in (0,1,2, 3,4, 10,20, 300,301, 90000,98765,98766,100000):
>   print bisect.bisect(rangenums, i) %2,
> 
> 0 1 0 1 0 1 1 1 0 1 1 0 0
> >>>

Hm. I haven't used bisect() before:


from bisect import bisect
from time import time
from random import random, randrange, shuffle

class Ranges:
    def __init__(self, ranges):
        self.ranges = []
        self.breakpoints = []
        for r in ranges:
            try:
                s, e = r
            except TypeError:
                s, e = r, r
            self.ranges.append((s, e))
        self.ranges.sort()
        self.breakpoints = map(lambda r: r[0], self.ranges)

    def has(self, n):
        s, e = self.ranges[bisect(self.breakpoints, n) - 1]
        return s <= n <= e

def isin(n, ranges):
    for r in ranges:
        try: 
            s, e = r
            if s <= n <= e:
                return 1
        except TypeError:
            if n == r:
                return 1
    return 0

def createRanges(n):
    # create a list of non-overlapping sequences and ints, there's probably a
faster way...
    ranges = []
    s = 0
    for i in range(n):
        l = int(random() * 1000) + 1
        e = s + l
        s = randrange(s, e)
        if int(random() * 10) % 2:
            ranges.append((s, e))
        else:
            ranges.append(s)
        s = e
        e += l
        
    shuffle(ranges)
    return ranges


# ------------
ranges = createRanges(10000)

print "Initialize Range:",
t1 = time()
r = Ranges(ranges)
print time() - t1

t2 = time()
print "Range():",
for i in range(0, 100, 3):
    r.has(i)
t = time()
print t - t2, "total", t - t1

t = time()
print "isin():",
for i in range(0, 100, 3):
    isin(i, ranges)
print time() - t


####################################
Initialize Range: 0.331060051918
Range(): 0.00136804580688 total 0.33601808548
isin(): 4.98800003529

Clearly faster although it does use a bit more memory...

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From claird at starbase.neosoft.com  Fri Jun  7 16:14:26 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 7 Jun 2002 15:14:26 -0500
Subject: Comparison: HTTP with Java, Perl or Python
References: <3D009911.9000904@gmx.de>
Message-ID: <5166C087A87C2AA7.39E7F099EDDE2E63.1EF2E28D1B41E320@lp.airnews.net>

In article <3D009911.9000904 at gmx.de>,
Ingo Linkweiler  <i.linkweiler at gmx.de> wrote:
>Hello,
>
>for a comparison of languages I need a complete "translation" of this short skript into Perl, Tcl and Java:
>
> myurl = "http://www.uni-dortmund.de"
> import urllib
> file  = urllib.urlopen(myurl)
> text  = file.read()
> file.close()
> print text
			.
			.
			.
How literal should this be?
  $myurl = 'http://www.uni-dortmund.de';
  use LWP::Simple;
  $text = get($myurl);
  print $text, '\n';
-- 

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


From nagarajagadi at yahoo.com  Tue Jun  4 04:00:01 2002
From: nagarajagadi at yahoo.com (nagaraj)
Date: 4 Jun 2002 01:00:01 -0700
Subject: accessing database of other applications
References: <56334c65.0206022208.363ade63@posting.google.com>
Message-ID: <56334c65.0206040000.38cefbe@posting.google.com>

nagarajagadi at yahoo.com (nagaraj) wrote in message news:<56334c65.0206022208.363ade63 at posting.google.com>...
> HI 
> I am a fresh developer in palmOS. i wanted to know how can I access
> the database of some other application (eg:datebook, memopad or any
> other user application) through my application. can I use the
> datamanager functions (APIs) given in sdk? if so how? if I use
> dmOpenDatabaseByTypeCreator how to get the creatorID of a perticular
> database?
> thanx in advance



to clarify it more.. if i dont know anything about the other
application's database (name, creator, type etc )... and even if I go
searching the databases one by one how to know a perticular database
belongs to the application? ( assuming that the appInfo block in
database is blank).. if it is possible then please anyone tell me.. if
not is it necessary to know the creator or name of the database?

thanks


From sholden at holdenweb.com  Wed Jun 12 23:27:44 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 12 Jun 2002 23:27:44 -0400
Subject: Medium to Large Scale Python Deployments
References: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com> <mailman.1023409393.3361.python-list@python.org> <3D003EC7.C72D27BE@engcorp.com> <3D006320.5040805@web.de> <3D00A209.EC24F671@engcorp.com> <adr2ao$bei$0@216.39.172.122> <3D0140BF.A239E39E@engcorp.com>
Message-ID: <7hUN8.529345$%u2.28370@atlpnn01.usenetserver.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3D0140BF.A239E39E at engcorp.com...
> Bengt Richter wrote:
> >
> > Is the "comments" number a count of _pure_ comment lines,
> > # (like this one) ?
>
> I believe so.  I haven't examined the pycount code yet.
>
> > I think it would be interesting to analyze comments further,
> > to see what the relative comment byte count is overall,
> > and for lines with both code and comments,
> > """like this line:""" # (this has code and comment).
>
> I consider it a mild stylistic no-no to use line-ending
> comments, personally.  I believe it's almost always better
> to put a block-comment of one or more lines above a block
> of code with related purpose.  A good comment there tends to
> be more maintainable, and useful, than lots of little comments
> on each line.
>
I wouldn't necessarily disagree, but I used (when programming (gulp) Basic
Plus) to make the line-end comments into a free-form narrative that rough;y
parallelled the code.

[...]
regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From seeger at sitewaerts.de  Wed Jun 12 06:15:08 2002
From: seeger at sitewaerts.de (Felix Seeger)
Date: Wed, 12 Jun 2002 12:15:08 +0200
Subject: Reload an imported file
Message-ID: <ae76vb$4l6iq$1@ID-79019.news.dfncis.de>

Hi

I have a config file. This is a normal python file.
Can I reimport this file during runtime ?
So changing the configuration will not need a restart of my program.

thanks
have fun
Felix


From robin.siebler at corp.palm.com  Mon Jun 10 15:58:46 2002
From: robin.siebler at corp.palm.com (Robin Siebler)
Date: 10 Jun 2002 12:58:46 -0700
Subject: No such directory?
Message-ID: <886c5e4b.0206101158.680c6278@posting.google.com>

When I try to change into a certain directory, I get an errorm message
that states the directory doesn't exist.  The directory does exist,
however, and the listdir function shows that it exists.  What gives?

>>> os.chdir('c:\\18_510GMC1\\18_510GMC1')
Traceback (most recent call last):
  File "<interactive input>", line 1, in ?
OSError: [Errno 2] No such file or directory:
'c:\\18_510GMC1\\18_510GMC1'
>>> os.getcwd()
'c:\\18_510GMC1'
>>> os.listdir(os.getcwd())
['18_510GMC']


From kragen at pobox.com  Sun Jun  2 21:33:22 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 02 Jun 2002 21:33:22 -0400
Subject: Python equivalent of shell's '$?'
References: <20020531100700.GN14959@software.plasmon> <1022843294.642.8.camel@walter> <mailman.1022847890.10145.python-list@python.org>
Message-ID: <831ybptacd.fsf@panacea.canonical.org>

Dave Swegen <dswegen at software.plasmon.com> writes:
> > > (I've learnt that catching
> > > errors from sub shells is highly unreliable, and ended up having not
> > > checking for errors, as I would sometimes get spurious results back).
> 
> I can't quite remember exactly what the problem was (I seem to recall
> that the errno returned was not what the called program returned). This
> was using both open and backticks. It was yet another nail in the coffin
> that is my opinion of perl ;)

FWIW wait(2) returns a 16-bit quantity with 8 bits of exit code and 8
bits of status information (normally 0 or a signal number).  $? and
the various Python and Perl methods of getting exit status return what
wait(2) returns.  Shift it right by eight bits to get what the program
passed to _exit(2).



From K.Rdt at TU-Berlin.DE  Thu Jun 27 10:23:07 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Thu, 27 Jun 2002 16:23:07 +0200
Subject: doc for threading?
Message-ID: <SOMHDAA9VS07KI8JXVXWM93PJ95EB.3d1b1fcb@FRITZweb>

---------------------------------------------------------------------
Hi
Is there a more elaborated doc (with examples)
for threading than the distribution is providing?
		K at Rdt
---------------------------------------------------------------------







From max at alcyone.com  Mon Jun  3 17:28:53 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 03 Jun 2002 14:28:53 -0700
Subject: Is there a Text to HTML conversion module?
References: <mailman.1023132485.29235.python-list@python.org> <adgn01$d18$1@slb5.atl.mindspring.net>
Message-ID: <3CFBDF95.92099C9E@alcyone.com>

Andrew Dalke wrote:

> <wink/>
> 
> >(sorry... couldn't resist)
> 
> Neither could I.  :)

Ah, but that's XML, not HTML.  :-)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Who'd ever think it / Such a squalid little ending
\__/ The American and Florence, _Chess_
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From martin at v.loewis.de  Thu Jun  6 03:21:35 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 06 Jun 2002 09:21:35 +0200
Subject: 64-bit Python?
References: <87elfl3yvu.fsf@mathdogs.com>
Message-ID: <m33cw0ub28.fsf@mira.informatik.hu-berlin.de>

Mike Coleman <mkc+dated+1024627496.6928dd at mathdogs.com> writes:

> Has anyone looked at making Python 64-bit clean?

I believe the interpreter itself is 64-bit clean; the Alpha port has
been working for a long time.

> I was looking at trying to use the new mmap module to map in a large
> (>2G) file, and the comments in the source note that the mmap
> regions are limited to int-size (31 bits), primarily because Python
> objects are also limited to this size.

I don't think this is the case. Yes, Python objects are limited to
int-size - whether or not this means 31 bits depends on your 64-bit
platform. It appears that you have a specific 64-bit platform in mind,
where sizeof(int) is 4.

> Does anyone know if there's an inherent reason why it would be
> impossible or difficult to make Python 64-bit clean (say as a
> compilation option)?

I would not make it a compilation option. I think the best approach
would be to use size_t as the object size; that may potentially cause
problems in case people want to put negative numbers into the size.
Because of this problem, it might be appropriate to write a PEP.

Notice that this change does not make it "64-bit clean" - there still
might be other problems that nobody has noticed so far.

Regards,
Martin


From derek at wedgetail.com  Fri Jun 21 20:51:55 2002
From: derek at wedgetail.com (Derek Thomson)
Date: Sat, 22 Jun 2002 10:51:55 +1000
Subject: I'd give up Perl tomorrow if only...
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au> <mailman.1024681648.27687.python-list@python.org>
Message-ID: <3D13CA2B.1000501@wedgetail.com>

Christopher A. Craig wrote:
> While there isn't any single Python repository as nice as CPAN, you
> can find a module to do just about anything.  Given your list I can
> find a module for everything but Tie on the first try using
> Google. (and I'd guess the fact that I have no idea what Tie is
> contributed to my not finding it)

Yes, but it's been my experience that they are not as well supported or 
complete as their Perl equivalents. Also, I often find that when a 
module does exist, it has been effectively desupported and is no longer 
usable. A flaw common among among open source software - it's 
interesting to "scratch that itch" and develop something cool, it's less 
interesting to support it, keep it up-to-date and generally fight 
against bit-rot.

It's no criticism of Python itself, that's just the way it is. I'm not 
entirely sure why. Perhaps Perl has some critical mass of of module 
contributors and maintainers willing to do less glamorous work. I wish I 
knew.

Regards,
Derek.



From Chris.Barker at noaa.gov  Tue Jun 25 12:19:08 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Tue, 25 Jun 2002 09:19:08 -0700
Subject: Suggestions for good programming practices?
References: <mailman.1024951532.5616.python-list@python.org> <af8n9i$3at$1@panix1.panix.com>
Message-ID: <3D18979A.D40B97BB@noaa.gov>

Aahz wrote:
> Depends on what you're testing for.

Exactly. I hardly ever use None to mean false, I tend to use it to mean,
well, "none", or nothing there, as it a value that is not defined yet or
doesn't exist.

> object model is that any "empty" object tests as false: None, 0, "", (),
> [], and {}.

That's part of the problem for me, when I am using numbers, 0 (or 0.0)
is as often as not a perfectly valid number, and should not be
interpreted as false. However as someone mentioned: 

if x is None

is preferabale to 

if x == None

What you want to know is whether x is bound to None so that's what you
should check for.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From debl2nononospammywhammy at bellatlantic.net  Sat Jun  1 18:22:13 2002
From: debl2nononospammywhammy at bellatlantic.net (David Lees)
Date: Sat, 01 Jun 2002 22:22:13 GMT
Subject: OverflowError: math range error ???
References: <ad8bkk$va3sb$1@ID-146704.news.dfncis.de> <ad8jv5$vb9es$1@ID-146704.news.dfncis.de>
Message-ID: <3CF94917.BD8CEC43@bellatlantic.net>

You are taking the log of zero, which is why you are getting an overflow
error.

david lees

Shagshag13 wrote:
> 
> So it's ok, i thing i should have work with real :o))))))))))
> 
> >>> math.log(15996 / 25859)
> Traceback (most recent call last):
>   File "<pyshell#7>", line 1, in ?
>     math.log(15996 / 25859)
> OverflowError: math range error
> >>> math.log(15996.0 / 25859.0)
> -0.48032001185119749
> 
> "Shagshag13" <shagshag13 at yahoo.fr> a ?crit dans le message de news:
> ad8bkk$va3sb$1 at ID-146704.news.dfncis.de...
> >
> > What does it mean "OverflowError: math range error" ???
> >
> > I get this:
> >
> > Traceback (most recent call last):
> >   File "./build_index.py", line 91, in ?
> >     ii.updateWeight()
> >   File "./invertedindex.py", line 109, in updateWeight
> >     idf = ( math.log( (N + 1) / (df + 1) ) / math.log(2) + 1 )
> > OverflowError: math range error
> >
> > (ps: i am on Solaris, if this helps)
> >
> > thanks in advance,
> >
> > s13.
> >
> >
> >


From gorny0 at zonnet.nl  Sat Jun 15 13:07:03 2002
From: gorny0 at zonnet.nl (Gorny)
Date: Sat, 15 Jun 2002 19:07:03 +0200
Subject: Args not set in extension class
References: <aefcav$8ef$1@reader05.wxs.nl>
Message-ID: <aefs7n$4cc$1@reader09.wxs.nl>

"Gorny" <gorny0 at zonnet.nl> schreef in bericht
news:aefcav$8ef$1 at reader05.wxs.nl...
> Hi,
>
> I'm having some difficulty writing an extension class. The Python
> core doesn't seem to set the args argument from the new_pkt_object
> function! It's allways set to NULL so the code beneath will allways
> end in a Segmentation Fault when importing this class from my libraries...

I've already dealt with the problem... Just create a function like:

static PyObject *
wrap_pkt_object(PyObject *self, PyObject *args)
{
        PyObject * return_object;
        return_object = new_pkt_object(args);
        return return_object;
}

But it would be very nice if someone can explain to me, why this
is happening and it isn't possible to *directly* call new_pkt_object even
if you prototype the latter function with two arguments like
wrap_pkt_object!

Gorny

--
"Spread the sources"
http://gorny.cjb.net





From eppstein at ics.uci.edu  Fri Jun 14 14:30:00 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Fri, 14 Jun 2002 11:30:00 -0700
Subject: I feel stoopid ... this code is to verbose
References: <3D0A1791.3060307@mxm.dk>
Message-ID: <eppstein-82BE2F.11300014062002@news.service.uci.edu>

In article <3D0A1791.3060307 at mxm.dk>, Max M <maxm at mxm.dk> wrote:

> I just cannot seem to find the nice 3-liner I expected from the 
> beginning. Has anybody got a better solution ? I thought somebody might 
> find it a fun exercise. Well I have...

My attempt:

def stringSlicer(string, chunkSize = 3, sep='.'):
   x = (len(string)-1)%chunkSize + 1
   C = [string[max(0,chunkSize*(i-1)+x):chunkSize*i+x] \
        for i in range((len(string)+chunkSize-1)//chunkSize)]
   return sep.join(C)

I've split it into four lines for better readability -- you could get a 
one-liner (without lambda or and-or tricks) by expanding the variables
(imagine the following all on one line):

def stringSlicer(string, chunkSize = 3):
   return sep.join([string[max(0,chunkSize*(i-1) + \
                             (len(string)-1)%chunkSize+1) : \
                           chunkSize*i+(len(string)-1)%chunkSize+1] \
                 for i in range((len(string)+chunkSize-1)//chunkSize)])

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From aahz at pythoncraft.com  Thu Jun 13 13:07:50 2002
From: aahz at pythoncraft.com (Aahz)
Date: 13 Jun 2002 13:07:50 -0400
Subject: why not "'in' in 'in'"?
References: <mailman.1023978378.11084.python-list@python.org> <aeadtt06rq@drn.newsguy.com>
Message-ID: <aeajh6$ij9$1@panix1.panix.com>

In article <aeadtt06rq at drn.newsguy.com>,
Grant Griffin  <not.this at seebelow.org> wrote:
>In article <mailman.1023978378.11084.python-list at python.org>, "Mark says...
>>
>>def is_in(search, search_for):
>>  """Return bool indicating whether search_for is in search."""
>>  return search.find(search_for) >= 0
>
>That's not a bad suggestion, Mark, but if I did that I guess I would
>be forever wondering if I was testing whether the first one was in the
>second or the second one was in the first.  Again, it doesn't read
>nearly as well as:
>
>   if x in subject:
>       ...
>
>which leaves no doubt.

class myStr(str):
    def contains(self, other):
        if self.find(other) >= 0:
            return True
        else:
            return False
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From max at alcyone.com  Wed Jun 26 14:23:20 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 26 Jun 2002 11:23:20 -0700
Subject: question od default args
References: <o3vjhuguen6gfm8vnq7rbk5volmi80shl5@4ax.com>
Message-ID: <3D1A0698.671F5FB1@alcyone.com>

"Gon?alo Rodrigues" wrote:

> When I want default args I usually do
> 
> def (arg = None):
>     ...
> 
> But what if I want to make None a reasonable argument too? That is, I
> want to know if the user has in fact passed an argument, and if not to
> do something special, with None (or whatever object) being a
> reasonable
> arg to be passed.

But then the question is, what would it makes sense for arg to have as a
value when the function is invoked without an argument?  It has to be
_something_.  If there are no "out-of-band" values that you can use to
signal "this default argument was not specified," then you're not going
to be able to distinguish the two.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ See the son in your bad day / Smell the flowers in the valley
\__/ Chante Moore
    Bosskey.net: Aliens vs. Predator 2 / http://www.bosskey.net/avp2/
 A personal guide to Aliens vs. Predator 2.


From sayles at cpsc.ucalgary.ca  Thu Jun  6 15:17:59 2002
From: sayles at cpsc.ucalgary.ca (Maxwell Sayles)
Date: Thu, 06 Jun 2002 13:17:59 -0600
Subject: creating an instance of a C extension type from another C extension type 
 (HOW?)
Message-ID: <3CFFB567.435254@cpsc.ucalgary.ca>

I have two C extension modules that would do the equivalent of the
following python classes:

class A:
  def __init__ (self, value):
    self.value = value

class B:
  def __init__ (self, value):
    self.value = value

  def makeA (self):
    return A (self.value)


now i've written class A and B into python C extensions.  I want class B
to create and return an instance of class A in the function makeA.  the
.so file for class A has a C function

PyObject* A_init_c (int value)
{
  A_struct* self = PyObject_New (A_struct, A_type);
  self->value = value;
  return (PyObject*)self;
}

this function is also called by the C function for A.__init__  :

PyObject* A_init (PyObject* self, PyObject* args)
{
  int value;
  if (!PyArg_ParseTuple ("i", &value)) return NULL;
  return A_init_c (value);
}

this works... but the function for B.makeA()  returns an object that
causes a core dump when i use the 'type' function with the object as a
parameter

PyObject* B_makeA (B_struct* self, PyObject* args)
{
  if (!PyArg_ParseTuple ("")) return NULL;
  return A_init_c (self->value);
}

now when i do

b = B(5)
a = b.makeA()
print type(a)

it causes a core dump.   is there anything wrong with my logic?  code?
also.. this isn't the real code from my program.. i just wrote it out
here in the window, so hopefully there aren't any syntactic errors

hope someone can help...  could you please reply to the newsgroup and to
my e-mail, as i don't check the newsgroup too often.

thanks!
max



From amuys at shortech.com.au  Thu Jun 27 20:29:00 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 27 Jun 2002 17:29:00 -0700
Subject: Help with an algorithm wanted
References: <rowen-CE3C0F.13020226062002@nntp1.u.washington.edu>
Message-ID: <7934d084.0206271628.227c8605@posting.google.com>

"Russell E. Owen" <rowen at cesmail.net> wrote in message news:<rowen-CE3C0F.13020226062002 at nntp1.u.washington.edu>...
> I'd like some help solving a problem in Python.
> 
> The basic idea is I have an item of data that can be represented in any 
> of four ways: A,B,C and D. I also have conversion functions between 
> certain pairs of representations, such that there's always a path 
> between any two data representations. In this case suppose I know A<->C, 
> B<->C and C<->D, or graphically:
> 
> A <--+
>      +--> C <---> D
> B <--+
> 
> I'd like to create objects that have one of these items (A,B,C,D) as a 
> given, then be able to query any such object for any data representation.
> 
> For instance, an object with A as the given would return D by solving 
> A->C->D.
> 
> What I'm trying to figure out is a simple and reasonably elegant way to 
> code such objects.
> 

So effectively you have a conversion graph, where each vertex
represents a known representation, and each edge represents a
conversion (presumably with a conversion function as an annotation to
the edge).

What you are asking for appears to be an algorthm to find the shortest
path between to representations in the graph.  Fortunately for you,
Dijkstra provided a solution to this problem a number of years ago,
and implementing it is a standard algorthms assignment (I do hope
you're not asking for help with a homework assignment?).

You'll find an implementation in the Python Cookbook @
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/119466

The recipie was submitted by David Eppstein.  It dosn't look
optimised, but that's ok because it is legible (much more important)
and the algorithm is efficient.

Of signifigant interest to you is that it represents the graph as a
dictionary of edges, and as you will want to annotate the edges with
conversion functions, this is a good thing.

Also note that Dijkstra's algorthm uses weighted edges, so you get to
consider differing conversion costs for free.

Andrae Muys


From chris.gonnerman at newcenturycomputers.net  Sun Jun  2 17:40:09 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sun, 2 Jun 2002 16:40:09 -0500
Subject: List
References: <20020602205610.31575.qmail@linuxmail.org>
Message-ID: <009b01c20a7e$12a5ad80$0101010a@local>

----- Original Message ----- 
From: "Gold Fish" <occeanlinux at linuxmail.org>

> What i would achieved is that I got the list of elements 
> say that 10 elements in the list, i want to seperate 
> this list into 2,or 3 sublists depends on user input the 
> number. Then these subject contain the number of 
> elements say 2, or 3 depend on user define. How can i do 
> this. For example. 
>
> BigList = [e1,e2,e3,e4,e5,e6,e7]
>
> this will be divided into 4 small sublists which is 
> sublist1,sublist2,sublist3,sublist4.These sublists will 
> have maximum of 2 elements inside therefore, sublist1 = 
> [e1,e2], sublist2=[e3,e4],sublist3=[e5,e6] and sublist7=
> [e7]
>
> I would appreciate that you could help me this problem. 
> I spend so much time to think about it but it's look 
> like i stuck in the maze when trying to divide biglist 
> into small list and small list to nano list. 

Given a list of unknown size, subdividing it into sublists
could go like this:

    Lists = []
    N = 2

    for i in range(len(BigList)/N):
        Lists.append(BigList[i*N:(i+1)*N])

so that given your BigList:

    BigList = [e1,e2,e3,e4,e5,e6,e7]

the result is:

    [['e1', 'e2'], ['e3', 'e4'], ['e5', 'e6']]

Note that e7 is missed; if the list you are subdividing 
does not divide evenly by N, an adjustment must be made:

    Lists = []
    N = 2

    adj = 0

    if len(BigList)%N:
        adj = 1

    for i in range(len(BigList)/N+adj):
        Lists.append(BigList[i*N:(i+1)*N])

(If anyone knows a more elegant solution please post it,
as I am getting headaches trying to think of one.)

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net






From logiplexsoftware at earthlink.net  Mon Jun 24 14:01:12 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Mon, 24 Jun 2002 11:01:12 -0700
Subject: wxPython question: window closing event
In-Reply-To: <3D175AD7.9CB28305@athensgroup.com>
References: <3D175AD7.9CB28305@athensgroup.com>
Message-ID: <20020624110112.3b0e7436.logiplexsoftware@earthlink.net>

On Mon, 24 Jun 2002 17:45:59 GMT
Mike Christie wrote:

> I have a specific and a general question about wxPython.  I have worked with
> GUIs and event-handlers before, but am new to both wxPython and Python
itself.
> 
> The specific question is: how do I attach code to the window closing event
> that occurs when (in MS Windows systems) you click the X in the top right
> corner of the screen?  It appears I need to place a "self.close(true)" there,
> since exiting via an exit menu (which does call "self.close(true)") shuts
> things down politely, but exiting via the X leaves the Python interpreter
> still executing my script, and forces me to close Python to shut it down.

from wxPython.wx import *

class MyFrame(wxFrame):
    def __init__(self):
        wxFrame.__init__(self, None, -1, "A Frame")
        EVT_CLOSE(self, self.OnClose)

    def OnClose(self, event):
        self.Destroy()
        print "Destroyed!"
        
app = wxPySimpleApp()
frame = MyFrame()
frame.Show(true)
app.MainLoop()


> The general question is: what's the best resource for finding this sort of
> thing out?  I've been working through the tutorial material at
> http://wiki.wxpython.org/index.cgi/Getting_20Started which has been very
> helpful indeed.  I've also made some use of the online documentation at
> http://wxpython.org/onlinedocs.php which is also great.  However, for a
> question like this, it seems like the right resource is something like a
> categorization of events, or a reference guide for event-handling, or
> something like that.   The online docs do have an "Event Handling Overview",
> but I couldn't find this information there.  Is there a book or online
> reference that I could use as a resource for this kind of question?

The best quick-reference is probably the demo that comes with wxPython (you can
download it from the wxPython website).

Also, questions like this are probably better answered on the wxPython mailing
list (which you can also subscribe to via the wxPython website).

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From skip at pobox.com  Tue Jun 11 14:42:58 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 11 Jun 2002 13:42:58 -0500
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
In-Reply-To: <20020611111101.420d7b9a.logiplexsoftware@earthlink.net>
References: <Xns9229EE722467Fgustafl@195.100.94.182>
        <mailman.1023746539.13373.python-list@python.org>
        <Xns922AD3C74280gustafl@195.100.94.182>
        <15622.26.84312.207270@12-248-41-177.client.attbi.com>
        <20020611111101.420d7b9a.logiplexsoftware@earthlink.net>
Message-ID: <15622.17586.341577.425403@12-248-41-177.client.attbi.com>

    Cliff> I do think this is a bit clearer, however I know you meant to write

    Cliff> except TypeError:
    Cliff>     s, e = r, r

Right you are...

Skip




From aahz at pythoncraft.com  Thu Jun 20 00:45:49 2002
From: aahz at pythoncraft.com (Aahz)
Date: 20 Jun 2002 00:45:49 -0400
Subject: Variable Interpolation - status of PEP 215
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com>
Message-ID: <aermlt$in5$1@panix1.panix.com>

In article <ouit4e63fr.fsf at andlx-anamika.cisco.com>,
Ganesan R  <rganesan at myrealbox.com> wrote:
>
>I am a relative newbie to Python. I have a reasonably good perl background
>(don't worry I am not about to start a language flame war :-), and one of
>the things I really miss is variable interpolation in strings. I searched
>the net and found PEP 215 (http://www.python.org/peps/pep-0215.html). This
>seems to be exactly what I need. The proposal seems to be quite old, so I'd
>like to know if it's still in consideration or has it been dropped
>completely?

Peter already told you what you really wanted to know, but for the
benefit of other people, PEP 215 is dead and PEP 292 is currently the
subject of a LARGE thread on python-dev.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From i.linkweiler at gmx.de  Sat Jun  8 11:43:48 2002
From: i.linkweiler at gmx.de (Ingo Linkweiler)
Date: Sat, 08 Jun 2002 17:43:48 +0200
Subject: Comparison: HTTP with Java, Perl or Python
References: <3D009911.9000904@gmx.de> <h99M8.11172$p56.3361428@newsb.telia.net> <3D0127D2.2060604@gmx.de> <B9273737.4982%js_nntp@nospam_blackrocksoftware.com>
Message-ID: <3D022634.9050304@gmx.de>

Thanks a lot for your versions.
Which of ALL solutions is the shortest, and which the best ?
Do you know good ways to measure the size? (LOC, irredundant or packed 
bytes....)
But even without this:

The Python version is quite short, even after adding exception handling. 
TCL seems to be same.
The Perl version seems to be the shortest, but can you read it without 
knowing Perl?

Can you estimate the time you needed for writing it?
Did you have any difficulties?
My Python version needed 2 Minutes, with Java I needed 20, because I did 
not unterstand the f******* manual.




PS:
This was my own Java solution:

package meintest;
import java.lang.Exception;
import java.net.*;
import java.io.*;
public class getURL {
  public static void main(String[] args) {
    try {
      URL url = new URL("http://www.uni-dortmund.de");
      InputStream in = url.openStream();
      for (;;)
      {
          int data = in.read();
          if (data == -1)
              break;
          else
              System.out.print ( (char) data);
      }
    }
    catch (Exception e) { }
  }
}



From kalle at lysator.liu.se  Sat Jun 29 13:24:12 2002
From: kalle at lysator.liu.se (Kalle Svensson)
Date: Sat, 29 Jun 2002 19:24:12 +0200
Subject: Newbie that don't understand
In-Reply-To: <AIlT8.13914$p56.4478653@newsb.telia.net>
References: <AIlT8.13914$p56.4478653@newsb.telia.net>
Message-ID: <20020629172412.GC507@i92.ryd.student.liu.se>

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

[Martin St?hl]
> "[...]The absolute values of 23 and 23 are equal"
> 
> how can 23 and -23 be equal?

Not 23 and -23, the absolute values of 23 and -23.  This is a
matemathical term.  The absolute value of x, where x is an integer, is
x if x >= 0, -x otherwise.

<swedish>
Jag kan ocks? passa p? att rekommendera tutor-listan
(http://mail.python.org/mailman/listinfo/tutor).  Det ?r nog ett
b?ttre st?lle f?r nyb?rjarfr?gor, med mer personlig atmosf?r och
betydligt f?rre meddelanden per dag ?n c.l.py.
</swedish>

Peace,
  Kalle
- -- 
Kalle Svensson, http://www.juckapan.org/~kalle/
Student, root and saint in the Church of Emacs.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)
Comment: Processed by Mailcrypt 3.5.6 <http://mailcrypt.sourceforge.net/>

iD8DBQE9He03dNeA1787sd0RAq//AKCi1tFcdTFhr0D54Jh2sV+JXxQ90ACeNTbM
j7pNfCp/O8OzdCOO3j1/aGE=
=Hk45
-----END PGP SIGNATURE-----




From bokr at oz.net  Fri Jun 14 19:57:36 2002
From: bokr at oz.net (Bengt Richter)
Date: 14 Jun 2002 23:57:36 GMT
Subject: filter2
References: <mailman.1023975618.6474.python-list@python.org> <lkptyvmdfz.fsf@pc150.maths.bris.ac.uk> <mailman.1023984463.10662.python-list@python.org> <lkhek7m9c8.fsf@pc150.maths.bris.ac.uk> <slrnagii89.55q.huaiyu@gauss.almadan.ibm.com> <lkd6uumci2.fsf@pc150.maths.bris.ac.uk>
Message-ID: <aedvtg$4ts$0@216.39.172.122>

On Fri, 14 Jun 2002 09:33:10 GMT, Michael Hudson <mwh at python.net> wrote:

>huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu) writes:
>
>> It's not that gross, but it only gives 14% speedup, though.
>
>Oh yeah.  What you've done is kind of obvious...
>
>> def filter2a(test, list):
>>     l1 = []
>>     l2 = []
>>     funcs = [l1.append, l2.append]
>>     map(lambda x: funcs[not test(x)](x), list)
>>     return l1, l2
>> 
>> BTW, what's the reason that using a list for funcs is faster than using a
>> tuple (as observed under Python 2.2.1 on Linux)?
>
>Look at BINARY_SUBSCR in ceval.c... list[int] is special-cased.
>
Try this one:

 >>> def filter3(test, alist): # not to use keyword
 ...     l2=[]
 ...     return [x for x in alist if test(x) or l2.append(x)], l2
 ...
 >>> print filter3(lambda x: x%3==0,range(16))
 ([0, 3, 6, 9, 12, 15], [1, 2, 4, 5, 7, 8, 10, 11, 13, 14])

Regards,
Bengt Richter


From marklists at mceahern.com  Thu Jun 13 09:43:52 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 13 Jun 2002 08:43:52 -0500
Subject: Poll: Readable languages
In-Reply-To: <Pine.LNX.4.44.0206130853001.20311-100000@adams.patriot.net>
Message-ID: <NCBBLFCOHHDIKCAFGCFBCEGNLEAA.marklists@mceahern.com>

> Before I'm accused of missing the point, I AM NOT SUGGESTING THAT PEOPLE 
> MUST PROGRAM IN ASSEMBLER. "THAT" would be ludicrous! I am suggesting 
> however that assembler also qualifies for the specific questions you 
> asked!

For an interesting take on programming in Assembly:

  http://grc.com/smgassembly.htm

Cheers,

// mark
-





From g0riaman at cdf.toronto.edu  Tue Jun 11 10:31:54 2002
From: g0riaman at cdf.toronto.edu (so very tired)
Date: Tue, 11 Jun 2002 14:31:54 GMT
Subject: Removing ^M
In-Reply-To: <mailman.1023546912.19781.python-list@python.org>
References: <mailman.1023546912.19781.python-list@python.org>
Message-ID: <Pine.GSO.4.30.0206111023300.6860-100000@eddie.cdf>

Easiest way to do this is with sed.
cat file.html | sed 's/^M//g' > newfile.html

the trick is you have to enter ^M as the proper control character.
You can do this at the command prompt by holding down CTRL+SHIFT and then
hitting V and M in sequence.
This should work with other control characters as well.
Hope this helps.
mRiaz

On Sun, 9 Jun 2002, Michael Hall wrote:

>
> I am trying remove ^M characters (some kind of newline character) from an
> HTML file. I've tried all sorts of string.replace and sed possibilities
> but the
> things just won't go away. Does anyone have a way of removing such
> characters?
>
> TIA
>
> Mick
>
>
>
>
>



From whisper at oz.net  Tue Jun 11 02:48:00 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 10 Jun 2002 23:48:00 -0700
Subject: Changing the path on Win2k/XP
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E20192F1A3F6@admin56.narex.com>
Message-ID: <GCEDKONBLEFPPADDJCOEOEANDIAA.whisper@oz.net>

You should be able to modify the path on Win2k/XP by modifying the
associated registry key(s). There is one for the "system" environment one
for the "per user" environment. This won't take effect for the current
process though, but only for new ones (at least comd.exe processes).

Take a look in (changes depending on user logged in):
	HKEY_CURRENT_USER\Environment
and:
	HKEY_LOCAL_MACHINE\SYSTEM\Current Control Set\Control\Session
Manager\Environment	HKEY_LOCAL_MACHINE\SYSTEM\ControlSet001\Control\Session
Manager\Environment	HKEY_LOCAL_MACHINE\SYSTEM\ControlSet002\Control\Session
Manager\Environment

Of the 3 in HKLM\SYSTEM, I think only "Current Control Set" is relevent.

You can modify these with the _winreg package ("import _winreg"). BE VERY
CAUTIOUS! Incorrect modifications of your registry can render your machine
unusable and possibly requiring an OS reinstall. There is a procedure for
creating a backup copy of your registry and a backup disk and I suggest you
do both until you are comfortable with what you're doing.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Bjorn Pettersen
> Sent: Monday, June 10, 2002 21:07
> To: python-list at python.org
> Subject: Changing the path on Win2k/XP
>
>
> Is there a way to change the path on Win2k/XP through Python? (I'm
> trying to write a .dll deployment script for our production department,
> and I'd like to give them a "one button" install...)
>
> -- bjorn
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From kragen at pobox.com  Mon Jun 10 18:11:18 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 18:11:18 -0400
Subject: prototyping good OOdesign in Python?
References: <XFMail.20020601131915.shalehperry@attbi.com> <Pine.LNX.4.33.0206030725420.5660-100000@rnd.onego.ru> <mailman.1023310034.25601.python-list@python.org>
Message-ID: <83u1oaokc9.fsf@panacea.canonical.org>

Geoff Gerrietts <geoff at gerrietts.net> writes:
> With UML, I generally find that it has TOO MANY features. If I use
> Visio, or Rose, or even the lovely pyut announced a few months ago on
> this very list, they'll ask for information about attribute and method
> visibility, data types, etc.
> 
> Since Python doesn't care about most of this stuff, modelling it is
> only useful if you expect to port the code later. And I prefer not to
> plan for an unfortunate event. ;)

Well, maybe I'm polluted by my previous experience with other
sort-of-OO languages like C++, but I still include method visibility
and data types in my Python designs.  

Attributes are never visible (although that will probably change when
I start using 2.2, because visible attributes will cease to cost
flexibility), most methods are visible, and occasionally I'll add some
private methods, usually named with a leading underscore.

WRT data typing, I usually try to know what kind of thing is in
particular attributes of my classes.  The answer might be "a number",
"a sequence", "a sequence of numbers", "a mapping from strings to
sequences of (string, number) tuples", but I try to keep a clear idea
in mind of what it is.

Python makes it very easy to change my mind about all of these things.



From david.abrahams at rcn.com  Mon Jun 17 10:18:12 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Mon, 17 Jun 2002 10:18:12 -0400
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de> <aeftmj$m29$1@bob.news.rcn.net> <3d0c6d67@nntp.server.uni-frankfurt.de> <3D0CE698.1050806@SPAMnwinternet.com> <aeiu3a$4ff$1@bob.news.rcn.net> <3D0D3223.5070500@SPAMnwinternet.com>
Message-ID: <aekrd1$4ao$1@bob.news.rcn.net>

"Jim" <jbublitzNO at SPAMnwinternet.com> wrote in message
news:3D0D3223.5070500 at SPAMnwinternet.com...
> David Abrahams wrote:

> > I'd really like to know what you find preferable about sip -- aside your
own
> > familiarity with it, of course, which for me usually counts more than
any
> > other single factor ;-). I'm actively developing v2 of Boost.Python and
> > would really like to make it the best tool possible for this job, so it
> > would be useful to know more.
>
> I inherited sip rather than choosing it after 'careful evaluation',
> and certainly familiarity is a big part of it. In fact I've only
> given other alternatives a cursory look, so what I like about sip
> may already be available in Boost or other packages.
>
> The biggest plus for sip is that the 'IDL' is the h file syntax
> with a few subtractions (variable names, forward declarations,
> inline code, but inline methods stay), so that most of the
> description file creation is easy to automate and still not
> that hard to do manually for a small number of files. sip
> supports most C++ h file syntax.


Hmm, I wonder about that. The C++ type system is pretty sophisticated (to
the point of being able to represent entire compile-time programs!) and all
of the parser-based wrapping systems I've seen eventually fail to handle
some constructs I use, especially if I'm wrapping template-based code.
That's part of the reason I chose to use the compiler itself in
Boost.Python. The compiler (in theory) has no problem with C++ syntax ;-)

On the other hand, I realize there's a large category of simple C++
interfaces which can be parsed will less-than-complete machinery.

> The other features that are helpful on large projects are the
> built-in support for versioning

How so? Can you say more about this?

> and embedded docs (both also fairly
> easy to automate),

Do you mean docstring generation?

> as well as support for 'platforms' and
> 'features'. So for example PyQt (not my project) can support Qt
> from 1.4.2 through 3.0.4 with different mixes of features (with
> SQL or not, QTable or not) across Windows, Linux/Unix/BSD, and
> (soon) Mac, with a single set of source files, and still be
> manageable by one person.

Boost.Python works on a similar range of platforms.
I'm not sure if this configuration stuff is an appropriate job for a
wrapping system to provide specific support for; what's the argument for
integrating it with the wrapping system?

> sip is flexible enough that you can handwrite code to "Pythonize"
> C++ code (return a Python list or dict, or for example,
> 'void rgb (int&, int&, int&)' returns (int, int, int) and takes
> no arguments in PyKDE), or create type substitutions/conversions
> ('mapped types') once and re-use them across a project. You can
> do C++ with anything of course, but sip has it well-integrated
> (manageable) and has support code to make it easier (arg unpacking,
> object conversions, etc). IIRC, Boost might not need handwritten
> C++ code quite as much.

Not sure about this; we'd need to compare specific functionality.

> I suppose those are a series of small advantages, but the end
> result seems to me that you write less code

Are you counting the IDL file? That's code, IMO.

> and you do it in
> fewer but well-known locations. It seems to add up to a big
> difference on a large project.
>
> Boost (IIRC) also has a big advantage in being able to bind much
> more selectively.

Can you explain what you mean by that?

> sip is easiest when binding everything in a lib
> and probably takes a little more effort to be selective - you
> can do it, but it isn't as 'mindless' (ie automatable) as most
> sip file creation is. And of course Boost is actually
> documented - no small thing.

Heh, yeah. Unfortunately, documentation is often harder than coding. Even
the current official release of Boost.Python is not very well-documented.
We're working hard to correct that for v2, which is in pre-release.

> How would you see Boost performing on a very large project -
> say 10 libs, 20 classes/lib, 10 exposed methods/class (2000
> methods overall)?

See http://cctbx.sourceforge.net/
Also
http://mambo.peabody.jhu.edu/~karlmac/publications/gamera_python_2001/gamera
.html

Of course, these don't prove anything, since wrapping is a nontrivial job
and people quickly get invested in whatever is working for them. You *can*
wrap a large project with Boost.Python, but whether it's easier than with
sip...?

> I haven't looked at Boost in a while -
> could you incorporate conditional compilation (eg versions,
> features)

It's C++; you can use #ifdef in just the same way that the rest of the C++
code surely does.

> and embed docs easily?

Docstring support is on the todo list for v2. It will be a matter of passing
a string literal to the def() function which wraps a function pointer:

class_<Foo>()
   .def_init(args<int>(), "Creates a Foo with the specified number of Bars")
   .def("get_name", &Foo::get_name, "get the Foo's name")
   .def("set_name", &Foo::set_name, "set the Foo's name")
   ;

> Could you automate code generation to some extent?

Not sure what you have in mind here. In a sense, the metaprogramming
techniques used by Boost.Python already *do* code generation.

> The automation for generating
> PyKDE is pretty crude at the moment, and in all fairness
> isn't generally available, so this example would be a lot
> of work using sip as it comes out of the box. Is Boost
> really intended for projects of this size (there probably
> aren't many), or do you have different goals for it?

Yes, Boost.Python is intended for large projects. One of the largest
projects I've ever heard is using it at Lawrence Livermore Labs, for
example.

-Dave





From larooy at xtar.co.nz  Wed Jun 26 20:18:11 2002
From: larooy at xtar.co.nz (John La Rooy)
Date: Thu, 27 Jun 2002 12:18:11 +1200
Subject: Problems using samples
References: <337619fa.0206260913.41e08940@posting.google.com>
	<Xns9239C44CFB16Emhueningzedatfuberli@130.133.1.4>
Message-ID: <20020627121811.40f616fd.larooy@xtar.co.nz>

On 26 Jun 2002 17:18:03 GMT
Matthias Huening <mhuening at zedat.fu-berlin.de> wrote:

> gusraan at terra.com.br (Raphael Ribeiro) wrote in 
> news:337619fa.0206260913.41e08940 at posting.google.com:
> 
> > this is my code, it says there's an error in line 6, but i can't find
> > what **** error is that, can anyone help me?
> 
> your line 6 is:
> > 6 if x = y:
> 
> should be:
> > 6 if x == y:
> 
> Matthias

Also lines 8 and 9 are indented too far

John


From emile at fenx.com  Tue Jun 25 09:27:41 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 25 Jun 2002 13:27:41 GMT
Subject: Suggestions for good programming practices?
References: <WEuR8.2074$5l4.75657@ozemail.com.au> <af8m9g$s5m$1@panix1.panix.com>
Message-ID: <hd_R8.306964$352.31513@sccrnsc02>

Aahz
> In Python 2.1.2 or higher, fire up the interactive interpreter and
type
> "import this".

Except ActiveState's

F:\Python22as>python
Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import this
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: No module named this

--

Emile van Sebille
emile at fenx.com

---------



From c.tee at verizon.net  Sun Jun 23 13:57:51 2002
From: c.tee at verizon.net (Curtis Taylor)
Date: Sun, 23 Jun 2002 17:57:51 GMT
Subject: Iterating through two lists
References: <mailman.1022234790.26965.python-list@python.org> <7934d084.0206031700.377c691e@posting.google.com>
Message-ID: <c.tee-FF2C9D.10574923062002@news.bellatlantic.net>

In article <7934d084.0206031700.377c691e at posting.google.com>,
 amuys at shortech.com.au (Andrae Muys) wrote:

> "Denis S. Otkidach" <ods at fep.ru> wrote in message 
> news:<mailman.1022234790.26965.python-list at python.org>...
> > On Fri, 24 May 2002, jb wrote:
> > 
> > j> I have two lists, x and y with the property len(x) = len(y).
> > j>
> > j> I should like to achive this (x is a list of class instances)
> > j>
> > j>   for (a,b) in (x,y): a.f(b)
> > j>
> > j> Is there a fancy way of doing this or have I to introduce an
> > j> auxillary
> > j> counter (that is very easy but maybe not very "lispy", that
> > j> is
> > j> "python-like").
> > 
> > If lists are not very lange try this:
> > 
> > for a, b in zip(x, y): a.f(b)
> 
> If the lists are large then you probably need something along the lines of 
> xzip.
> 
> def xzip(*args):
>     iters = [iter(a) for a in args]
>     while 1:
>         yield tuple([i.next() for i in iters])
> 
> Andrae Muys


Hi Andrae,

I've got a similar situation I'm dealing with, in that I have 2 lists of 
dictionaries, each of which is of equal length. What's different 
however, is that I need to do an equality check for the values of 2 of 
the keys (both of which exist in each dictionary), then set the value of 
the 2nd dictionary's key to the value of that of the first's, e.g.:

if dict2[key1] == dict1[key1]:
   dict2[key2] = dict1[key2]

I'm concerned with overhead, as the 2 lists could potentially be very 
large. Can your xzip function be extended to do such a thing?

Thanks,

Curtis


From keilj_33 at yahoo.com  Wed Jun 19 14:26:46 2002
From: keilj_33 at yahoo.com (Jeffrey Keil)
Date: 19 Jun 2002 11:26:46 -0700
Subject: Sending email from a Python script
Message-ID: <67f6bbe5.0206191026.72e14e8e@posting.google.com>

Greetings:

Is it possible to send an email from a Python script?

Thanks,
Jeff Keil


From gorny0 at zonnet.nl  Sat Jun 15 08:21:57 2002
From: gorny0 at zonnet.nl (Gorny)
Date: Sat, 15 Jun 2002 14:21:57 +0200
Subject: Args not set in extension class
Message-ID: <aefcav$8ef$1@reader05.wxs.nl>

Hi,

I'm having some difficulty writing an extension class. The Python
core doesn't seem to set the args argument from the new_pkt_object
function! It's allways set to NULL so the code beneath will allways
end in a Segmentation Fault when importing this class from my libraries...

pkt_object *
new_pkt_object(PyObject *args) {
        pkt_object *self;
        char * prot;

        if (!PyArg_ParseTuple(args, "s:new_pkt_object", &prot))
                return NULL;

        self = PyObject_New(pkt_object, &pkt_object_type);

[.....]


Gorny

--
"Spread the sources"
http://gorny.cjb.net






From peter at engcorp.com  Sun Jun 23 16:57:06 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 23 Jun 2002 16:57:06 -0400
Subject: mail-problem over t-online
References: <mailman.1024859369.26585.python-list@python.org>
Message-ID: <3D163622.E7733360@engcorp.com>

Klaus Reinhardt wrote:
> 
> Testing python and email I adapted this script:
> 
>         host=smtplib.SMTP('mailszrz.zrz.tu-berlin.de')
>         host.sendmail(zwi_eck(From),zwi_eck(Tp),al)
>         host.quit
> 
> This is working, when I'm directly connected to my
> old ISP TU-Berlin via ISDN. But the same is failing,
> when I'm connected at t-online via ADSL, with
> authentification required .. Providing my log and pass
> produces errors, too.
> 
> Pop3 is working in both cases.
> 
> Who can give me a hint?

Try connecting manually with TELNET first, to ensure you
have a clear channel to that server and port.  Many ISPs
will block port 25 except to their own mail server, to 
prevent you using them to send spam through other systems
and having them suffer the consequences...

-Peter


From rnd at onego.ru  Sun Jun  2 23:25:22 2002
From: rnd at onego.ru (Roman Suzi)
Date: Mon, 3 Jun 2002 07:25:22 +0400 (MSD)
Subject: prototyping good OOdesign in Python?
In-Reply-To: <EFEFIHAINIINNELLHFAIKEHPCHAA.kseehof@neuralintegrator.com>
Message-ID: <Pine.LNX.4.33.0206030722500.5660-100000@rnd.onego.ru>

On Sat, 1 Jun 2002, Ken Seehof wrote:

>> prototype certain design? Developing in C++ looks so unnecessary hard
>> after things were done in Python...
>> Or, it may be put this way: what discipline a Python programmer must
>> obey to allow it's prototype to be conveniently rewriten in C++?
>> The above thoughts aren't probably well-formed. But I hope
>> my concern is understood.
>>
>>
>> Sincerely yours, Roman Suzi
>
>I agree with what you are saying.  Here's my approach :-)
>
>0. Write a project schedule for management; something like this:
>   1. Prototype in python
>   2. Convert to C++
>1. Prototype in python
>2. Give a nice demo, and convince management to skip step 2
>3. Re-implement a few cpu intensive classes in C++

Interesting ;-)

>Since most of the code never /actually/ gets rewritten in C++,
>the need for special C++ "discipline" in the python code goes
>away.  Weeks are slashed from the schedule, and the program is
>more maintainable.
>
>Another nice advantage is that the python programmer gets to ask
>for a raise, and has excellent job security!

...because only he knows the magic powers of Python?

>There may be a few oddball examples of projects that actually
>need to end up written in C++ after prototyping in python, but
>for the most part, the most common reason to target C++ is to
>alleviate fear of the unknown.  (Hmm, is this also why people,
>(including myself), still buy from Microsoft? :-)

Fear of unknown? That is why I DO NOT buy from Microsoft ;-)

>- Ken Seehof

Sincerely yours, Roman Suzi
-- 
\_ Russia \_ Karelia \_ Petrozavodsk \_ rnd at onego.ru \_
\_ Monday, June 03, 2002 \_ Powered by Linux RedHat 7.2 \_
\_ "Bugs are Sons of Glitches!" \_





From max at alcyone.com  Mon Jun 10 04:50:20 2002
From: max at alcyone.com (Erik Max Francis)
Date: Mon, 10 Jun 2002 01:50:20 -0700
Subject: Looking for a list subclassing example...
References: <adi4dv$118tko$1@ID-146704.news.dfncis.de> <83bsaj4kgn.fsf@panacea.canonical.org>
Message-ID: <3D04684C.BA67E64F@alcyone.com>

Kragen Sitaker wrote:

> Why do you think it's a good idea to subclass 'list'?
> 
> I don't have one to show, but I think it's generally not a good idea.

Why do you think it was made possible in 2.2, then?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Who'd ever think it / Such a squalid little ending
\__/ The American and Florence, _Chess_
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From peter at engcorp.com  Fri Jun 21 21:43:07 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 21 Jun 2002 21:43:07 -0400
Subject: urllib POST question
References: <644f6688.0206200930.48110837@posting.google.com> <slrnah4dbq.r77.insyte-clp@petra.bitstream.net> <644f6688.0206210900.68dd930c@posting.google.com>
Message-ID: <3D13D62B.D7AE6554@engcorp.com>

rdack wrote:
> 
> Ben Beuchler <insyte-clp at emt-p.org> wrote in message news:<slrnah4dbq.r77.insyte-clp at petra.bitstream.net>...
> > In article <644f6688.0206200930.48110837 at posting.google.com>, rdack wrote:
> > > i am doing a POST to a server using urllib and it works.
> > > is there some way i can see exactly what is being posted?
> > > a way to get the post object out of urllib - the exact lines it is sending?
> >
> > I'm a big fan of Ethereal...
> > "Sniffing the glue that holds the internet together!"
> >
> > http://www.ethereal.com/
> 
> i am looking into it. seems overkill for wanting to see a few bytes
> that python is generating.
> No way within python to see what urllib intends to send out for the
> POST?

Howzabout this?

>>> import httplib
>>> httplib.HTTPConnection.debuglevel = 1
>>> import urllib
>>> page = urllib.urlopen('http://web.engcorp.com:8080/?test=1').read()
connect: (web.engcorp.com, 8080)
send: 'GET /?test=1 HTTP/1.0\r\n'
send: 'Host: web.engcorp.com:8080\r\n'
send: 'User-agent: Python-urllib/1.15\r\n'
send: '\r\n'
reply: 'HTTP/1.0 200 OK\r\n'
header: Server: Zope/(Zope 2.5.1 (binary release, python 2.1, linux2-x86), pytho
n 2.1.3, linux2) ZServer/1.1b1
header: Date: Sat, 22 Jun 2002 01:25:46 GMT
header: Connection: close
header: Content-Type: text/html
header: Etag:
header: Content-Length: 174
>>> len(page)
174

-python-rocks-ly yr's,
 Peter


From joe+usenet at sunstarsys.com  Sat Jun  1 12:44:01 2002
From: joe+usenet at sunstarsys.com (Joe Schaefer)
Date: 01 Jun 2002 12:44:01 -0400
Subject: Migrating from PHP to Python...
References: <xWyk7.4824$v%4.492814@news20.bellglobal.com> <l4ihfu0nhp1a9p53vgch3n3kcc25s8c00p@4ax.com> <slrnafhj0d.dp4.gerhard@lilith.my-fqdn.de> <m3znyf3x2o.fsf@mumonkan.sunstarsys.com> <slrnafhr3p.e0j.gerhard@lilith.my-fqdn.de>
Message-ID: <m3ptzb3q5q.fsf@mumonkan.sunstarsys.com>

Gerhard =?iso-8859-15?Q?H=E4ring?= <gerhard at bigfoot.de> writes:

> Sorry. I should have just PLONKed him and been done with it.

No harm, no foul.  IMO the clp.misc denizens have recently
elevated troll management to near artform, and it has been
quite a pleasure to witness the collective triage effort.  
For instance, check out the subthread in clp.misc that 
revolves around this post

Message-ID: <3CF686D1.608DE457 at NO-SPAM-at-2host.com>

That 2Host.com guy can draw blood from a stone.

Cheers.
-- 
Joe Schaefer



From gerhard.haering at gmx.de  Thu Jun 27 16:10:00 2002
From: gerhard.haering at gmx.de (=?ISO-8859-1?Q?Gerhard=20H=E4ring?=)
Date: Thu, 27 Jun 2002 22:10:00 +0200 (CEST)
Subject: self destruction :)
In-Reply-To: <f4KS8.1326$qq1.14832@castor.casema.net>
References: <f4KS8.1326$qq1.14832@castor.casema.net>
Message-ID: <20020627200752.8C85763A@gargamel.hqd-internal>

Am Thu, 27 Jun 2002 19:54:51 GMT schrieben Sie (Guyon Mor?e <gumuz at looze.net>):
> i've got a list of dynamicaly instantiated objects of a class of my own.
> how
> can i let one of the objects delete itself?
> 
> i've tried: 'del self', but somehow it didn't work. afterwards it should
> also _not_ be in the list anymore.

This might work for you:

Mark the object as closed by setting self.closed to 1. Then, periodically,
scan your objects, and delete any that have closed == 1. You cannot if any
of the objects's methods are still in the call stack. What I mean, you can
not call the cleanup function from within any of the object's methods.

I have no idea what you're trying to do, but if you want to make some sort
of simulation, either multithreading or a state machine that's run from a
main loop doing the cleanup should work.

Gerhard
--
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))





From theller at python.net  Wed Jun 26 09:08:09 2002
From: theller at python.net (Thomas Heller)
Date: Wed, 26 Jun 2002 15:08:09 +0200
Subject: Pythoniac: Thoughts on a hardware Python processor
References: <afc52t$2pj$1@sirius.dur.ac.uk> <eKiS8.43021$mF.33425@atlpnn01.usenetserver.com>
Message-ID: <afcebv$d6n8a$1@ID-59885.news.dfncis.de>

> > Dear All,
> >
> > I have seen the occasional post about creating a specific hardware 'Python
> > Processor', that could then be built and used in programmable logic or the
> > like (Valves, relays, core memory - Pythoniac ;-) - in other words, create
> > a processor that executes Python bytecode directly.  There would be
> > various different reasons for doing this (the 'neat' factor etc...)
> >
> [... Pythoniacity ...]
> >
> > So, Pythoniac: a hardware (well, FPGA(2)...) processor that directly
> > executes Python bytecode.  What do people think about the idea?  Is it
> > really feasible?  I have doubts it would be particularly quick (in the
> > same way I have my doubts about anti-gravity generators... ;-), but it
> > should be fun!
> >
> You will, of course, have to stop the development team from tinkering with
> the bytecode when moving from one major version to another.
The processor will propably programmed in VHDL, but why not program
it in Python, and translate this to VHDL. Pass this through synthesis, and
you get a byte-stream which the host CPU (if there is one) uses to initialize
the FPGA (FPGA's are reprogrammable).
So, basically, a reconfigurable Python Processor hardware in a file
on the host system should be possible.

Thomas




From dale at riverhall.NOTHANKS.co.uk  Thu Jun 20 07:09:26 2002
From: dale at riverhall.NOTHANKS.co.uk (Dale Strickland-Clark)
Date: Thu, 20 Jun 2002 12:09:26 +0100
Subject: Transparent GIFs out of Piddle?
Message-ID: <rnd3hugios9cnmhb1a5qtpv25tsfb9568n@4ax.com>

We need to produce some simple bar graphs for a web site. They need to
be GIFs with a transparent background. We are considering Piddle but
it doesn't seem to support transparency.

It may be possible to post-process the GIF with PIL but I wondered if
anyone had any suggestions on the best way to do this?

Thanks
--
Dale Strickland-Clark
Riverhall Systems Ltd


From niemeyer at conectiva.com  Sun Jun 16 15:00:24 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Sun, 16 Jun 2002 16:00:24 -0300
Subject: ConfigParser
In-Reply-To: <7647A9F4B298E74DB6793865DA67285004B354@exchange.adrembi.com>
References: <7647A9F4B298E74DB6793865DA67285004B354@exchange.adrembi.com>
Message-ID: <20020616160024.A12074@ibook.distro.conectiva>

> The problems is: when I use this file from GUI I don't want variables
> to be expanded, but when I use it from "run" program I do want it. How
> I do it?

Use the raw parameter:

parser.get("section", "option", raw=1)

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From i.linkweiler at gmx.de  Thu Jun 13 07:24:59 2002
From: i.linkweiler at gmx.de (Ingo Linkweiler)
Date: Thu, 13 Jun 2002 13:24:59 +0200
Subject: Poll: Readable languages
References: <3D05BC38.2020400@gmx.de>
Message-ID: <3D08810B.2040707@gmx.de>

Poll:
How readable is the code of this languages?

I do not want to give an exact definition of "readable", but it should 
include each of this questions with ~25%:

- Can someone who has never seen the language before understand what the 
program is doing without reference material?
- Can a developer write code that they can return to after six months 
and still understand it?
- Can you use this language for teaching basics of programming and for 
showing, how an easy algorithm works?
- What is your general optionion about his language?

1.)
-----
ABC:
ADA:
APL:
BETA:
C++:
COBOL:
DELPHI:
EIFFEL:
FORTH:
FORTRAN:
JAVA:
LISP:
MODULA:
PASCAL:
PERL:
PROLOG:
PYTHON:
SML:
SCHEME:
SMALLTALK:
TCL:
VISUAL BASIC:
other  (  *name of language * ):
-----
2.)
My most used language is:
-----

Please vote....
1 = Easy to read, like pseudocode :-)
2 = Good
3 = Average
4 = difficult
5 = nearly impossible, when not knowing the language :-(
blank = I do not know this language or have no opinion.

Please only vote, if you know at lease 2 languages.
Please reply to : i.linkweiler at gmx.net

This poll is done for my diploma paper at university of dortmund, germany.
I will not use your e-mail for any purpose.

Thanks for your help.
Ingo Linkweiler

PS: I know, that a NG is not the best place for a poll, and some people 
do not like questions like this in a NG.
But as I want to get opinions of many programmers, including users of 
NGs, I hope you can accept it.



From james.kew at btinternet.com  Thu Jun 27 17:55:37 2002
From: james.kew at btinternet.com (James Kew)
Date: Thu, 27 Jun 2002 22:55:37 +0100
Subject: Be the first on your block to register your computer.
References: <mailman.1025211444.21440.python-list@python.org>
Message-ID: <afg1ks$e6lj5$1@ID-71831.news.dfncis.de>

"David LeBlanc" <whisper at oz.net> wrote in message
news:mailman.1025211444.21440.python-list at python.org...
> Yes! exciting times are coming again as Intel, AMD and Microsoft (among
> quite a few others) are banding together to once more put ID numbers (and
> now basic encryption) onto next generation silicon.

Not that I _want_ to sound crabby, but: this has relevance to Python _how_
exactly?

James





From mgilfix at eecs.tufts.edu  Mon Jun 10 18:53:26 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Mon, 10 Jun 2002 18:53:26 -0400
Subject: Newbie OOP Question.(diff between function, module and class)
In-Reply-To: <15620.49909.679355.726113@12-248-41-177.client.attbi.com>; from skip@pobox.com on Mon, Jun 10, 2002 at 10:17:09AM -0500
References: <B9279B52.6AEB%sarmstrong13@mac.com> <15620.49909.679355.726113@12-248-41-177.client.attbi.com>
Message-ID: <20020610185326.B5683@eecs.tufts.edu>

  Just a small addition...

On Mon, Jun 10 @ 10:17, Skip Montanaro wrote:
> 
>     SA> I'm still trying to grasp this OOP concept in Python. Can someone
>     SA> tell me the difference between a function, a module, and a class?
>     SA> Since Python defines them all as objects, how do they differ outside
>     SA> of being called from inside the script or outside the script?
> 
> You posted this midday Saturday and here it's Monday morning but you still
> have no replies to this question?  I would have thought people would jump
> all over the opportunity to respond to a question like this.
> 
> Here's my take on things:
> 
>  * A function is an action to perform using a set of input arguments and
>    returning a set of output arguments.  For example:
> 
>    >>> print sin(47)
>    0.123573122745
> 
>  * A class associates a chunk of data with a set of functions which operate
>    on that data.  The first argument to each function in a class (typically
>    called a "method" in Python) is a reference to the data stored in that
>    instance of the class.  For example:
> 
>     import math
>     class Point:
>         def __init__(self, x, y):
>             self.x = x
>             self.y = y
> 
>         def polar(self):
>             return math.sqrt(self.x**2+self.y**2), math.atan2(self.y, self.x)

  A class is what is usually used to embody an object. An object
can be though of as a data structure whose methods adhere to the
"protocol" of that data structure. Python looks for some special
methods to implement behaviors, such as object creation and deletion,
addition, etc. In the OOP world, methods are operations which act on
the object and usually bring about a change in the object, either
internally or externally. You can get as fancy as you want in the
philosophy but there's a start :)

>  * A module is an object that partitions the namespace in a hierarchy.  For
>    example, you can have a function named "sin" which is distinct from the
>    sin function in the math module ("math.sin").  Modules used in a
>    straightforward way only add a single extra level of hierarchy to the
>    namespace.  Packages generalize this concept to multiple levels (modules
>    inside modules).

                     -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From eddie at holyrood.ed.ac.uk  Fri Jun 14 15:09:45 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Fri, 14 Jun 2002 19:09:45 +0000 (UTC)
Subject: a possibly foolish question about slices
References: <yu99sn3prbqm.fsf@europa.research.att.com> <aedd8k$brf$1@scotsman.ed.ac.uk> <m3znxxu21l.fsf@mira.informatik.hu-berlin.de>
Message-ID: <aedf1p$bv1$1@scotsman.ed.ac.uk>

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

>eddie at holyrood.ed.ac.uk (Eddie Corns) writes:

>> It frequently bugs me that you can't supply any number for b in <string>[a:b]
>> to indicate spanning to the end of the string.
>[...]
>> So if the designers could fix this for version 2.3 I think we'll all be happy
>> :)

>Let me introduce Guido's time machine: you can always use sys.maxint
>as the value of b.

Hmm, I was actually trying to suggest (non-seriously) fixing zero based
indexing! but I realise now that wasn't clear.

I never noticed using maxint, now we can say [p:p+n or sys.maxint] Hurrah!

Eddie


From not.this at seebelow.org  Wed Jun 12 17:12:39 2002
From: not.this at seebelow.org (Grant Griffin)
Date: 12 Jun 2002 14:12:39 -0700
Subject: why not "'in' in 'in'"?
Message-ID: <ae8dg70217g@drn.newsguy.com>

Hi Gang,

One of the things I like about Python is that it works as expected.  However, I
sometimes accidently expect it to allow one to use "in" to determine if one
string is "in" another, e.g.:

>>> 'in' in 'in'

which yields a traceback:

   Traceback (most recent call last):
     File "<interactive input>", line 1, in ?
   TypeError: 'in <string>' requires character as left operand

Evidently one is invited to use "count" or "find" instead:

>>> 'in'.count('in') > 0
1
>>> 'in'.find('in') >= 0
1

I prefer "find" over "count" because it (presumably) punts after the first
occurrence.  However, neither one reads nearly as well as the (illegal) "in"
form.

So why is the "in" operator limited to single characters when used as a
membership operator?  This peculiar lack of generality seems a bit (dare I say
it?) "non-Pythonic".  (That said, "Pythonic" seems to be a bit like
"enlightenment": those who claim to have mastered it wisely avoid spelling out
any concrete criteria by which others can judge their mastery <wink>.)

Perhaps its somehow due to the overloading of "in", which is used both as a
membership operator and as a separator gizmo inside "for" loops.  ("It's a floor
wax/No, it's a desert topping/No it's both!")  But is that any good reason to
exclude the handy, sensible, and downright intuitive use of "in" as a substring
detector?

Or maybe support for multiple characters would penalize the performance of the
single-character "in".

Or maybe Tim just doesn't want to burden us with a third way to do it <wink>.

gotten-spoiled-to-Python's-habits-of-working-as-expected
   -and-reading-as-pseudo-code-ly y'rs,

=g2 

_________________________________________________________________________

Grant R. Griffin                                           g2 at dspguru.com
Publisher of dspGuru                               http://www.dspguru.com
Iowegian International Corporation                http://www.iowegian.com



From deathtospam43423 at altavista.com  Fri Jun 28 15:57:06 2002
From: deathtospam43423 at altavista.com (netvegetable)
Date: 28 Jun 2002 19:57:06 GMT
Subject: Problem installing PIL 1.1.3
Message-ID: <Xns923C386473E29AusYourStandingInIt@130.133.1.4>

Keep getting this error everytime I try to run setup.py

File "D:\PYTHON22\lib\site-
packages\Pythonwin\pywin\framework\scriptutils.py", line 301, in RunScript
    exec codeObject in __main__.__dict__
  File "D:\Python22\Imaging-1.1.3\setup.py", line 52, in ?
    for line in open(os.path.join("libImaging", "ImConfig.h")).readlines():
IOError: [Errno 2] No such file or directory: 'libImaging\\ImConfig.h'

Anyone seen this happen before? Or even better, know how I can get it to 
work? 

-- 
netvegetable at excite.com


From geoff at gerrietts.net  Wed Jun  5 16:30:51 2002
From: geoff at gerrietts.net (Geoff Gerrietts)
Date: Wed, 5 Jun 2002 13:30:51 -0700
Subject: prototyping good OOdesign in Python?
In-Reply-To: <Pine.LNX.4.33.0206030725420.5660-100000@rnd.onego.ru>
References: <XFMail.20020601131915.shalehperry@attbi.com> <Pine.LNX.4.33.0206030725420.5660-100000@rnd.onego.ru>
Message-ID: <20020605203051.GC1455@isis.gerrietts.net>

Quoting Roman Suzi (rnd at onego.ru):
> On Sat, 1 Jun 2002, Sean 'Shaleh' Perry wrote:
> 
> Related topic is UML with Python. Does UML has everything to reflect
> Python OO model or does it need to add features?
> (Or maybe in order to interoperate, there is a need not to use some
> of Python features).

With UML, I generally find that it has TOO MANY features. If I use
Visio, or Rose, or even the lovely pyut announced a few months ago on
this very list, they'll ask for information about attribute and method
visibility, data types, etc.

Since Python doesn't care about most of this stuff, modelling it is
only useful if you expect to port the code later. And I prefer not to
plan for an unfortunate event. ;)

> And if in order to interoperate with less expressive
> languages/systems, how to keep Python pythonic and not fall back to
> some standard universal subset, found in many programming languages.

Certainly you could use UML and thorough documentation as a hedge
against relying too heavily on Python's dynamic nature. A startling
percentage of the things we take for granted in python can actually be
expressed similarly in other languages, though -- it just takes more
work, more infrastructure, and a few more lines of code.

I've written code with the thought that I'd eventually rewrite it in
another language, and I shared many of your concerns about how well
ideas would "port".

For the most part, I find myself doing one of two things: either
writing the other (generally Java) language's idioms (like Singleton,
like stateless classes) in Python when it's not too inconvenient, or,
when it is, just writing it off as "the burden of porting".

For the most part, the really "interesting" parts of your system are
going to port relatively cleanly, because you're doing calculations
and computations. It's the boilerplate crap that you can skip over
pretty effortlessly, and that's going to be the irritation of porting
anyway.

You can, of course, take special steps to avoid hard-to-port code.
That's where you spend time writing foreign idioms into your Python
code. How much time you waste doing that will largely depend on how
quickly you need the "convert to another language" step to go.

--G.

-- 
Geoff Gerrietts             "I am always doing that which I can not do, 
<geoff at gerrietts net>     in order that I may learn how to do it." 
http://www.gerrietts.net                    --Pablo Picasso




From ianb at colorstudy.com  Mon Jun 24 23:41:51 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 24 Jun 2002 22:41:51 -0500
Subject: Web templating/db tool with best designer/coder separation?
In-Reply-To: <af8dtp$5qp$0@216.39.172.122>
References: <af2gh6$2c2$0@216.39.172.122>  <af8dtp$5qp$0@216.39.172.122>
Message-ID: <1024976511.542.189.camel@lothlorien.colorstudy.net>

On Mon, 2002-06-24 at 19:36, Bengt Richter wrote:
> On 22 Jun 2002 18:43:50 GMT, bokr at oz.net (Bengt Richter) wrote:
> and from the description, this would seem to have a real advantage over all of
> the above, in that it uses attributes of html tags as opposed to special tags
> that browsers or design tools won't generally understand and render usefully.

Cheetah avoids using these tags for this reason.  It still requires
invalid HTML for some cases, though, particularly:

<table>
#for $item in $list #
<tr><td>$item</td></tr>
#end for #
</table>

(Note that you can use # to end statements, as you should do if you want
the template to work in WYSIWYG tools, which won't preserve newlines)

There's ideas for fixing this, but it's very annoying from the parsing
perspective... right now Cheetah doesn't have to understand markup at
all, but anything that will fix this will likely have to parse the
markup in order to work well.

In most other cases, though, I think Cheetah will be more usable if you
want to expose programming logic to designers.  I personally don't feel
that designers should be insulated from the logic -- it is essential to
know, for instance, that something will be repeating, or exactly what
variable gets inserted where.  Using dummy information to express this
isn't, IMHO, as transparent or reliable (since the dummy information
isn't necessarily in sync with the programming code).  

But this depends on how you view the designer, and what you expect of
them.  If you want the designer to be able to code the template, I don't
know how well ZPT will do -- it doesn't seem like an intuitive way to
think of the template (though I would expect a programmer to adapt
easily enough).  If you want the designer to only edit documents that
are first created by programmers, then it might work well.

But this is only my intuition, I haven't actually tried to use Cheetah
or ZPT with designers like this.


You might be interested to look at XMLC (http://xmlc.enhydra.org/) --
similar to ZPT, but with greater WYSIWYG support, and with considerably
less transparency (no actual code is contained in the template).  It's
kind of the extreme of that direction -- like accessing a DOM where the
markup is annotated and the API is much more friendly.  It's not in
Python, though maybe someone has created a clone (which shouldn't be
hard).

> BTW, which is the most appropriate Python version to assume everyone's servers
> will be converging on? 1.5, 2.1, 2.2, or ? Could I reasonably plan on using 2.2?
> E.g., I think 2.1 is being included in most newer linux distributions. But more
> to the point, what version will most likely be available via server cgi etc.?
> I'd like the answer to be 2.2 ;-)

Servers tend not to be very up-to-date.  I think 1.5.2 is still quite
common -- at least, it's likely to be /usr/bin/python on lots and lots
of servers (even if python2 is available somewhere).

  Ian






From marklists at mceahern.com  Thu Jun 13 10:24:22 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 13 Jun 2002 09:24:22 -0500
Subject: why not "'in' in 'in'"?
In-Reply-To: <3D08A161.A9F3136F@seebelow.org>
Message-ID: <NCBBLFCOHHDIKCAFGCFBCEGPLEAA.marklists@mceahern.com>

I haven't followed this thread.  Someone else has probably already suggested
this, but in case they haven't, it's fairly easy to define your own function
that does what you want:

def is_in(search, search_for):
  """Return bool indicating whether search_for is in search."""
  return search.find(search_for) >= 0

subject = "i like spam, i do."
x = "spam"
y = "ham"

print is_in(subject, x)
print is_in(subject, y)

// m

-





From graz at mindless.com  Wed Jun  5 15:05:49 2002
From: graz at mindless.com (Graham Ashton)
Date: Wed, 05 Jun 2002 20:05:49 +0100
Subject: Detecting OS and Window Manager
References: <ufsnh8pmmde6cf@corp.supernews.com>
Message-ID: <7ftL8.1441$VM4.132098@newsfep1-win.server.ntli.net>

On Wed, 05 Jun 2002 19:53:21 +0100, Andrew Wilkinson wrote:

> I'd like to know if
> they are running KDE or Gnome, does anyone know a (preferably) simple
> test for this?

I suspect your best bet is to resort to environment variables. I just
typed "env | grep -i gnome" and found out that (as I'm running GNOME) I
have an env variable called GNOME_SESSION_NAME (accessible via os.environ)

I've no idea how portable that is though. I'm running Ximian GNOME 1.4.

-- 
Graham


From sholden at holdenweb.com  Mon Jun 17 19:29:51 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 17 Jun 2002 19:29:51 -0400
Subject: socket.inet_aton - bug?
References: <MPG.1771a71d361193fc989683@news.rz.uni-karlsruhe.de>
Message-ID: <TaMP8.165343$Fp1.44927@atlpnn01.usenetserver.com>

"Uwe Mayer" <merkosh at hadiko.de> wrote ...
> Hi,
>
> socket.inet_ntoa() accepts the string '\xff\xff\xff\xff' which is
> excactly the same i'd expect socket.inet_aton('255.255.255.255') to
> produce if it accepted the string '255.255.255.255' as an IP adress.
> socket.inet_aton() dies with socket.error "illegal IP adress".
>
> i know this is no valid ip adress, but it is a valid netmask. the
> purpose of converting the netmask to a binary format was that i wanted
> to isolate the subnet of an ip adress.
> i wanted to use test_ip & netmask == ip_range to allow subnets to access
> a tcp server.
> is there another way of controlling this or why does socket.inet_aton()
> behave like that?
>

I can't really tell you why socket does that, but it seems to be a feature
of the underlying C socket library, which is where inet_aton() is defined.

The address 255.255.255.255 is, in fact, a valid destination for a
restricted broadcast (one which should be sent out only to hosts on the
local LAN).

 regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From pedrobados at eresmas.net  Fri Jun 28 09:39:29 2002
From: pedrobados at eresmas.net (Pedro Bados)
Date: Fri, 28 Jun 2002 15:39:29 +0200 (CEST)
Subject: ImportError: cannot import name mm_cfg
Message-ID: <XFMail.20020628153929.pedrobados@eresmas.net>

Hi : 

 I am not a python user so I have no idea at all of how I can resolve this
problem with mailman. Currently, I get this error each minute ...

   File "/home/mailman/cron/qrunner", line 90, in ?
    from Mailman.Bouncers import BouncerAPI
  File "/home/mailman/Mailman/Bouncers/BouncerAPI.py", line 32, in ?
    from Mailman import mm_cfg
ImportError: cannot import name mm_cfg

Any ideas ?

Thanks. Pedro.

-----------------------------------------------
E-Mail: Pedro Bados <pedrobados at eresmas.net>
Date: 28-Jun-2002 
-----------------------------------------------




From R.Barrett at ftel.co.uk  Mon Jun 24 05:20:04 2002
From: R.Barrett at ftel.co.uk (Richard Barrett)
Date: Mon, 24 Jun 2002 10:20:04 +0100
Subject: Socket Disconnection
In-Reply-To: <7934d084.0206232354.52c2163d@posting.google.com>
References: <3d1304d0$0$224$4d4ebb8e@news.nl.uu.net>
 <3d16817e$1@news.nz.asiaonline.net>
Message-ID: <5.1.0.14.2.20020624092656.00ad9e30@pop.ftel.co.uk>

At 00:54 24/06/2002 -0700, Andrae Muys wrote:
>"Colin Brown" <cbrown at metservice.com> wrote in message 
>news:<3d16817e$1 at news.nz.asiaonline.net>...
> > Hi Guyon
> >
> > I assume here that you are talking TCP/IP.
> >
> > If you are waiting on a receive [ data = connection.recv() ] then you will
> > have an empty data string returned when the client disconnects.
> >
> > If you are sending you will get an error telling you the link was shut 
> down.
> >
>
>Also note that the only reliable way to detect a client disconnect is
>the latter.  If you want to demonstrate the failure of the recv()
>approach, just yank out the ethernet cable on the client machine, and
>notice that although the client is now disconnected, the server has no
>idea.

That's because making a physical disconnect doesn't immediately cause a 
disconnection of connection oriented services over the medium. Which is 
probably what one wants to happen.

If you reinsert the ethernet cable then you will often find TCP/IP 
connections being carried over it that haven't attempted to pass traffic 
and failed, and disconnected for that reason, during the period that cable 
was out, are still active. A good demonstration of the robust nature of the 
TCP/IP protocols.

Disconnection in a TCP/IP sense is an exchange of segments that fits the 
protocol specification. TCP/IP supports asynchronous, full duplex 
communication and the forward and backward 'channels' each have to be 
terminated individually. TCP/IP itself is a symmetrical protocol i.e. it 
provides a peer-peer communication, not an inherently client-server 
communication. The client/serverness derives from whatever protocol is 
layered on top of TCP/IP, e.g. HTTP.

Simply because one of the two peers closes their outbound 'channel' doesn't 
mean, in a TCP/IP sense, that their inbound 'channel' will not/cannot 
deliver more data.

In the majority of cases, protocols layered on top of TCP/IP tend to be 
synchronous and often inherently create a client/sever relationship between 
the two TCP/IP  peers. It is common practice for higher level protocols, 
when receiving EOF (indicated, for instance, by receipt of an empty string 
from a recv call on a Python socket object), for the recipient to 
immediately respond by closing their outbound connection; this leading to 
termination of the underlying TCP/IP connection. But it doesn't have to be 
so. The recipient can. as far as TCP/IP is concerned, continue to transmit 
data until it chooses to stop.

If you are going to create your own higher level protocol layered on top of 
TCP/IP (rather than using an existing one such as HTTP) then part of the 
protocol design and specification is to decide on such issues. People often 
forget these issues when rolling their own higher level protocols, and can 
then experience considerable difficulty sorting out the problems that arise.

>Might I suggest taking a look at TCP/IP Illustrated Vol 1, by
>W.Richard Stevens.  Definately a book to read if you are doing any
>TCP/IP network programming.
>
>Andrae Muys
>--
>http://mail.python.org/mailman/listinfo/python-list





From benjl at cse.unsw.edu.au  Thu Jun 13 01:15:52 2002
From: benjl at cse.unsw.edu.au (Benno)
Date: 12 Jun 2002 22:15:52 -0700
Subject: map -> class instance
References: <m2r8jcp3pv.fsf@video.paradise.lost>
Message-ID: <332c7c11.0206122115.3a734b12@posting.google.com>

John Hunter <jdhunter at nitace.bsd.uchicago.edu> wrote in message news:<m2r8jcp3pv.fsf at video.paradise.lost>...
> I want to convert dictionary instances to instances of a class that
> has the keys of the map as attributes.  I naively tried this:
> 
> class Map2Class:
>     def __init__(self,m):
>         self.m = m
> 
>     def __getattr__(self, key):
>         return self.m[key]
> 
>     def __setattr__(self, key, val):
>         self.m[key] = val
> 
>     def __delattr__(self, key):
>         if self.m.has_key(key):
>             del self.m[key]
> 
> m = {'first' : 'John',
>      'last' : 'Hunter',
>      'age' : 34}
> 
> c = Map2Class(m)
> print c.first, c.last, c.age

<snip>

Depending on the exact semantics that you require a much simpler way would be:

class Map2Class: 
    def __init__(self, dikt): 
        self.__dict__ = dikt 
 
 
dikt = {'first':'John', 'last':'Hunter', 'age':34} 
 
c = Map2Class(dikt) 
print c.first, c.last, c.age 
 

Cheers,

Benno


From tdelaney at avaya.com  Wed Jun 12 00:19:54 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 12 Jun 2002 14:19:54 +1000
Subject: Cross platform spawning
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A37C@natasha.auslabs.avaya.com>

> From: sketerpot at chase3000.com [mailto:sketerpot at chase3000.com]
> 
> It works perfectly, and that solved the final bug in my 
> program.

That's what you think *now* ... ;)

Tim Delaney




From imbosol at vt.edu  Fri Jun 21 16:47:39 2002
From: imbosol at vt.edu (Carl Banks)
Date: Fri, 21 Jun 2002 16:47:39 -0400
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com> <lkznxqjqx2.fsf@pc150.maths.bris.ac.uk> <3D11A21D.9010809@mxm.dk> <uh3e1usajle42c@news.supernews.com> <3D11DA9B.163A7BD0@info.unicaen.fr> <uh446elev7af68@news.supernews.com>
Message-ID: <bd30fa.sm1.ln@grey.aerojockey.localdomain>

John Roth wrote:
> At the time he designed the language, I suspect that if anyone had
> brought it up, he'd have been laughed out of the room. We're talking
> late 80's, early 90s here, not the '60s or '70s.

Perl was also designed late 80's, early 90's, but it does have goto in
it.  I doubt Larry Wall was ever laughed out of a room.

Then again, one of the Perl manpages presents this as an "interesting
approach to a switch statement":

$amode = do {
    if     ($flag & O_RDONLY) { "r" }       # XXX: isn't this 0?
    elsif  ($flag & O_WRONLY) { ($flag & O_APPEND) ? "a" : "w" }
    elsif  ($flag & O_RDWR)   {
        if ($flag & O_CREAT)  { "w+" }
        else                  { ($flag & O_APPEND) ? "a+" : "r+" }
    }
};

So maybe these people thought goto was a little boring.

:)


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From brueckd at tbye.com  Sat Jun 22 11:38:24 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Sat, 22 Jun 2002 08:38:24 -0700 (PDT)
Subject: Python hits the spot
In-Reply-To: <3D1482AE.8FA299D8@kfunigraz.ac.at>
Message-ID: <Pine.LNX.4.44.0206220829330.16335-100000@ns0.tbye.com>

On Sat, 22 Jun 2002, Siegfried Gonzi wrote:

> My conclusion: I will rewrite my simulation in Fortran 90 and will
> abadon Python. Python is maybe good for some small scripts but not
> appropiate for serious programming.

Whatever works best for you, go for it. Be advised though that your
conclusion in the general sense is obviously bogus as for every one of you
there's a hundred or a thousand others who use Python quite successfully,
even for "serious programming" (see, for example, the recent thread on
large production systems in Python).

"It's the poor craftsman who blames his tools..." But who knows, maybe
Python is not the right tool for the problem you're solving... but to
infer from that that it's not usable for "serious programming" tasks is
either really poor reasoning or a big, steaming heap o' FUD.

Have fun,
-Dave





From aahz at pythoncraft.com  Tue Jun 18 13:47:19 2002
From: aahz at pythoncraft.com (Aahz)
Date: 18 Jun 2002 13:47:19 -0400
Subject: Tkinter and threads
References: <3D0DCC8F.9A6D3B81@ipm.fhg.de>
Message-ID: <aenrn7$7fl$1@panix1.panix.com>

In article <3D0DCC8F.9A6D3B81 at ipm.fhg.de>,
Markus von Ehr  <markus.vonehr at ipm.fhg.de> wrote:
>
>I'd like to have a progress bar increasing while a calculation process
>calculates some results.
>Unfortunately there is no mainloop processing while the calculation is
>running and so the progress bar begins after the calculation.
>Another thread neither can access the Tkinter resources, how can
>I manage this problem?

See the OSCON 2001 thread slides on my web site.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From Chris.Barker at noaa.gov  Wed Jun 19 15:52:09 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Wed, 19 Jun 2002 12:52:09 -0700
Subject: How to Store Binary data in a *.py module
Message-ID: <3D10E0E9.906DF517@noaa.gov>

I'd like to be able to store binary data in a python module file.

The motivation behind this is that I'd like to be able to distribute an
app with some of it's support files included in python modules, rather
than as separate files. This makes it easier to tao pas around, saves
some path issues, etc. It would also allow Py2exe to put it all in one
package. At the moment, if you have some data files to be read, you have
to distribute them separately.

I know a Python string can contain an arbitrary set of bytes, but I
don't know how to express binary data as a literal. If I were to try to
use r" ..." and I happened to have a byte that matched the ", it would
be mis-parsed. I'm imaging that UUencoding or something might solve my
problem, but I don't know anything about that.

Anyone have an idea??

BTW: this brings up the idea of having some kind of "resource file"
simulation in python. What I'm imagining is a way to build a module that
could contain a bunch of file-like objects that could contain any
arbitrary data, and be accessed like a file from Python. I suppose if I
figure out the above, the rest of it would be pretty easy...

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From aahz at pythoncraft.com  Tue Jun 25 12:26:38 2002
From: aahz at pythoncraft.com (Aahz)
Date: 25 Jun 2002 12:26:38 -0400
Subject: Suggestions for good programming practices?
References: <WEuR8.2074$5l4.75657@ozemail.com.au> <af8m9g$s5m$1@panix1.panix.com> <hd_R8.306964$352.31513@sccrnsc02>
Message-ID: <afa5ju$p64$1@panix1.panix.com>

In article <hd_R8.306964$352.31513 at sccrnsc02>,
Emile van Sebille <emile at fenx.com> wrote:
>Aahz
>> In Python 2.1.2 or higher, fire up the interactive interpreter and type
>> "import this".
>
>Except ActiveState's
>
>F:\Python22as>python
>Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
>Type "help", "copyright", "credits" or "license" for more information.
>>>> import this
>Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>ImportError: No module named this

Interesting.  That's an ActiveState bug, pure and simple; this.py is a
normal module in Lib/.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From peter at engcorp.com  Wed Jun 19 21:14:08 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 19 Jun 2002 21:14:08 -0400
Subject: Unicode hexadecimal characters
References: <3eeda89d.0206191615.71b0629d@posting.google.com>
Message-ID: <3D112C60.9C3D2F9F@engcorp.com>

Thinkit wrote:
> 
> It's a bit OT but does Unicode support the digits ten through fifteen
> (for hexadecimal)?

Do you mean "does Unicode have unique glyphs representing the
hexadecimal digits ten through fifteen other than the standard
A, B, C... ?"

Unicode of course supports various glyphs for A through F, but
it has nothing to do with other representations of bases other
than ten and under, as far as I know.

Not sure if that's exactly what you meant...

-Peter


From peter at engcorp.com  Fri Jun  7 08:48:54 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 07 Jun 2002 08:48:54 -0400
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com> <Lz1M8.123774$352.6153@sccrnsc02>
Message-ID: <3D00ABB6.6FAFD567@engcorp.com>

Emile van Sebille wrote:
> 
> Peter Hansen
> > You forgot the even more important first thing for a beginner:
> > get it correct!  The only good way to do this is to write unit
> > tests that prove it.  If you don't have tests, you don't have
> > working code (or how can you prove it?).
> 
> We always used absence of bug reports as a gauge... and always
> questioned if that indicated usage or correctness.  ;-)

For those who are serious when they say that, it's all well
and good if you can get your customer to test the code for 
you.... ;-)

-Peter


From ken at hotmail.com  Sat Jun  1 23:56:54 2002
From: ken at hotmail.com (Ken)
Date: Sun, 2 Jun 2002 13:56:54 +1000
Subject: Python and HTML frames
References: <ada6hf$vico3$1@ID-49758.news.dfncis.de> <r08ada.b7b.ln@10.0.0.1> <ada9ur$vcigh$1@ID-49758.news.dfncis.de> <rk3cda.nbb.ln@10.0.0.1>
Message-ID: <adc4u4$vrdda$2@ID-49758.news.dfncis.de>

"Jiri Baum" <jiri at baum.com.au> wrote in message
news:rk3cda.nbb.ln at 10.0.0.1...
> >> Ken:
> >> > How do I output HTML onto frames? This code below doesn't work. Can
> >> > someone help me?
>
> Jiri:
> >> This isn't really a python question, it's a HTML question (or possibly
> >> HTTP; are you ouputting the right headers?).
> ...
>
> Ken:
> > Why you think it isn't python question? I want that HTML frame output
> > from the python script. Can you tell me how?
>
> 1) Get your HTML right first, using static pages.
>
> 2) Output those same static pages from your python script.
>
> 3) Then worry about the dynamic stuff.
>
>

Solved..... printed too many <body> tags

Thanks




From russblau at hotmail.com  Tue Jun 11 14:51:20 2002
From: russblau at hotmail.com (Russell Blau)
Date: Tue, 11 Jun 2002 14:51:20 -0400
Subject: Sorting list (maybe stupid)
References: <mailman.1023789674.28761.python-list@python.org> <3D05F3FD.7020202@mxm.dk> <ae574q$42cjd$1@ID-76829.news.dfncis.de>
Message-ID: <ae5grc$43f78$1@ID-76829.news.dfncis.de>

"Russell Blau" <russblau at hotmail.com> wrote in message
news:ae574q$42cjd$1 at ID-76829.news.dfncis.de...
> "Max M" <maxm at mxm.dk> wrote in message news:3D05F3FD.7020202 at mxm.dk...
> > afile = open('myfile.txt')
> > tuplelist = []
> > for aline in afile:
> >      nummericValue = int(aline[:aline.find('.')])
> >      tuplelist.append((nummericValue, aline))
> > tuplelist.sort()
> > sortedLines = [tup[1] for tup in tuplelist]
> > for line in sortedLines:
> >      print line
>
> If the programmer is *certain* that each line will be in the format
> given by the OP, which is a line number followed by a period and a
> space, followed by arbitrary text, then the solution can be even
> simpler:
>
> linelist = []
> for aline in afile:
>     linelist.append(aline.split(".", 1))
> # each entry in linelist is a list of two elements: [line_number,
> line_body]
> linelist.sort()
> resultlist = []
> for item in linelist:
>     resultlist.append(".".join(item))
> # used resultlist in case the OP wants to do something with these
lines
> besides just print them
> # alternately, you could join the strings in-place in linelist
> for line in resultlist:
>     print line
>
D'oh!  I forgot to convert the line numbers to ints before sorting.  How
about this?

linelist = []
for aline in afile:
    linenum, linebody = aline.split(".", 1)
    linelist.append((int(linenum), linebody))
linelist.sort()
resultlist = []
for linenum, linebody in linelist:
    resultlist.append(str(linenum)+"."+linebody)
for line in resultlist:
    print line

--
I don't actually have a hotmail account; but I do have one on excite.com
if you really want to get in touch with me.





From brian at rk-speed-rugby.dk  Thu Jun 20 02:25:49 2002
From: brian at rk-speed-rugby.dk (Brian Elmegaard)
Date: 20 Jun 2002 08:25:49 +0200
Subject: SNIP IT: tk.Canvas Drag Items Around
References: <63604d2.0206031518.2da29d3b@posting.google.com>
Message-ID: <wkit4eh3f6.fsf@mail.afm.dtu.dk>

Hi Phlip,

I am not very experienced in tkinter, so I would like to understand
what is the advantage of your solutions compared to what I have in
Matt Conways example:
c:/Python22/Demo/tkinter/matt/canvas-moving-w-mouse.py  



phlip_cpp at yahoo.com (Phlip) writes:

> I have upgraded it to actually move the target items. This also 
> demonstrates canvasx and canvasy, the subject of recent mysteries.



-- 
Brian (remove the sport for mail)
http://www.rk-speed.dk http://fiduso.dk http://sunsite.auc.dk/dk-tug
\TeX, tak.


From frankm at hiwaay.net  Sun Jun 16 12:20:12 2002
From: frankm at hiwaay.net (Frank McIngvale)
Date: 16 Jun 2002 09:20:12 -0700
Subject: Source formatting for long expressions.
References: <mailman.1023570972.27430.python-list@python.org> <m3it4srhq5.fsf@mira.informatik.hu-berlin.de> <874rgcxy99.fsf@twistedmatrix.com> <mailman.1023668414.25769.python-list@python.org> <6bd9f01b.0206110543.168b0f83@posting.google.com> <878z5lx4vl.fsf@twistedmatrix.com> <mailman.1023816076.400.python-list@python.org>
Message-ID: <efbc8677.0206160820.31fc550d@posting.google.com>

Michael Gilfix <mgilfix at eecs.tufts.edu> wrote in message news:<mailman.1023816076.400.python-list at python.org>...
> On Tue, Jun 11 @ 10:32, Christopher Armstrong wrote:
> > XML is a terrible format for object persistence, due to the lack of robust
> > mappings. Twisted's XML uses attributes in some cases, but that is only
> > possible when you have string:string mappings. In *any* other case, you
> > have to resort to children nodes, with such ugly syntax as::
> > 
> > <instance class="twisted.some.Class">
> > <dictionary>
> >   <string role="key" value="someAttr" />
> >   <list>
> >     <string value="foo" />
> >   </list>
> > </dictionary>
> > 
> > Compare that with the equivalent AOT format::
> > 
> > Instance("twisted.some.Class", 
> >   someAttr=["foo"],
> > )
> > 
> > *That* is why I don't just use XML.
> 
>   I'm not disagreeing with you because I don't think you're wrong in
> this case and well, XML isn't for everything. I think XML can work
> here just as well but it's just more work. 
> 
>                     -- Mike

Hi, I'm not arguing the relative merits of using XML for persistance, but
just fyi, if you want a straight XML pickler, check out:

http://www.gnosis.cx/download/Gnosis_Utils-current.tar.gz
http://freshmeat.net/projects/gnosisxml/

frank


From phlip_cpp at yahoo.com  Mon Jun  3 19:18:31 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 3 Jun 2002 16:18:31 -0700
Subject: SNIP IT: tk.Canvas Drag Items Around
Message-ID: <63604d2.0206031518.2da29d3b@posting.google.com>

Hyp Ont:

Many moons ago one Tom Vrankar posted a snippet to this forum 
regarding dragging Tkinter Canvas items around.

I have upgraded it to actually move the target items. This also 
demonstrates canvasx and canvasy, the subject of recent mysteries.

The snippet does not make allowances for the offset between the 
mouse point and the anchor point of the draggee; this should be 
easy to add. But the symptom is clicking an object and starting 
to drag makes it "snick" over to the mouse location.

import sys, os, string, time
import Tkinter
tk = Tkinter

# A Python example of drag and drop functionality within a single Tk widget.
# The trick is in the bindings and event handler functions.
# Tom Vrankar twv at ici.net

# empirical events between dropee and target, as determined from Tk 8.0
# down.
# leave.
# up, leave, enter.

class CanvasDnD (tk.Frame):
  def __init__ (self, master):
    self.master =master
    self.loc =self.dragged =0
    tk.Frame.__init__ (self, master)
    canvas =tk.Canvas (self, width =256, height =256,
      relief =tk.RIDGE, background ="white", borderwidth =1)
    self.defaultcolor =canvas.itemcget (canvas.create_text (30, 25,
      font =("Helvetica", 14), text ="Item 1", tags ="DnD"), "fill")
    canvas.create_text (75, 75,
      font =("Helvetica", 14), text ="Item 2", tags ="DnD")
    canvas.create_text (125, 125,
      font =("Helvetica", 14), text ="Item 3", tags ="DnD")
    canvas.create_text (175, 175,
      font =("Helvetica", 14), text ="Item 4", tags ="DnD")
    canvas.create_text (225, 225,
      font =("Helvetica", 14), text ="Item 5", tags ="DnD")
    canvas.pack (expand =1, fill =tk.BOTH)
    canvas.tag_bind ("DnD", "<ButtonPress-1>", self.down)
    canvas.tag_bind ("DnD", "<ButtonRelease-1>", self.chkup)
    canvas.tag_bind ("DnD", "<Enter>", self.enter)
    canvas.tag_bind ("DnD", "<Leave>", self.leave)

  def down (self, event):
    print "Click on %s" %event.widget.itemcget (tk.CURRENT, "text")
    self.loc =1
    self.dragged =0
    event.widget.bind ("<Motion>", self.motion)

  def motion (self, event):
    root.config (cursor ="exchange")
    cnv = event.widget
    cnv.itemconfigure (tk.CURRENT, fill ="blue")
    x,y = cnv.canvasx(event.x), cnv.canvasy(event.y)
    got = event.widget.coords (tk.CURRENT, x, y)
  
  def leave (self, event):
    self.loc =0

  def enter (self, event):
    self.loc =1
    if self.dragged ==event.time:
      self.up (event)

  def chkup (self, event):
    event.widget.unbind ("<Motion>")
    root.config (cursor ="")
    self.target =event.widget.find_withtag (tk.CURRENT)
    event.widget.itemconfigure (tk.CURRENT, fill =self.defaultcolor)
    if self.loc: # is button released in same widget as pressed?
      self.up (event)
    else:
      self.dragged =event.time

  def up (self, event):
    event.widget.unbind ("<Motion>")
    if (self.target ==event.widget.find_withtag (tk.CURRENT)):
      print "Select %s" %event.widget.itemcget (tk.CURRENT, "text")
    else:
      event.widget.itemconfigure (tk.CURRENT, fill ="blue")
      self.master.update()
      time.sleep (.1)
      print "%s Drag-N-Dropped onto %s" \
        %(event.widget.itemcget (self.target, "text"),
   event.widget.itemcget (tk.CURRENT, "text"))
      event.widget.itemconfigure (tk.CURRENT, fill =self.defaultcolor)

root =tk.Tk()
root.title ("Drag-N-Drop Demo")
CanvasDnD (root).pack()
root.mainloop()

-- 
  Phlip
                http://www.greencheese.org/YaAw
  --  I'l have my Web site call your Web site...  --


From whisper at oz.net  Fri Jun 14 17:38:30 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 14 Jun 2002 14:38:30 -0700
Subject: [ANN] PyReverse 0.3
In-Reply-To: <mailman.1024086967.9145.clpa-moderators@python.org>
Message-ID: <GCEDKONBLEFPPADDJCOECEECDJAA.whisper@oz.net>

On windows, "setup install" terminates with:
error: file 'logilab\pyreverse\bin\py2tests.bat' does not exist

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Sylvain Thenault
> Sent: Friday, June 14, 2002 8:06
> To: python-list at python.org
> Subject: [ANN] PyReverse 0.3
>
>
>
> What's new in 0.3 ?
> -------------------
>         * full architecture reorganisation. Not yet perfect but much more
> 	  clean code :-)
> 	* begin unit tests generator using the unittest framework
> 	* apply patch from cognite at zianet.com fixing typos and KeyError in
> 	  pystats
> 	* add warning for functions parameters with default to a list or a
> 	  dictionnary
> 	* add a command line tool to generate xmi without argo files
> 	* add the --ignore option which permits to filter files analyzed
>         * use 4Suite cDomlette for XMI-UML generation if it's available
>
> What's PyReverse ?
> ------------------
>
> Pyreverse is a set of utilities to reverse enginering Python code. So
> far, it features dependency analysis tools, unittest generation,
> and XMI generation for importation in a UML modeling tool. A special
> module can be used to generate files readable by Argo UML.
> It uses a representation of a Python project in a class hierarchy
> which can be used to extract any information (such as generating UML
> diagrams and make a few statistics from the Python code, as "pyargo"
> and "pystats")
>
>
> Home page
> ---------
> http://www.logilab.org/pyreverse/
>
> Download
> --------
> ftp://ftp.logilab.org/pub/pyreverse/
>
> Mailing list
> ------------
> mailto://xml-logilab at logilab.org
>
>
> --
> Sylvain Th?nault,  LOGILAB
>
> http://www.logilab.org
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From peter at engcorp.com  Wed Jun 19 19:41:31 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 19 Jun 2002 19:41:31 -0400
Subject: Newbie Q: File accessing
References: <aeqb18$k10$1@phxtst22.phx1.aro.allied.com>
Message-ID: <3D1116AB.8A2D5BA4@engcorp.com>

Todd Harris wrote:
> 
> For those of you who use Microsoft <gulp...flame shield on> Word, when a
> document is opened by one user, and another user tries to access it, they
> can open it for read only. Obviously, Word has made some file attribute
> where it is locked to that first user, and only that first user can write to
> it while they have it open. Is there a way to do this in Python? Or what
> other possible options do I have? I've looked been using the open(
> filePath ) function as well as the os.open( ) function and os.access( ) to
> check my progress, but can't seem to figure out how to get around this.
> Thanks,

Possibly msvcrt.locking() for an MS-only solution... no idea how it
works, but the docs give you a bit of a start...

-Peter


From eddie at holyrood.ed.ac.uk  Thu Jun 13 13:22:37 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Thu, 13 Jun 2002 17:22:37 +0000 (UTC)
Subject: Poll: Readable languages
References: <mailman.1023982001.21846.python-list@python.org>
Message-ID: <aeakct$4a6$1@scotsman.ed.ac.uk>

>	'Good programming cannot be taught by preaching generalities.
>	The way to learn to program well is by seeing, over and over,
>	how real programs can be improved by the application of a few
>	principles of good practice and a little common sense.
>	Practice in critical reading leads to skills in rewriting,
>	which in turn leads to better writing.'


>If I understand the excerpt, I think it's saying that experience through
>practice from good examples makes for good programming skills.

Surely that's exactly what it's not saying for examples provide only
generalities.  The key word in the quote is 'critical' (critical: 'involving
careful judgement or judicious evaluation' from my dictionary) ie the essence
is to look at the code and see WHY certain principles will IMPROVE the
original.

The reason why most programmers do not produce good code is because, once they
have learned *A* method for accomplishing a task they simply repeat that method
ad infinitum without reflecting on whether there might be a better way.

Even if you whack them on the head with a better method they will at best just
take that new one and use it in future without thinking.  People mostly just
don't like thinking (even smart ones sadly).

And of course there are other aspects to 'think' about to become a good
programmer.  Understanding doesn't come easily - you have to keep looking and
thinking until the day you die (OK so I've been watching too many martial arts
movies lately).

The reason why *I* don't produce good code is entirely different, honest.

(Trying very hard not to turn this into a rant about the general uselessness
of people lest I give the impression of being old and embittered! :)

Eddie


From m2 at plusseven.com  Sun Jun 16 07:29:47 2002
From: m2 at plusseven.com (Alex Polite)
Date: Sun, 16 Jun 2002 13:29:47 +0200
Subject: Profiling stuff like urlopen and connection.execute()
Message-ID: <20020616112947.GA5420@matijek.plusseven.com>

In my little webspider project I've just realized something about the
python profiler. It counts only CPU seconds not "wall clock"-seconds.
So profiling database queries or urlopens is kind of pointless. One
could measure the time manually on every urlopen and conn._execute()
(which is what I did to find out that the db was the bottleneck and
not the DSL pipe), but for larger projects this is tiresome.

Is there anyway to do something about this?

-- 
Alex Polite
http://plusseven.com/gpg/




From maxm at mxm.dk  Fri Jun 21 06:47:12 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 21 Jun 2002 12:47:12 +0200
Subject: application development - separating content from function
References: <yroQ8.260949$i6.17986369@amsnews02.chello.com> <3D1289F8.58A17BA4@engcorp.com>
Message-ID: <3D130430.6050004@mxm.dk>

Peter Hansen wrote:

> Welcome to the club! :-)  Undertaking an overwhelming project is pretty
> common, especially with many open source projects.  Check out the number
> of idle projects on SourceForge, for example.  I've made many many 
> false starts in my life as a programmer, and I still regret much of the
> time I spent on some of those hopeless projects.


I absolutely agree. I have heard of people doing detailed design and 
architecture for big projects. I have never been able to work 
efficiently that way.

Getting a good basic project up, showing it to the customer, and then 
modify it, is definitely the way I prefer to work.

I don't mean thah I just sit down blindly and code out of the blue. I 
usually do interface mock-up's, designs etc. as much as I can. But even 
then I keep them absolutely minimal. This has also had the unexpected 
side effect that my interfaces usually are very easy to understand.

The problem with doing designs up-front is that the designs are very 
limited by your experience as a programmer.

Sometimes you need to use new technology you don't know, or you need to 
develop for an area of expertice you don't have etc. Then it is really 
hard to do good designs.

The XP method makes it much easier to learn as you go along.

regards Max M



From rnd at onego.ru  Tue Jun  4 15:38:05 2002
From: rnd at onego.ru (Roman Suzi)
Date: Tue, 4 Jun 2002 23:38:05 +0400 (MSD)
Subject: Kalman Filters, Particle Filters and SVM's
In-Reply-To: <671ybne7zm.fsf@mail.uni-oldenburg.de>
Message-ID: <Pine.LNX.4.33.0206042337190.11291-100000@rnd.onego.ru>

On 4 Jun 2002, Johannes Nix wrote:

>
>Are there any implementations of Kalman Filters, Particle Filters or
>Support Vector Machines in (Numerical) Python ?

LIBSVM has Python hooks.

(google find it for you)

>Especially the latter would be fine, I'm dreaming to implement a SVM
>based script with the capability to distinguish spam from personal
>mails.... but I'm way to lazy to write my own SVM.
>
>Johannes
>
>
>
>

Sincerely yours, Roman Suzi
-- 
\_ Russia \_ Karelia \_ Petrozavodsk \_ rnd at onego.ru \_
\_ Tuesday, June 04, 2002 \_ Powered by Linux RedHat 7.2 \_
\_ "Shin: A device for finding furniture in the dark." \_





From alf at logilab.fr  Mon Jun 10 10:44:16 2002
From: alf at logilab.fr (Alexandre Fayolle)
Date: Mon, 10 Jun 2002 14:44:16 +0000 (UTC)
Subject: [ANN] constraint-0.1
References: <20020606135835.GF26159@orion.logilab.fr> <mailman.1023712335.14288.python-list@python.org>
Message-ID: <slrnag9eq0.109.alf@virgo.logilab.fr>

Niki Spahiev a ?crit :
> 6/6/2002, 16:58:35, Alexandre wrote:
> 
> A> Hello,
> 
> A> Logilab has released constraint-0.1.
> 
> Is it usable as Geometry Solver?

I don't know what a Geometry Solver is. Could you tell me what you mean
more precisely? What kind of problem do you want to solve?

constraint is not yet able to deal with continuous domains, so this may
be a limitation when dealing with geometry. 

Alexandre Fayolle
-- 
LOGILAB, Paris (France).
http://www.logilab.com   http://www.logilab.fr  http://www.logilab.org
Narval, the first software agent available as free software (GPL).


From gerhard at bigfoot.de  Sat Jun  1 11:47:03 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 1 Jun 2002 15:47:03 GMT
Subject: Migrating from PHP to Python...
References: <xWyk7.4824$v%4.492814@news20.bellglobal.com> <l4ihfu0nhp1a9p53vgch3n3kcc25s8c00p@4ax.com> <slrnafhj0d.dp4.gerhard@lilith.my-fqdn.de> <m3znyf3x2o.fsf@mumonkan.sunstarsys.com>
Message-ID: <slrnafhr3p.e0j.gerhard@lilith.my-fqdn.de>

Joe Schaefer wrote in comp.lang.python:
> Gerhard =?iso-8859-15?Q?H=E4ring?= <gerhard at bigfoot.de> writes:
> 
> [f'ups disobeyed]
> 
> [...]
> 
>> X-Post and F'up to appropriate group. Please continue trolling over
>> there.
> 
> Please don't spam clp.misc with garbage like this.
> IMO you've done a disservice to both communities.

Sorry. I should have just PLONKed him and been done with it.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From joakim at abargon.com  Thu Jun  6 12:48:44 2002
From: joakim at abargon.com (Joakim Ziegler)
Date: Thu, 06 Jun 2002 11:48:44 -0500
Subject: MS SQL connectivity
Message-ID: <pan.2002.06.06.16.47.25.714491.1989@abargon.com>

I'm currently faced with the... interesting task of talking to an MS SQL
Server 8.0 database running on W2k, from Python running on GNU/Linux. I've
looked at Dave Cole's MSSQL module, but I'm having problems getting it to
compile (using FreeTDS), it seems there are defines used in the C
extension module that are not defined in the freetds header files, such as
INT_CONTINUE and DBRPCRECOMPILE, and when I track the values of those
defines down on the web and fix that problem, the build continues a bit
further, but then fails with lots of undeclared functions, such as
dbcaltbind, dbclropt, DBCURCMD, DBCURROW, dbisopt, etc.

Additionally, according to the TODO file of that module,
Connection.commit() is unimplemented, which seems to hint that it's not
possible to use transactions.

Dave Cole hasn't answered my email, so I'm a bit at a loss for what to do
with that module (although it seems the simpler alternative).

The other alternative I've looked at is mxODBC with the FreeTDS ODBC
driver. This seems to be a more mature solution, but to be honest, I don't
know ODBC at all, and I'm having problems making it work. Does someone
know of a step-by-step guide to getting mxODBC with the FreeTDS ODBC
driver up and running on GNU/Linux, talking to an MS SQL Server? I've
defined the data source in /etc/odbc.ini, looking like this:

[ODBC Data Sources]
SUNTEST = Sunsystems test server

[SUNTEST]
Driver          = /usr/lib/libtdsodbc.so
Description     = Sunsystems test server
Trace           = No
Servername      = 192.168.1.39
Database        = SUNDB
UID             = <whatever>
PWD             = <whatever>

[Default]
Driver          = /usr/lib/libtdsodbc.so


(Yeah, I'm trying to talk to a SunSystems server. I know this is pretty
masochistic.) Anyway, I'm not at all sure about what these fields should
be, though, the database server is on the IP above, and the UID and PWD
are valid. Yet, when I try to do:

connection = mx.ODBC.iODBC.connect("DSN=SUNTEST")

I get:

LookupDSN: LoadDSN retured FALSETraceback (most recent call last):
[...]
mxODBC.InterfaceError: failed to retrieve error information(line 5939,
rc=-1)


I suspect there's some semi-easy way to make this work, as I said, I'm
totally unknowledgable when it comes to ODBC, I just want to get a
connection up to said database so I can do queries. :) Any hints,
recipes, HOWTOs, etc.?

-- 
--Joakim Ziegler


From omega at temple-baptist.com  Sun Jun  2 23:49:54 2002
From: omega at temple-baptist.com (Erik Walthinsen)
Date: Sun, 02 Jun 2002 20:49:54 -0700
Subject: Problems with exec scoping
Message-ID: <adep6r01ntg@enews1.newsguy.com>

I'm working on a project that involves significant use of python embedded
in HTML.  I've modified an existing package (psp.py) to work in my
environment, and cleaned up its syntax a little (more to do still).  It
returns a string containing the entire page with verbatim html in
stringIO.write() statements and the code out in the clear.

My problem is that all the pages written in this style need to have
access to variables in other parts of the program (the http server
functionality is only a small part of it the whole system).  My solution
was to optionally pass a dictionary with the variables I need to have in
the page's scope:

def send_page(....,filename,....,dict={}):
  ...process filename into execdata...

  localdict = locals()
  for key in dict.keys():
    localdict[key] = dict[key]
  exec execdata in globals(), localdict

class someclass:
  objects = {}
  def somefunction(filename):
    send_page(....,filename,....,{ objects = someclass.objects })

The problem I'm running into now is that if I print out the repr() of the
'objects' variable from within the python in the HTML page, I get a dict
with the right keys, but the values of each of the keys are non-existant.
This obviously puts a bit of a crimp in my plans, since the whole point of
making the dict available is to use the objects contained.

I've seen the other threads on exec scope, and they seem like they might
be applicable, but the thread mostly covers the black magic of most
efficiently creating the right dictionary for exec.  That doesn't seem to
be relevant to this problem, but I might be wrong.  I've only been doing
python for a couple weeks now, so there's a lot I don't know yet.

The entire project will be available in CVS shortly, at
http://savannah.gnu.org/projects/metanet/, with captive_send_page residing
in captivewebserver.py and the someclass/somefunc being CaptiveClient/ 
captive_http_request in captiveclient.py.

In the meantime I'll code up the pages I need manually, but that can't
scale very far.

TIA,
   Omega

      Erik Walthinsen <omega at temple-baptist.com> - System Administrator
        __
       /  \                GStreamer - The only way to stream!
      |    | M E G A        ***** http://gstreamer.net/ *****
      _\  /_


From jeremy at zope.com  Tue Jun 18 18:45:53 2002
From: jeremy at zope.com (Jeremy Hylton)
Date: Tue, 18 Jun 2002 18:45:53 -0400
Subject: any users of ExtensionClass out there?
Message-ID: <15631.47137.259544.229900@slothrop.zope.com>

We've made a number of small changes to ExtensionClass over the last
year or two, but never issued a release.  The most significant change
that I can remember is that the ExtensionClass in the Zope CVS will
work with a debug build of Python.  It used to segfault almost
immediately. 

There has been little perceived interest in ExtensionClass releases
because Zope is moving away from ExtensionClass.  I wonder, though, if
there are other users of ExtensionClass who would be interested in a
new release.  Please let me know if you are.

Jeremy





From thor__00 at yahoo.com  Thu Jun 13 14:05:21 2002
From: thor__00 at yahoo.com (Thor)
Date: Thu, 13 Jun 2002 20:05:21 +0200
Subject: system command
References: <aeajuh$5b5eo$1@ID-108351.news.dfncis.de> <3d08dba0_2@hpb10302.boi.hp.com>
Message-ID: <aeamjt$5eflk$1@ID-108351.news.dfncis.de>

Daniel Fackrell wrote:

> Assumption: You are referring to the "source" command as provided by the
> bash shell on some *nix.
> 
> The man page for bash suggests that you probably want:
> 
> print "source "+root+".in"
> os.system("bash "+root+".in")

Changing to bash does the same.. i mean, nothing, no error, just finishes 
the script 
-- 
Thor -- Stockholm -- Sverige


From baas at ira.uka.de  Thu Jun 27 17:05:40 2002
From: baas at ira.uka.de (Matthias Baas)
Date: Thu, 27 Jun 2002 23:05:40 +0200
Subject: Python Documentation generation
References: <mailman.1024599266.494.python-list@python.org>
Message-ID: <72umhusf2i7cbevgs5e4muademk7dkdl32@4ax.com>

On Thu, 20 Jun 2002 14:49:14 -0400, "Mahrt, Dallas"
<dallasm at aiinet.com> wrote:

>I am looking for an application which can generate HTML documentation from
>Python modules.
>(Note: some examples based on JavaDoc syntax)

You can use doxygen (http://www.doxygen.org) to create such a
documentation. Basically, it's a tool like JavaDoc but for the C/C++
language. It has no built-in support for Python but it allows to pass
files through filters before they're actually processed.
I've written a filter that converts Python code into something that
looks like a C++ header files (see below). That way doxygen can also
be used to document Python code.

Every comment starting with ## and every docstring is literally
translated into a doxygen comment block. 

>I have a few desires from such a program:
>	- Ability to exclude elements based on naming conventions or
>keywords 
>		(ex. Exclude names starting with a single underscore '_foo'
>or containing the keyword @private)

My filter understands the special comments:

## public: 
## protected:
## private:

which will insert the corresponding keyword in the C++ code. Thus, you
can mark methods as protected or private and tell doxygen to exclude
them from the documentation. (However, this won't work for functions).

>	- Ability to have inter module linking. (If module A uses module B,
>then there should be an '@see B' like link [JavaDoc-ish])

You can add such commands into the doxygen comments (just like
JavaDoc). Here you should know that my filter puts everything into a
namespace with the same name as the file name. This means, if you
manually link to other parts of the documentation you have to include
the namespace.

>	- The program should also allow for either:
>		- generating module documentation from C extension modules
>(Should be tricky)

Well, that's what doxygen was actually written for. :)

>		- allow for other documentation sections to be linked
>automatically by registration of module name

Uhm, I'm not sure if I know what you mean here. Maybe it's this: If
you mention a class name in your documentation this class name will
automatically appear as a link to the documentation of the class
(however, the "namespace feature" often prevents the automatic linking
from working properly).

- Matthias -

---------------------------------------------------------------------------------------------------
Here's the filter (pythfilter.py), I hope the word wrapping didn't
mangle the script....

#! /usr/bin/python

# Doxygen filter which can be used to document Python source code.
# Classes (incl. methods) and functions can be documented.
# Every comment that begins with ## is literally turned into an
# Doxygen comment. Consecutive comment lines are turned into
# comment blocks (-> /** ... */).
# All the stuff is put inside a namespace with the same name as
# the source file.

# Conversions:
# ============
# ##-blocks                  ->  /** ... */
# "class name(base): ..."    ->  "class name : public base {...}"
# "def name(params): ..."    ->  "name(params) {...}"

# Written by Matthias Baas (baas at ira.uka.de)

# Changelog:
# 18.06.2002: Es gibt jetzt auch ein ## public:
# 21.01.2002: from ... import wird umgesetzt in "using namespace ...;"
#             TODO: "from ... import *" vs "from ... import names"
#             TODO: Bei normalem Import: name.name -> name::name
# 20.01.2002: #includes werden VOR den Namespace gesetzt

######################################################################

# The program is written as a state machine with the following states:
#
# - OUTSIDE               The current position is outside any comment,
#                         class definition or function.
#
# - BUILD_COMMENT         Begins with first "##".
#                         Ends with the first token that is no "##"
#                         at the same column as before.
#
# - BUILD_CLASS_DECL      Begins with "class".
#                         Ends with ":"
# - BUILD_CLASS_BODY      Begins just after BUILD_CLASS_DECL.
#                         The first following token (which is no
comment)
#                         determines indentation depth.
#                         Ends with a token that has a smaller
indendation.
#
# - BUILD_DEF_DECL        Begins with "def".
#                         Ends with ":".
# - BUILD_DEF_BODY        Begis just after BUILD_DEF_DECL.
#                         The first following token (which is no
comment)
#                         determines indentation depth.
#                         Ends with a token that has a smaller
indendation.

import sys,os.path,string,token,tokenize

OUTSIDE          = 0
BUILD_COMMENT    = 1
BUILD_CLASS_DECL = 2
BUILD_CLASS_BODY = 3
BUILD_DEF_DECL   = 4
BUILD_DEF_BODY   = 5
IMPORT           = 6
IMPORT_OP        = 7
IMPORT_APPEND    = 8

# Output buffer
outbuffer = []

out_row = 0
out_col = 0

# Variables used by rec_name_n_param()
name         = ""
param        = ""
doc_string   = ""
record_state = 0

# Tuple: (row,column)
class_spos  = (0,0)
def_spos    = (0,0)
import_spos = (0,0)

# Which import was used? ("import" or "from")
import_token = ""

# Comment block buffer
comment_block = []
comment_finished = 0

# Imported modules
modules = []

# Program state
stateStack = [OUTSIDE]


######################################################################
# Output string s. '\n' may only be at the end of the string (not
# somewhere in the middle).
#
# In: s    - String
#     spos - Startpos
######################################################################
def output(s,spos, immediate=0):
  global outbuffer, out_row, out_col

  os = string.rjust(s,spos[1]-out_col+len(s))
  if immediate:
    sys.stdout.write(os)
  else:
    outbuffer.append(os)
  if (s[-1:]=="\n"):
     out_row = out_row+1
     out_col = 0
  else:
     out_col = spos[1]+len(s)


######################################################################
# Records a name and parameters. The name is either a class name or
# a function name. Then the parameter is either the base class or
# the function parameters.
# The name is stored in the global variable "name", the parameters
# in "param".
# The variable "record_state" holds the current state of this internal
# state machine.
# The recording is started by calling start_recording().
#
# In: type, tok 
######################################################################
def rec_name_n_param(type, tok):
  global record_state,name,param,doc_string
  s = record_state
  # State 0: Do nothing.
  if   (s==0): 
     return
  # State 1: Remember name.
  elif (s==1):
     name = tok
     record_state = 2
  # State 2: Wait for opening bracket or colon
  elif (s==2):
     if (tok=='('): record_state=3
     if (tok==':'): record_state=4
  # State 3: Store parameter (or base class) and wait for an ending
bracket
  elif (s==3):
     if (tok==')'): 
        record_state=4
     else:
        param=param+tok
  # State 4: Look for doc string
  elif (s==4):
    if (type==token.NEWLINE or type==token.INDENT or
type==token.SLASHEQUAL):
      return
    elif (tok==":"):
      return
    elif (type==token.STRING):
      while tok[:1]=='"':
        tok=tok[1:]
      while tok[-1:]=='"':
        tok=tok[:-1]
      doc_string=tok
    record_state=0

######################################################################
# Starts the recording of a name & param part.
# The function rec_name_n_param() has to be fed with tokens. After
# the necessary tokens are fed the name and parameters can be found
# in the global variables "name" und "param".
######################################################################
def start_recording():
  global record_state,param,name, doc_string
  record_state=1
  name=""
  param=""
  doc_string=""

######################################################################
# Test if recording is finished
######################################################################
def is_recording_finished():
  global record_state
  return record_state==0

######################################################################
## Gather comment block
######################################################################
def gather_comment(type,tok,spos):
  global comment_block,comment_finished
  if (type!=tokenize.COMMENT):
     comment_finished = 1
  else:
    # Output old comment block if a new one is started.
    if (comment_finished): 
       print_comment(spos)
       comment_finished=0
    if (tok[0:2]=="##" and tok[0:3]!="###"):
       comment_block.append(tok[2:])

######################################################################
## Output comment block and empty buffer.
######################################################################
def print_comment(spos):
  global comment_block,comment_finished
  if (comment_block!=[]):
     output("/**\n",spos)
     for c in comment_block:
       output(c,spos)
     output("*/\n",spos)
  comment_block    = []
  comment_finished = 0

######################################################################
def set_state(s):
  global stateStack
  stateStack[len(stateStack)-1]=s

######################################################################
def get_state():
  global stateStack
  return stateStack[len(stateStack)-1]

######################################################################
def push_state(s):
  global stateStack
  stateStack.append(s)

######################################################################
def pop_state():
  global stateStack
  stateStack.pop()  


######################################################################
def tok_eater(type, tok, spos, epos, line):
  global stateStack,name,param,class_spos,def_spos,import_spos
  global doc_string, modules, import_token

  rec_name_n_param(type,tok)
  if (string.replace(string.strip(tok)," ","")=="##private:"):
     output("private:\n",spos)
  elif (string.replace(string.strip(tok)," ","")=="##protected:"):
     output("protected:\n",spos)
  elif (string.replace(string.strip(tok)," ","")=="##public:"):
     output("public:\n",spos)
  else:
     gather_comment(type,tok,spos)

  state = get_state()
  # OUTSIDE
  if   (state==OUTSIDE):
    if   (tok=="class"):
       start_recording()
       class_spos = spos
       push_state(BUILD_CLASS_DECL)
    elif (tok=="def"):
       start_recording()
       def_spos = spos
       push_state(BUILD_DEF_DECL)
    elif (tok=="import") or (tok=="from"):
       import_token = tok
       import_spos = spos
       modules     = []
       push_state(IMPORT)
  # IMPORT
  elif (state==IMPORT):
    if (type==token.NAME):
       modules.append(tok)
       set_state(IMPORT_OP)
  # IMPORT_OP
  elif (state==IMPORT_OP):
    if (tok=="."):
       set_state(IMPORT_APPEND)
    elif (tok==","):
       set_state(IMPORT)
    else:
       for m in modules:
         output('#include "'+m+'.py"\n', import_spos, immediate=1)
         if import_token=="from":
           output('using namespace '+m+';\n', import_spos)
       pop_state()
  # IMPORT_APPEND
  elif (state==IMPORT_APPEND):
    if (type==token.NAME):
       modules[len(modules)-1]+="."+tok
       set_state(IMPORT_OP)
  # BUILD_CLASS_DECL
  elif (state==BUILD_CLASS_DECL):
    if (is_recording_finished()):
       s = "class "+name
       if (param!=""): s = s+" : public "+param
       if (doc_string!=""): comment_block.append(doc_string)
       print_comment(class_spos)
       output(s+"\n",class_spos)
       output("{\n",(class_spos[0]+1,class_spos[1]))
       output("  public:\n",(class_spos[0]+2,class_spos[1]))
       set_state(BUILD_CLASS_BODY)
  # BUILD_CLASS_BODY
  elif (state==BUILD_CLASS_BODY):
    if (type!=token.INDENT and type!=token.NEWLINE and type!=40 and
type!=tokenize.NL\
       and type!=tokenize.COMMENT \
       and (spos[1]<=class_spos[1])):
          output("}; // end of class\n",(out_row+1,class_spos[1]))
          pop_state()
    elif (tok=="def"):
       start_recording()
       def_spos = spos
       push_state(BUILD_DEF_DECL)
  # BUILD_DEF_DECL
  elif (state==BUILD_DEF_DECL):
    if (is_recording_finished()):
       s = name+"("+param+");\n"
       if (doc_string!=""): comment_block.append(doc_string)
       print_comment(def_spos)
       output(s,def_spos)
#       output("{\n",(def_spos[0]+1,def_spos[1]))
       set_state(BUILD_DEF_BODY)
  # BUILD_DEF_BODY
  elif (state==BUILD_DEF_BODY):
       if (type!=token.INDENT and type!=token.NEWLINE \
          and type!=40 and type!=tokenize.NL \
          and (spos[1]<=def_spos[1])):
#            output("} // end of
method/function\n",(out_row+1,def_spos[1]))
            pop_state()
#       else:
#            output(tok,spos)


def dump(filename):
  f = open(filename)
  r = f.readlines()
  for s in r:
    sys.stdout.write(s)

def filter(filename):
  path,name = os.path.split(filename)
  root,ext  = os.path.splitext(name)
  
  output("namespace "+root+" {\n",(0,0))

  sys.stderr.write('Filtering "'+filename+'"...')
  f = open(filename)
  tokenize.tokenize(f.readline, tok_eater)
  f.close()
  print_comment((0,0))

  output("\n",(0,0))
  output("}  // end of namespace\n",(0,0))

  for s in outbuffer:
    sys.stdout.write(s)
  

######################################################################
######################################################################
######################################################################

try:
  filename = string.join(sys.argv[1:])
  root,ext  = os.path.splitext(filename)

  if ext==".py":
    filter(filename)
  else:
    dump(filename)

  sys.stderr.write("OK\n")
except IOError,e:
  sys.stderr.write(e[1]+"\n")




From romany at actimize.com  Wed Jun 26 09:27:51 2002
From: romany at actimize.com (Roman Yakovenko)
Date: Wed, 26 Jun 2002 15:27:51 +0200
Subject: system-output to wxpython-window
Message-ID: <7647A9F4B298E74DB6793865DA67285004AD94@exchange.adrembi.com>

According to what you define as simple ?
You need to open pipe and then all you read is going to the output.

 streamInput, streamOutput, streamError = popen3('C:\\run.bat')
 while 1:
            sData = streamOutput.readline()
            if not sData:
                break
            writeToGUI(sData) # <-- your function
answer = streamOutput.close()

Also I think after a few minutes of running this code you will want to write multi threading. 
It is also simple. I have working example. I think exactly what you need. 

Roman

-----Original Message-----
From: Klaus Reinhardt [mailto:K.Rdt at TU-Berlin.DE]
Sent: Wednesday, June 26, 2002 7:20 AM
To: Python
Subject: system-output to wxpython-window


---------------------------------------------------------------------
Hi

I have some batch-files, which I want to
invoke from a nice wxpython-frame. So
I did   system("0m.bat");   but the output
is coming to the DOS-shell-window. Is
there a simple way to do the job. Best case
would be one menue-window and another
for output.

		K at Rdt
---------------------------------------------------------------------





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




From david.bear at asu.edu  Mon Jun  3 20:56:12 2002
From: david.bear at asu.edu (David Bear)
Date: Mon, 3 Jun 2002 17:56:12 -0700
Subject: using getopt
Message-ID: <adh37d$dj8$1@news.asu.edu>

Found a couple of frustrating things with getopt.  I am writing a filter 
that will be called by lprng -- or other spool services.  rather than 
trying to code for every possible cmd line arg I may be given, in there a 
way to use getopt to grab just a few options that I care about and through 
the rest away?

-- 
David Bear
College of Public Programs/ASU
480-965-8257
...the way is like water, going where nobody wants it to go


From dialtone#/N0SPAM/# at aruba.it  Sat Jun 15 19:20:09 2002
From: dialtone#/N0SPAM/# at aruba.it (Dialtone)
Date: Sun, 16 Jun 2002 01:20:09 +0200
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com>
Message-ID: <3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com>

On Sat, 15 Jun 2002 22:58:36 GMT, "George Hester"
<hesterloli at hotmail.com> wrote:

>I think I need a script to do this.

>>> import sys
>>> sys.version
'2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)]'
>>> sys.version_info
(2, 2, 1, 'final', 0)
-- 
try: troll = "Brain"
except TypeError, data: 
   troll.plonk()
Co-Responsabile Hardware Information & Technology http://hit.edengames.net


From news at lindbergs.org  Tue Jun  4 00:09:14 2002
From: news at lindbergs.org (VanL)
Date: Mon, 03 Jun 2002 22:09:14 -0600
Subject: Which is fastest dict or list.index() ?
References: <mailman.1023078005.25452.python-list@python.org>
Message-ID: <3CFC3D6A.5010706@lindbergs.org>

Hello,

I am curious about your experience with these BTrees.  I am 
writing an outliner with a ZODB backend in which I was going 
to use these as nodes for storage (IOBTrees, probably) as 
opposed to the supplied PersistentDict class.

I was planning on using a data structure that (with 
extraneous stuff ommitted) looked like this:

(Note: ObjID is a unique int.)

BTree(ObjID_1: {'Content': <Stuff>,
           'Connections': [Parent_ObjID,
            ObjID_child1, ObjID_child2, etc]
           'ID': ObjID},
ObjID_2: {'Content': <Stuff>,
           'Connections': [Parent_ObjID,
            ObjID_child1, ObjID_child2, etc]
           'ID': ObjID},
etc.

(This is an extension of the 'ideal python graph model' as 
expounded by Guido at 
http://www.python.org/doc/essays/graphs.html)

The content of the various nodes (the <stuff> above) could 
potentially be quite large (500 k -> 1MB per node), so I was 
thinking that the BTree might be better, as it would only 
bring the necessary nodes into memory.

However, your mention that these Btrees are 'sluggish' 
grabbed my attention.  Would it be better to pay the memory 
price for speedier performance?

I guess I am asking what you mean by sluggish. Sluggish as 
in "I get 10,000 ops per second on a Btree, but 100,000 ops 
per second on a dict" or sluggish as in "I only get 10 ops 
per second on a Btree because I have to go to disk every 
time I pull a node?"

Thanks,

VanL

> 
> 




From tdelaney at avaya.com  Mon Jun 17 20:59:12 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 18 Jun 2002 10:59:12 +1000
Subject: what *is* a class?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A396@natasha.auslabs.avaya.com>

> From: holger krekel [mailto:pyth at devel.trillke.net]
> 
> I agree that this is a good thing. But it wouldn't hurt to have
> 
>     reverse, sort, extend
> 
> as standalone functions somewhere in the stdlib.  They would return
> a copy, so that you can write:
> 
>     # iterate over a sorted copy of a list
>     for item in sort(l):
>         ...
> 
> reverse could probabally be made an iterator so that it gets really
> cheap to write:
> 
>     for item in reverse(l):
>         ...
> 
> And i think these functions should work with tuples and strings also.

Personally, I think this is a great idea. But I would exclude extend.

I think there should be a library of "iterable" functions (perhaps in
builtins ;) which return an iterator, and accept any iterable (where
appropriate). Let's call it builtin module iterable (or maybe iterator ...).
Functions defined in this module would be specified to not modify the
original iterable (indeed, should all call iter(iterable) before doing
anything). Of course, whee appropriate a function passed in could modify the
original.

Examples of such functions would be:

iterable.sort(iterable [, func])
iterable.reverse(iterable)
iterable.map(func, iterable, ...)
iterable.filter(func, iterable, ...)
iterable.reduce(func, iterable, ...)
iterable.zip(iterable, ...)
iterable.range([start,] stop[, step])

There have been suggestions like this before, but I don't recall one for
sort() and reverse(), and these two have fired me :)

Tim Delaney




From jadestar at idiom.com  Wed Jun 12 11:52:25 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 12 Jun 2002 15:52:25 GMT
Subject: Is 'isinstance()' the right thing?
References: <xduznzl1xci.fsf@leto.informatik.uni-freiburg.de> <3CCE5F2A.1103E744@alcyone.com> <xdur8kx1i1i.fsf@leto.informatik.uni-freiburg.de> <uttC8.40554$zW3.440516@news1.tin.it> <efb63dac.0205280500.6f9ed1e5@posting.google.com> <83u1orsvxj.fsf@panacea.canonical.org>
Message-ID: <ae7qnp$1m2c$1@news.idiom.com>

Kragen Sitaker <kragen at pobox.com> wrote:

> juenglin at informatik.uni-freiburg.de (Ralf Juengling) writes:
>> Why aren't types not the right machinery for specifying the support 
>> of a protocol? ...

> The problem is that once you provide a mechanism for requiring
> arguments to be derived from a particular type, people will start
> requiring them to be derived from lots of different types, and some of
> those types won't be interface types.

> The result, in languages that support this, like C++ and Java, is
> tight coupling between classes: class A requires that arguments to its
> bletch() method be derived from class B, which typically means they
> have to inherit class B's bugs and stupid design mistakes, and also
> means that they have to be written by somebody who knew about class B.

> In one Python application I currently work on, we've been able to
> switch between using lists and Numeric arrays in several places with
> very little difficulty.

 However, we still want *some* way to use the introspection features
 in appropriate ways (sometimes for sanity checking, more often for
 application specific reasons).

 It seems like we can usually replace type tests with try: except:
 blocks or we can use hasattr() to test for the method or attribute
 that implements the desired object feature.  For example, the only
 sequence types for which hasattr(x,'isalpha') is true are the string
 and user string types.  Thus, if I want to write a generalized function
 that can recursively iterate over nested containers (lists , tuples, etc
 possibly CONTAINING lists, tuples, etc ad nauseum) but DON'T wish to 
 iterate over strings (which are iterable but are NOT really containers
 in most applications) then I can use hasattr(x,'isalpha') as a way to 
 stop my recursion.

 The problem with such generality is that there are always corner
 or unaccounted for cases.  For my "recursive container iteration" 
 example, I don't know what I'd use to avoid recursing into mmap objects
 if one of those were passed to me.  Maybe I'd want to interate on 
 file or database containers but NOT on chars in an mmap?  Obviously 
 this depends greatly on the task at hand.  If at all possible I'll
 restructure a specific case into an exception handling block.

 From what I can tell hasattr() should work through any form of GoF 
 Decorator, Adapter, Proxy or similar delegated composition objects 
 *iff* you can find the appropriate attribute to test for.

 Any thoughts on that?  Counterexamples?



 

 


From marklists at mceahern.com  Thu Jun 13 12:02:47 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 13 Jun 2002 11:02:47 -0500
Subject: why not "'in' in 'in'"?
In-Reply-To: <aeadtt06rq@drn.newsguy.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEHHLEAA.marklists@mceahern.com>

> That's not a bad suggestion, Mark, but if I did that I guess I 
> would be forever wondering if I was testing whether the first one 
> was in the  second or the second one was in the first.

Since you insist on torturing yourself, then, why not name the function:

	def is_the_second_in_the_first(a, b):
        ...

;-)

// mark
-





From martin at v.loewis.de  Wed Jun 19 17:00:29 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 19 Jun 2002 23:00:29 +0200
Subject: Extending long
References: <69310ed7.0206191220.3bbf68d8@posting.google.com>
Message-ID: <m3ofe7q902.fsf@mira.informatik.hu-berlin.de>

estama at dblab.ntua.gr (Freeman Stopjohn) writes:

> class bitnum(long):
>     def getbit(self,bitn):
>         if self & (1 << bitn) >0:
>             return 1
>         else:
>             return 0
>     def setbit(self,bitn):
>         self=self ^ ( 1 << bitn )
> -----
> 
>  The getbit method works ok.
>  The setbit doesn't (it doesn't change the number).
> 
>  What am i doing wrong?

You are assigning to self. self is a parameter of setbit, just like
bitn. Assignments to formal parameters have no effect.

What you want to do cannot be done; long objects are immutable. I
recommend you implement a wrapper, rather than inheriting.

Regards,
Martin



From egbert at bork.demon.nl  Fri Jun 14 04:26:58 2002
From: egbert at bork.demon.nl (Egbert Bouwman)
Date: Fri, 14 Jun 2002 10:26:58 +0200
Subject: system command
In-Reply-To: <aeajuh$5b5eo$1@ID-108351.news.dfncis.de>; from thor__00@yahoo.com on Thu, Jun 13, 2002 at 07:19:48PM +0200
References: <aeajuh$5b5eo$1@ID-108351.news.dfncis.de>
Message-ID: <20020614102657.A1077@bork.demon.nl>

    
On Thu, Jun 13, 2002 at 07:19:48PM +0200, Thor wrote:
> Having this code
>  
>   print "source "+root+".in"
>   os.system("source "+root+".in")
> 
> the script does not run the root+".in" file, but copying and pasting the 
> output form teh 1st line work.. what i'm doing wrong? Thanks in advance

I cannot reproduce your error, all goes well.
First I made a bash script called hallo.in, with only one line in it:

echo Hello, this hello.in

and next in the interactive python shell:

>>> import os
>>> root='hello'
>>> print "source "+root+".in"
source hello.in
>>> os.system("source "+root+".in")
Hello, this hello.in

egbert

-- 
Egbert Bouwman - Keizersgracht 197 II - 1016 DS  Amsterdam - 020 6257991
========================================================================




From bokr at oz.net  Wed Jun  5 19:39:12 2002
From: bokr at oz.net (Bengt Richter)
Date: 5 Jun 2002 23:39:12 GMT
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com>
Message-ID: <adm7f0$1cc$0@216.39.172.122>

On Tue, 4 Jun 2002 08:03:30 -0400, "John Roth" <johnroth at ameritech.net> wrote:

>
>"Huaiyu Zhu" <huaiyu at gauss.almadan.ibm.com> wrote in message
>news:slrnafo7qv.37c.huaiyu at gauss.almadan.ibm.com...
>> Gustavo Cordova <gcordova at hebmex.com> wrote:
>
>>
>> Great explanation!  The problem stems from the fact that the current
>+=
>> operator implements two semantically distinct operations, chosen
>according
>> to some properties of the operands.  This is not unlike the situation
>of the
>> division operator with distinct behaviors on floats and ints.
>
>This is not an issue with '+='. It's an issue with assignment,
>regardless
>of whether it is standalone or combined with some other operation.
>
>And I disagree that assignment implements two distinct operations.
>Assignment always rebinds. Always. In the examples cited, it
Not so fast with that "Always." ;-)

 >>> class P(object):
 ...     def __init__(self,v): self._v = v
 ...     def get_v(self): return self._v
 ...     def set_v(self,v): return # ;-)
 ...     v = property(get_v, set_v)
 ...
 >>> p=P('orig value')
 >>> p.v
 'orig value'
 >>> id(p.v)
 8209984
 >>> p.v = 'try a change'
 >>> id(p.v)
 8209984
 >>> p.v
 'orig value'
 >>> p.v += ' try an increment'
 >>> id(p.v)
 8209984
 >>> p.v
 'orig value'

>attempted to rebind into a tuple, because that's what was asked
>for when the left hand side had a subscript.
>
>The reason that the examples given at the top of the thread fail
>in such strange ways is simply that '+=' does not check whether
>the assignment is possible before doing the addition (or concatination,
>in the specific cases cited.)
>
>To see this, you could try the first example, using explicit '+' and
>'='. I suspect it would fail in exactly the same way, with exactly
>the same side effect.
>
You can make a tuple that is more forgiving, in the spirit of Jeff Epler's
suggestion:

 >>> class T(tuple):
 ...     def __setitem__(self,i,thing):
 ...         if self[i] is thing: return
 ...         raise TypeError,"object doesn't support item assignment of new item"
 ...
 >>> t = T( ([0],) )
 >>> t
 ([0],)
 >>> t[0] += [1]
 >>> t
 ([0, 1],)
 >>> t[0] = t[0] + [2]
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "<stdin>", line 4, in __setitem__
 TypeError: object doesn't support item assignment of new item

And:

 >>> a=t[0]
 >>> a
 [0, 1]
 >>> a += [2]
 >>> a
 [0, 1, 2]
 >>> t[0] = a
 >>> t
 ([0, 1, 2],)

But:

 >>> a = a + [3]
 >>> a
 [0, 1, 2, 3]
 >>> t[0] = a
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "<stdin>", line 4, in __setitem__
 TypeError: object doesn't support item assignment of new item

Regards,
Bengt Richter


From peter at engcorp.com  Sat Jun 22 23:59:07 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 22 Jun 2002 23:59:07 -0400
Subject: ConfigParser & converting strings to lists
References: <3D152A28.2177B054@tds.net> <3D152C69.CC41ADB0@tds.net> <3D1531DD.ACF06E31@engcorp.com> <3D15335C.8CB1558D@tds.net>
Message-ID: <3D15478B.EE5BA0D2@engcorp.com>

"Edward K. Ream" wrote:
> Here is what I did:
> 
> s = config.get(section,"recentFiles")
> files = [x[1:-1] for x in s[1:-1].split()]
> 
> Is this what you intended?  This doesn't work for me: it "doubles" every
> backspace.

I have no idea how you're getting backspaces (\b) anywhere.  I tried 
a file that looked like your example and the above worked fine for
me.

> I am already regretting trying to be clever.  It is easy enough to set
> file individual filenames like this:
[...]
> 
> This is bulletproof: filenames can contain any character at all.  The
> only drawback is that the config file looks like:
> 
> [recent files]
> file0 = filename0
> file1 = filename1
> ...
> 
> But this isn't so bad, is it?

No, that's not bad.  In fact, I think it's pretty common.  Doing
The Simplest Thing That Could Possibly Work is generally a good strategy. :-)

-Peter


From kosh at aesaeion.com  Fri Jun 21 03:26:46 2002
From: kosh at aesaeion.com (kosh at aesaeion.com)
Date: Fri, 21 Jun 2002 01:26:46 -0600 (MDT)
Subject: compile python witout tkinter
Message-ID: <Pine.LNX.4.44.0206210124060.585-100000@aesaeion.com>

On a machine I need to build python on I need to build it without tkinter.
How can I do that. I didn't notice any configure options to do that and I
also looked at Modules/Setup and that seems to have _tkinter commented
out. I am compiling python 2.1.3 on linux.

Thanks

Designing and building web applications http://webme-eng.com





From cliechti at gmx.net  Sun Jun 23 17:07:32 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 23 Jun 2002 23:07:32 +0200
Subject: mail-problem over t-online
References: <mailman.1024859369.26585.python-list@python.org>
Message-ID: <Xns9236EBAE3BEA4cliechtigmxnet@62.2.16.82>

Klaus Reinhardt <K.Rdt at TU-Berlin.DE> wrote in 
news:mailman.1024859369.26585.python-list at python.org:
> Testing python and email I adapted this script:
>      
>      host=smtplib.SMTP('mailszrz.zrz.tu-berlin.de')
>      host.sendmail(zwi_eck(From),zwi_eck(Tp),al)
>      host.quit
> 
> This is working, when I'm directly connected to my
> old ISP TU-Berlin via ISDN. But the same is failing,
> when I'm connected at t-online via ADSL, with
> authentification required .. Providing my log and pass
> produces errors, too.
> 
> Pop3 is working in both cases.
> 
> Who can give me a hint?

spam protection? serious admins only allow sending from inside their 
network and block send requests from the outside. (spammers search for such 
machines so that someone else has the trouble with sending thousands of 
emails)

POP3 on the other hand is no problem as reading emails is no problem 
(except for some small privacy issues, like unencrypted passwords)

so i think your mailserver is properly configured :-)

sometimes mailservers are configred that they let you send emails from the 
outside after you logging in to POP3. then for a couple of minutes you can 
use SMPT. guessing from the error message, that could be that case here.

chris

-- 
Chris <cliechti at gmx.net>



From martin at v.loewis.de  Thu Jun 13 02:57:08 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 13 Jun 2002 08:57:08 +0200
Subject: Something faster then sgmllib for sucking out URLs
References: <mailman.1023913936.26808.python-list@python.org>
Message-ID: <m34rg71xa3.fsf@mira.informatik.hu-berlin.de>

Alex Polite <m2 at plusseven.com> writes:

> I'm working on a webspider to fit my sick needs. The profiler
> tells me that about 95% of the time is spent in sgmllib. I use sgmllib
> solely for extracting URLs. I'm looking for a faster way of doing
> this. Regular expressions, string searches? What's the way to go? I'm
> not a python purist. Calling some fast C program with the html as
> argument and getting back a list of URLs would be fine by me.

I recommend to use sgmlop, which is distributed both as part of PyXML,
and separately by Fredrik Lundh. It is the fastest SGML/XML parser I
know of, for use within Python.

Regards,
Martin



From cliechti at gmx.net  Thu Jun 20 19:27:10 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 21 Jun 2002 01:27:10 +0200
Subject: Raw String Output
References: <aetnkf$tot$1@news.umbc.edu>
Message-ID: <Xns9234F35131B7cliechtigmxnet@62.2.16.82>

"mike" <gringro1 at umbc.edu> wrote in news:aetnkf$tot$1 at news.umbc.edu:

> If I have a string variable with the value '\n' how do I output it to
> a file with the value of '\012' instead ?

i assume that its clear to you that '\n' == '\012' so i guess you want the 
string r'\012' in the file and not the newline character?!?

then you must do it by hand, i think. this one replaces all characters < 32 
with their octal value:
>>> print ''.join([ c>=' ' and c or '\\%03o' % ord(c) for c in 'hello\n'])
hello\012

chris

-- 
Chris <cliechti at gmx.net>



From stephen at xemacs.org  Mon Jun  3 02:51:04 2002
From: stephen at xemacs.org (Stephen J. Turnbull)
Date: 03 Jun 2002 15:51:04 +0900
Subject: Iterating through two lists
References: <3cee0ae6_1@news3.newsgroups.com> <mailman.1022234742.24014.python-list@python.org> <3cee2338_2@news3.newsgroups.com> <mailman.1022241402.21254.python-list@python.org> <1022297508.675568@yasure> <831ybrwl1v.fsf@panacea.canonical.org>
Message-ID: <87sn44lusn.fsf@tleepslib.sk.tsukuba.ac.jp>

>>>>> "Kragen" == Kragen Sitaker <kragen at pobox.com> writes:

    Kragen> "Donn Cave" <donn at drizzle.com> writes:

    >> The function that carries data is cool enough.  Lisp does it
    >> easier than Python, and in fact does "functional programming"
    >> better than Python by a long shot.

    Kragen> That's debatable.

If you're going to take the Python side of that, I think you've found
yourself a real challenge.  After all, Lisp was invented by writing an
automatic interpreter for a purely functional logic notation.

While you can do functional programming in Python, the syntax and the
ubiquity of side effects makes it hard to emphasize that style.

At least for me.

Hm.  Maybe I should take a look at M. Pinard's topy tool, and see if
maybe I shouldn't be writing my Python programs in Lisp (Scheme, to be
precise)<wink>.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
 My nostalgia for Icon makes me forget about any of the bad things.  I don't
have much nostalgia for Perl, so its faults I remember.  Scott Gilbert c.l.py


From franklin at NO.SPAM.statix.net  Wed Jun 19 13:24:42 2002
From: franklin at NO.SPAM.statix.net (Franklin Kingma)
Date: Wed, 19 Jun 2002 17:24:42 GMT
Subject: PIL composite
Message-ID: <3D10BE5E.3020608@NO.SPAM.statix.net>

Hi,

i'm trying to 'merge/overlay' two transparant gifs using the composite 
function.. but whatever i try i cant get the job done...

has anyone tried something similair before?

thanks,

franklin



From peter at engcorp.com  Mon Jun 24 20:55:14 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 24 Jun 2002 20:55:14 -0400
Subject: How to stop a SocketServer?
References: <7xhejv9af2.fsf_-_@ruckus.brouhaha.com> <3d1744d4$1_1@hpb10302.boi.hp.com> <3D1757DF.776A428F@engcorp.com> <3d176a0c$1_3@hpb10302.boi.hp.com>
Message-ID: <3D17BF72.2E1CCF40@engcorp.com>

Daniel Fackrell wrote:
> 
> "Peter Hansen" <peter at engcorp.com> wrote:
> 
> Any chance I could see a very simple example of this in action?  That would
> help my little project out quite a bit.

This is just pulled out of a little server somebody around here wrote
the other week.  Out of context it might not help, but maybe the idea
would give you a start.  By the way, look into "timeoutsocket.py"
which is a plug-in replacement for the standard socket.py and which
might do the trick for you.

    def run(self):
        import threading

        try:

            try:
                # handle requests, waking up often to check if we should terminate
                while not self.sockserver.terminated:
                    readySocket = select([self.sockserver.socket], [],
                        [self.sockserver.socket], SELECT_TIMEOUT)

                    if readySocket[0] or readySocket[2] :
                        if threading.activeCount() < self.threadLimit:
                            self.sockserver.handle_request()
                        else:
                            time.sleep(0.5)

            finally:
                self.sockserver.socket.close()

        except:
            if self.exceptionHandler:
                self.exceptionHandler.handle()

        return


> Would this be something that would allow me to guarantee that the all
> threads of the server terminated within, say, five seconds of the server
> shutdown request?

I believe it would allow you to guarantee that.

I also think Skip's suggestion in another thread is even better if you're
just trying to terminate the whole application.

-Peter


From CousinStanley at HotMail.Com  Wed Jun 12 11:25:23 2002
From: CousinStanley at HotMail.Com (Cousin Stanley)
Date: 12 Jun 2002 15:25:23 GMT
Subject: PIL _imaging
References: <Xns922AA829A8C5FHotMailcs@130.133.1.4> <3D06E694.4040103@mxm.dk> <Xns922B3150A6B32HotMailcs@130.133.1.4> <3D074529.E848AA6@ipm.fhg.de>
Message-ID: <Xns922B561EDB212HotMailcs@130.133.1.4>

Cousin Markus ... 

Thanks for the reply ... 

Moved the PIL folder up a 2 levels :
  ... From K:\Python\Lib\PIL\PIL 
  ...   To K:\Python\PIL           # contains xxxxx.py modules   

I also copied  PIL.pth : 
  ...  To K:\Python\PIL.pth
  ...  This file contains 3 characters ... PIL 
  ...  as originally shipped with the distribution 

There are copies of the  .pyd  files in :
  ... K:\Python\DLLs\_imaging.pyd
  ... K:\Python\DLLs\_imagingtk.pyd

However, the result is still the same : 
  ... See trace below ...

I tried with and without K:\Python\DLLs 
in the PYTHONPATH variable and without 
PYTHONPATH set at all ...

Cousin Stanley

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

< K:\Python\PIL >
Wed 06-12-2002  7:23:51.33 > python Samples/viewer.py ButtonD.jpg
Traceback (most recent call last):
  File "Samples/viewer.py", line 46, in ?
    UI(root, im).pack()
  File "Samples/viewer.py", line 25, in __init__
    self.image = ImageTk.PhotoImage(im)
  File "K:\Python\PIL\PIL\ImageTk.py", line 86, in __init__
    self.paste(image)
  File "K:\Python\PIL\PIL\ImageTk.py", line 108, in paste
    im.load()
  File "K:\Python\PIL\PIL\ImageFile.py", line 126, in load
    self.load_prepare()
  File "K:\Python\PIL\PIL\ImageFile.py", line 180, in load_prepare
    self.im = Image.core.new(self.mode, self.size)
  File "K:\Python\PIL\PIL\Image.py", line 37, in __getattr__
    raise ImportError, "The _imaging C module is not installed"
ImportError: The _imaging C module is not installed



From ulbi at ivs.tu-berlin.de  Fri Jun 28 11:11:28 2002
From: ulbi at ivs.tu-berlin.de (Andreas Ulbrich)
Date: Fri, 28 Jun 2002 17:11:28 +0200
Subject: feeding a prompt with python
References: <3D1C5341.9070001@huno.net>
Message-ID: <3D1C7CA0.7060201@ivs.tu-berlin.de>

thomas wrote:
> hello,
> 
> is there a way with python to feed a prompt with data? i call sth. with 
> system($some_cmd) and the cmd gives me a prompt which of course waits 
> for user input. however i want to automate this, is there a way?
> i browsed through the essential reference 2nd edt. but didn't find 
> anything.
> 
> thx,
> thomas
> 

What you actually want to do is using pipes. See os.popen and related 
stuff for documentation.



From sabren at manifestation.com  Sun Jun  2 17:46:49 2002
From: sabren at manifestation.com (Michal Wallace)
Date: Sun, 2 Jun 2002 17:46:49 -0400 (EDT)
Subject: Python Turing machine
In-Reply-To: <ade0mc$j4r$1@newsg3.svr.pol.co.uk>
Message-ID: <Pine.LNX.4.44.0206021745430.887-100000@mercury.sabren.com>

On Sun, 2 Jun 2002, G. Willoughby wrote:

> > hint: post some URLs of those implementations to the list. that way we
> > don't need to look up what it is and people here have fun trying to solve
> > things pythonic...
> 
> I belive he means a chat bot that tries to fool humans into beliving its a
> real person.


:) That's the Turing test, another of Alan Turing's
inventions. A Turing machine is sort of an extremely simple
computer:

http://plato.stanford.edu/entries/turing-machine/



Cheers,

- Michal   http://www.sabren.net/   sabren at manifestation.com 
------------------------------------------------------------
Switch to Cornerhost!             http://www.cornerhost.com/
       High Powered Hosting - With a Human Touch. :)
------------------------------------------------------------





From jeff at sasmor.com  Mon Jun 17 14:32:55 2002
From: jeff at sasmor.com (Jeff Sasmor)
Date: Mon, 17 Jun 2002 18:32:55 GMT
Subject: If you use Python -OO is it a permanent condition??
Message-ID: <rXpP8.32471$ks6.25445@nwrddc02.gnilink.net>

I tried to find an answer to this in the comp.lang.python archives
on google, but could only find vague references:

I have an app that works fine until I run compileall on it when
Python is run with the -OO flag, e.g.,

python -OO makedist.py

where part of makedist.py has

compileall.compile_dir('.',20,force=1)


I then take all the .pyo files and put them in a different
directory; I'm trying to make a smaller distribution and was
trying this out as an option.

The only problem is that when I run this application that is
only .pyo files,  imports do not work:

python P1.pyo  (this is the entry point)

produces errors at the first import:

Traceback (most recent call last):
  File ".\P1.py", line 132, in ?
ImportError: No module named plat_init

So  I tried:

python -OO P1.pyo

and it works just fine.  Does this mean that the interpreter must always
be started with -OO to execute files compiled with -OO?  I can't imagine
why this would be the case, but it seems to be -- at least as far as I can
see
from my little bit of experimentation.

My app uses the 'compile' built-in within it; as the interpreter is running
with the -OO flag it would seem that what I compile would be compiled
that way too - no good for what I'm trying to do.

Can some expert in the deep innards of Python-dom comment on this
and help clarify this for me?

TIA,

Jeff Sasmor
jeff at sasmor.com




From jjl at pobox.com  Sun Jun  9 17:33:40 2002
From: jjl at pobox.com (John J. Lee)
Date: Sun, 9 Jun 2002 22:33:40 +0100
Subject: Source formatting for long expressions.
In-Reply-To: <m3it4srhq5.fsf@mira.informatik.hu-berlin.de>
References: <mailman.1023570972.27430.python-list@python.org> <m3it4srhq5.fsf@mira.informatik.hu-berlin.de>
Message-ID: <Pine.SOL.4.30.0206092230510.15771-100000@mimosa.csv.warwick.ac.uk>

On 9 Jun 2002, Martin v. Loewis wrote:

> Christopher Armstrong <radix at twistedmatrix.com> writes:
>
> > Does anyone know of a tool that will beak up a very, very long
> > expression like this into multiple lines with sane indentation?
>
> Emacs indentation gives this:
>
> Ref(1, Instance('twisted.internet.app.Application',
>                 persistenceVersion=7, services={}, udpPorts=[], gid=1000,
>                 connectors=[], uid=1000, tcpPorts=[(9080,
>                                                     Instance('twisted.web.server.Site', sessions={},
[...]

Emacs python-mode will actually align only one indent level in (instead of
aligning with the opening bracket) if you have a newline after the
bracket:

Ref(1, Instance(
    'twisted.internet.app.Application',
    persistenceVersion...

The tcpPorts bit should obviously be moved to another line, though.


John



From shalehperry at attbi.com  Sat Jun 29 20:03:59 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Sat, 29 Jun 2002 17:03:59 -0700 (PDT)
Subject: newbie Q: delete list few lines in file
In-Reply-To: <7fe97cc4.0206291519.1c95ce31@posting.google.com>
Message-ID: <XFMail.20020629170359.shalehperry@attbi.com>

> 
> 
> I'm much more interested in learning python idiom and correctness
> here. I'm not interested in just "get it to work". For the latter, i
> can do well with perl. (also, is there a site or mailing list that
> collect very small trivial python programs for learning purposes?)

1) the python tutor mailing list.  Designed for those new to python.

2) useless python, do not have the URL handy, do a search. 




From chandan_mathur at hotmail.com  Mon Jun  3 23:05:25 2002
From: chandan_mathur at hotmail.com (chandan)
Date: 3 Jun 2002 20:05:25 -0700
Subject: Login Authentication with cookies...
Message-ID: <e271e93c.0206031905.73796d31@posting.google.com>

Hi.., 

Problem: - write a python script to access pages that are password
protected, but DONT return a 401 forbidden code. i.e. Login to a
website like hotmail and check you mail message or in my particular
case, http://www.fool.com/login.asp
and then download the discussion board messages for a particular
stock.

Attempts so far: - The login and password protection is based on
cookies, so I used ClientCookie.py as follows (although, i am not sure
how to use it properly)

import ClientCookie
import urllib2
import urllib
params = urllib.urlencode({'login':'login','password':'password'})
request = urllib2.Request("http://www.fool.com/login.asp", params)
# note we're using the urlopen from ClientCookie, not urllib2
result = ClientCookie.urlopen(request)

Error message: 
Traceback (innermost last)
  File "c:\python22\programs\demos\trial", line 7, in ?
    result = ClientCookie.urlopen(request)
NameError: global name 'types' is not defined

Program disconnected.

I dunno what to do about the types in ClientCookie.py... even the
example code given in the README file for this library returns the
same error.., i am using windows XP, python 2.2

ClientCookie can be obtained from here -
http://wwwsearch.sourceforge.net/ClientCookie/

Thanx for the help
Chandan

P.S : Newbie...


From mcherm at destiny.com  Mon Jun 24 16:57:23 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Mon, 24 Jun 2002 16:57:23 -0400
Subject: SOLVED (Re: Python hits the spot)
Message-ID: <3D1787B3.1070003@destiny.com>

One of us, ranting:
 > My bet is that once you cool off and sits down with some debugging tools
 > like GDB, Boundschecker or Electric Fence you will indeed find the
 > programming error and fix it.
 >
 > My bet is also that you will not tell anyone either.


Siegfried Gonzi:
> RPM1 wrote:
> 
>> Hey!  That's why I was asking if your laptop was a Hewlitt
>> Packard.  I had a Hewlitt Packard Pavillion laptop that would
>> die when the fan kicked on
 >
> It was really the laptop processor. I tried the calculation on a 
> stationary machine (Pentium II 450 MHz and SuSE 8.0), and it works.
> 
> Die Typen k?nnen jetzt was erleben.

Looks like we would have lost BOTH bets! Debugging tools don't find 
hardware problems, and Siegfried DID let us know when it worked.

Sigfried, I hope Python feels a little better to you now that you know 
it wasn't the cause of your problems. As you can see, this newsgroup has 
an ASTONISHING amount of knowledge on ALL kinds of things. But you'll 
find that you get even MORE help (and more polite help) if you are a 
little slower to blame Python next time. Approach the denizens of this 
newsgroup with the respect they deserve, and they'll impress you every time.

-- Michael Chermside






From mgerrans at mindspring.com  Sat Jun 29 01:07:12 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Fri, 28 Jun 2002 22:07:12 -0700
Subject: I'm an idiot
References: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16> <Xns923BD424CF53Cdavidgriswold1netsca@204.127.68.17>
Message-ID: <afjfdg$mh2$1@slb7.atl.mindspring.net>

With such a small file, I don't know why no one mentioned this:

outfile = open( r'c:\path\output.txt', 'w' )
for line in open( r'c:\path\input.txt' ).readlines():
   outfile.write( line.strip() + '\n' )
outfile.close()

(or was I not paying attention?)

- Matt

"David Griswold" <david_griswold1 at yahoo.com> wrote in message
news:Xns923BD424CF53Cdavidgriswold1netsca at 204.127.68.17...
> Thank you, everyone, that took the time to help a newbie.
>
> David
>




From rajarshi at presidency.com  Thu Jun 13 16:14:03 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Thu, 13 Jun 2002 16:14:03 -0400
Subject: what is the return type of a SOAP function call?
Message-ID: <pan.2002.06.13.16.14.02.333921.29371@presidency.com>

Hi,
  I have a registered SOAP function which puts some data into  a
dictionary and returns the dictionary. However when the client recieves
this result from the function call, the return value is a string,  - the
dict seems to be mangled.

IS there any way to ocnvert a string represenation of a dict to a dict?
Or is there some special way to use the return value from a SOAP func
call?

TIA,


From janto_d at hotmail.com  Fri Jun 28 11:14:41 2002
From: janto_d at hotmail.com (Janto Dreijer)
Date: Fri, 28 Jun 2002 17:14:41 +0200
Subject: squared functions--most Pythonic way?
References: <mailman.1025207218.19740.python-list@python.org> <3D1C207A.74F24164@gol.ge>
Message-ID: <afhu82$15ed$1@news.adamastor.ac.za>

I don't have time to refine this, so I'm just throwing it out there. (I've
already spent 20min trying to post this through google)

def addNumbers(k):
    def f(x):
        a = addNumbers(x + k)
        a.val = x+k
        return a
    return f

>>> addNumbers(9)(5)(2)(4)(6).val
26

ooooh!!! Now _that's_ pretty :-)

Now if only I could figure out how to use __repr__() so I don't need that
".val"
It also fails when passed only one number. i.e addNumbers(5)
Help?


"Giorgi Lekishvili" <gleki at gol.ge> wrote in message
news:3D1C207A.74F24164 at gol.ge...
> One of the biggest advantages of Python is its nice and elegand mixture of
> procedural, OO and functional programming... I would prefer lambda stile
as the
> most concise, but it is a matter of personal "phylosophy";)
>
> Grtz,
> Giorgi
>
> Sean 'Shaleh' Perry wrote:
>
> > >
> > > functional programming:
> > >
> > >     # with an internal named function
> > >     def fsquare(f):
> > >         def f2(x):
> > >             return f(x)**2
> > >         return f2
> > >
> >
> > this seems to be the most pythonic (lambda is not the preferred approach
> > usually).
> >
> > Of course someone will mention taking this to the next step with a
compose()
> > function which take f(x) and g(x).
>




From rkr1410 at poczta.onet.pl  Sun Jun 23 11:20:42 2002
From: rkr1410 at poczta.onet.pl (hellboy)
Date: Sun, 23 Jun 2002 17:20:42 +0200
Subject: Newbie question (anyone programming with curses?)
Message-ID: <af4ovd$1bt3$1@pingwin.acn.pl>

Hello,

I'm quite new to Python, but I'm positive that version 2.2 (for Windows)
comes without curses module, though documentation states it is included.
import curses gives me "Import Error: no module named curses" message.
Anyone knows where to download curses module (for Windows)?

-- rkr




From peter at engcorp.com  Tue Jun 18 21:03:58 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 18 Jun 2002 21:03:58 -0400
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <D7NP8.46301$GY.14487785@twister.nyroc.rr.com>
Message-ID: <3D0FD87E.E7E62A85@engcorp.com>

George Hester wrote:
> 
> I just cannot believe what I ran into in this newsgroup.  If you people want to
> be paid for your suggestions then that is fine.  But this newsgroup does not
> have the facility set up for that to happen.  Then since it don't it is
> courteous not to give advice if that is what you really want. It almost seems
> that I was given half an answer so that the other half could be provided by
> someone that would try to make a buck off me.
> 
> Thank you Duncan for finally seeing the error of thier ways.
> 
> Don't worry I won't return here.  This is not a group to get help but a group of
> individuals that think the world owes them something when they participate here.

Troels Therkelsen kindly pointed George to a helpful page by guru Eric Raymond.  
I found the "On Not Reacting Like a Loser" section very apropos for this thread:

   Remember: When that hacker tells you that you've screwed up, and 
   (no matter how gruffly) tells you not to do it again, he's acting out 
   of concern for (1) you and (2) his community. It would be much easier 
   for him to ignore you and filter you out of his life. If you can't manage 
   to be grateful, at least have a little dignity, don't whine, and don't expect 
   to be treated like a fragile doll just because you're a newcomer with a 
   theatrically hypersensitive soul and delusions of entitlement.

-Peter


From warlock at eskimo.com  Fri Jun 21 13:11:23 2002
From: warlock at eskimo.com (Jim Richardson)
Date: Fri, 21 Jun 2002 10:11:23 -0700
Subject: converting an array of chars to a string
References: <3d12dda2_6@news.newsgroups.com> <3D12E228.6030501@wedgetail.com>
Message-ID: <rnmvea.r5f.ln@127.0.0.1>

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

On Fri, 21 Jun 2002 08:22:02 GMT,
 Derek Thomson <derek at wedgetail.com> wrote:
> JB wrote:
>> How can I convert an array of chars to string?
>> 
>> The problem is this: I should like to perform 8 bit 
>> calculations in a buffer. After this has been done, the 
>> buffer has to ba written to a binary file (and file.write 
>> takes strings instead of lists).
>> 
> 
> Easy, just use the "join" function to concatenate a sequence of strings 
> into a single string. For example:
> 
> 
> import string
> 
> list = ['a', 'b', 'c', 'd']
> 
> s = string.join(list, '')
> 
> print s
> 
> 
> Regards,
> Derek.
> 

huh! 

cool, thanks, that solved a problem I was having. 
<muttter> when did that show up? or was I not paying attention?</mutter>


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)

iD8DBQE9E147d90bcYOAWPYRAl91AJ44se3CAT5EoLfRoF9vUBVbdPuqVACg03Qs
HWQP8Sqoh+OMF1xSn2b28Ew=
=1+ZI
-----END PGP SIGNATURE-----

-- 
Jim Richardson
	Anarchist, pagan and proud of it
http://www.eskimo.com/~warlock
Linux, from watches to supercomputers, for grandmas and geeks. 


From donn at drizzle.com  Sun Jun  9 12:56:26 2002
From: donn at drizzle.com (Donn Cave)
Date: Sun, 09 Jun 2002 16:56:26 -0000
Subject: newbie system() question
References: None <3d035180$0$7017$bc7fd3c@news.sonofon.dk>
Message-ID: <1023641774.112216@yasure>

Quoth "Dan Larsen" <dl at bestpriceeu.com>:
| Im trying to run a script on a FreeBSD 4.0 system.
| It works as long as I start it like "python myprogram.py", but as soon as I
| run it in the background, it is stopped, and I have to make it run in the
| foreground to continue i.e. "fg %1"... I know it is a call to a system(""),
| that makes it stop... The system() call runs a program that outputs some
| text... how do I stop it doing so? It doesn't help telling it to do stuff
| like "program >& /dev/null &"... Isn't there a function, where all output
| from the command, is parsed to a variable?

Yes, there is.

Your background program stops because of a Berkeley job control
function in the tty driver (all modern UNIX implementations do the
same - as with sockets, everyone is Berkeley in this respect.)
To see exactly what happened, type "jobs" or "jobs -l", and you'll
see something like "[1] + Stopped(SIGTTOU)"  That means your background
job tried to write to the terminal driver, and got put to sleep.
Or it might say "(SIGTTIN)" instead, which means the job tried to read.
The latter is more or less inevitable, you don't want background jobs
reading input from the tty, but the former is optional:  try "stty tostop",
and see if it makes any difference.

The redirection syntax "program >& /dev/null &" won't help in system(),
because it isn't valid for the Bourne shell.  If you want to redirect
both standard and diagnostic output to /dev/null in the Bourne shell,
try "program >&2 /dev/null".

The shell notation for output to a variable is X="`program`".

	Donn Cave, donn at drizzle.com


From beej at piratehaven.org  Fri Jun 28 17:55:32 2002
From: beej at piratehaven.org (Beej Jørgensen)
Date: 28 Jun 2002 21:55:32 GMT
Subject: OT: Hayes AT commands (was Re: Dialing out on MODEM)
References: <mailman.1025270142.10532.python-list@python.org> <Xns923BE7FFF29D3cliechtigmxnet@62.2.16.82>
Message-ID: <afim0k0g71@enews1.newsguy.com>

In article <Xns923BE7FFF29D3cliechtigmxnet at 62.2.16.82>,
Chris Liechti  <cliechti at gmx.net> wrote:
>somtimes it it's a '\r\n' and maybe you need to send the escape sequence 
>first, to enter the command mode. if i remember correctly thats 3 times ESC 
>and then wait 100ms or so.

Not really python anymore, but it's actually '+++' (three plusses) if
_I_ remember correctly. 8-)  And I'm pretty sure it's longer than
100ms...but this is just my guess from waiting for the 'OK' after
sending.  Maybe 500-800ms.  In any case, 'OK' should appear after some
time of no activity after sending '+++'.

-Beej



From jonathan at onegoodidea.com  Sun Jun 30 15:44:22 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Sun, 30 Jun 2002 20:44:22 +0100
Subject: newbie Q: delete list few lines in file
References: <7fe97cc4.0206291519.1c95ce31@posting.google.com> <3D1E5123.FDEFADDF@engcorp.com> <3D1E83C0.8060700@earthlink.net> <3D1E8A08.53DD2272@engcorp.com>
Message-ID: <B9451E26.D35B%jonathan@onegoodidea.com>

On 30/6/2002 5:33, in article 3D1E8A08.53DD2272 at engcorp.com, "Peter Hansen"
<peter at engcorp.com> wrote:

> Hans Nowak wrote:
>> 
>> Peter Hansen wrote:
>> 
>>> You cannot write files in place.
>> 
>> <nitpick>
>> 
>> Sure you can:
> 
> [snip proof]
> 
> Of course, but we were answering a newbie's questions, right?

I think that it is unclear from Xah's original post whether the files or the
lines are to be deleted. Certainly, asking how to write to files in place
suggests that the lines are to be deleted.
 
> -trying-not-to-complicate-things-ly yr's,

If you want to say "that's complicated, are you sure you want to do that?",
go ahead. But it's not helpful to tell someone that something cannot be done
when it can. That's just misdirection.

-newbies-aren't-children-ly y'rs,

Jonathan



From jiba at tuxfamily.org  Sun Jun 16 11:54:21 2002
From: jiba at tuxfamily.org (Lamy Jean-Baptiste)
Date: Sun, 16 Jun 2002 17:54:21 +0200
Subject: What If.....    Strong Types
References: <3D0C0F0D.4030900@bgb.cc>
Message-ID: <20020616.175421.861584889.1250@localhost.localdomain>

> I'm not suggesting any real changes to any, only proposing a thought
> experiment.
> 
>    What if Python had been built from the start with strong types? And I
>    mean
> strong types, not strong typing. The distinction I am trying to draw is
> that variables would not be typed, but that all types would have rigid
> interfaces.
> 
>    Primarily, what if classes always had rigidly defined interfaces. I
>    mean
> that public members had to be declared to exist, and that methods
> couldn't be modified on an instance. Private members could exist but
> would be really private.

I've done really nice hack by changing object class, adding method or
properties, and so on... !

The object concept says that all the data of an object should be
encapsulated in a single location (=the object) rather that being spread
everywhere. Can you really speak about object if, for a given object, you
suddenly discover that you have a new property ? The right location for
that property is IN the object, so you need to add it INSIDE !

E.g. changing the class to a new one that inherit the former and add a
__setattr__ method can be used to get a callback for object modification.

There is already (too) many static language (Java, C, C++, ...) so i
believe Python must really stay dynamic !

BTW you can already have non-modifiable classes with __slots__.

Jiba


From thenault at nerim.net  Thu Jun 13 05:27:37 2002
From: thenault at nerim.net (Sylvain Thenault)
Date: 13 Jun 2002 11:27:37 +0200
Subject: How to call a function using apply with keyword args?
References: <ae9fud$u23$00$1@news.t-online.com> <mailman.1023956476.1179.python-list@python.org>
Message-ID: <87u1o7h6k6.fsf@tux.ntw23.fr>

Anoop P B <anoop79 at myrealbox.com> writes:

> hi Achim,
> 
> try this:
> 
> def MyFkt(A=0,B=0,C=0): pass
> params = {'A':5,'B':8,'C':9}
> MyFkt(params)
> 
> the reason behind the error is coz u can only pass a dictionary using
> 'apply' for keyword arguments. and keyword args are of the form
> keyword=value (see docs)

I don't think this give the desired result:

>>> def MyFkt(A=0,B=0,C=0):     
...     print 'A=',A
...     print 'B=',B
...     print 'C=', C
... 
>>> MyFkt(params)
A= {'B': 8, 'C': 9, 'A': 5}
B= 0
C= 0

Actually, you have to tell that your dictionnary is the param list and not the
value of the first param:


>>> MyFkt(**params)
A= 5
B= 8
C= 9
>>> apply(MyFkt, [], params)
A= 5
B= 8
C= 9


> Achim Domma wrote:
> > Hi,
> > I have a function like this:
> > def MyFkt(A,B,C): pass
> > and some parameters from a config file:
> > params = {'A':5,'B':8,'C':9}
> > how can I call MyFkt with this parameters? I tried
> > apply(MyFkt,[],params)
> > but get this error:
> > TypeError: MyFkt() takes exactly 3 non-keyword arguments (0 given)
> > any idea?
> > greetings
> > Achim

-- 
Sylvain Th?nault 


From thenault at nerim.net  Thu Jun 13 04:46:09 2002
From: thenault at nerim.net (Sylvain Thenault)
Date: 13 Jun 2002 10:46:09 +0200
Subject: How to check version of Python under CGI
References: <rbmggus7ln1uiutktf5libf4bbngu29ji7@4ax.com>
Message-ID: <873cvrin1q.fsf@tux.ntw23.fr>

Johann <observer at NOSPAM.space.pl> writes:

> I would like to write a simple CGI script to get as much info about
> installed Python as possible. But I even don't know how to check
> version of Python. :( From shell command  it is easy, but how to do it
> from CGI script? Eg.
> 
> #!/usr/bin/env python
import sys
> print "Content-type: text/html\n\n"
print "Python version: %s" % '.'.join([str(x) for x in sys.version_info[:3]])
or
print "Python version: %s" % sys.version


-- 
Sylvain Th?nault 


From whisper at oz.net  Tue Jun 11 10:40:27 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 11 Jun 2002 07:40:27 -0700
Subject: Problem with wxPython
In-Reply-To: <b7226ba.0206110618.272c3a98@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEKECLDIAA.whisper@oz.net>

Assuming that you're using the exact same distributions of python, wxwindows
and wxpython on all 4 computers, what might be different at the OS level? I
would make sure that the build levels on the C/C++ runtime dll's are the
same and that the same build and service packs for IE are on all 4
machines - IE has a way of quietly updating OS components without bothering
to mention it. Another potential problem is that the OS's may be not _quite_
the same - MS also has the habbit of changing what's on the install CD in
mid-production without mentioning it or changing the SKU. If none of those
apply, do all 4 machines have the same Intel or AMD brand CPU (not worried
about speed difference, only vendor). If you compiled on one brand with
brand-specific optimizations (they do exist - there are even CPU step
optimizations for a single chip!). This is somewhat of a wild shot though...

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Stephane Larouche
> Sent: Tuesday, June 11, 2002 7:19
> To: python-list at python.org
> Subject: Re: Problem with wxPython
>
>
> "Roman Yakovenko" <romany at actimize.com> wrote in message
> news:<mailman.1023629412.29376.python-list at python.org>...
> > Could you post small piece of code. I have some computer configuration
> > and it works fine to me. Also try Boa IDE. It is very helpful.
> >
> > Roman
>
> As I said in my previous message, I do run wxPython on the same
> configuration on 3 other computers. But, on one of my computers, the
> fastest one, wxPython crashes on wxFrame.__init__ if the style
> includes wxMAXIMIZE_BOX, wxRESIZE_BOX, wxSYSTEM_MENU or
> wxDEFAULT_FRAME_STYLE. For exemple, the wxPython demo crashes on
>
> wxFrame.__init__(self, parent, -1, title, size = (800, 600),
> style=wxDEFAULT_FRAME_STYLE|wxNO_FULL_REPAINT_ON_RESIZE)
>
> in the file Main.py. This demo works perfectly well on the other
> computers. Compiled software (with py2exe) also crashes.
>
> Maybe i'll try Boa, but I don't think it will solve the problem. The
> problem isn't the code. The code runs on every other computer I tried.
> The problem is probably coming from an incompatibility of wxPython
> with something else I have on my computer, I just can't figure out
> what! Anybody's help is welcome, I'm close to reformat my hard-drive
> and verify between every software installation if wxPython still works
> in order to identify the conflict.
>
> Thank you for your help!
>
> Stephane Larouche
> steph_lar at videotron.ca
> --
> http://mail.python.org/mailman/listinfo/python-list





From gerhard at bigfoot.de  Tue Jun 11 19:55:46 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 11 Jun 2002 23:55:46 GMT
Subject: lists and sequences
References: <mailman.1023833476.10249.python-list@python.org>
Message-ID: <slrnagd3g4.rtc.gerhard@lilith.my-fqdn.de>

Collin Monahan wrote in comp.lang.python:
> What is a python list? Is it a linked list or a random access structure
> in contiguous memory?

This type of questions can be easily answered by looking into its
implemenation (Objects/listobject.c).

> What type of situations should be avoided when using it?

Checking for containment using the 'in' operator.

> E.g. when does n^2 time happen using it? What sorting algorithm is
> used by its sort member function?

Read the Source, Luke. Looks like 'samplesort' (never heard of), and
quicksort are used, among others. IOW, it's pretty damn well
optimized.

> Are the answers to these questions the same for Jython?

Read the Source, Luke. Looks like it's a combination of quicksort and
insertion sort.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From christophe.delord at free.fr  Wed Jun 19 13:07:44 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Wed, 19 Jun 2002 19:07:44 +0200
Subject: how tranform in String??
References: <mailman.1024437202.15280.python-list@python.org>
Message-ID: <20020619190744.4e7d9a90.christophe.delord@free.fr>

Hi,

You may want to use split.

For example:

In [1]: s='xavante brasil'

In [2]: s=s.split()

In [3]: s
Out[3]: ['xavante', 'brasil']

In [4]: s='xavante,brasil'

In [5]: s=s.split(',')

In [6]: s
Out[6]: ['xavante', 'brasil']

In [7]:


Christophe.


On Tue, 18 Jun 2002 18:49:37 -0300
jubafre at zipmail.com.br wrote:

> i have s='xavante brasil', i want to transform in
> 
> s='xavante', 'brasil' with two indexes
> 
> how i make this???????????
> 
> 
> Juliano Freitas
> www.gebrasil.hpg.com.br
> 
> 
> 
> ------------------------------------------
> Use o melhor sistema de busca da Internet
> Radar UOL - http://www.radaruol.com.br
> 
> 
> 
> 
> 


-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From robin at jessikat.fsnet.co.uk  Thu Jun  6 08:00:39 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Thu, 6 Jun 2002 13:00:39 +0100
Subject: inspect under 2.1.3
Message-ID: <pwrrrKAn70$8EwXT@jessikat.demon.co.uk>

is there any hope of getting inspect to work reasonably under 2.1. In
2.2 I see lots of methods from the following, but in 2.1 it's empty. 

C:\Python\rlextra\ers>python
Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import inspect
>>> import framework
>>> b=framework.BasicProcessor()
>>> inspect.getmembers(b)
[('__doc__', None), ('__module__', 'framework'), ('actionEcho', <bound
method BasicProcessor.actionEcho of <framework.Ba
sicProcessor instance at 0x007D90A8>>), ('actionPdf', <bound method
BasicProcessor.actionPdf of <framework.BasicProcesso..........






C:\Python21\rlextra\ers>..\..\python.exe
Python 2.1.3 (#35, Apr  8 2002, 17:47:50) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> import inspect
>>> import framework
>>> b=framework.BasicProcessor()
>>> inspect.getmembers(b)
[]
>>> exit
'Use Ctrl-Z plus Return to exit.'
>>> ^Z



-- 
Robin Becker


From maxm at mxm.dk  Wed Jun 12 17:27:29 2002
From: maxm at mxm.dk (Max M)
Date: Wed, 12 Jun 2002 23:27:29 +0200
Subject: PIL _imaging
References: <Xns922AA829A8C5FHotMailcs@130.133.1.4> <3D06E694.4040103@mxm.dk> <Xns922B3150A6B32HotMailcs@130.133.1.4> <3D074529.E848AA6@ipm.fhg.de> <Xns922B561EDB212HotMailcs@130.133.1.4>
Message-ID: <3D07BCC1.2000002@mxm.dk>

Cousin Stanley wrote:
> Cousin Markus ... 

> I tried with and without K:\Python\DLLs 
> in the PYTHONPATH variable and without 
> PYTHONPATH set at all ...


Perhaps you have more than one version of Python installed and you are 
using different versions?

regards Cousin M



From tjreedy at udel.edu  Mon Jun  3 12:31:10 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 03 Jun 2002 16:31:10 GMT
Subject: Efficient posting-list
References: <adffdk$10mep9$1@ID-146704.news.dfncis.de> <3iKK8.99777$%y.10468246@bin4.nnrp.aus1.giganews.com> <adfu52$10t9pp$1@ID-146704.news.dfncis.de>
Message-ID: <iRMK8.113037$%o.10726820@bin3.nnrp.aus1.giganews.com>

"Shagshag13" <shagshag13 at yahoo.fr> wrote in message
news:adfu52$10t9pp$1 at ID-146704.news.dfncis.de...
> By now i use a node with : 2 int and 1 float. Node objects are
handled
> in a python list. ...(i can't use a tuple, i need to update values)

I would think that the data for a given keyword in a given document
would be fixed ... or do docs change without changing id?

> > For pure Python, tuple is most space efficient.  If
> > each node really is (integer id, integer count), one could devise
a
> > system using a pair of array (module) objects or 2-dimensional
> > numerical python arrays; which store actual integers rather than
> > pointers to int PyObjects.  When filled, they would have to be
copied
> > to larger arrays in much the same manner as done automatically by
> > Python lists.
>
> So here you think that i should use only one list  containing :
>
> [id_x, count_x, float_x, id_x+1, count_x+1, float_x+1, ...]
>
> and so on ?

With modification, yes.  To reword a bit: Python is a bit proliflagate
with space, especial when boxing numbers.  A tuple of three numbers +
the three number objects take up roughtly ten times as much space as
the raw numbers themselves.  A list is a bit worse.  So, to save space
and (usually) time when working with 'lots' of numbers, one needs to
switch to (boxed) arrays of raw (unboxed) numbers.  I believe both the
array and numpy modules make such available but I have not yet worked
with either, so I will let you explore.  With your additional
information, I suggest that a node be a list or class instance
wrapping three arrays - idx, countx, floatx  - which you copy/extend
as needed.

Terry J. Reedy





From rotundo52 at hotmail.com  Wed Jun 12 00:36:28 2002
From: rotundo52 at hotmail.com (Jon J. Morin)
Date: Wed, 12 Jun 2002 04:36:28 GMT
Subject: newbie: copying dictionaries
Message-ID: <gdAN8.8432$nn1.3195949@news1.news.adelphia.net>

Hi.  I'm a newbie to python and have a question about copying dictionary 
objects.  I'm working out of Learning Python by Mark Lutz & David Ascher 
(O'Reilly).  The following bit of code was an exercise in the book and it 
works, but what I'm asking is why?  I see where the dictionary keys get 
copied from one object to the other, but how do the values get copied over 
as well?

#!/usr/bin/python
# copyDict.py - copy a dictionary argument and return it

def copyDict(aDict):
    newDict = {}
    for key in aDict.keys():
        newDict[key] = aDict[key]
    return newDict

D = {'jon':1, 'allison':2, 'zachary':3}
print copyDict(D)

Output:
[jmorin at cisx1 python]$ ./copyDict.py
{'zachary': 3, 'jon': 1, 'allison': 2}

Any help with this is appreciated.  Thanks.

Jon J. Morin


From brueckd at tbye.com  Wed Jun 26 15:18:21 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 26 Jun 2002 12:18:21 -0700 (PDT)
Subject: Installation problems DCOracle2 & W2000
In-Reply-To: <talon34-E85D1F.10285926062002@netnews.attbi.com>
Message-ID: <Pine.LNX.4.44.0206261211140.29594-100000@ns0.tbye.com>

On Wed, 26 Jun 2002, Talon wrote:

> OK, here you go:
> 
> Path = C:\Python21\;  (I don't have a specific PYTHONPATH variable set. 
> This is in the general PATH variable)
> 
> The error I get is
> 
> File "C:\Python21\myfiles\htsa_data_loader_v2.py", line 9, in ?
>   import os,sys,re,string,DCOracle2
> File "C:\Python21\DCOracle2\__init__.py", line 91, in ?
>   import DA
> File "C:\Python21\DCOracle2\DA.py", line 90, in ?
>   from db import DB
> File "C:\Python21\DCOracle2\db.py", line 89, in ?
>   import DCOracle2, DateTime
> ImportError: No module named DateTime
> 
> DateTime is located in C:\Program Files\ZopeDev\lib\python\DateTime

Aha! Here's what I think the problem is: the DCOracle2 bundle off the Zope 
site comes all wrapped up in the Zope database adapter (ZOracleDA), so 
your above script is actually trying to import the DA and not the 
DCOracle2 package itself (I can see why this is confusing - both 
directories are named DCOracle2 but one is a subdirectory of the other!).

So... if you look in c:\python21\DCOracle2 you should find a subdirectory 
in there also called DCOracle2. This is the only one you should need for 
doing standalone (none-Zope) work with Oracle. The easiest thing to do is 
rename c:\python21\DCOracle2 to DCOracle2bak or something, and then move 
its DCOracle2 subdirectory into c:\python21 and then try your script 
again. 

Does this make sense? Please let me know if it doesn't or if it still 
doesn't work after that.

-Dave





From msoulier at nortelnetworks.com  Thu Jun 20 14:53:18 2002
From: msoulier at nortelnetworks.com (Michael P. Soulier)
Date: Thu, 20 Jun 2002 14:53:18 -0400
Subject: do...while
In-Reply-To: <XFMail.20020620111603.shalehperry@attbi.com>
References: <20020620181230.GB2402@nortelnetworks.com> <XFMail.20020620111603.shalehperry@attbi.com>
Message-ID: <20020620185318.GC2402@nortelnetworks.com>

On Thu, Jun 20, 2002 at 11:16:03AM -0700, Sean 'Shaleh' Perry wrote:
> 
> not every language has had a do .. while construct.  You code with the tools
> given.

    If we always just code with the tools given, Python would never have been
born. 

    Cheers,

    Mike

-- 
Michael P. Soulier, QX41, SKY  Tel: 613-765-4699 (ESN: 39-54699)
Optical Networks, Nortel Networks, SDE Pegasus
"...the word HACK is used as a verb to indicate a massive amount
of nerd-like effort."  -Harley Hahn, A Student's Guide to Unix




From aahz at pythoncraft.com  Sat Jun 22 12:35:05 2002
From: aahz at pythoncraft.com (Aahz)
Date: 22 Jun 2002 12:35:05 -0400
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <3d149f19$0$53133$edfadb0f@dspool01.news.tele.dk>
Message-ID: <af28vp$j57$1@panix1.panix.com>

In article <3d149f19$0$53133$edfadb0f at dspool01.news.tele.dk>,
Frihtiof Andreas Jensen <nomailhere at lookitup.yourself.com> wrote:
>
>My bet is that once you cool off and sits down with some debugging tools
>like GDB, Boundschecker or Electric Fence you will indeed find the
>programming error and fix it.
>
>My bet is also that you will not tell anyone either.

Ya know, you could have saved yourself a lot of typing by simply saying
PBCAK.  ;-)
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From dvnu at mega.ist.utl.pt  Wed Jun 19 13:01:19 2002
From: dvnu at mega.ist.utl.pt (David Nunes)
Date: 19 Jun 2002 10:01:19 -0700
Subject: python __main__ is similar to java main()
Message-ID: <3015ca5e.0206190901.3cb0b726@posting.google.com>

Well I've read the docs and searched in the newsgroup archives but
I'm still confused about __main__

        I  expected it to startup the program, like main in java but
when I feed it to the interpreter he just "eats" the file loads wathever
is in the file but doesn't run main.

        So am I suposed to explicitly call it like in the file below or is
there a better (pythonic) way to do it??

        TIA

        David Nunes


import string
import PedLan
import GLOBAL


class Lancamento:
    ReqID = 0

    #PDNConsulta

    def ___init__(self):
        #inicializacao do PDNConsulta
        pass

    def trataPedido(self):#nao faz verificacao de input
        linha = raw_input('>')
        parse = string.splitfields(linha, GLOBAL.INPUTSEP)
        disc = parse[0]
        nomes = []
        notas = []
        for indice in range(1,len(parse), 2):
            nomes.append(parse[indice])
        for indice in range(2,len(parse), 2):
            notas.append(parse[indice])
        pedido = PedLan.PedLan(ReqID, nomes, notas)
        #<PDNConsulta.Lancamento(pedido)

    def __main__(self):
        print('running')
        interface = Lancamento()
        while(1):
            interface.trataPedido()

Lancamento().__main__() #to run main ????


From chris at cmb-enterprises.com  Sat Jun  1 15:30:51 2002
From: chris at cmb-enterprises.com (Chris Dutton)
Date: Sat, 01 Jun 2002 15:30:51 -0400
Subject: Python/Scripting language performance
References: <3CF65A62.2070705@bioeng.ucsd.edu> <mailman.1022812010.18367.python-list@python.org>
Message-ID: <chris-F10481.15305001062002@corp.supernews.com>

In article <mailman.1022812010.18367.python-list at python.org>,
 Skip Montanaro <skip at pobox.com> wrote:

> I believe the Parrot project is also working on a more efficient virtual
> machine for Perl (and eventually other languages).

Just as a small note, Parrot is being developed from the start to work 
with languages besides Perl6, though that understandably is the 
emphasis.  Whether language designers target the Parrot platform is 
their choice.

Already a number of languages are being developed for Parrot.

Jako : C-ish
Cola : C#-type language
Scheme
BASIC
miniperl
And of course ou can always write programs in Parrot Assembly.


From peter at engcorp.com  Fri Jun  7 08:12:57 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 07 Jun 2002 08:12:57 -0400
Subject: common approach in socket programming
References: <MPG.176a98d24f0689e1989680@news.rz.uni-karlsruhe.de>
Message-ID: <3D00A349.5E4091EA@engcorp.com>

Uwe Mayer wrote:
> 
> now my problem is how to know that
> 
> a) a connection is still open
> 
> i couldn't find a opened() or isopen() function or a similar attribute
> in the documentation. not for the socket object. for the file object
> there should be the open attribute.

It's open until it is closed, at which point a blocking recv() will 
return "" (once).

> b) how do i know the size of each packet?

You do not!  There are no guarantees in TCP/IP that the original
send() arrives in a single recv().  Data can be split, combined,
delayed in many ways.  You _must_ have some kind of termination
sequence (e.g. \r\n) to identify the end of a sequence, or you
must send the length ahead of time so you know how much more data
to read (e.g. "net strings").

Unless you are trying to do some really low level stuff where
you actually care about the individual packets, in which case 
you do not care about SocketServer (or probably sockets, either).

-Peter


From jepler at unpythonic.net  Fri Jun 28 16:16:16 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Fri, 28 Jun 2002 15:16:16 -0500
Subject: Introspection: determining parent classes?
In-Reply-To: <3D1CC1CA.70004@onsitetech.com>
References: <3D1CC1CA.70004@onsitetech.com>
Message-ID: <20020628201610.GB14119@unpythonic.net>

On Fri, Jun 28, 2002 at 08:07:46PM +0000, Robb Shecter wrote:
> Hi,
> 
> Is there any way to determine the parent classes given an instance?
> 
> I'm writing unit test code, and I want to assert that a particular 
> result is a subclass of some other class.  In Java, you can do something 
> like:
> 
> anInstance.getClass().isSubclassOf( aClass )
> 
> I looked through the type(), __class__, etc. info, but these seem to 
> just return string-like objects that don't allow further introspection.

They're not string-like objects.

>>> x = 1
>>> type(x)
<type 'int'>
>>> type(type(x))
<type 'type'>
>>> help(isinstance)
Help on built-in function isinstance:

isinstance(...)
    isinstance(object, class-or-type-or-tuple) -> bool
    
    Return whether an object is an instance of a class or of a subclass thereof.
    With a type as second argument, return whether that is the object's type.
    The form using a tuple, isinstance(x, (A, B, ...)), is a shortcut for
    isinstance(x, A) or isinstance(x, B) or ... (etc.).
>>> help(issubclass)
Help on built-in function issubclass:

issubclass(...)
    issubclass(C, B) -> bool
    
    Return whether class C is a subclass (i.e., a derived class) of class B.
>>> isinstance(x, type(x))
True
>>> isinstance(x, type(3.14159))
False

Jeff




From mfranklin1 at gatwick.westerngeco.slb.com  Tue Jun 11 09:32:08 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Tue, 11 Jun 2002 13:32:08 +0000
Subject: Making an About Box...
In-Reply-To: <Xns9229BE9B416A2endymion84charternet@216.168.3.40>
References: <Xns9229BE9B416A2endymion84charternet@216.168.3.40>
Message-ID: <200206111235.g5BCZgB26749@helios.gatwick.geco-prakla.slb.com>

On Monday 10 Jun 2002 11:44 pm, you wrote:
> Hello.  I'd like to make an About Box using Tkinter.  I'm using a grid
> inside a Toplevel widget, but I don't really like this solution.  I'd like
> a window that did not have the minimize and maximize buttons, like the
> tkMessageBox derived boxes showinfo, showwarning, etc.  Is it possible to
> make a window that has this type of title bar and put my own widgets in it?


Take a look at the SimpleDialog.py file in the standard distribution 
(python2.2/lib-tk)  It uses  the transient() method that I'm not sure I 
understand (but thats just me!) anyhow it should be enough to get you started

Martin




From bernie at 3captus.com  Tue Jun 11 12:35:20 2002
From: bernie at 3captus.com (Bernard Yue)
Date: Tue, 11 Jun 2002 16:35:20 GMT
Subject: tk gui for python
References: <e0b1d6f6.0206110802.7e38179c@posting.google.com>
Message-ID: <3D06251D.979A2CE@3captus.com>

Martin Krallinger wrote:
> 
> Dear all,
> 
> I would appreciate any sample code to generate tk gui for python.
> i just need to load a file, of past instead a text (protein sequence)
> and run on it a c programm and return the content of it's output
> 
> any sample code is welcomed,
> 

Try http://www.python.org/doc/life-preserver/.

> thanks
> 
> Martin

Bernie


From osuchw at ecn.ab.ca  Fri Jun 14 16:58:30 2002
From: osuchw at ecn.ab.ca (waldekO)
Date: 14 Jun 2002 13:58:30 -0700
Subject: Python and ADSI
References: <slrnagjuud.15g.gerhard@lilith.my-fqdn.de>
Message-ID: <c5b3dbee.0206141258.742d5565@posting.google.com>

I am not sure if that is what you want but I am using something
similar to access user on WinNT domain

from win32com.client import Dispatch
adsi = Dispatch('ADsNamespaces')
nt = adsi.GetObject('','WinNT:')
for domain in nt:
 	domain.Filter = ("User",)
 	for usr in domain:
 		print domain.name, usr.name, usr.fullname

In your case you probably want:
ldap = adsi.GetObject('','LDAP://correct ldap path goes here') and
then you would do your stuff.
http://www.microsoft.com/ntserver/downloads/bin/nts/adsi25.chm
should get you started

waldekO

Gerhard Haering <gerhard at bigfoot.de> wrote in message news:<slrnagjuud.15g.gerhard at lilith.my-fqdn.de>...
> Does anybody have an example script using win32com and ADSI to access the users
> in an active directory?
> 
> Gerhard


From xx758 at cam.ac.uk  Fri Jun  7 11:52:52 2002
From: xx758 at cam.ac.uk (Xiao-Qin Xia)
Date: Fri, 07 Jun 2002 16:52:52 +0100
Subject: Detect events positively
Message-ID: <adqkna$67o$1@pegasus.csx.cam.ac.uk>

Hello,

The following code display two button:

from Tkinter import *
def Calculate(event = None):
        while 1:
                pass
def Exit(event = None):
        import sys
        sys.exit(0)
root = Tk()
btn_cal = Button(root, text="calculate", command = Calculate)
btn_cal.pack()
btn_exit = Button(root, text="exit", command = Exit)
btn_exit.pack()
root.mainloop()

if clicked the "calculate" button, the program will fall in a loop that 
will never end. So program will not response to clicking to the "exit" 
button, so it cannot be terminated normally. If we don't use multithread, 
can we add some code in the loop of the function Calculate to let it exit 
the loop after the user clicks the "exit" button or press a "Esc" key?

Many thanks from, 
Xiao-Qin Xia


From bokr at oz.net  Sun Jun 30 14:37:04 2002
From: bokr at oz.net (Bengt Richter)
Date: 30 Jun 2002 18:37:04 GMT
Subject: arguments to functions
References: <MPG.178955a4fc8bfc1c989680@news.rz.uni-karlsruhe.de>
Message-ID: <afnj4g$4l5$0@216.39.172.122>

On Sun, 30 Jun 2002 18:52:10 -0700, Uwe Mayer <merkosh at hadiko.de> wrote:

>hi,
>
>I wanted to pass one named and a collection of unnamed (an arbitrary 
>number) to a function:
>
>def test(expected=None, *misc):
>    print(expected)
>    print(misc)
>
>Now, unluckyly 'test' interprets the first unnamed parameter to test as 
>the value for 'expected':
>
>test(5, expected=10)
>TypeError: test() got multiple values for keyword argument 'expected'
>
>on the other side any keyword arguments must follow non-keyword 
>arguments:
>
>test(expected=5, 10)
>SyntaxError: non-keyword arg after keyword arg
>
>Which is a pitty, because I think this version is clear.
>
>Is that *it*, or am I doing something wrong here?
>
There is something wrong. If you want to call with keyword args, you have to provide
a keyword dict arg (a name at the end of the arg list prefixed with **) for receiving
them, and you didn't.

I think the multiple values message is misleading, though, since keywords
are in a separate name space (a dict parameter), and I don't see why you
couldn't have equivalent names (other than implementation inconvenience)
to those in the named parameter list. In fact, it could be convenient
for implementing selective override of defaults.

Nut not only is the error message misleading, it's the message is generated
before checking whether _any_ keyword arguments _at all_ are allowed.
Note what happens with another kw name:

 >>> test(10, other=5)
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
 TypeError: test() got an unexpected keyword argument 'other'
                          ^^^^^^^^^^

Which shows the real error in the call. You could fix that part:
 >>> def test(expected=None, *misc, **kwargs):
 ...     print expected
 ...     print misc
 ...     print kwargs
 ...
 >>> test(10, 11, 12, other=5)
 10
 (11, 12)
 {'other': 5}

But unfortunately, it still won't let you use 'expected' as keyword for kwargs:

 >>> test(10, 11, 12, expected=5)
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
 TypeError: test() got multiple values for keyword argument 'expected'

I'd be inclined to call that a bug, actually.

You could get what I guess is the effect you want (keyword value takes priority) something like:

 >>> def test(_expected=None, *misc, **kwargs):
 ...     expected = kwargs.get('expected',None) or _expected
 ...     print 'expected:%s, misc:%s, kwargs:%s' % (`expected`,`misc`,`kwargs`)
 ...
 >>> test()
 expected:None, misc:(), kwargs:{}
 >>> test(10)
 expected:10, misc:(), kwargs:{}
 >>> test(10, expected=5)
 expected:5, misc:(), kwargs:{'expected': 5}
 >>> test(expected=5)
 expected:5, misc:(), kwargs:{'expected': 5}
 >>> test(10, 11, 12, expected=5)
 expected:5, misc:(11, 12), kwargs:{'expected': 5}
 >>> test(10, 11, 12)
 expected:10, misc:(11, 12), kwargs:{}

 >>> test(10, other=5)
 expected:10, misc:(), kwargs:{'other': 5}

If you want to check first that kwargs is either empty or has only
'expected' as a key, that's easy too.

Regards,
Bengt Richter


From shalehperry at attbi.com  Fri Jun 28 10:55:20 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 28 Jun 2002 07:55:20 -0700 (PDT)
Subject: squared functions--most Pythonic way?
In-Reply-To: <3D1C207A.74F24164@gol.ge>
Message-ID: <XFMail.20020628075520.shalehperry@attbi.com>

On 28-Jun-2002 Giorgi Lekishvili wrote:
> One of the biggest advantages of Python is its nice and elegand mixture of
> procedural, OO and functional programming... I would prefer lambda stile as
> the
> most concise, but it is a matter of personal "phylosophy";)
> 

in recent discussions (around a month ago) it was revealed that the powers that
be question lambda's existence in python and if they had it to do over again
may not have included it.  No they do not plan to remove it now (-:




From jiba at tuxfamily.org  Thu Jun 20 15:18:26 2002
From: jiba at tuxfamily.org (Lamy Jean-Baptiste)
Date: Thu, 20 Jun 2002 21:18:26 +0200
Subject: [ANN] Tree widget for Tkinter
Message-ID: <mailman.1024601110.13018.clpa-moderators@python.org>

TreeWidget 0.2

http://oomadness.tuxfamily.org/p-treewidget.php

This is a tree widget for Tkinter, in pure-Python.

It is inspired on Guido's IDE tree widget, but adds some features :
 - highly optimized (supports more than 10000 nodes !)
 - asynchronous rendering
 - Linux-friendly look-and-feel (and PGM icons instead of gif)
 - inheritance-based interface (inherit the Node class)
 - ...

The same package also provides a scrollpane, that can be used to add
scrollbars for a Tkinter Frame.

For documentation, look at the doc strings and the 2 demo scripts !

TreeWidget is Free Software and is available under the GNU GPL.

Enjoy the tree !

Jiba -- Jean-Baptiste LAMY -- jiba at tuxfamily.org
                              http://oomadness.tuxfamily.org




From db3l at fitlinxx.com  Wed Jun  5 16:25:38 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 05 Jun 2002 16:25:38 -0400
Subject: self
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu> <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com> <mailman.1023294728.2322.python-list@python.org>
Message-ID: <uy9dt79rx.fsf@ctwd0143.fitlinxx.com>

holger krekel <pyth at devel.trillke.net> writes:

> the problem is that the values in the dict have to be 'evaluated' if
> accessed.  For every name in such a value you need to evaluate its
> value again.  IOW it's recursive. The obvious solution is to
> intercept the namevalue-lookups during evaluation of an
> equation. But this doesn't work for exec/eval: They ignore the
> interception points.

Right, so even if you use Python 2.2 and subclass dict, while you'll
get the first lookup properly the recursive lookups within the eval go
directly to the dictionary.

I think your earlier solution posted here, which was to start with an
empty dictionary for the eval, and then trap lookup errors and slowly
populate the dictionary by evaluating the needed symbol, looked the
most promising.  It ends up doing a bunch of eval() calls, but using
the exception processing in lieu of needing the __getattr__ override
on the dictionary.  I think it might still need some checking to
avoid infinite loops in cyclical references though.

> IOW you cannot successfully wrap or inherit a dictionary type and pass it 
> to exec/eval :-(

Yes, I did also find a posting by Guido (didn't save the reference,
but it was via a google search) where he seemed to indicate that
subclassing list/dict other than add additional functionality (rather
than override core operations like lookups) could be problematic -
presumably because of cases like this where the special Python-exposed
functions are used.  The message was during one of the later betas of
2.2, so I don't know if the viewpoint changed before final release.

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From hwcowan at hotmail.com  Fri Jun  7 10:54:58 2002
From: hwcowan at hotmail.com (Hugh Cowan)
Date: 7 Jun 2002 07:54:58 -0700
Subject: Creating Dynamic Web Pages (ZOPE)
References: <46ca81a0.0206062016.789d3af6@posting.google.com> <mailman.1023437832.11588.python-list@python.org>
Message-ID: <46ca81a0.0206070654.787ac0f7@posting.google.com>

Nicola,

Thanks for the reply and all the good information and links.  As I
mentioned to Mike, I initially just want to make sure that I wasn't
completely wasting my time.  I don't have a lot of time to research /
learn various tools and options before I need to start implementing a
solution and therefore I really wanted to minimize the time I spent
researching and looking at various solutions instead of implementing
one.

Thanks for the tips on using the ZOPE mailing lists.  I will sign-up
as a user and post there.  I also appricate the advice in setting up
access to MS-SQL Server and also using the "DataQueryKit Product".

You have not only helped in making my decision easier (not that it
seems I need to decide in the end) but also given me excellant
starting points.

Management here is not exactly that computer literate and don't really
understand Open Source software (like a lot of people) and as we run
primarily a Microsoft shop it's good to know in advance that ZOPE will
work nicely with what's already in place (I know that the Tech guys
will raise concerns for sure).

Thanks again,

Hugh

P.S. Thanks also for the weekend reading -- I will grab the ZOPE book
and look through it this weekend!!


From emile at fenx.com  Wed Jun  5 09:19:02 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 05 Jun 2002 13:19:02 GMT
Subject: python cookbook at oreilly's web site
References: <dCnL8.5$566.1250@news.ecrc.de> <adl2nq$3hrm$1@ID-41653.news.dfncis.de>
Message-ID: <adoL8.111072$352.5618@sccrnsc02>

Tobias Klausmann
> Markus Jais <mjais at web.de> wrote:
> > just found
> > 
> > http://www.ora.com/catalog/pythoncook/
> 
> Does this have any relation to ASPNs Python cookbook?
> http://aspn.activestate.com/ASPN/Cookbook/Python
> 


Yes.  

--

Emile van Sebille
emile at fenx.com

---------



From yvon.thoraval at free.fr  Sat Jun  8 02:27:30 2002
From: yvon.thoraval at free.fr (Yvon Thoraval)
Date: Sat, 8 Jun 2002 08:27:30 +0200
Subject: Tkinter help on MacOSX please.
References: <B9266129.6930%sarmstrong13@mac.com>
Message-ID: <1fdga3f.1kt9ridf2v94yN%yvon.thoraval@free.fr>

SA <sarmstrong13 at mac.com> wrote:

> I've installed python, tcl, tk on my MacOSX 10.1.5 system using the
> --enable-frameworks configuration. I have a framework for tcl,tk,and python
> under my /Library/Frameworks directory.

i've read somewhere that tkinter isn't - yet - compatible with Carbon ,
don't know if that apllies to os X but to os 9 for sure !


From deltapigz at telocity.com  Thu Jun 27 12:40:56 2002
From: deltapigz at telocity.com (Adonis)
Date: Thu, 27 Jun 2002 12:40:56 -0400
Subject: SimpleHTTPServer + Multithreading
Message-ID: <3d1b4010$1_3@nopics.sjc>

i was googling around and was able to come up with cannibalized code, and
was wondering if there was anyway to make it more "clean" looking, or as it
is, is the best way:

(main idea: a multithreading SimpleHTTPServer)

import BaseHTTPServer, SimpleHTTPServer, SocketServer, socket, webbrowser

class XFSPWebServer(SocketServer.ThreadingMixIn, BaseHTTPServer.HTTPServer):
    pass

class ReqHandler(SimpleHTTPServer.SimpleHTTPRequestHandler):

    def do_POST(self):
        data = self.rfile.read(int(self.headers["content-length"]))
        data = data.split('&')[0].split('=')[1]
        self.wfile.write("HIH")

PORT = 80

httpd = XFSPWebServer((socket.gethostname(), PORT), ReqHandler)

httpd.serve_forever()

any help would greatly be appreciated.

Adonis




From shalehperry at attbi.com  Tue Jun  4 14:38:33 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 04 Jun 2002 11:38:33 -0700 (PDT)
Subject: os.path.expanduser ignores os.seteuid
In-Reply-To: <20020604182255.97926.qmail@web10306.mail.yahoo.com>
Message-ID: <XFMail.20020604113833.shalehperry@attbi.com>

On 04-Jun-2002 Graham Guttocks wrote:
> Why doesn't os.path.expanduser pay attention to the fact that I've
> changed the process's effective user id?
> 
> As root:
> 
> Python 2.1.3 (#1, Apr 11 2002, 18:30:30) 
> [GCC 2.95.3 20010315 (release) [FreeBSD]] on freebsd4
> Type "copyright", "credits" or "license" for more information.
>>>> import os
>>>> os.seteuid(667)
>>>> os.path.expanduser("~")
> '/root'
>>>> 
> 
> Other functions like os.getlogin() don't have this problem.
> 

def expanduser(path):
    """Expand ~ and ~user constructions.  If user or $HOME is unknown,
    do nothing."""
    if path[:1] != '~':
        return path
    i, n = 1, len(path)
    while i < n and path[i] != '/':
        i = i + 1
    if i == 1:
        if not os.environ.has_key('HOME'):
            return path
        userhome = os.environ['HOME']
    else:
        import pwd
        try:
            pwent = pwd.getpwnam(path[1:i])
        except KeyError:
            return path
        userhome = pwent[5]
    if userhome[-1:] == '/': i = i + 1
    return userhome + path[i:]

is from python 2.2.  It uses short circuit eval for the '~' case and simply
reads the environment $HOME.  If you could pass '~user/' it would give you the
answer you want.




From sholden at holdenweb.com  Thu Jun  6 09:51:02 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 6 Jun 2002 09:51:02 -0400
Subject: HttpServer help!
References: <3CFF4C14.450850D8@millfilm.co.uk> <P1JL8.312164$%u2.176575@atlpnn01.usenetserver.com> <3CFF6568.7EE613CE@millfilm.co.uk>
Message-ID: <ERJL8.313108$%u2.202230@atlpnn01.usenetserver.com>

"Eric Texier" <erict at millfilm.co.uk> wrote in message
news:3CFF6568.7EE613CE at millfilm.co.uk...
> The server and the client run from the same directory.
> and the cgi file is in the same directory.
>
> all the following failled.
> req.putrequest("POST", "favquote.cgi")
> req.putrequest("POST", "/favquote.cgi")
> req.putrequest("POST", "./favquote.cgi")
>
>
> What is the meaning of  *relative to the server root*
>
Well, I was kind of assuming that the server was written to run CGI scripts,
but I think on further examination I was wrong. The "server root" is the
directopry that the server looks in when you request a document like
http://server:port/index.html -- it's the top-level content directory of a
web server, which may have subdirectories, and so on.

>From your server code:

# start the server on port 2112, requires browser URLs
# to be addressed as http://servername:2112/pathtofile
myServer = HTTPServer(('', 2112), myRequestHandler)


This has one obvious implictaion you seem to have overlooked. Your call in
the client

    req = HTTP("10.16.3.132")

is trying to connect to the default port, port 80. You may need to change
this to

    req = HTTP("10.16.3.132:2112")

It looks, from the code of the server, as though *any* POST request will
result in the same output, so the path you request maybe isn't that
important. Although I haven't read Fred's book yet, I am guessing that the
.putrequest() method assumes you are already connected to the server, and
that the HTTP() object will allow you to put a port number after the IP
address.

If this isn't the case, try writing an HTML page containing a form whose
action attribute is something like

    ACTION="10.16.3.132:2112/something"

and see if submitting that displays a response in the browser.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------





From emile at fenx.com  Wed Jun 12 00:44:41 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 12 Jun 2002 04:44:41 GMT
Subject: newbie: copying dictionaries
References: <gdAN8.8432$nn1.3195949@news1.news.adelphia.net>
Message-ID: <ZkAN8.158536$352.7787@sccrnsc02>

Jon J. Morin
> The following bit of code was an exercise in the book and it
> works, but what I'm asking is why?  I see where the dictionary keys
get
> copied from one object to the other, but how do the values get copied
over
> as well?

> def copyDict(aDict):
>     newDict = {}
>     for key in aDict.keys():
>         newDict[key] = aDict[key]

This statement does both:

    newDict[key] ... creates key in newDict

and

    = aDict[key] ... places the object pointed at in it


>>> d = {1:11,2:22,3:33}
>>>
>>> d[1]
11

HTH,

--

Emile van Sebille
emile at fenx.com

---------



From pyth at devel.trillke.net  Sun Jun 30 13:13:41 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 30 Jun 2002 19:13:41 +0200
Subject: Parsing
In-Reply-To: <3D14CEBC.C6655F86@vip.fi>; from krissepu@vip.fi on Sat, Jun 22, 2002 at 10:23:40PM +0300
References: <3D14CEBC.C6655F86@vip.fi>
Message-ID: <20020630191341.D10625@prim.han.de>

Pekka Niiranen wrote:
> Has anybody used python's parser -module to parse user defined
> text structure? It is possible ?

you might want to give simpleparse a try. There is a nice article
and a link in it:

    http://www-106.ibm.com/developerworks/linux/library/l-simple.html

hth, holger




From mcfletch at rogers.com  Sun Jun 16 14:54:35 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Sun, 16 Jun 2002 14:54:35 -0400
Subject: [Extending] Py_BuildValue( "((OiiO))",...) creates empty string?
Message-ID: <3D0CDEEB.2070208@rogers.com>

I'm working on a C module, and have come across something I don't 
understand.  Basically, I have a method definition/specification which 
says I will pass my clients a 4-item tuple to their append method.

So, I want to call:
	PyObject_CallObject( objectToCall, parameter );

With a parameter equal to:
	(("this",1,2,[]),)
That is, a singleton tuple with a 4-item tuple as the first (and only) 
item.  Now, the docs for Py_BuildValue say I can create arbitrarily 
nested tuples, so I would expect this:

	parameter = Py_BuildValue( "((OiiO))", tagobj, childStart, childPosition, 
childResults );

to create the appropriate tuple-structure.  Unfortunately, it doesn't. 
Instead it creates an empty string as far as I can tell (that's what my 
Python test code receives, anyway).

I also tried it like this:

	parameter = Py_BuildValue( "(N)", Py_BuildValue( "(OiiO)", tagobj, 
childStart, childPosition, childResults ));

And again, got the same result (an empty string passed to my method).

Any clues from the C gurus as to what I'm doing wrong?

With thanks,
Mike
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From merkosh at hadiko.de  Mon Jun 17 18:51:36 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Tue, 18 Jun 2002 00:51:36 +0200
Subject: inherrited private class varriables
Message-ID: <MPG.17788662b0a1b67a98968c@news.rz.uni-karlsruhe.de>

hi,

why can't i access a class private varriable from within a predecessor?
f.e.

class parent:
  __length = 5
  def __init__(self): pass
  def action(self):
    print self.__class__.__length

class child(parent):
  __length = 10
  def __init__(self): pass

test = child()
test.action()

will print 5, instead of 10. why?

Ciao
Uwe  


From look at replyto.address.invalid  Tue Jun  4 01:35:37 2002
From: look at replyto.address.invalid (Greg Ewing)
Date: Tue, 04 Jun 2002 17:35:37 +1200
Subject: prototyping good OOdesign in Python?
References: <mailman.1023075244.23766.python-list@python.org>
Message-ID: <3CFC51A9.954C721E@replyto.address.invalid>

"Delaney, Timothy" wrote:
> 
> Unfortunately, this will only work for 1 or 2 projects ... after that they
> will *expect* you to bring everything in under time and budget ...

So, don't tell them it's finished until just before the
deadline... and find something plausible-sounding to
spend the extra money on... :-)

-- 
Greg Ewing, Computer Science Dept, 
University of Canterbury,	  
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg


From whisper at oz.net  Wed Jun 19 13:27:22 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 19 Jun 2002 10:27:22 -0700
Subject: CPU Usage Windows
In-Reply-To: <ad381f5b.0206190233.4fa1bf58@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEEEDFDKAA.whisper@oz.net>

Although billed as a true SMP OS, in fact NT, 2K and XP aren't quite. They
pretty much split I/O and system onto on CPU and applications get the other
one. I have seen the pattern you describe repeatedly on my old NT 4.0 dual
pentium pro. While it was never all that fast (200 MHZ CPU's), it took some
real effort to get it to slow down much, so I think there is some better
load sharing then what MS suggests (what I said above).

There are actually some obscurely documented programatic ways to force the
OS to actually use both (or more then 2 even) processors for one process, or
to get the OS to use both processors for applications, but I only knew about
them in passing and don't recall exactly where the information was in MSDN.

Regards,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Rad
> Sent: Wednesday, June 19, 2002 3:34
> To: python-list at python.org
> Subject: CPU Usage Windows
>
>
> Hi,
>
> I've noticed that when I run Python script on
> Windows 2k Pro
> Dual Intel Xeon (2.2GHz)
> CPU Usage for pythonw.exe is only 50%?!
>
> Nothing else is running on the machine (apart form the OS). Some of
> the python programs are running for hours (large data sets) and even
> though python shell freeze's (doesn't refresh) CPU usage is still only
> 50%.
>
> System Idle Process is about 40%-50%
>
> Any ideas why is this and can I do anything about it?
>
> Is it something to do with python not being able to use the power of
> both processors?
>
> thanks,
> --
> http://mail.python.org/mailman/listinfo/python-list





From rnd at onego.ru  Sat Jun  8 22:21:50 2002
From: rnd at onego.ru (Roman Suzi)
Date: Sun, 9 Jun 2002 06:21:50 +0400 (MSD)
Subject: Comparison: HTTP with Java, Perl or Python
In-Reply-To: <3D02A187.5070505@gmx.de>
Message-ID: <Pine.LNX.4.33.0206090618400.9726-100000@rnd.onego.ru>

WHat you want to do reminds me of this (open source) project:

http://www.bagley.org/~doug/shootout/

- it already has a framework/examples for many things you 
are measuring. (and for MUCH MUCH more languages).
The project seem to be frozen but everybody can
advance it further.

On Sun, 9 Jun 2002, Ingo Linkweiler wrote:

...
>For comparison:
>- Time for coding
>- Time for testing and debugging
>- Number of changes during debugging
>- Lines of Code / Packed bytes (without comments)
>- Is code readable by a programmer who not has used the language before?
>
>Are you interested? Any suggestions?
>We need at least 2 programmers for each language (Python, Perl, Java, TCL)
>
>Ingo
>
>

Sincerely yours, Roman Suzi
-- 
\_ Russia \_ Karelia \_ Petrozavodsk \_ rnd at onego.ru \_
\_ Sunday, June 09, 2002 \_ Powered by Linux RedHat 7.2 \_
\_ "Will Rogers never met the Clintons." \_





From bokr at oz.net  Tue Jun 25 20:53:22 2002
From: bokr at oz.net (Bengt Richter)
Date: 26 Jun 2002 00:53:22 GMT
Subject: Suggestions for good programming practices?
References: <mailman.1024951532.5616.python-list@python.org> <af8n9i$3at$1@panix1.panix.com> <afafee$fc$0@216.39.172.122> <afaohi$682$1@panix1.panix.com>
Message-ID: <afb3a2$skm$0@216.39.172.122>

On 25 Jun 2002 17:49:38 -0400, aahz at pythoncraft.com (Aahz) wrote:

>In article <afafee$fc$0 at 216.39.172.122>, Bengt Richter <bokr at oz.net> wrote:
>>
>>Do you recommend using def foo(*args) and testing len(args) instead?
>
>Only if you really want a list of unnamed parameters.  Personally, I
>usually prefer to pass a sequence directly in such cases, to distinguish
>between None (i.e. no list) and the empty list.

Well, I meant using *args to allow detecting actual "no default-overriding
argument passed", in place of x=None. The arg list items (only x here)
don't have to remain unnamed long. E.g.,

 >>> def foo(*args):
 ...     if len(args)==1:
 ...         print "We know we have one non-default arg: %s" % `args[0]`
 ...         x = args[0]
 ...     elif len(args)==0:
 ...         print "We can supply a default arg, even None"
 ...         x = None
 ...     else: raise TypeError,'foo() takes 0 or 1 arguments (%d given)' % len(args)
 ...     print 'Effective arg was: %s' % `x`
 ...
 >>> foo('one arg')
 We know we have one non-default arg: 'one arg'
 Effective arg was: 'one arg'
 >>> foo(None)
 We know we have one non-default arg: None
 Effective arg was: None
 >>> foo()
 We can supply a default arg, even None
 Effective arg was: None
 >>> foo(None,'too many')
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "<stdin>", line 8, in foo
 TypeError: foo() takes 0 or 1 arguments (2 given)

Hm. It occurs to me that *args also lets you put defaults at the front if you like,
and specify not just min and max number of args. E.g., a textout method could
use current x,y by default or allow override (obviously this example is just
about parameter passing, not graphics output. And no prizes for speed ;-)

 >>> class GX:
 ...     def __init__(self, x, y): self.x, self.y = x,y
 ...     def textout(self, *args):
 ...         if len(args) not in (1,3):
 ...             raise TypeError,"Use textout('string') or textout(x,y,'string') (%d args given)" % len(args)
 ...         if len(args)==1:
 ...             x, y, a = self.x, self.y, args[0]
 ...         else:
 ...             x, y, a = args
 ...         print 'output "%s" at %s,%s' % (a, x, y)
 ...
 >>> g=GX(123, 456)
 >>> g.textout('using curr pos')
 output "using curr pos" at 123,456
 >>> g.textout(333, 444, 'using spec pos')
 output "using spec pos" at 333,444

And accurate arg count validation can be had (ignoring types):
 >>> g.textout()
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "<stdin>", line 5, in textout
 TypeError: Use textout('string') or textout(x,y,'string') (0 args given)
 >>> g.textout(1,2)
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "<stdin>", line 5, in textout
 TypeError: Use textout('string') or textout(x,y,'string') (2 args given)
 >>> g.textout(1,2,3,4)
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "<stdin>", line 5, in textout
 TypeError: Use textout('string') or textout(x,y,'string') (4 args given)

Regards,
Bengt Richter


From phr-n2002b at NOSPAMnightsong.com  Fri Jun 28 15:31:12 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 28 Jun 2002 12:31:12 -0700
Subject: C API for Python's RNG?
References: <3D1CBB3A.E1C43D8A@astro.cornell.edu> <3D1CBBED.58EAC61E@astro.cornell.edu>
Message-ID: <7xeleri4jj.fsf@ruckus.brouhaha.com>

Tom Loredo <loredo at astro.cornell.edu> writes:
> Oops!  I should have dug first---the standard lib RNG is pure
> Python.  So I guess it's Numeric or some third-party RNG if
> I want to do this quickly.  Any suggestions?

The Python RNG is simple enough that you should be able to re-code it
in C for your extension easily.


From michael at damaru.com  Sun Jun  9 11:51:19 2002
From: michael at damaru.com (Michael Davis)
Date: Sun, 09 Jun 2002 11:51:19 -0400
Subject: libftp.size function gives wrong results!
Message-ID: <RNKM8.560$Yw3.96921@news20.bellglobal.com>

Hi,

I'm writing a python program which will upload or download files via ftp. 
It's intended to be used to deploy a web site from a local machine to a 
server. It tries to determine which files are modified on the local machine 
by comparing file sizes and modification times.

I try to compare files using os.stat for the local file and FTP.size for the 
remote. The problem is, FTP.size is giving me incorrect results. If I use 
the ftp command line client, and look at one of my files on the server (the 
output looks like 'ls -l') then I can see that the file size, is, say, 
1200. But the FTP.size function returns, say, 1242. This happens for all 
the files, they're all reported to be about 5% larger than they really are.

Is there a better way of doing this, aside from capturing the output of 
FTPLIB.dir (which goes to stdout) and parsing it?

Thanks in advance!
-- 
Michael Davis
Damaru
Custom Programming - Web Development - Database Design
http://www.damaru.com
416-540-1284


From Administrator at python.org  Thu Jun 20 14:59:44 2002
From: Administrator at python.org (Administrator at python.org)
Date: Thu, 20 Jun 2002 12:59:44 -0600
Subject: ScanMail Message: To Recipient file blocking settings matched and action taken.
Message-ID: <00d801c2188c$a3cf6000$0e0fbb82@corp.es.com>

ScanMail for Microsoft Exchange has blocked an attachment.

Sender = Sylvain Th?nault
Recipient(s) = David LeBlanc;python-list at python.org; XML Logilab Mailing list
Subject = Re: [ANN] PyReverse 0.3
Scanning Time = 06/20/2002 12:59:43

Action on file blocking:
The attachment py2tests.bat matches the file blocking settings. ScanMail has Deleted it. 






From fperez528 at yahoo.com  Mon Jun 24 18:15:07 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Mon, 24 Jun 2002 16:15:07 -0600
Subject: Numerical Methods in Python
References: <%32R8.853$xv1.36390126@newssvr15.news.prodigy.com>
Message-ID: <af85lm$6dt$1@peabody.colorado.edu>

Russ Dilley wrote:

> I am interested in the numerical capability of Python as well as it's
> applicability to system administration  tasks.

[snip]

> I would like to focus more attention to become proficient with a single
> language and, based on what I've seen, it seems that Python might be
> appropriate for use in system administration AND numerical programming.
> 
> Does anyone have any insight to the applicability of Python to either, or
> both, of these tasks?

One reason I love python is _exactly_ what you said. Now I only need to 
remember ONE syntax and I can do what I used to need 
shell+awk+sed+perl+IDL/matlab+C to do. That's a _huge_ gain. These days I 
have it down to Python + C (but much less C than before, now it's only for 
tight loops) + Mathematica (symbolics). And with those three I can get by 
whether I'm writing numerical code or a script to fix my network 
configuration when I move my laptop. I love it.

On the topic of python for science, I gave someone else a bunch of pointers 
you may find useful at:
http://groups.google.com/groups?selm=aeqfuh$88n$1 at peabody.colorado.edu

Cheers,

f.


From rchopra at myrealbox.com  Mon Jun 17 19:49:31 2002
From: rchopra at myrealbox.com (Rajat Chopra)
Date: Mon, 17 Jun 2002 19:49:31 -0400
Subject: Python Executable Code
Message-ID: <1024357771.569c0ff8rchopra@myrealbox.com>

I have a utility program written in Python that I wish to install on several different machines. I do not want to install the Python interpreter onto each of the machines, but would rather have a Windows Executable (*.exe) version of my Python code so that any user can execute the script.

Does anybody know how to convert a Python file (*.py) to a Windows Executable format (*.exe) ?

Rajat





From uncleenzo at hushmail.com  Wed Jun 26 19:29:34 2002
From: uncleenzo at hushmail.com (Uncle Enzo)
Date: 26 Jun 2002 16:29:34 -0700
Subject: COM server threads/unloading
Message-ID: <10d944cb.0206261529.741da082@posting.google.com>

Hello, 

I have noticed that my multithreaded COM servers are not unloading
when their clients have unloaded.

To reproduce this in a simple case, I am using the py2exe COM server
recipe. Running the server example with the client code below results
in the testserver.exe loading after the dispatch call, and unloading
after the server reference is set to None.

Then I added the googlelogger import statement and the two lines
initializing and calling the log method. The logger that has two
threads. Using the same client code results in the testserver.exe
loading after the dispatch call, but it does not unload after the
server reference is set to None, nor does it unload after the client
terminates.

What is the proper way to get a multithreaded python COM server to
unload when there are no external references to it?

Thanks in advance,

Enzo

ps, comments about my logger design are welcome, but I'm mostly
interested in the threading issues :)

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

Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on
win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import win32com.client
>>> server = win32com.client.Dispatch("Python.TestServer")
>>> print server.Hello('enzo')
Hello enzo
>>> server = None
>>> ^Z

----------------------------
file: testcomserver.py

import googlelogger
import sys
import pythoncom

if hasattr(sys, 'importers'):
    # we are running as py2exe-packed executable
    pythoncom.frozen = 1

class HelloWorld:
    _reg_clsctx_ = pythoncom.CLSCTX_LOCAL_SERVER
    if hasattr(sys, 'importers'):
        # In the py2exe-packed version, specify the module.class
        # to use. In the python script version, python is able
        # to figure it out itself.
        _reg_class_spec_ = "__main__.HelloWorld"
    _reg_clsid_ = "{592C15CE-894D-11D6-9A35-0050DA05A582}"
    _reg_desc_ = "Python Test COM Server"
    _reg_progid_ = "Python.TestServer"
    _public_methods_ = ['Hello']
    _public_attrs_ = ['softspace', 'noCalls']
    _readonly_attrs_ = ['noCalls']
    def __init__(self):
        self.softspace = 1
        self.noCalls = 0
        self.mylogger = googlelogger.getLogger()
    # __init__()

    def Hello(self, who):
        self.noCalls = self.noCalls + 1
        # insert "softspace" number of spaces
        self.mylogger.log('caller', str(who))
        return "Hello" + " " * self.softspace + str(who)

if __name__ == '__main__':
    if hasattr(sys, 'importers'):
        # running as packed executable.
        if '--register' in sys.argv[1:] or '--unregister' in
sys.argv[1:]:
            # --register and --unregister work as usual
            import win32com.server.register
            win32com.server.register.UseCommandLine(HelloWorld)
        else:
            # start the server.
            from win32com.server import localserver
            localserver.main()
    else:
        import win32com.server.register
        win32com.server.register.UseCommandLine(HelloWorld)

----------------------
file: setup.py

from distutils.core import setup
import py2exe

setup(name='testCOMserver',
      scripts=['testCOMserver.py'])

------------------------
file: googlelogger.py

import threading
import Queue
import sys
import pavlnchcfg
import time
import win32com.client

# This thread waits for ("name", "value") items to appear in the
queue.
# It then checks for a log named "name", creates it if it doesn't
exist,
# and appends the message to that log, with a timestamp.
class Writer(threading.Thread):
    def setLogger(self, logger):
        #print "setting self._logger to ", logger
        self._logger = logger
        
    def run(self):
        print "starting Writer.run thread"        
        while 1 :
            # blocks until there are items in the queue
            event = self._logger._logeventqueue.get(1) 
            #get returns a tuple, (name, message)            
            if event[0] == 'QUIT': #magic quit event
                print "logger.Writer.run: got magic event: returning"
                return
            
            logname = event[0] 
            message = event[1]
            try:
                filename = self._logger._dir + '\\' + logname + ".txt"
                logfile = open(filename, 'a')    
                logfile.write(time.asctime() + " " + message + '\n')
                logfile.flush()
                logfile.close()
                print "logger.Writer.run: logged ", message, " to ",
logname
            except:
                print "logger.Writer.run: with to log: ", logname, "
failed with message: ", message

class Logger:
    # class variables
    _logeventqueue = Queue.Queue() #one thread safe multiple user
queue
    _dir = "" 
    
    def __init__(self):          
        try:            
            Logger._dir="C:\\" # for our example
            print "logger.Logger.init: got log directory",
Logger._dir, "from config"
        except:
            Logger._dir = "C:\\"
            print "logger.Logger.init: No log directory specified in
config, _dir set to  C:\\"
            
        writer = Writer()
        writer.setName("writerthread")
        writer.setLogger(self)
        writer.start()
        
    def log(self, logname, message):
        event = logname, message  #making the event tuple
        # event tuple is tuple[0] = name, tuple[1] = event
        Logger._logeventqueue.put(event)      
            

# factory pattern as module instance
_loggerInstance = None
       
def getLogger():
    import logger
    if logger._loggerInstance == None:
        logger._loggerInstance = Logger()
        return logger._loggerInstance
    else:
        return logger._loggerInstance

if __name__ == '__main__':  #test main
    import logger
    print "starting main"
    l = getLogger()
    print l
    l.log("DEBUG", "l, debug message 1")
    l.log("USAGE", "l, usagestat 1")
    l.log("DEBUG", "l, debug message 2")
    lg = getLogger()
    print lg
    lg.log("DEBUG", "lg, debug message 1")
    lg.log("USAGE", "lg, usagestat 1")
    lg.log("DEBUG", "lg, debug message 2")    
    l.log("QUIT", "QUIT")
    print "done"


From rnd at onego.ru  Tue Jun 18 00:51:11 2002
From: rnd at onego.ru (Roman Suzi)
Date: Tue, 18 Jun 2002 08:51:11 +0400 (MSD)
Subject: unicode strings and strings mix 
Message-ID: <Pine.LNX.4.33.0206180848090.28845-100000@rnd.onego.ru>

Is there any way to define "default" encoding so the following error
will not happen?

>>> unichr(0x3345) + "?????"
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
UnicodeError: ASCII decoding error: ordinal not in range(128)
>>> unichr(0x3345) + "babab"
u'\u3345babab'
>>>


Sincerely yours, Roman Suzi
-- 
\_ Russia \_ Karelia \_ Petrozavodsk \_ rnd at onego.ru \_
\_ Monday, June 17, 2002 \_ Powered by Linux RedHat 7.2 \_
\_ "I have a rock garden. 3 of them died last week." \_





From siegfried.gonzi at kfunigraz.ac.at  Mon Jun 24 08:37:37 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Mon, 24 Jun 2002 14:37:37 +0200
Subject: SOLVED (Re: Python hits the spot)
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at> <af4vrl$ffe$0@216.39.172.122> <3D16DD5C.1050708@kfunigraz.ac.at> <uhdume2n1mhbba@corp.supernews.com>
Message-ID: <3D171291.2060905@kfunigraz.ac.at>

RPM1 wrote:

> Hey!  That's why I was asking if your laptop was a Hewlitt
> Packard.  I had a Hewlitt Packard Pavillion laptop that would
> die when the fan kicked on.
> 
> Any chess program will keep your CPU at 100%.
> 
> GnuChess is probably on your system now.  If not, it is a free
> program.  Ask around on rec.games.chess.computer if you need
> help.

It was really the laptop processor. I tried the calculation on a 
stationary machine (Pentium II 450 MHz and SuSE 8.0), and it works.

Die Typen k?nnen jetzt was erleben.




From sholden at holdenweb.com  Sun Jun  2 14:23:10 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Sun, 2 Jun 2002 14:23:10 -0400
Subject: HTML and python
References: <ad94md$viphf$1@ID-49758.news.dfncis.de>
Message-ID: <gstK8.148299$%u2.116279@atlpnn01.usenetserver.com>

"Ken" <ken at hotmail.com> wrote in message
news:ad94md$viphf$1 at ID-49758.news.dfncis.de...
> How do you call to a static HTML page from the script?
>
> <FORM METHOD="POST" ACTION="frame2a.html" target=_parent>
> This doesn't seem to be working..... Can someone help?
>

It really makes no sense to use method POST to access a static page. Using
method GET will probably owrk, though you will see the for contents in the
URL.

regards
 Steve
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From gcordova at hebmex.com  Thu Jun  6 18:45:52 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 6 Jun 2002 17:45:52 -0500 
Subject: win32com map attributes to methods
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062330116F367@mail.hebmex.com>

Have you tried using property() in py2.2?

# YOU MUST DERIVE FROM "object"
class SimpleContainer(object):
   def __init__(self, default):
      self.__value = default

   def __get(self):
      print "<<< inside get() >>>"
      return self.__value

   def __set(self,value):
      print "<<< inside set() >>>"
      self.__value = value

   def __del(self):
      raise AttributeError("NO CAN DO, PAL!!!")

   value = property(__get, __set, __del, "docstring")

>>> sc = SimpleContainer(3)
>>> sc.__value
3
>>> sc.value
<<< inside get() >>>
3
>>> sc.value = 666
<<< inside set() >>>
>>> sc.__value
666
>>> sc.value
<<< inside get() >>>
666
>>> del sc.value
argh!!! (traceback)
NO CAN DO, PAL!
>>>

So, using property objects as class attributes,
you can gate access to internal data members,
or return calculated attributes, or create
read-only attributes.

Good luck!

-gustavo




From gustav at morpheus.demon.co.uk  Thu Jun 27 09:29:08 2002
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: Thu, 27 Jun 2002 15:29:08 +0200
Subject: sys.argv[0] - Returns path and filename.
References: <mailman.1025182895.13299.python-list@python.org>
Message-ID: <n0tg6eal.fsf@morpheus.demon.co.uk>

"Mark McEahern" <marklists at mceahern.com> writes:

>> How do I get the current path of the script I am running, on the mac
>> osx and the linux box ? (Any work around would be great, I don't want
>> to have to hardcode the path into the script, or even give the path as
>> an argument,I want the script to know where it is.)
>
> import os
> print os.getcwd()

Or os.path.abspath(sys.argv[0]).

Run this before you change directory...

Paul.



From mickey at tm.informatik.uni-frankfurt.de  Sun Jun 16 10:38:10 2002
From: mickey at tm.informatik.uni-frankfurt.de (Michael 'Mickey' Lauer)
Date: 16 Jun 2002 16:38:10 +0200
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de> <aeftmj$m29$1@bob.news.rcn.net> <3d0c6d67@nntp.server.uni-frankfurt.de> <aei645$o66$1@bob.news.rcn.net>
Message-ID: <3d0ca2d2@nntp.server.uni-frankfurt.de>

David Abrahams <david.abrahams at rcn.com> wrote:
> 
> "Michael 'Mickey' Lauer" <mickey at tm.informatik.uni-frankfurt.de> wrote in
> message news:3d0c6d67 at nntp.server.uni-frankfurt.de...
>> David Abrahams <david.abrahams at rcn.com> wrote:
>> > With Boost.Python, what you write looks like a kind of IDL (interface
>> > description language), which just happens to be built out of C++ source
>> > code. Lots of jobs (function argument type checking, conversion, arity
>> > checking) are done automatically, but the library doesn't decide which
>> > elements to wrap. Unlike SWIG and a few others, there's no attempt to
> parse
>> > your source code. Instead, the compile-time introspection capabilities
> of
>> > C++ are exploited to automatically build wrappers from function and
> member
>> > function pointers.
>>
>> I see. So it seems in order to get a nice object oriented Python
> interface,
>> I can 1.) use SWIG to parse the header file and automatically produce a
> wrapper
>>       1.1.) and then use this low-level wrapper to write some high-level
> python classes around it or
>>       2.) use Boost.Python and do the two steps in one by writing a
> "handcrafted" interface?
> 
> I have no deep familiarity with SWIG, so I really don't know it takes to
> produce high-level classes on the Python side with that tool. If you use
> Boost.Python, you do get true classes on the Python side which are wrapped
> around your C++ classes.
> 
>> How would sip come into this picture? Is it more like SWIG for C++ or more
> like Boost.Python?
>> (Apart from not beeing concerned about the total lack of documentation ;)
> 
> I wrote this page more than a year ago; it's surely out-of-date:
> 
> http://www.boost.org/libs/python/doc/comparisons.html
> 
> Someone needs to freshen it up for the upcoming release of Boost.Python v2;
> Since you seem to be doing a survey, would you care to volunteer?

Sounds reasonable. I will use your document as a basis and will gather
updated data. However, I won't have time to do this before mid-july, since
the current semester is very busy :)

Yours,

Mickey.

:M:
--
|----------------------------------------------------------------------------|
| Dipl.-Inf. Michael 'Mickey' Lauer    mickey at tm.informatik.uni-frankfurt.de |
|   Raum 10b - ++49 69 798 28358        Fachbereich Informatik und Biologie  |
|----------------------------------------------------------------------------|


From jadestar at idiom.com  Mon Jun 10 17:45:40 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 10 Jun 2002 21:45:40 GMT
Subject: Why does Python mix OO concepts and non OO concepts for operation  s on basic types?
References: <mailman.1022091281.4506.python-list@python.org> <slrnaenqq4.98v.jajvirta@sirppi.helsinki.fi> <YHfI8.103006$ut3.25086@atlpnn01.usenetserver.com>
Message-ID: <ae36m4$6r3$2@news.idiom.com>

Steve Holden <sholden at holdenweb.com> wrote:

> "Jarno J Virtanen" <jajvirta at cc.helsinki.fi> wrote ...
>> Wed, 22 May 2002 13:13:06 -0500 Michael Bauers wrote:
>>> Why do you say x = []; x.append('a'), but get the length with len(a) ?
>>> Is there a reason for this sort of inconsistency?

>> why not check the FAQ?
>> http://www.python.org/cgi-bin/faqw.py?req=all#6.5

> I have updated this entry, and would welcome any feedback about the
> changes -- it's just a first attempt, and I was trying not to be too
> defensive about the inconsistencies. After all, history is important even in
> technologies.

> regards
>  Steve

	Personally I think it would be generally nice if I could
	call most of the list and string methods in both ways:

	list.append(somelist,someitem) would be the same as:
	somelist.append(someitem).

	And the obvious:

	string.join('',something) would be as: ''.join(something).

	(Personally I find the latter rendering to be ugly; a hobgoblin
	of consistency).

	


From mcherm at destiny.com  Mon Jun 24 10:27:22 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Mon, 24 Jun 2002 10:27:22 -0400
Subject: Generic Python
Message-ID: <3D172C4A.5070307@destiny.com>

Uwe writes:
> is it possible to to write a class which, f.e. takes an argument in
> __init__() and that doesn't return an instance object but a new class
> object?
> 
> The problem is that  [.....]
        [Detailed explanation follows]

So far I've seen suggestions that:
   (1) Create a factory function which returns the appropriate class.
   (2) Somehow use exec to build the classes at runtime.
   (3) Use metaclass programming.

I would suggest that (1) is by far the best solution, and that you stay 
away from (2) and (3) unless you KNOW what you are doing, or you are 
more interested in playing around (and learning) than getting your 
problem solved.

But just to be confusing, I'm going to add a (4)th solution, completely 
different from the other three!


(4) In python, you can change individual methods of an instance, so you 
may not even NEED to have many separate classes, if they differ only in 
that one method.

Here is an illustration of how it would work:

Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
 >>> class Base:
...     def output_0(self):
...         print 'zero'
...     def output_1(self):
...         print 'one'
...     def output_2(self):
...         print 'two'
...     def __init__(self, n):
...         self.output = (
...             [self.output_0, self.output_1, self.output_2] [n] )
...
 >>> b = Base(0)
 >>> b.output()
zero
 >>> b = Base(2)
 >>> b.output()
two


HOWEVER... if your subclasses need more than just the one function 
defined in them, then you are better off going with solution (1) above.

-- Michael Chermside





From gerhard at bigfoot.de  Tue Jun 11 21:56:37 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 12 Jun 2002 03:56:37 +0200
Subject: (no subject)
In-Reply-To: <EFEFIHAINIINNELLHFAIEECACJAA.kseehof@neuralintegrator.com>
References: <EFEFIHAINIINNELLHFAIEECACJAA.kseehof@neuralintegrator.com>
Message-ID: <20020612015637.GA29113@lilith.my-fqdn.de>

* Ken Seehof <kseehof at neuralintegrator.com> [2002-06-11 18:55 -0700]:
> Where are the windows binaries for python22_d.lib?
> Do I need to build the sources to get it?
 
Yes.

You could get install them via pyppm if you used ActiveState's
distribution. But you learn more if you build Python from source
(requires Visual C++ or insane amounts of hacking for other compilers).

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From hamish_lawson at yahoo.co.uk  Thu Jun  6 12:33:54 2002
From: hamish_lawson at yahoo.co.uk (Hamish Lawson)
Date: 6 Jun 2002 09:33:54 -0700
Subject: Emulating classmethod in Python 2.1
Message-ID: <915a998f.0206060833.12ad76a8@posting.google.com>

Is it possible to define a 'classmethod' class or function for Python
2.1 that would effectively provide the functionality of Python 2.2's
built-in classmethod function and have the same usage pattern? I'd
like to be able run code like that below against both 2.2 (using its
builtin classmethod function) and 2.1 (using some added-in wrapper).

    class SomeClass:
        def hello(cls, name):
            print "Hi there", cls, name
        hello = classmethod(hello)

    SomeClass.hello("Peter")

I've come across Thomas Heller's recipe in the online Python Cookbook,
but this would require SomeClass to be modified, and so fails my
criteria.

I've also come across Alex Martelli's recipe for emulating
staticmethod, but of course that doesn't deal with passing a class to
the wrapped method.


Hamish Lawson


From mwh at python.net  Tue Jun 18 06:47:11 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 18 Jun 2002 10:47:11 GMT
Subject: [development doc updates]
References: <mailman.1024057266.26182.python-list@python.org> <ugk341aivqpl27@news.supernews.com> <lku1o2l062.fsf@pc150.maths.bris.ac.uk> <ugriqqbhlt4u15@news.supernews.com>
Message-ID: <lklm9clv9g.fsf@pc150.maths.bris.ac.uk>

"John Roth" <johnroth at ameritech.net> writes:

> Possibly I didn't make myself clear. Item 5 in the What's New
> document contains two paragraphs. It's not clear that the second
> paragraph is talking about something completely different than the
> topic in the header (Extended Slices). Hence my original question.

Ah yes.  Drafts documents and all that :)

There's a section break commented out.  Not sure why.

Cheers,
M.

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


From jonathan at onegoodidea.com  Thu Jun 27 08:21:35 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Thu, 27 Jun 2002 13:21:35 +0100
Subject: Python 2.2 __slots__ Bug?
References: <mailman.1025089474.17780.python-list@python.org> <B93F7990.CECB%jonathan@onegoodidea.com> <mailman.1025105074.31122.python-list@python.org> <afdlv6$bu6$0@216.39.172.122> <B940C17E.D101%jonathan@onegoodidea.com>
Message-ID: <B940C1DF.D104%jonathan@onegoodidea.com>

On 27/6/2002 13:19, in article B940C17E.D101%jonathan at onegoodidea.com,
"Jonathan Hogg" <jonathan at onegoodidea.com> wrote:

> It appears that GjR is on the case ;-)

...and his alter-ego GvR.

oops.

Jonathan



From guy at lightwork.co.uk  Thu Jun 27 08:32:19 2002
From: guy at lightwork.co.uk (Guy)
Date: 27 Jun 2002 05:32:19 -0700
Subject: sys.argv[0] - Returns path and filename.
Message-ID: <c07d0edc.0206270432.faf1fd3@posting.google.com>

Hi

I work on multiple platforms, at the moment, I have python installed
on my win32 box which I mainly work on, a MacOSX box and a linux box
which has a very old version installed on it.

The only major problem that I have come across at the moment is the
argv[0], on my win32 box it retruns the path and a filename, on the
mac osx and the linux only the filename are returned.

How do I get the current path of the script I am running, on the mac
osx and the linux box ? (Any work around would be great, I don't want
to have to hardcode the path into the script, or even give the path as
an argument,I want the script to know where it is.)

I've tryed searching for this but can't find it anywhere. 

Example script called arguments.py :
import sys
print sys.argv[0]

Result Win32 :
>>> D:\TestArg\arguments.py

Result Mac OSX : 
MacOSX% python arguments.py
arguments.py

Result Linux : 
linux% python arguments.py
arguments.py


From cliechti at gmx.net  Sun Jun 30 13:16:19 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 30 Jun 2002 19:16:19 +0200
Subject: arguments to functions
References: <MPG.178955a4fc8bfc1c989680@news.rz.uni-karlsruhe.de>
Message-ID: <Xns923DC4803F1FDcliechtigmxnet@62.2.16.82>

Uwe Mayer <merkosh at hadiko.de> wrote in 
news:MPG.178955a4fc8bfc1c989680 at news.rz.uni-karlsruhe.de:
> I wanted to pass one named and a collection of unnamed (an arbitrary 
> number) to a function:

def f(*args, **kwargs):
    	print "positional args:", args
    	print "keyword args:", kwargs
    	if kwargs.has_attr("expected"):
    	    	print "Expected was set to", kwargs["expected"]

does that help somehow? :-)

chris

-- 
Chris <cliechti at gmx.net>



From sholden at holdenweb.com  Thu Jun  6 10:20:10 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 6 Jun 2002 10:20:10 -0400
Subject: array, list, performance...
References: <3CFF5F98.2060406@destiny.com> <mailman.1023373030.31871.python-list@python.org>
Message-ID: <_gKL8.314196$%u2.248214@atlpnn01.usenetserver.com>

"Christopher Armstrong" <radix at twistedmatrix.com> wrote in message
news:mailman.1023373030.31871.python-list at python.org...
>
> > Michael Chermside <mcherm at destiny.com> writes::
> >   li[n] = x      assignment to the middle is O(len(li))
>
> This doesn't seem right. Shouldn't that be O(1), just like
> access-time?
>

(I should have specifically said I was talking about INSERTION. You are
correct about assignment over an existing element).

--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From sheershion at mailexpire.com  Thu Jun 20 11:26:07 2002
From: sheershion at mailexpire.com (Robert Amesz)
Date: Thu, 20 Jun 2002 15:26:07 -0000
Subject: Writing to Excel
References: <79b54d28.0206191151.52ee78a3@posting.google.com>
Message-ID: <Xns9233B0570702Brcamesz@amesz.demon.nl>

Chris wrote:

> Hi I'm using a Python cgi script to write the contents of a table
> to an Excel sheet. I retrieve all the information fine but it
> seems like Excel doesnt want to print more than one line for me.
> It prints 'headerLine' but nothing more. Why???
> 
> ... 
> headerLine = "some crap"
> 
> sys.stdout.write( Html.header( 'application/x-msexcel' ) )
> print string.strip(headerLine)
> allText = ""
> 
> ...
> # set allText = "some other crap"
> ...
> 
> print allText

My oh my, what *are* you trying to do here? Hmmm, are you trying to 
convert an ASP script by any chance? That won't work. VBS accesses 
Excel via the COM interface, and you'd need Mark Hammond's Win32all 
extension to do the same.

Not too hard, but... You'd need to have Excel installed on the web 
server, along with Python and Win32all. If that web server isn't a 
Windows machine, that's the end of the story.

But why bother with Excel at all? In most cases .CSV files will work 
equally well, and those files are a lot simpler, definitely more 
portable, and dead easy to make with Python.

Or is it that IE is set up to display Excel files directly using 
ActiveX? But what, then, would be the advantage over using HTML-tables?


Robert Amesz (slightly confused)


From eric.brunel at pragmadev.com  Wed Jun 19 14:34:40 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Wed, 19 Jun 2002 18:34:40 +0000
Subject: tkinter menu (newbie question)
References: <Xns9232778FCD335stefninnospamyahoofr@213.228.0.138> <Xns9232967929C7Fstefninnospamyahoofr@213.228.0.138> <aeq0cv$p5g$1@wanadoo.fr> <Xns9232B9BE55D67stefninnospamyahoofr@213.228.0.138>
Message-ID: <aeqbfl$2ka$1@wanadoo.fr>

Stephane Ninin wrote:

> Eric Brunel <eric.brunel at pragmadev.com> a tapote dans
> news:aeq0cv$p5g$1 at wanadoo.fr:
> 
>> 
>>> If I want to use several of these windows,
>>> can still use for base class Tk, instead of Frame ?
>> 
>> Nope: the instance of Tk is your main window. There should only be
>> one, and closing it will end your application. If you want to create
>> other windows, you should create instances of Toplevel.
>> 
> 
> So, what would be the easiest way to attach a menu to a window which is
> not the main window ? (that is which doesnot inherit from Tk class)
> 
> It seems that Frame class has no menu key,
> so I cannot use config(menu=...)
> for a class that would inherit for Frame...

Do not inherit from Frame. A frame is not a window: it's just an 
all-purpose container. Creating a frame (or an instance of a sub-class of 
frame) will create it in the main window (your instance of Tk). If you want 
to create your own classes for windows, create sub-classes of Toplevel.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From thenault at nerim.net  Thu Jun 13 05:22:03 2002
From: thenault at nerim.net (Sylvain Thenault)
Date: 13 Jun 2002 11:22:03 +0200
Subject: How to check version of Python under CGI
References: <rbmggus7ln1uiutktf5libf4bbngu29ji7@4ax.com> <873cvrin1q.fsf@tux.ntw23.fr> <cgnggu40ethjhpqcmltnfb06d5qvuf078k@4ax.com> <t1ogguoua7b9ne9n3mb3n7rkp8mrqe2nee@4ax.com>
Message-ID: <87y9djh6tg.fsf@tux.ntw23.fr>

Johann <observer at NOSPAM.space.pl> writes:

> On Thu, 13 Jun 2002 10:57:21 +0200, Johann <observer at NOSPAM.space.pl>
> wrote:
> 
> >Do you know how to check what modules are installed? E.g. I
> >would like to check if PYANA or other xml modules are installed.
> 
> for x in dir(__builtins__): print x
> 
> Is it all i can get?
this shows you all the built-in methods, not modules.
I think the best way to know if a module is available is to try to import it:

try:
    import pyana
except ImportError:
    print 'pyana not available'

-- 
Sylvain Th?nault 


From theller at python.net  Fri Jun  7 02:58:37 2002
From: theller at python.net (Thomas Heller)
Date: Fri, 7 Jun 2002 08:58:37 +0200
Subject: constructive critisism on pythonwin IDE
References: <6bd9f01b.0206050736.f469082@posting.google.com> <admuke$gu8$1@newsreaderg1.core.theplanet.net> <3CFFCA79.4050905@bgb.cc>
Message-ID: <adpllv$13keg$1@ID-59885.news.dfncis.de>

"Don Garrett" <garrett at bgb.cc> wrote in message news:3CFFCA79.4050905 at bgb.cc...
>    Thomas Heller wrote a wonderful script named 'autoreload' that helps with
> this problem quite a bit. I don't think this is an official distribution
> location (and thus not current?), but I found a link to it here:
>
>    http://starship.python.net/crew/zack/

No, this is something different.
The module I wrote is archived here:
http://aspn.activestate.com/ASPN/Mail/Message/907876,
but note the correction in this message:
http://aspn.activestate.com/ASPN/Mail/Message/907945

Thomas




From picpost at mailbox.hu  Mon Jun 17 22:08:12 2002
From: picpost at mailbox.hu (picpost)
Date: Tue, 18 Jun 2002 04:08:12 +0200
Subject: Képeslapod érkezett!
Message-ID: <02081218243635@relay.pronet.hu>

Szia ,

picpost k?peslapot k?ld?tt neked a  http://www.picpost.hu oldalr?l.

A k?peslapod megtekint?s?hez klikkelj a k?vetkez? cimre vagy m?sold
be a b?ng?sz?dbe:

http://www.picpost.hu/pickup.php?message_id=020613220636C4tZW3DzWdEB
 

picpost.hu team




From eric.brunel at pragmadev.com  Mon Jun 10 11:48:55 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 10 Jun 2002 15:48:55 +0000
Subject: Program to an interface, not an implementation
References: <D54BF26DB5A7D511B36B00508B9596A402F8EC63@sdgcaexch2.peregrine.com> <mailman.1023704594.18063.python-list@python.org>
Message-ID: <ae2ad4$s36$1@wanadoo.fr>

Egbert Bouwman wrote:
[snip]
> In Gordon's reply the venom is in the end:
>> For C++ interface inheritance REQUIRES class inheritance.  For
>> Python/Smalltalk, it does not.
> 
> This I read as:
> For Python interface_inheritance does not require class_inheritance.

Nope: it simply means that there's no such thing as an interface in Python, 
or more exactly that the interface of an object is figured out dynamically. 
For example, this code works:

class A:
  def m(self):
    print "I'm A"
class B:
  def m(self):
    print "I'm B"
def f(o): o.m()
f(A())
f(B())

The function "f" uses the method "m" in both classes. However, classes "A" 
and "B" are not related at all: neither do any inherits from the other, nor 
do they have a common ancestor. In C++ (or Java), the Python code above 
would be unwritable if the method "m" wasn't declared in a common ancestor 
of A and B. If you write:

public class A {
  public void m() {
    System.out.writeln("I'm A");
  }
}
public class B {
  public void m() {
    System.out.writeln("I'm B");
  }
}

you would not be able to write the equivalent of the method/function "f" 
than can act on either an instance of A or an instance of B. In C++ or 
Java, the interfaces *must* be declared, so you'll have to write:

public class I {
  public abstract void m();
}
public class A inherits I {
  public void m () {
    System.out.writeln("I'm A");
  }
}
public class B inherits I {
  public void m () {
    System.out.writeln("I'm B");
  }
}
...
public void f(I o) {
  o.m();
}

Now you can call f on an instance of A or an instance of B, because you 
have been able to *type* the object with the type (= interface) I.

> So you can inherit an interface without inheriting the class,
> but interface_inheritance by way of class_inheritance is possible as well.

Again, nope: you cannot *inherit* an interface without inheriting from the 
class. In fact, I find Gordon's sentence a bit unclear (sorry 
Gordon...). Maybe it would be a little clearer if you spoke about 
"interface *sharing*" instead of "interface inheritance":

"For C++ interface sharing REQUIRES class inheritance.  For 
Python/Smalltalk, it does not."

In Python or SmallTalk, you may have the same (dynamically figured out) 
interface without having to make classes inherit from a common ancestor. In 
C++ or Java, it would be impossible, as we saw above...

In fact, things are a bit unclear because the word "interface" may be used 
to represent two things:
- the actual set of methods with their signature (in the example, the 
single method "m" with no parameters and no return value);
- the type used to define this set (in the Java flavour of the example, the 
type "I")

The first exists in every OO language, but the second may not exist, like 
in Python or SmallTalk. But both are always inherited by class inheritance 
(at least in every OO language I know). In the GoF book, they only speak 
about the second. But you may of course choose to force yourself to always 
do interface sharing via inheritance. You may write the example above as 
follows:

class I:
  def m(self):
    raise NotImplementedError
class A(I):
  def m():
    print "I'm A"
class B(I):
  def m(self):
    print "I'm B"
def f(o):
  # Here, we may say the o has the "type" I
  o.m()
f(A())
f(B())

In fact, making things as if Python had interfaces is just a matter of 
discipline: in the function f, you choose to see the object o as an 
instance of class I, and not to call any method not defined in class I, 
even if you know the method will exist. For example, you may consider the 
following code incorrect:

class I:
  def m1(self):
    raise NotImplementedError
class A(I):
  def m1():
    print "I'm A"
  def m2(self):
    pass
class B(I):
  def m1(self):
    print "I'm B"
  def m2(self):
    pass
def f(o):
  # o is considered as an instance of I
  o.m1()  # is correct...
  o.m2()  # is incorrect...

Of course, the "incorrectness" of the second line only happens in your 
head, because the method "m2" exists on both classes A and B, so the Python 
interpreter would not complain at all. But you may want to consider it 
incorrect, because the method m2 does not exist in I... But to be able to 
achieve that, you'll need to carefully document your code to know what 
method calls are "legal" on the objects you handle. And this can only be 
achieved by comments in your code and/or code documentation via UML 
diagrams, for example. But it may be worth the effort, since, maybe, 
someday you'll want to write a class C whose instances can be passed to f. 
And at that time, it will be really easier to know which methods should be 
implemented on C just looking at I...

Hope it's a bit clearer now...
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From gcordova at hebmex.com  Fri Jun 14 12:00:46 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Fri, 14 Jun 2002 11:00:46 -0500
Subject: Type subclassing: bug or feature
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F0623301263BF1@mail.hebmex.com>

> 
> * Gustavo Cordova <gcordova at hebmex.com> [2002-06-13 16:18 -0500]:
> > maybe all methods of builtin types should do:
> > 
> > 	return self.__class__(return_value)
> > 
> > when returning a new object (like str methods).
> 
> Looks like a waste of performance in about 99 % of the time. 
> Hopefully,
> this can be optimized at the C level.
> 
> Gerhard
> 

Well, it *was* supposed to be pseudo-code, I'm sure
that at the C level it can be done in a much more
efficient way, without having to resort to Python code.

:-P

:-)

-gustavo

pd: TGIF!!




From gerhard.haering at gmx.de  Fri Jun 21 22:17:58 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 22 Jun 2002 04:17:58 +0200
Subject: Help with MySQLdb insert with data extracted from odbc <-- Oracle
In-Reply-To: <3d13d5fb.651829078@netnews.attbi.com>
References: <3d13d5fb.651829078@netnews.attbi.com>
Message-ID: <20020622021757.GA1372@lilith.my-fqdn.de>

* candiazoo at attbi.com <candiazoo at attbi.com> [2002-06-22 01:43 +0000]:
> Can someone tell me what this means?  I mean, I guess I *know* what it means,
> but why?
> 
> The insert statement into a mysql database is...
> 
> 'insert into works values(?, ?, ?, ?, ? ,?)'

Wrong. paramstyle = 'format' in MySQLdb, not 'qmark'. Look into the
DB-API specs (PEP 0249) for details about the various paramstyles that
are possible.

> The invocation of the dynamic sql is...
> 
>     status = cur_mysql.execute(sql, 
>         (wrk_inst, ZERO, trs_inst, aas_inst, title, NADA))

Simply use %s instead of ?. And read the fine MySQLdb docs that are
included :-)

Gerhard

PS: I don't like these sort of spam protection attempts.
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From peter at engcorp.com  Sat Jun 29 20:30:27 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 29 Jun 2002 20:30:27 -0400
Subject: newbie Q: delete list few lines in file
References: <7fe97cc4.0206291519.1c95ce31@posting.google.com>
Message-ID: <3D1E5123.FDEFADDF@engcorp.com>

Xah Lee wrote:
> 
> i have a folder full of individual email text file, whose ending may
> be the following:
> _______________________________________________
> some mailing list
> some at list.com
> http://some.com/ttt
> 
> I want to delete them, but only if they are the last few lines in the
> file.
> 
> if you can show me a script that'll be great otherwise my question can
> be broken down as:
> 
> * how to write to file inplace? (or, do i write to a new file, then
> delete original, and rename the new file?)

You cannot write files in place.  I don't understand why you would
want to, however, since your purpose was to delete them, not change
them.  (Or did I misunderstand?)

If you did have to modify, however, the tempfile module could be helpful 
for you here, as well as os.remove and os.rename.

> * as far as a newbie, i use "xreadlines" module. I wasn't able to get
> the total number of lines in a file.
> li=xreadlines.xreadlines(f)
> num=len(li)
> fails.

You shouldn't need to read the individual lines this way if you 
are looking for *exactly* that pattern.  Just do something like this:

f = open(filename)
data = f.read()
f.close()
if data.find('some mailing list\nsome at list.com\nhttp://some.com/ttt\n') >= 0:
    print 'Removing evil file %s' % filename
    os.remove(filename)

> I'm much more interested in learning python idiom and correctness
> here. I'm not interested in just "get it to work". For the latter, i
> can do well with perl. (also, is there a site or mailing list that
> collect very small trivial python programs for learning purposes?)

Well, Python idioms are all about just "getting it to work" :-),
but doing it cleanly and readably.  Other than that, it's just like
Perl programming. :-)

-Peter


From martin at came.sbg.ac.at  Tue Jun 11 11:59:36 2002
From: martin at came.sbg.ac.at (Martin Krallinger)
Date: 11 Jun 2002 08:59:36 -0700
Subject: python-gui-tk
Message-ID: <e0b1d6f6.0206110759.125af56d@posting.google.com>

Dear all,

I am interested in developing a gui to a programm written in c and
python.
i would like to construct a gui for it using tk.

what i need is basically a widget which has an option to load a file
or to past a certain text (protein sequence) in a proper field and
take it as an input run on it the c programm ( which needs just the
name of the input file and the name of the outputfile). there should
be an option to get a gnuplot out of the results,..(but this is at
this stage resudial).

if anyone has a  sample code similar to my problem (python-tkgui)it
would be very nice to have a look at it to get some hints, as i am not
very familiar to tk under python.

thanks a lot,

Martin


From karczma at info.unicaen.fr  Tue Jun 25 11:57:59 2002
From: karczma at info.unicaen.fr (Jerzy Karczmarczuk)
Date: Tue, 25 Jun 2002 17:57:59 +0200
Subject: Integrals with Python
References: <3D188551.6BA00DED@lanl.gov>
Message-ID: <3D189307.1417ECD5@info.unicaen.fr>

Nathan Given wrote:


> I'm somewhat familiar with MATLAB, and I'm trying to move completely
> over to python... and I was wondering if there is a function out there
> that will compute integrals for me...

> I already have Numeric... what else do I need?

Perhaps you might need to know WHAT integrals you need to compute.

* Adaptive algorithms?
* Singularities possible? Integration interval always finite or not?
* Multi-dimensional cases? Recursive algorithms?
* Monte-Carlo and other random generator based integration tools?
* etc.

Anyway, I don't want to sound harsh, but NO FUNCTION OUT THERE will
"compute integrals for you". *You* may compute some integrals with
the aid of some procedures, appropriately chosen. And if there isn't
any, then Matlab, Python, etc. are good tools to implement them 
yourself. Romberg, all Gaussian quadratures, etc. - nothing really
difficult, and in several cases it is wiser to make a specific, problem-
oriented procedure (e.g. in some cases of wildly oscillating functions)
than to use generic, off-shelf tools.


Jerzy Karczmarczuk
Caen, France


From mbelanger at pixelsystems.com  Tue Jun 25 10:00:02 2002
From: mbelanger at pixelsystems.com (Mathieu Belanger)
Date: 25 Jun 2002 07:00:02 -0700
Subject: Switch from 2.1 to 2.2
Message-ID: <f29c2586.0206250600.48e0a8e3@posting.google.com>

Hi all!

I was trying to switch my server from python 2.1 to python 2.2 when I
began to have some compatibility problems.

I'm a newbie in python and the following code was not written by me,
I'm only try to make it works!

	docType = implementation.createDocumentType('','','')
	doc = implementation.createDocument('',None,docType)
	parentNode = doc.createElementNS(None,'ACTION')
	parentNode.setAttribute("name", "RESULT")
	parentNode.setAttribute("DCId", DCSystemId)
	doc.appendChild(parentNode)
	RegSection1(doc,parentNode)
	FileSection1(doc,parentNode)
	SrvSection1(doc,parentNode)
	StationSection1(doc,parentNode)
	strDoc = StringIO()
	PrettyPrint(doc, strDoc)

I got an error in the PrettyPrint call. This call worked well before!!
Whats wrong now??

Thanks for your help!

Mathieu


From whisper at oz.net  Mon Jun  3 14:40:54 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 3 Jun 2002 11:40:54 -0700
Subject: HTML links in a GUI tkinter
In-Reply-To: <83y9dwqpw2.fsf@panacea.canonical.org>
Message-ID: <GCEDKONBLEFPPADDJCOEEEFKDGAA.whisper@oz.net>

>From the sound of his original post, he doesn't want to just open a site in
an external browser, he wants to have clickable links in a tk widget. Maybe
i'm missing something here though.

BTW, there is a Python binding for Richard Hipp's Tk html widget around
somewhere if one really wants webbishness in a tkinter app.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Kragen Sitaker
> Sent: Monday, June 03, 2002 9:38
> To: python-list at python.org
> Subject: Re: HTML links in a GUI tkinter
>
>
> "Steve Holden" <sholden at holdenweb.com> writes:
> > "Kragen Sitaker" <kragen at pobox.com> wrote in message
> > news:83elfprs1x.fsf at panacea.canonical.org...
> > > Sorry about the empty reply --- in Python 2.x, the webbrowser module
> > > has a webbrowser.open() method that does what you want.
> >
> > So, all Aurel needs to know now is how to run wxPython widgets inside
> > tkinter programs...
>
> Why would Aurel want to do that?  webbrowser.open() doesn't use
> wxPython.  At least, I don't have wxPython installed, and it works for
> me.
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From karim at nemetschek.net  Fri Jun  7 11:34:45 2002
From: karim at nemetschek.net (Karim Shehadeh)
Date: Fri, 7 Jun 2002 11:34:45 -0400
Subject: Session Cookies
Message-ID: <000d01c20e38$da94fdb0$4100a8c0@KARIMXP1>

I apologize if this has been covered recently, but I can't seem to find
any good examples on the web.  Can someone point me to some example code
for handling session cookies in python.  Or, even better, a generalized
user log-in system with python.  
 
Thanks for your help,
Karim
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020607/531193b6/attachment.html>

From cpbotha at dutidad.twi.tudelft.nl  Sun Jun  2 16:01:00 2002
From: cpbotha at dutidad.twi.tudelft.nl (Charl P. Botha)
Date: Sun, 2 Jun 2002 20:01:00 +0000 (UTC)
Subject: List
References: <3cfa6ed8_1@news.iprimus.com.au> <addrjh$imq$1@news.tudelft.nl> <3cfa7619_1@news.iprimus.com.au>
Message-ID: <addths$jb9$1@news.tudelft.nl>

In article <3cfa7619_1 at news.iprimus.com.au>, Gold Fish wrote:
> Charl P. Botha wrote:
> 
>> In article <3cfa6ed8_1 at news.iprimus.com.au>, Gold Fish wrote:
>>> Assume we got 2 list
>>> list 1 = ['ABC','CDG','DAV','FE','FECA','FEAFA','FEA']
>>> list 2 = []
>>> how can we copy element from list 1 to list 2 say i just want to copy the
>>> 3 element 'CDG','DAV','FE' to list2 only.
>> 
>> list1 = ['ABC','CDG','DAV','FE','FECA','FEAFA','FEA']
>> list2 = list1[1:4]
>> 
>> Is that what you meant?
>> 
> No i mean if i got from list1 i wanna take 3 elements in any position to 
> list2 but i don't know how many item in that list.

Could you rephrase that?  The example I posted does exactly what your
example showed.  Maybe if you gave another example and tried to explain more
clearly, I'd know what you were trying to do.

-- 
charl p. botha http://cpbotha.net/ http://visualisation.tudelft.nl/


From aahz at pythoncraft.com  Sat Jun 29 16:33:50 2002
From: aahz at pythoncraft.com (Aahz)
Date: 29 Jun 2002 16:33:50 -0400
Subject: can't import generators
References: <m2hejlc060.fsf@mother.paradise.lost>
Message-ID: <afl5je$4iu$1@panix1.panix.com>

In article <m2hejlc060.fsf at mother.paradise.lost>,
John Hunter  <jdhunter at nitace.bsd.uchicago.edu> wrote:
>
>I am having trouble working with generators.  Anyone know what might
>cause this behavior:
>
>Python 2.2.1 (#1, Apr 22 2002, 21:20:54) 
>[GCC 3.0.4] on linux2
>Type "help", "copyright", "credits" or "license" for more information.
>>>> from __future__ import generators
>Traceback (most recent call last):
>  File "<stdin>", line 1, in ?
>ImportError: cannot import name generators

Can't do this in interactive mode.  Use a script.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From ark at research.att.com  Tue Jun 25 11:57:59 2002
From: ark at research.att.com (Andrew Koenig)
Date: Tue, 25 Jun 2002 15:57:59 GMT
Subject: GOTO w/ Python?
References: <mailman.1024964672.10880.python-list@python.org>
Message-ID: <yu99y9d3jqpk.fsf@europa.research.att.com>

Tim> COME FROM is a joke. Literally. It is a parody of GOTO whose only
Tim> redeeming feature is that it is not GOTO.

The "Come From" statement sucks.  Literally.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From ark at research.att.com  Mon Jun 24 10:04:03 2002
From: ark at research.att.com (Andrew Koenig)
Date: Mon, 24 Jun 2002 14:04:03 GMT
Subject: Python hits the spot
References: <mailman.1024868430.3289.python-list@python.org>
Message-ID: <yu99sn3chiy4.fsf@europa.research.att.com>

Tim> If it is, it's a serious misfeature of Linux's overcommitment
Tim> policies.  Python does a malloc() sufficient for 50*1000*1000
Tim> list elements at the start (about 200Mb), malloc doesn't return
Tim> NULL, and Python believes it can use the memory.  What would you
Tim> like it to do instead?

Don't some implementations of malloc offer a way to ask the operating
system to commit to the memory being available when you allocate it?
If so, I can imagine testing for that facility at config time and
using it if available.

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From m.laranjeira at datacraft.com.br  Thu Jun 13 11:55:17 2002
From: m.laranjeira at datacraft.com.br (Marcus Vinicius Laranjeira)
Date: Thu, 13 Jun 2002 12:55:17 -0300
Subject: For and Dates
Message-ID: <67F0CEAF5F0ED411B53900508BC5C8BC700195@admin.datacraft.com.br>

All,

I need to do a loop that goes from 2002-6-1 to 2002-7-31, but I don't know
how to do that...

Is there a way to do such range using for !?

Do you have any suggestions !?

Thanks in advance,

Marcus Vinicius Laranjeira
MON Center

Datacraft do Brasil - (http://www.datacraft.com.br)





From hesterloli at hotmail.com  Wed Jun 19 04:19:45 2002
From: hesterloli at hotmail.com (George Hester)
Date: Wed, 19 Jun 2002 08:19:45 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com> <mailman.1024460666.24801.python-list@python.org>
Message-ID: <B8XP8.47925$GY.15259265@twister.nyroc.rr.com>

I don't want to learn the language.  At least not right now.  This started
becuase I read there was a new version of Python.  Then I saw an ASP at
Microsoft that used Python.  So I thought I would run an ASP to determine my
version of Python.  And then I thought I would ask how to do that here.  WRONG!

--
George Hester
_________________________________
"Chris Gonnerman" <chris.gonnerman at newcenturycomputers.net> wrote in message
news:mailman.1024460666.24801.python-list at python.org...
> ----- Original Message -----
> From: "George Hester" <hesterloli at hotmail.com>
>
>
> > The whole issue started from here.  I did not deserve to be told to "crawl
> > before I walk."  I may have responded to that inappropriately I confess.
> But
> > that does not excuse the implied derogatory.
>
> Have you ever used USENET before?  You MUST be kidding!  comp.lang.python is
> the
> nicest group of people anywhere on USENET... we regularly help users such as
> yourself.  You received EXCELLENT assistance with no derogatory comments, at
> least until you lost your senses and started whining.
>
> On most programming newsgroups you would have been told RTFM; if you were
> lucky,
> a web page reference or manual page would be mentioned, but in many cases
> you'd
> be blown off.  You got a correct, but incomplete, answer, because we
> naturally
> assumed you were a programmer and just needed a pointer.  Note that when you
> whined that you needed a complete answer you got one; you could have gotten
> the
> same thing without whining.
>
> You ARE a nerd, right?  Nerds (programmers, engineers, scientists, etc.)
> tend to
> be among the most blunt and direct of all people.  Expect if you ask a
> simple
> question, you'll get an answer which directs you to the place where you
> would have
> found the answer if only you had looked.  If you can't take the heat, get
> out of
> the kitchen.
>
> > Since that the lambasting crecscendoed.
>
> Since you started to call us unfair, actually.  Get a thicker skin or a more
> humble attitude before you ask questions in the future.
>
> Read the manual, already.  Learn to use Python from the interactive
> prompt...
> it's the best place to learn the basics.  Read that page over on ESR's site
> about
> asking questions on USENET, because you really need to understand it:
>
> http://www.tuxedo.org/~esr/faqs/smart-questions.html
>
> That page contains the RULES.  Learn them.  Expect to be treated just like
> everyone
> else.
>
> And, stop whining.  The correct way for you to respond to this posting, and
> all
> the others, is to NOT RESPOND.  The second-best thing would be a SINGLE
> apology.
> Don't keep telling us how we misunderstand, or how we are unfair; that just
> marks
> you as a troll, and frankly that IS derogatory.  Don't live up to it (or
> down to
> it, whichever) and you'll never hear it.
>
>
> Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
> http://newcenturycomputers.net
>
>
>
>




From gupsgrotto at no-spam.netscape.net  Tue Jun 25 10:37:45 2002
From: gupsgrotto at no-spam.netscape.net (grant)
Date: Tue, 25 Jun 2002 14:37:45 GMT
Subject: os.kill for win32
Message-ID: <pan.2002.06.24.14.40.03.222610.147@no-spam.netscape.net>

Anyone know the same functionality of the os.kill module for the win32
platform.  All examples that I've seen are just for Unix.  In the program
that I've written I'm lauching an external program via a wxCheckbox
switch and would like to kill the program when the switch is unchecked
however wxKill only works on Unix, and os.kill is the same.  Is there any
relatively simple way to do this same thing on win32 (NT platform)

Thanks

Grant (Da Gup)


From BPettersen at NAREX.com  Sat Jun 29 17:24:07 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Sat, 29 Jun 2002 15:24:07 -0600
Subject: private
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158F11@admin56.narex.com>

> From: Dave Reed [mailto:dreed at capital.edu] 
> 
> > From: "Bjorn Pettersen" <BPettersen at narex.com>
> > 
> > > From: Dave Reed [mailto:dreed at capital.edu]
> > > 
> > [snip private members]
> > > 
> > > Kind of off-topic, but this got me curious. The only way I
> > > could think of is to assume (which I think is always true) 
> > > that the private data members are in order from the starting 
> > > address so based on the offset from the address of the object 
> > > you could access each one.
> > 
> > .. and of course #define private public, or just go in and 
> edit the .h 
> > file...
> > 
> > -- bjorn
> 
> Ok, I was thinking about doing it w/o recompiling the object 
> or lib file containing the class.

No reason to recompile as long as you make binary compatible changes to
the header file (which all of the suggestions above are).

-- bjorn




From imcmeans at home.com  Fri Jun 28 02:23:08 2002
From: imcmeans at home.com (Ian McMeans)
Date: Fri, 28 Jun 2002 06:23:08 GMT
Subject: metaclasses vs. inheritance
Message-ID: <ghTS8.58749$mh.1826422@news1.telusplanet.net>

Can someone give an example of a problem that is solved with metaclasses,
which would be difficult or impossible to solve with inheritance?

I (very barely) understand metaclasses, but to me it just seems like
inheritance, where the classes who get the benefit of the meta-class get the
metaclass' features.

PS - what's the antonym of meta?




From observer at NOSPAM.space.pl  Mon Jun 17 08:59:40 2002
From: observer at NOSPAM.space.pl (Johann)
Date: Mon, 17 Jun 2002 14:59:40 +0200
Subject: PythonWorks Pro problem!
Message-ID: <q3nrgu019nlnc2bc19jaouhfo0833fknrj@4ax.com>

At 
http://www.pythonware.com/products/works/articles/editprefs.htm

I read the following:

"Once your settings (editprefs) are registered you can change the
current settings by clicking on the small colored icon next to the
bindings symbol. This will cycle through all available settings.
Alternatively right-click in the editor view and select the desired
setting from the popup menu."

I registered some examples from above link. The problem is, I cannot
find  this "small, coloured icon" , so I cannot cycle through avalable
settings. :-(

I tried to clink on everything, but without results :( Anybody can
help?

--
Johann




From tim.one at comcast.net  Fri Jun 21 18:25:54 2002
From: tim.one at comcast.net (Tim Peters)
Date: Fri, 21 Jun 2002 18:25:54 -0400
Subject: How to represent the infinite ?
In-Reply-To: <slrnah6oc4.amp.huaiyu@gauss.almadan.ibm.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEDIPPAA.tim.one@comcast.net>

[Huaiyu Zhu]
> ...
> Now I'm confused.  I thought from previous discussions that you
> intended to disable inf and nan until full ieee floating point
> compliance is implemented in Python.

I intended that x**y act the same as math.pow(x, y).

> And I infered that that's in order to make Windows work.

It has nothing to do with Windows specifically; Python's 754 behavior "is
random" on all platforms, in platform-specific ways, because Python is
implemented in C, platform C 754 behavior "is (still) random", and the only
code ever contributed to address this sorry state of affairs is the fpectl
module (whose purpose is in a direction opposite the one you favor).

> Since it could be made to work on Linux, I simply commented out
> the checking in pyport.h.  Voila, instant ieee compliance!

You happened to get one particular 754 behavior you like; full IEEE
compliance is a large and difficult task; subset compliance is worse than
none (if we're to believe Kahan's virulent rants about Java <wink>).

> Your example appears to show that Windows can also live with nan and inf.

Virtually all platforms can, given enough platform-specific support code,
almost none of which Python has.

> So my question is, are there any significant platforms out there
> that still really need such checkings which make nan and inf almost
> useless?

If Python were 100% 754-conforming, Guido and I would still favor enabling
exceptions on overflow, divide-by-zero, and invalid operation, by default.
As things are, you've got no choice at all, unless you want such exceptions
enabled and are on a platform where the fpectl module happens to work.





From emile at fenx.com  Wed Jun  5 21:59:25 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 06 Jun 2002 01:59:25 GMT
Subject: Thought on PEP 204 and 276
References: <mailman.1023308781.5126.python-list@python.org>
Message-ID: <1mzL8.120091$cQ3.3861@sccrnsc01>

<James_Althoff at i2.com>
> [phil hunt]
> >I used to think that
> >
> >   fir i in x:
> >
> >where (x) is an intreger, was confusing. But lately, I've
> >found myself accidently spontaneously typing it myself
> >instead of range(x) so i guess it makes intuitive sense
> >after all.
>
> I "accidently spontaneously" type it too.  But of course, I'm not very
> objective (oops, dangerous wording <wink>) about this.
>

Isn't this how we got into it last time... ;-)


--

Emile van Sebille
emile at fenx.com

---------



From mickey at tm.informatik.uni-frankfurt.de  Sun Jun 16 06:45:38 2002
From: mickey at tm.informatik.uni-frankfurt.de (Michael 'Mickey' Lauer)
Date: 16 Jun 2002 12:45:38 +0200
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <mailman.1024166060.26968.python-list@python.org>
Message-ID: <3d0c6c52@nntp.server.uni-frankfurt.de>

Jean Pierre DENYS <stochastic at eresmas.net> wrote:
> Do you know where I could find the GRAD system?

IMHO it has been superseded by the Basil project:
http://wildideas.org/basil/

Yours,

:M:



From sasoft at gmx.de  Sun Jun 23 19:52:44 2002
From: sasoft at gmx.de (Stefan Antoni)
Date: Sun, 23 Jun 2002 23:52:44 +0000
Subject: Python hits the spot
In-Reply-To: <yppfsn3dlpco.fsf@hotlips.cs.chalmers.se>
References: <mailman.1024772248.9399.python-list@python.org> <yppfsn3dlpco.fsf@hotlips.cs.chalmers.se>
Message-ID: <20020623235244.GA13370@blackscarab>

* P?r Kurlberg <kurlberg at math.chalmers.se> [020623 22:46]:

 
> for i in range(50*1000*1000):
>    pass
 
> Can someone else reproduce this behaviour?  (Make sure to quit all
> important applications before you try!)

following:

>>> for i in range(50*1000*1000):
...  pass
... 
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  MemoryError


> Some trivia:
> Python 2.2.1 (#1, Jun 12 2002, 17:57:17)
> [GCC 2.96 20000731 (Mandrake Linux 8.1 2.96-0.62mdk)] on linux2

Python 2.3a0 (#1, Jun 23 2002, 13:54:31) 
[GCC 2.96 20000731 (Linux-Mandrake 8.0 2.96-0.48mdk)] on linux2




From paul1brian at yahoo.com  Wed Jun 19 08:03:02 2002
From: paul1brian at yahoo.com (Paul Brian)
Date: Wed, 19 Jun 2002 13:03:02 +0100
Subject: Python to MS SQL Server
References: <3D0F8099.8080106@matteicos.com>
Message-ID: <1024488185.1604.0.nnrp-14.c1c3e11b@news.demon.co.uk>

Richard,

I can thoroughly recommend mx.ODBC from egenix.com
It is a robust package at a reasonable price, and just works. You also get
mx.datetime which is v good as well.
I think it is a 30 day evaluation.

Below is code snippet on connecting to an sql server, assuming you have ODBC
drivers loaded up on your (windows) machine already. (the server address
could be DNS, or netbios, depending on how you connect now. does not really
matter, but is nice to be consistent)
#note the string os on continuos string, formatting a bit off here.

---------------------
#ODBC connection string - see msdn for more details, but this covers most
eventualities
testdsn = r''driver={SQL Server};
server=mySQLServer;uid=webserver;pwd=webserver;database=dbtest;app=Example"

db = mx.ODBC.Windows.DriverConnect(testdsn)

c = db.cursor()
c.execute('select * from tblTest')
z = c.fetchall()

pprint.pprint(z)

c.close()
db.close()

del(c)
del(db)
---------------------------


"Richard Kessler" <richardkessler at matteicos.com> wrote in message
news:3D0F8099.8080106 at matteicos.com...
> I am interested in using Python in a MS SQL Server environment, but do
> not see any database modules out there, especially for native versus
> using ODBC. Is there such a module?
>
> Thanks in advance.
>
> - Richard
>




From johnroth at ameritech.net  Tue Jun 25 08:07:56 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 25 Jun 2002 08:07:56 -0400
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com> <lkznxqjqx2.fsf@pc150.maths.bris.ac.uk> <3D11A21D.9010809@mxm.dk> <uh3e1usajle42c@news.supernews.com> <3D11DA9B.163A7BD0@info.unicaen.fr> <uh446elev7af68@news.supernews.com> <3D12EEAA.9C3A3D29@info.unicaen.fr> <6v30fa.sm1.ln@grey.aerojockey.localdomain> <3d1730f2$1_1@hpb10302.boi.hp.com>
Message-ID: <uhgn6pd0h2ve1e@news.supernews.com>

"Daniel Fackrell" <dfackrell at DELETETHIS.linuxmail.org> wrote in message
news:3d1730f2$1_1 at hpb10302.boi.hp.com...
> After skimming the linked page to get a feel for the CAME FROM
statement, is
> it conceptually different from short procedures?  Does the only
difference
> lie in the fact that for compiled languages one of the instances will
be
> inline?  I do understand that it would have some effect on variable
scoping,
> but this can be easily handled in other ways.
>
> So what's the benefit that might outweigh the opacity of the code
created?

It's a joke! Really.

John Roth





From quinn at lira.ugcs.caltech.edu  Sun Jun 16 15:12:25 2002
From: quinn at lira.ugcs.caltech.edu (Quinn Dunkan)
Date: 16 Jun 2002 19:12:25 GMT
Subject: Persistence
References: <90cc01ee.0204110958.6945c776@posting.google.com> <a94qk5$22mg@r02n01.cac.psu.edu>
Message-ID: <slrnagpooo.heq.quinn@lira.ugcs.caltech.edu>

On Thu, 11 Apr 2002 16:10:17 -0400, Rajarshi Guha <rxg218 at psu.edu> wrote:
>On Thursday 11 April 2002 13:58 in comp.lang.python Riccardo de Maria wrote:
>
>> Is there a way to save in a file the status, objects, functions of
>> interpreter
>> in order to restore it after the interpreter has been closed?
>> 
>> If it is possible, one can work on a project stop and then continue
>> later adding functions, objects and so on.
>
>I have a related question - does the OP refer to object persistence or 
>object serialization? Is there a difference between the two?
>
>I looked at the shelve and pickle modules - the former uses a dbm database 
>whereas the latter just dumps an ASCII representation of the object. The 
>fact that shelve uses a dbm indicates that it should be more efficient. Is 
>this true? When would I choose one over the other?

That's only true if you leave bin=0 when you call dump().  For best efficiency,
use cPickle with bin=1.

A quick glance at shelve.py reveals that it uses pickle, so it can hardly
be more efficient.

The difference is that pickle takes a single object and serializes it to a
string, which can then be written to a file.  shelve does the same thing only
it writes it to a dbm, so you could have many pickles and retrieve them by
name.

If you don't have lots of pickles or if you're using ReiserFS, you could get
the dbm effect by just sticking a bunch of pickles in a directory.  Lately,
I've been "pickling" by just doing 'f=open(obj_name); pprint.pprint(obj, f)'
and eval(open(obj_name).read()), which has the advantage of giving easy-to-
read, easy-to-edit files.  Works nicely if your data is simple.

> On Thu, 11 Apr 2002 16:10:17 -0400, Rajarshi Guha <rxg218 at psu.edu> wrote:

Of course, everything said is with the utmost respect.  As ever, I am y'r
humblest servant.  If you please, and if it is not any trouble, I would
appreciate a kind word to some of the higher-ups.  If there are any.

Operation Rear-view Mirror has been completed and I await further instructions.


From plugin at supercable.es  Sun Jun 23 18:55:56 2002
From: plugin at supercable.es (Ugo García)
Date: Mon, 24 Jun 2002 00:55:56 +0200
Subject: Redirecting StdOut
References: <1024856420.145085@seux119> <Xns9236D1FF016AAcliechtigmxnet@62.2.16.82> <1024860416.580309@seux119> <3D163550.E525D840@engcorp.com>
Message-ID: <1024873013.652528@seux119>

"Peter Hansen" <peter at engcorp.com> escribi? en el mensaje
news:3D163550.E525D840 at engcorp.com...
> "Ugo Garc?a" wrote:
> >
> > perfect. This is what I was looking for. Now I'd make a C module to
handle
> > the messages written to stdout & stderr in order to show them in the
screen.
> > Thanks a lot
>
> Why would you bother with C for a game that relies on simple print
> statements for its interface?
The game not only consist of printing message (that's not a game... :-) ).
What I'm making is a engine to make Graphical Adventures.

 Can't you implement the game entirely in Python?
Probably it would be implemented entirely in Python. Why not? SPEED. I'm
familiarized with C and prefer to make the core of the engine in C, which is
faster and more reliable for this kind of things, and leave Python be ONLY
the script language of the engine.

Thanks,
--ugo

>
> -Peter




From peter at engcorp.com  Tue Jun  4 23:26:44 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 04 Jun 2002 23:26:44 -0400
Subject: writing a chat program
References: <d4c18195.0206040840.7ece1b8c@posting.google.com>
Message-ID: <3CFD84F4.D50D5BBC@engcorp.com>

Shaun Whelden wrote:
> 
> I'm writing this on a Windows machine, so curses is out of the
> question.  Also, I want this to be a text based program, so I don't
> want to have to mess around with Tkinter or any other forms of GUI
> unless I have to.  I'm pretty sure that I need to do something
> involving thread.py and/or threading.py, but I've never used a thread
> before in my life.  Can anyone help?

Check out Aahz's tutorial(s) at http://starship.python.net/crew/aahz/

All you need to know, and more. :)

-Peter


From aldo at nullcube.com  Sat Jun 29 00:36:54 2002
From: aldo at nullcube.com (Aldo Cortesi)
Date: Sat, 29 Jun 2002 14:36:54 +1000
Subject: metaclasses vs. inheritance
In-Reply-To: <3D1C7EF1.7060002@earthlink.net>
References: <ghTS8.58749$mh.1826422@news1.telusplanet.net> <mailman.1025259037.9256.python-list@python.org> <3D1C7EF1.7060002@earthlink.net>
Message-ID: <20020629043654.GA28445@nullcube.com>

Thus spake Hans Nowak (wurmy at earthlink.net):

> Aldo Cortesi wrote:
> 
> >>PS - what's the antonym of meta?
> >
> >Hmmm... the meta- in "metaclass" or "metaphysics" really
> >has no relation to the original Greek, which means
> >"with", or "among" (as in "metatarsus" - the bones in the
> >foot which lie next to the tarsus). So whatever antonym
> >you come up with can't have anything to do with the
> >original etymology of the prefix...  Erm... How about
> >mundane? Or actual?  Intrinsic? Essential? Or even "per
> >se"....?
> 
> In ancient Greek, it also meant "over", like Latin
> "trans"... e.g. "over the river". I think the "meta" in
> "metaclass", "metadata" etc is related to this meaning.


Hi Hans,


Are you sure about that? 


In actual fact, the interesting thing about the meta- prefix
as used in modern English, is that it is entirely based on a
misapprehension of the Ancient Greek - indeed, on the exact
same error you made above... If you have a copy of the OED
handy, look up the etymology of "metaphysics" (which is the
precursor to all of the other common meta- usages in
English). My copy (Second Edition, 1989) says that the term
derives from an Aristotelean phrase "ta meta ta physika"
(excuse the clumsy transliteration), which means "that which
comes after the Physics". It then goes on to note that:


	"[This term] came to be misinterpreted as meaning
	'the science of things transcending what is physical
	or natural'. This misinterpretation is found, though
	rarely, in Greek writers, notwithstanding the fact
	that "meta" does not admit of any such sense as
	'beyond' or 'transcending'. In scholastic Latin
	writers the error was general (being helped,
	perhaps, by the known equivalence of the prefixes
	meta- and trans- in various compounds); and in
	English its influence is seen in the custom,
	frequent down to the 17th c., of explaining
	metaphysical by words like 'supernatural',
	'transnatural', etc."



Cheers,



Aldo




-- 
Aldo Cortesi
aldo at nullcube.com




From irmen at NOSPAMREMOVETHISxs4all.nl  Sat Jun 15 17:10:37 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Sat, 15 Jun 2002 23:10:37 +0200
Subject: type(socket) != SocketType
Message-ID: <3D0BAD4D.5050105@NOSPAMREMOVETHISxs4all.nl>

Hello

I have to detect if a certain socket object is a regular socket,
or a M2Crypto SSL socket. I do this by checking type(sock),
and then checking if this is a regular socket type. Why this way?
because the systems that run my code do not necessarily have
M2Crypto installed, and thus I cannot compare the socket to
any class/type from M2Crypto, nor just go ahead and catch SSL
exceptions (simply because they are not defined!).

There is a problem, however.
With Python 2.2.1 on Windows, type(socket) != SocketType:

 >>> from socket import *
 >>> s=socket(AF_INET,SOCK_STREAM)
 >>> type(s)
<type 'instance'>
 >>> SocketType
<type '_socket.socket'>
 >>> type(s) is SocketTye
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
NameError: name 'SocketTye' is not defined
 >>> type(s) is SocketType
0
 >>> type(s)== SocketType
0
 >>>


I find this most unfortunate, because the docs on the socket module
explicitly state that type(socket)=SocketType.

BTW on Linux, all is fine.

Any clues how to fix things on windows?

Thanks a lot,
Irmen



From bbeck at REMOVE-THIS-TO-EMAIL-ME.austin.rr.com  Fri Jun 28 14:40:05 2002
From: bbeck at REMOVE-THIS-TO-EMAIL-ME.austin.rr.com (Brandon Beck)
Date: Fri, 28 Jun 2002 18:40:05 GMT
Subject: Plugins in Python
References: <afgeol$sgd$1@mulga.cs.mu.OZ.AU> <3D1BD350.8F0C74FF@alcyone.com>
Message-ID: <3D1CAD56.6070500@REMOVE-THIS-TO-EMAIL-ME.austin.rr.com>

Erik Max Francis wrote:
> Lars Yencken wrote:
> 
> 
>>I'm trying to do plugins for a program I'm writing in Python, and I'm
>>not
>>sure how exactly to do it. Basically what I want it to be able to
>>write a
>>module which loads every python file in a directory and stores an
>>instance
>>of each one that loaded successfully.
> 
> 
> The most obvious implementation is simply to have them each imported as
> modules with __import__, and define some kind of simple
> initialization/shutdown API that will be called upon import.
> 

I've recently written something which pretty much does exactly this 
using the imp module.

The source is browsable and HTML viewable at:
	http://isomorphism.org/~bbeck/bot

You probably want to specifically look at the PluginManager class which 
is a class inside of bot.html.

Hope this helps some.



From pyth at devel.trillke.net  Tue Jun 18 14:26:11 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 18 Jun 2002 20:26:11 +0200
Subject: Pop a list from beginning ? and memory saving...
In-Reply-To: <BIEJKCLHCIOIHAGOKOLHAELEDEAA.tim.one@comcast.net>; from tim.one@comcast.net on Tue, Jun 18, 2002 at 01:51:53PM -0400
References: <aem8e8$4kq$1@news.idiom.com> <BIEJKCLHCIOIHAGOKOLHAELEDEAA.tim.one@comcast.net>
Message-ID: <20020618202611.F15079@prim.han.de>

[Tim Peters]
> [James T. Dennis]
> > ...
> >  So, the critical question is:
> >
> > 	Is the list's reverse() method done in place
> 
> Yes.
> 
> > and does it run in linear time
> 
> Yes.
> 
> > (or better?
> 
> No.  It does len(list)/2 swaps.

in another thread we had some discussion about a
reverse-iterator.  You could write

    for item in iters.reverse(list):
        # item iterates over list's items backwards
        # but the list itself stays unmodified. 

Do you agree that this would be a good idea?

please-share-your-brain-ly y'rs, holger




From mdk at mdk.com  Thu Jun 20 15:14:02 2002
From: mdk at mdk.com (mdk)
Date: 20 Jun 2002 19:14:02 GMT
Subject: Proper number alignment
References: <aet5hu$9leln$1@ID-98166.news.dfncis.de>
Message-ID: <aet9hp$9vodv$1@ID-98166.news.dfncis.de>

mdk <mdk at mdk.com> wrote in news:aet5hu$9leln$1 at ID-98166.news.dfncis.de:

Both options work great.  Thanks for your time.



From fredrik at pythonware.com  Thu Jun 13 18:11:59 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 13 Jun 2002 22:11:59 GMT
Subject: PIL _imaging
References: <Xns922AA829A8C5FHotMailcs@130.133.1.4> <3D06E694.4040103@mxm.dk> <Xns922B3150A6B32HotMailcs@130.133.1.4> <3D074529.E848AA6@ipm.fhg.de> <Xns922B561EDB212HotMailcs@130.133.1.4> <3D07BCC1.2000002@mxm.dk> <Xns922BB340B788BHotMailcs@130.133.1.4> <9C_N8.11930$p56.3660854@newsb.telia.net> <Xns922C48A439D35HotMailcs@130.133.1.4>
Message-ID: <PM8O8.43167$n4.9913522@newsc.telia.net>

"Cousin Stanley" wrote:

> However, I'm confused as to whether or not
> part of  PIL v1.1.3  must be re-compiled/re-built
> before using it ...

like all other binary extensions (PYD files) on windows,
the _imaging module must be compiled for the Python
version you're using.

the PIL/*.py part of the library works for all versions
on all platforms.

</F>




From siegfried.gonzi at kfunigraz.ac.at  Tue Jun 25 03:28:34 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Tue, 25 Jun 2002 09:28:34 +0200
Subject: SOLVED (Re: Python hits the spot)
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at> <af4vrl$ffe$0@216.39.172.122> <3D16DD5C.1050708@kfunigraz.ac.at> <uhdume2n1mhbba@corp.supernews.com> <3D171291.2060905@kfunigraz.ac.at> <af8ib7$k51$0@216.39.172.122>
Message-ID: <3D181BA2.3090009@kfunigraz.ac.at>

Bengt Richter wrote:

> Well, lucky guess on the heat. That _probably_ accounts for laptop
> going to 87% and then shutting down, but how about the other huge
> slowdowns in your big loop of 8? Did you get that diagnosed too?
> Just curious ;-)

On my stationary machine definitely not. The 8 steps remain constant. I 
am contemplating to buy a special cooler for may laptop (I saw some in a 
computer magazine).


Regards,
S. Gonzi



From eric.brunel at pragmadev.com  Tue Jun 11 08:06:05 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Tue, 11 Jun 2002 12:06:05 +0000
Subject: Sorting list (maybe stupid)
References: <slrnagbh7m.h03.jarvin24@lehtori.cc.tut.fi> <ae4gnp$l18$1@news.tudelft.nl>
Message-ID: <ae4hn9$jm4$1@wanadoo.fr>

Charl P. Botha wrote:

> In article <slrnagbh7m.h03.jarvin24 at lehtori.cc.tut.fi>, Jarvinen Petri
> wrote:
>> I have a problem and can't solve it:
>> 
>> There is lines of text like:
>> 
>> 2. Header2
>> 1. Header1
>> 102. Header102
>> 
>> and so on...
>> 
>> And I need to sort them according to the number (1,2,3,4...)
> 
> Put each line in a tuple, first element the number, second element the
> header.  Make a list of these tuples, and then use list.sort() with a comp
> function that compares using the first element in the tuple.

I think it's useless. Just try:

l = [(102, 'foo'), (2, 'bar'), (1, 'baz')]
l.sort()
print l

Apparently, the default sort function already sorts tuples according to 
their first element...
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From aahz at pythoncraft.com  Wed Jun 19 17:15:31 2002
From: aahz at pythoncraft.com (Aahz)
Date: 19 Jun 2002 17:15:31 -0400
Subject: OS X debugger?
References: <aeqotj$dm$1@panix1.panix.com> <aeqp4m$c09$2@peabody.colorado.edu>
Message-ID: <aeqs9j$cfn$1@panix1.panix.com>

In article <aeqp4m$c09$2 at peabody.colorado.edu>,
Fernando =?ISO-8859-1?Q?P=E9rez?=  <fperez528 at yahoo.com> wrote:
>Aahz wrote:
>>
>> Anyone know a good visual debugger for OS X???I'd?originally?been
>> planning to just use shell windows for the demos of my OSCON tutorial,
>> but it occurs to me that a single-stepping debugger with watch variables
>> would be even better
>
>You can probably build ddd for OS X. I've used it for python in the past and 
>it's quite nice (but I've only used it in Linux, so I don't know how tricky 
>the osX build would be. It's Motif based).

>From some quick Googling, it appears that a bigger issue would be lack
of support for Python 2.2.  Have you used DDD with 2.2?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From aahz at pythoncraft.com  Fri Jun 14 22:04:10 2002
From: aahz at pythoncraft.com (Aahz)
Date: 14 Jun 2002 22:04:10 -0400
Subject: Using Python to script a game?
References: <cdc2589b.0206141718.7dc2ae13@posting.google.com>
Message-ID: <aee7aq$iaf$1@panix1.panix.com>

In article <cdc2589b.0206141718.7dc2ae13 at posting.google.com>,
Ben Blonder <ben at sourcecod.com> wrote:
>
>I'm writing a 3D naval tactics in C++ for the Mac. I'm planning on
>having all the object behaviors, movements, and AI controlled through
>a scripting layer. The scripts would be text files the game would
>parse and then carry out actions. All the scripts would be read at
>runtime, not compile time, so they'd all have to be interpreted.

Is this real-time or turn-based?  If the latter, I'd recommend scrapping
your C++ code and going with straight Python.

>Can I use Python to accomplish this, and if so, can you point me to
>some tutorials on integratring Python with my existing code?

Integrating might be a bit difficult if you haven't already designed
your application specifically for that purpose.  If it's a real-time
game, check out Pygame, http://www.pygame.org/
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From stuq at dstoys.com  Thu Jun  6 20:39:15 2002
From: stuq at dstoys.com (Stuart Quimby)
Date: 6 Jun 2002 17:39:15 -0700
Subject: Medium to Large Scale Python Deployments
References: <KHkL8.21383$eD2.2399923271@newssvr10.news.prodigy.com>
Message-ID: <34b82fda.0206061639.67964684@posting.google.com>

You may be interested in the gnuenterprise project.  They are
developing a multi-tier db independant ERP that is open-source. 
Details can be found at http://www.gnuenterprise.org.

Stuart Quimby


"Domenic R. Merenda" <dom at edgereport.put_a_c_o_m_here> wrote in message news:<KHkL8.21383$eD2.2399923271 at newssvr10.news.prodigy.com>...
> Greetings,
> 
>    I am running a home-grown Enterprise Resource Planning system, written
> entirely in Python, to manage the operations and manufacturing aspects of a
> medium ($100,000,000 annual sales) manufacturing organization.  I would be
> interested in hearing from other individuals who are likewise "betting the
> farm" on Python as an ERP or MRP solution, and some of the experiences they
> may have had.
> 
>    We have found Python to be a robust solution for the problems we've
> encountered to date, and pass approximately half a gigabyte of daily data
> through our custom systems.  We have merged the AS/400 (Python 2.1, soon to
> go to 2.2.1) with midrange (PIII700) PC's to create an interesting
> heterogenous environment.
> 
> 
> Cheers!


From todd.harris at honeywell.com  Wed Jun 19 12:21:40 2002
From: todd.harris at honeywell.com (Todd Harris)
Date: Wed, 19 Jun 2002 11:21:40 -0500
Subject: Newbie Q: File accessing
Message-ID: <aeqb18$k10$1@phxtst22.phx1.aro.allied.com>

Hi, I apologize if this question has been beaten to death or if there's just
something that I haven't been seeing, but here's my problem (with a little
bit of background):

I have a file out on the network somewhere that stores info as text. I'm
writing a Python script that updates that file. However, it's a multi-user
application, and so my problem is that one user's script may be in the
middle of writing to that file while another user's script is attempting to
read that file or <even worse> writing to that file as well.

For those of you who use Microsoft <gulp...flame shield on> Word, when a
document is opened by one user, and another user tries to access it, they
can open it for read only. Obviously, Word has made some file attribute
where it is locked to that first user, and only that first user can write to
it while they have it open. Is there a way to do this in Python? Or what
other possible options do I have? I've looked been using the open(
filePath ) function as well as the os.open( ) function and os.access( ) to
check my progress, but can't seem to figure out how to get around this.
Thanks,

Todd




From spam at melkor.dnp.fmph.uniba.sk  Wed Jun 19 05:52:28 2002
From: spam at melkor.dnp.fmph.uniba.sk (Radovan Garabik)
Date: Wed, 19 Jun 2002 11:52:28 +0200
Subject: procmail replacement in Python
References: <mailman.1024444823.19953.python-list@python.org> <3u8pea.617.ln@127.0.0.1> <mailman.1024476024.32540.python-list@python.org>
Message-ID: <s8kpea.992.ln@127.0.0.1>

Gerhard H?ring <gerhard at bigfoot.de> wrote:
 : * Radovan Garabik <spam at melkor.dnp.fmph.uniba.sk> [2002-06-19 08:38 +0200]:
 :> Gerhard H?ring <gh_pythonlist at gmx.de> wrote:
 :>  : Does such a thing exist? I've recently looked into the procmail sources
 :> 
 :> <shameless advertising plug>
 :> http://melkor.dnp.fmph.uniba.sk/~garabik/pycmail.html
 :> </shameless advertising plug>

 : Whoa! _Exactly_ what I was searching. Would you mind registering it at
 : Freshmeat, so other people can find it more easily?

http://freshmeat.net/projects/pycmail/
has been there for a long time:-)

 : Looks like a good project for my next weekend to convert my
 : fetchmail/procmail setup to getmail/pycmail.

I would welcome especially suggestion about possible syntax
changes. Since I cannot help myself, but the syntax (chosen functions
and procedures dealing with mail) does feel a bit clumsy to me
(even though I am the author :-))


-- 
 -----------------------------------------------------------
| Radovan Garabik  http://melkor.dnp.fmph.uniba.sk/~garabik |
| __..--^^^--..__         garabik @ fmph . uniba . sk       |
 -----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!


From ark at research.att.com  Fri Jun  7 12:50:01 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 7 Jun 2002 12:50:01 -0400 (EDT)
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
	(message from Skip Montanaro on Fri, 7 Jun 2002 11:42:19 -0500)
References: <yu99vg8v6orb.fsf@europa.research.att.com> <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
Message-ID: <200206071650.g57Go1Y00711@europa.research.att.com>

PS: I can't find a "ldconfig" command on my system.




From plugin at supercable.es  Sun Jun 23 15:25:59 2002
From: plugin at supercable.es (Ugo García)
Date: Sun, 23 Jun 2002 21:25:59 +0200
Subject: Redirecting StdOut
References: <1024856420.145085@seux119> <Xns9236D1FF016AAcliechtigmxnet@62.2.16.82>
Message-ID: <1024860416.580309@seux119>

perfect. This is what I was looking for. Now I'd make a C module to handle
the messages written to stdout & stderr in order to show them in the screen.
Thanks a lot

--ugo
"Chris Liechti" <cliechti at gmx.net> escribi? en el mensaje
news:Xns9236D1FF016AAcliechtigmxnet at 62.2.16.82...
> "Ugo Garc?a" <plugin at supercable.es> wrote in
> news:1024856420.145085 at seux119:
>
> > Hi all. I'm making a game and I'm using Python as the script language.
> > I'm including a console to communicate with Python. I can
> > interactively send commands to the interpreter but I can't obtain the
> > response from it. Is there any way to redirect stdout and stderr in
> > order to print its contain to the screen?
>
> sure. just assign a file like object to sys.stderr and stdout.
>
> class DummyFileForWrites:
>     def write(self, text):
>         #write 'text' to GUI here
>
> sys.stdout = sys.stderr = DummyFileForWrites()
>
> so you can do the same for sys.stdin but then you would need read,
> readline, and readlines.
>
> chris
>
> --
> Chris <cliechti at gmx.net>
>




From gerhard at bigfoot.de  Fri Jun 14 12:54:43 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 14 Jun 2002 16:54:43 GMT
Subject: executemany & UPDATE (MySQL & Python)
References: <mailman.1024071438.5095.python-list@python.org>
Message-ID: <slrnagk7tu.114.gerhard@lilith.my-fqdn.de>

In article <mailman.1024071438.5095.python-list at python.org>, Nick Arnett wrote:
> Can anyone confirm that it is *not* possible to pass a list to the Python
> MySQLdb module in UPDATE operations?

executemany accepts a sequence of sequences, like:

self.dbh.execute("UPDATE Foo SET bla=%s, blorg=%s WHERE snork=%s",
    [(1,2,3), (4,5,6)])

> Doing the latter with a loop in Python is much slower than I'd expect that
> doing it with a list would be.

For database modules that don't support prepared statements, the db module
implements executemany with multiple calls to execute, anyway.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From kragen at pobox.com  Thu Jun  6 13:30:21 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 06 Jun 2002 13:30:21 -0400
Subject: 64-bit Python?
References: <87elfl3yvu.fsf@mathdogs.com> <C0FL8.41941$n4.9496108@newsc.telia.net> <j41ybk69ac.fsf@informatik.hu-berlin.de>
Message-ID: <83660wgvrm.fsf@panacea.canonical.org>

loewis at informatik.hu-berlin.de (Martin v. L?wis) writes:
> I believe there is another limit: the maximum length of a string,
> list, or tuple is 2**31-1 on 64-bit platforms where a C int is 4
> bytes. That, in particular, includes mmapped objects as well.

Yes, this is correct.  Even the buffer interface, which my
arrayfrombuffer package[0] uses to provide access to mmapped files as
Numeric arrays inexplicably uses int instead of size_t.

The problem is that all of the 64-bit platforms I have access to
(various Linuxes on Alpha and IA-64, Tru64 on Alpha) have 4-byte ints
and 8-byte longs, at least by default.  This is called "LP64": longs
and pointers are 64, but ints are 32, so all the code that assumes
ints are 32 will continue to work.

[0] http://pobox.com/~kragen/sw/arrayfrombuffer



From sholden at holdenweb.com  Tue Jun 18 16:54:26 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 18 Jun 2002 16:54:26 -0400
Subject: Creating Dynamic Web Pages (ZOPE)
References: <46ca81a0.0206062016.789d3af6@posting.google.com> <J29M8.358696$%u2.156726@atlpnn01.usenetserver.com> <46ca81a0.0206141201.23db4a8c@posting.google.com> <kh3tgus5d6hlrt4qjsdg1odr6j9ft0nue2@4ax.com> <3D0ED3F3.2020008@mxm.dk>
Message-ID: <u4NP8.166082$Fp1.153799@atlpnn01.usenetserver.com>

"Max M" <maxm at mxm.dk> wrote ...
> robin at execulink.com wrote:
>
> > I would very much advise against DreamWeaver and similar tools that
> > pretend designing web pages is a WYSIWYG activity, which it certainly
> > is not.
>
>
> I beg to differ Sire!
>
> Dreamweaver is a very fast and convenient way to edit html. If you know
> what you do, your code is just as clean as handcrafted html. There is
> nothing in html that you cannot tweak in Dreamweaver.
>

That's been my experience too. I have certainly been no fan of WTSIWYG
tools, FrontPage having made my life a misery for some short time before I
abandoned it. DreamWeaver does seem to be in a rather different class,
though not perfect. Don't knock it 'til you've tried it.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From benjl at cse.unsw.edu.au  Sun Jun 16 07:05:01 2002
From: benjl at cse.unsw.edu.au (Benno)
Date: 16 Jun 2002 04:05:01 -0700
Subject: What If.....    Strong Types
References: <3D0C0F0D.4030900@bgb.cc>
Message-ID: <332c7c11.0206160305.304b7cb3@posting.google.com>

Don Garrett <garrett at bgb.cc> wrote in message news:<3D0C0F0D.4030900 at bgb.cc>...
> I'm not suggesting any real changes to any, only proposing a thought 
> experiment.
> 
>    What if Python had been built from the start with strong types? And I mean 
> strong types, not strong typing. The distinction I am trying to draw is that 
> variables would not be typed, but that all types would have rigid interfaces.

<snip>

>    I'm just wondering if the change would be good or bad. Would it really 
> break anything important in Python today? 

Well, it would break my stuff, which is pretty damned important to me :).
Dynamically creating/modifying things on the fly can produce nicer, cleaner
code than would otherwise be possible. 

> Would it prevent any common errors? 

Possibly but not that many imho.

> Would it help with automated code analysis (compilation and optimization 
> included) as much as I think it would?

Possibly, it is probably rather difficult to follow through all the 
implications of such as change however.

Cheers,

Benno


From spam at melkor.dnp.fmph.uniba.sk  Wed Jun 19 05:55:28 2002
From: spam at melkor.dnp.fmph.uniba.sk (Radovan Garabik)
Date: Wed, 19 Jun 2002 11:55:28 +0200
Subject: zlib/gzip + anydbm or ...
References: <aeoufa$rkk$1@news.idiom.com> <mailman.1024470203.17467.python-list@python.org>
Message-ID: <gekpea.992.ln@127.0.0.1>

Gerhard H?ring <gerhard at bigfoot.de> wrote:
 : * James T. Dennis <jadestar at idiom.com> [2002-06-19 03:40 +0000]:
 :> 
 :>  Is there an obvious and simple way to use the existing standard
 :>  libraries to support transparently compressed dbm files?

 : I don't think so.

 : If at all, you'll need to hack dumbdbm to achieve this goal, I think
 : only two approaches are feasible:

 : - compress the whole file with zlib, this means you need to keep the
 :  whole file in memory
 : - compress only values with zlib

or three:
 - compress the whole file with dictzip, then you can have random access
(see how dictd servers and tools do it, of course not with dbm, but with
their own database format)



-- 
 -----------------------------------------------------------
| Radovan Garabik  http://melkor.dnp.fmph.uniba.sk/~garabik |
| __..--^^^--..__         garabik @ fmph . uniba . sk       |
 -----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!


From lt6789remove at hotmail.com  Wed Jun 12 16:14:15 2002
From: lt6789remove at hotmail.com (Larry Taitt)
Date: Wed, 12 Jun 2002 13:14:15 -0700
Subject: Post installation problem with Python
Message-ID: <9dafguci28vdiv5lr3qbhset2jko78t6lm@4ax.com>

Installed ActivePython 2.2.1 build 222 on my NT4.0 (sp6a) workstation.
After the installation I'm unable to execute any .exe, .com and .bat.
program (e.g. notepad).  It tells me that it can't find the files.
However, if I add the proper extension (e.g., notepad.exe), the
programs execute without a problem.  I checked the associations and
filetypes, and everything appears normal.  Any ideas?

Thanks.


-----------== Posted via Newsfeeds.Com - Uncensored Usenet News ==----------
   http://www.newsfeeds.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Ulimited Fast Downloads - 19 Servers =-----


From whisper at oz.net  Sat Jun  8 13:59:29 2002
From: whisper at oz.net (David LeBlanc)
Date: Sat, 8 Jun 2002 10:59:29 -0700
Subject: Where is Berkeley DB on your system? Can you live without DB 1.85?
In-Reply-To: <20020608152226.E31320@phd.pp.ru>
Message-ID: <GCEDKONBLEFPPADDJCOECEIPDHAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Oleg Broytmann
> Sent: Saturday, June 08, 2002 4:22
> To: David LeBlanc
> Cc: skip at pobox.com; python-list at python.org
> Subject: Re: Where is Berkeley DB on your system? Can you live without
> DB 1.85?
>
>
> On Fri, Jun 07, 2002 at 01:14:59PM -0700, David LeBlanc wrote:
> > If Sleepycat will allow it, i'm all in favor of updating to the latest
> > version! However, they want money for commercial use of versions > 1.85
> > AFAIK.
>
>    They don't. You can freely incorporate BSDDB into an
> opensource project.
>    http://www.sleepycat.com/faq.html#A22
>
> Oleg.
> --
>      Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
>            Programmers don't die, they just GOSUB without RETURN.

>From the Sleepycat FAQ you cite:

14. Do current Berkeley DB releases have the same license as Berkeley DB
1.85?
Here's the non-commercial license that we're using for current Berkeley DB
releases: http://www.sleepycat.com/license.net. It's different from the one
we used for Berkeley DB 1.85. Generally speaking, the difference is that the
current license for Berkeley DB requires that software using Berkeley DB be
freely redistributable if Berkeley DB is going to be redistributed. If
Berkeley DB is not being redistributed in any way, no commercial license is
ever required. Please see the license for more specific information.
----------------------------------------------------------------------------
----
17. Do I have to license Berkeley DB to create an application for a single
customer?
Yes. Because you are redistributing your application off-site, you must
either make your application freely available in source code form or
purchase a license from Sleepycat Software for use of Berkeley DB in a
proprietary application. Sleepycat generally offers site-based licensing for
small deployments at prices that are lower than we charge for large-scale
redistribution. Please contact us for details.
----------------------------------------------------------------------------
----
18. Do I have to license Berkeley DB to use it in Perl or Python scripts?
No. The Berkeley DB license requires that software that uses Berkeley DB be
freely redistributable. In the case of Perl or Python, that software is Perl
or Python, and not your scripts. Any scripts you write are your property,
including scripts that make use of Berkeley DB. None of the Perl, Python or
Berkeley DB licenses place any restrictions on what you may do with them.
----------------------------------------------------------------------------
----
----------------------------------------------------------------------------
----

AFAIK, this is true for any version of Sleepycat > 1.85, and has been true
for years. Including a later version of Sleepycat BSDDB imposes a
constraint, not on Python itself, which is free, but on someone who would
like to distribute Python along with their _commercial_for_sale_ app (think
Py2Exe for example). That would be a distribution of BSDDB and and so "you
must either make your application freely available in source code form or
purchase a license from Sleepycat Software for use of Berkeley DB in a
proprietary application."

If you want to change Python's market dynamics and require _all_ Python apps
to be free and open source, then you can include any version of Sleepycat
BSDDB you like. I doubt it will have a positive impact on Python's future
prospects. It didn't work for Perl in it's early years and that's why the
Artistic License was change to allow "for fee" distributions that included
Perl.

Dave LeBlanc
Seattle, WA USA





From cliechti at gmx.net  Sat Jun 22 12:42:43 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 22 Jun 2002 18:42:43 +0200
Subject: shallow copy vs. deep copy
References: <MPG.177ec42fc19fcbf598968f@news.rz.uni-karlsruhe.de>
Message-ID: <Xns9235BEC865DCDcliechtigmxnet@62.2.16.82>

Uwe Mayer <merkosh at hadiko.de> wrote in 
news:MPG.177ec42fc19fcbf598968f at news.rz.uni-karlsruhe.de:
> am am wondering where to draw the line between a shallow copy and a deep 
> copy of compound objects.
> 
> when creating a shallow copy i could most easily create a new object and 
> say 
> 
> new.__dict__ = old.__dict__

hat way both object have the same namespace, don't they?. thats more like 
the Borg pattern (Singleton replacement). if you change an attribute in one 
object it will apear in the other object too.

> on the other side i could also:
> 
> new.__dict__ = old.__dict__.copy()

that's what i would call a swallow copy.
 
> which creates a shallow copy of the old instance dictionary.
> if all attributes of the class are primitive (like int, long, string) 
...
> also: if i create a deep copy and my instance attributes only contain 
> immutable objects there's no difference if i create another copy or just 
> pass the reference. if one of the objects attribute gets "changes" it'll 
> be reassigned a new object and thats it.
> what's the "fine" way here?

not copying immutable objects is good to save memory, but it might not be 
so easy to know if an object is immutable. ok you can check types for 
tuples, strings, numbers etc. but you can't do much (at least simple) about 
instances of user objects. it might be overkill for a first shoot to 
optimize this.

viel spass,
chris

-- 
Chris <cliechti at gmx.net>



From peter at engcorp.com  Mon Jun 10 23:55:11 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 10 Jun 2002 23:55:11 -0400
Subject: getting files from a windows network
References: <6bd9f01b.0206101523.7af66fc8@posting.google.com> <Xns922A1317B34C7cliechtigmxnet@62.2.16.82>
Message-ID: <3D05749F.ED64D509@engcorp.com>

Chris Liechti wrote:
> 
> sameer_ at email.com (sameer) wrote in
> news:6bd9f01b.0206101523.7af66fc8 at posting.google.com:
> > is there a way to read files in python that are found on a windows
> > network, i.e. files that users choose to share on their computers on a
> > windows network.
> 
> do you mean by e.g.
> >>> open(r'\\computername\share\test.txt').read()
> 
> yes :-)
> 
> ( r'' is a raw string that doesn't touches backslashes unlike normal
> strings)

Or slightly simpler, just use forward slashes all the time:

>>> open('//computername/share/test.txt').read()

Except for at the DOS command line, Windows generally likes
forward slashes pretty well.

-Peter


From CousinStanley at HotMail.Com  Wed Jun 12 20:34:40 2002
From: CousinStanley at HotMail.Com (Cousin Stanley)
Date: 13 Jun 2002 00:34:40 GMT
Subject: PIL _imaging
References: <Xns922AA829A8C5FHotMailcs@130.133.1.4> <3D06E694.4040103@mxm.dk> <Xns922B3150A6B32HotMailcs@130.133.1.4> <3D074529.E848AA6@ipm.fhg.de> <Xns922B561EDB212HotMailcs@130.133.1.4> <3D07BCC1.2000002@mxm.dk>
Message-ID: <Xns922BB340B788BHotMailcs@130.133.1.4>

|| Perhaps you have more than one version of Python installed and you are 
|| using different versions?
|| 
|| regards Cousin M

Cousin Max ... 

I only have a single installation of Python :
... < K:\Python >
... Wed 06-12-2002 16:41:48.86 > python
... Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
... Type "help", "copyright", "credits" or "license" for more information.
... >>>

Setting the  PYTHONPATH  environment variable
was only an experiment, really just grasping at straws,
and I now have it  REM'd  out in the  autoexec.bat  file ...   

I do have an installation of  Magenta  which uses Python :
... http://www.magentammt.com/
... Could this interfere ???  
... I don't see how ...

I also have the wxPython / wxWindows  
and Pmc packages that seem to function OK ...

My latest experiment was to stick the 2 pyd files,
_imaging.pyd and _imagingtk.pyd, into the Samples folder 
along with the  viewer.py  program that I'm trying to test :
... Same Result, No luck ... 

Cousin Stanley


From ken at hotmail.com  Sun Jun  2 10:18:01 2002
From: ken at hotmail.com (Ken)
Date: Mon, 3 Jun 2002 00:18:01 +1000
Subject: How to delete file?
Message-ID: <add9ai$107clv$1@ID-49758.news.dfncis.de>

How do you delete file?

If you write to file from beginning position of an existing file, do you
overwrite the existing data or insert from it?

Thanks




From occeanlinux at linuxmail.org  Sat Jun  8 22:48:03 2002
From: occeanlinux at linuxmail.org (Gold Fish)
Date: Sun, 09 Jun 2002 02:48:03 GMT
Subject: Html
Message-ID: <3d02c220_1@news.iprimus.com.au>

Anyone can tell me how can i run the script with the argumemt such as python 
example.py -o 
What i have to write in script. And if i want to render the script in html 
format hov can i do that


From christophe.delord at free.fr  Thu Jun 20 13:32:30 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Thu, 20 Jun 2002 19:32:30 +0200
Subject: How to represent the infinite ?
References: <3d11ee9b$1@news.swissonline.ch>
Message-ID: <20020620193230.53d58f8f.christophe.delord@free.fr>

There is another simple solution. Just use a float that is bigger than any 64-bit float. For example 10^1000. This float has a special encoding meaning +oo !
So no need to redefine anything else ;-)

>>> x=1e1000
>>> x
inf
>>> x>10
True
>>> x<10
False
>>> 10>x
False
>>> 10<x
True
>>> x-x
nan
>>> -x
-inf

Warning: +oo==+oo so don't be surprised by this example:

>>> 1e1000==1e2000
True


Christophe.


> Not need to say to me that I can use 99999999999999 or -99999999999999 or I
> do not know what.
> Nor from a MAX_INT of the machine... bus Python can go higher.
> 




-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From fredrik at pythonware.com  Thu Jun  6 14:52:21 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 06 Jun 2002 18:52:21 GMT
Subject: Medium to Large Scale Python Deployments
References: <KHkL8.21383$eD2.2399923271@newssvr10.news.prodigy.com> <833cw0gvg3.fsf@panacea.canonical.org>
Message-ID: <FbOL8.42027$n4.9527047@newsc.telia.net>

Kragen Sitaker wrote:

> >    I am running a home-grown Enterprise Resource Planning system, written
> > entirely in Python, to manage the operations and manufacturing aspects of a
> > medium ($100,000,000 annual sales) manufacturing organization.
>
> That's fascinating!

indeed.

> Are you running Python natively on the AS/400?  How hard was that?

chances are he's using Per Gummedals port:

    http://home.no.net/pgummeda/

</F>




From shalehperry at attbi.com  Sat Jun  1 16:19:15 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Sat, 01 Jun 2002 13:19:15 -0700 (PDT)
Subject: prototyping good OOdesign in Python?
In-Reply-To: <Pine.LNX.4.33.0206012225290.16212-100000@rnd.onego.ru>
Message-ID: <XFMail.20020601131915.shalehperry@attbi.com>

> 
> Well, if you still follow this thread, the question is - how well
> Python could serve as an ARCHITECTURAL prototype if it has too rich
> OOP facilities? That is, is Python of help when trying to
> prototype certain design? Developing in C++ looks so unnecessary hard
> after things were done in Python... 
> Or, it may be put this way: what discipline a Python programmer must
> obey to allow it's prototype to be conveniently rewriten in C++?
> 
> The above thoughts aren't probably well-formed. But I hope
> my concern is understood.
> 

architectural prototyping is what I usually use python for when I prototype,
the other use is rapid class creation for behaviour testing.  You have to keep
in mind the goal of eventually porting to C++ but it is a prototype so there is
not a lot of code there anyway.




From eric.brunel at pragmadev.com  Thu Jun  6 10:15:41 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Thu, 6 Jun 2002 14:15:41 +0000
Subject: Program to an interface, not an implementation
References: <mailman.1023358448.9091.python-list@python.org>
Message-ID: <adnjee$p88$1@wanadoo.fr>

Egbert Bouwman wrote:
> Hello,
> Can someone shed some light on the GoF adagium:
> 'program to an interface, not an implementation'
> and especially what it means for programming in Python ?

The difference is quite simple: for a given class, its interface is the set 
of its public methods, with their syntax (parameters, return value(s)) and 
semantics (what they should perform). In other terms, the interface of a 
class defines _what_ the class should do. The implementation for a class is 
_how_ things are done, i.e. the actual code of the methods, performing the 
"what" defined in their semantics, including algorithms, call to 
protected/private methods, and so on...

For example, for a "sort" method on a list, you may define like in Python 
that it has no parameters, no return value and that it sorts the list "in 
place". This is the interface, i.e. the "what". The "how" will be the 
actual algorithm, which may be any of the well-known algorithms for sorting 
lists.

"Program to an interface, not an implementation" means that the caller of a 
method on a class should only rely on "what" the method should do, not on 
"how" it's done. In the example, it just means that you'll use the "sort" 
method for what it's done (sorting a list), and nothing more. When told 
like that, it may seem quite obvious, but imagine for example that you 
coded the "sort" algorithm and that during this algorithm, you had to set 
an attribute of the list to the list length. If after doing the "sort", you 
use this attribute to get the list length, you're using a side effect of 
your algorithm that should have remained hidden from the outside. That's 
what is called "programming to an implementation": it's *because* your 
"sort" method is coded like you did that you can use the attribute.

In fact, saying "program to an interface, not an implementation" is just a 
way to describe the traditionnal OO concept of "encapsulation": the exposed 
part of a class is its interface; how methods are implemented inside the 
class should remain hidden. The aim of this is to allow future changes in 
the implementation of your class without affecting its interface. In the 
example, imagine that you coded a simple but quite inefficient sort 
algorithm, and that someday, somebody comes and wishes to use a more 
efficient quick-sort: given the interface of the "sort" method, he has no 
way to know that you used the attribute giving the list length. So his 
modifications may break your code if he doesn't set the list length 
attribute anymore.

In fact, I often say that a good OO programmer should be schizoid: the 
person who codes the actual algorithms should not be the same than the one 
who uses them, even if it's the same physical person. You should always 
define the interface of your classes, implement them, then forget 
everything you did during the implementation phase, and finally write the 
code that uses the classes, just looking at the interface you defined in 
the first place.

Hope it's a bit clearer now...
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From marklists at mceahern.com  Wed Jun 19 10:20:44 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 19 Jun 2002 09:20:44 -0500
Subject: python version?
In-Reply-To: <CT%P8.33113$uk2.13710586@twister.nyroc.rr.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBOEGILFAA.marklists@mceahern.com>

[George Hester]
> Yes basically I wanted it handed to me on a platter.

What I can't figure out is how come you didn't put 10% of the energy you
spent in defending yourself into solving your problem?  Heck, if you'd have
done that you might actually have been able to write a useful tutorial or
mini-FAQ.  Instead, you've probably got people wondering whether we will yet
Rue the day.  ;-)

I mean, fer chrissakes, this thread is how many posts long when all of the
Python-related knowledge it contains can be summed up like this:

Q:  How do I tell what version of Python I'm using?
A:  Here's how:

  import sys
  print sys.version

Of course, who am I to suggest there might be better ways to spend your
time.  I mean, it's not like my reply here is actively helping me solve my
own problems.  <wink>

You mentioned you're not interested in learning Python.  That's unfortunate.
I personally think it's a great language and has a lot to offer.  I won't
hold it against you if you change your mind.

To each his own, though.

Cheers,

// m

-





From Christopher.Saunter at durham.ac.uk  Fri Jun 21 05:49:32 2002
From: Christopher.Saunter at durham.ac.uk (Christopher Saunter)
Date: Fri, 21 Jun 2002 09:49:32 +0000 (UTC)
Subject: Flame fests (was Re: python version?)
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <3D107948.B3FEE4BB@engcorp.com> <aer53s$7is$1@panix1.panix.com> <3D112BA6.4CA5F584@engcorp.com> <aerosm$p18$1@panix1.panix.com>
Message-ID: <aeusrc$lge$1@sirius.dur.ac.uk>

: >I think we'd all object strongly to the characterization that we were 
: >"kicking a newbie".

: Well, then, if you're not kicking a newbie, you're swallowing trolled
: bait.  Which makes you feel better?

I wonder if the above should be classed as a 'meta-troll' ;-)))


From bokr at oz.net  Tue Jun  4 18:11:27 2002
From: bokr at oz.net (Bengt Richter)
Date: 4 Jun 2002 22:11:27 GMT
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <3CFCF3E7.AE891EBE@noaa.gov> <adiumj$22g2$1@nntp6.u.washington.edu>
Message-ID: <adjduf$92g$0@216.39.172.122>

On 4 Jun 2002 17:51:15 GMT, "Donn Cave" <donn at u.washington.edu> wrote:

>Quoth Chris Barker <Chris.Barker at noaa.gov>:
>| Huaiyu Zhu wrote:
>|> Great explanation!  The problem stems from the fact that the current +=
>|> operator implements two semantically distinct operations, chosen according
>|> to some properties of the operands.
>|
>| exactly, and it wa kind of a weird decision. Personally, I feel the only
>| compelling reason to add The += (and friends) operator was to support
>| in-place addition, most notably for Numeric arrays. Frankly it ought to
>| just be disallowed for non-mutable types, but all the folks that want it
>| for integers would scream. to be honest, I've kind of lost track of why
>| it is a good idea for numbers to be immutable, but that's another topic!
>
>That would be almost as cool as the old implementations of FORTRAN
>where you could change the value of 1.
>
>|> I'd imagine that at some future time there would be a desire to split += to
>|> two distinct operators, one for true in-place operator and one for the fake
>|> one currently implemented for immutables.  That would remove the spurious
>|> exceptions for mutable members of immutable structures.
>|
>| Do we really need a new notation? is a = a + b so bad?
>
>No, it's fine.
>
But see below...

>| ... In fact, are the
>| current semantics of += s bad, or should this bug just be fixed?
>
>Not really - the current semantics are actually better than the fixed
>version, inasmuch as they are at least consistent.  As I interpret the
>proposal, the fixed += would sometimes work without rebinding (if the
>object supports it) and sometimes would be an assignment (falling back
>to +.)  To me, to allow an object to dictate its relationship to its
>computational environment like that is anathema.
>
>...
>| By the way, the += and friends operators also cause similar confusion
>| with name scopes in functions, += is assumed to be a re-bind, when it
>| isn't necessarily intended to:
>|
>| >>> l = [1,2,3]
>| >>> def test():
>| ...     l += [4]
>| ... 
>| >>> test()
>| Traceback (most recent call last):
>|   File "<stdin>", line 1, in ?
>|   File "<stdin>", line 2, in test
>| UnboundLocalError: local variable 'l' referenced before assignment
>|
>| but:
>|
>| >>> def test2():
>| ...     l.extend([4])
>|
>| >>> test2()
>| >>> l
>| [1, 2, 3, 5]
>|
>| Works just fine. Perhaps  Huaiyu is right, we need different operators.
>| However, then the operator for incrementing a non-mutable would just be
>| a synonym for a = a + b, which looks like unnecessary syntactical sugar
>| to me.
>
>Yes.
>

Here is a contrived case where += lets you use a
mutable-lhs-target-expression-with-side-effect and
get only one side effect. There are obviously other
ways, but at least += is convenient (and presumably
faster in a case like this).

 >>> a=[0]
 >>> def foo():
 ...     print 'side effect'
 ...     return a
 ...
 >>> foo()
 side effect
 [0]
 >>> foo()[0]+=1
 side effect
 >>> a
 [1]
 >>> foo()[0] = foo()[0]+1
 side effect
 side effect
 >>> a
 [2]

Regards,
Bengt Richter


From mcfletch at rogers.com  Wed Jun 12 13:02:32 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Wed, 12 Jun 2002 13:02:32 -0400
Subject: Python Virtual File System: A spike test
References: <3D070F27.9050608@rogers.com> <20020612142015.X6609@prim.han.de>
Message-ID: <3D077EA8.3090004@rogers.com>

Hi Holger,

I really like the idea of the filter package (I'm guessing it would work 
without alteration even with the spike-test class).

As for providing methods with precisely the same names as the os module; 
I'm not particularly thrilled with the names used there, based as they 
are on UNIX commands, rather than descriptions of what the method does; 
with that said, I'm not pathologically averse to including such method 
names.

Creating an abstraction for the various URL designations is definitely 
attractive.  Being able to pass around typed URL objects, rather than 
raw strings, with rich methods on each type should make the Python 
programming experience feel a little more consistent.

Is your module available somewhere to take a peek?

Enjoy,
Mike

holger krekel wrote:
> Hi Mike,
...
> Interesting. I have written a similar 'filename' class. 
> It is not based on the str-type but accepts strings in many places. 
> Also the filename class offers all of the methods found in 'os' or 'os.path'
> under the exact name. I think this is important to make using it easy. 
> Many people already know and use the stdlib-methods.
> 
> One of the nice features is that you can do stuff like:
> 
> fn=filename('/usr/src')
> cfilter=filters.OR(filters.hasext('.cpp'), filters.hasext('c'))
> mfilter=filters.rmatch('.*/[Mm]akefile.*')
> recursefilter=rignore('.*/CVS')
> 
> for makefiles, cppfiles in fn.filterwalk([mfilter,cfilter], recursefilter):
>     # work with list of makefiles
>     # work with cppfiles
> 
> 'filterwalk' is a generator which honors a recursion filter and one or
> more filename-filter.  It allows to recurse once and get back categories of files
> (as lists). No further checking like 'islink' etc. needed. You makeup
> your filters and call filterwalk and afterwards work with the lists.
...
> Having an abstraction for 'Names'/URLs and working from there would 
> be very interesting IMO. At the moment we only work with (relatively
> untyped) strings which places the burden of calling the right methods
> (urllib/httplib/urllib2/os.path/ftplib/webdav...) on the programmer
> (who just has a string).  
...
> i am interested but we should discuss if our approaches/viewpoints 
> work well together :-) 
...





From mwarden22 at hotmail.com  Thu Jun 13 13:33:08 2002
From: mwarden22 at hotmail.com (nealj)
Date: 13 Jun 2002 10:33:08 -0700
Subject: Hap debugger released
References: <mailman.1023825435.5394.python-list@python.org>
Message-ID: <d3768569.0206130933.46c1d236@posting.google.com>

Great comments, I really appreciate your feedback - I've been looking
at it too long so it's easy for things to slip by.

Neal

"Bjorn Pettersen" <BPettersen at NAREX.com> wrote in message news:<mailman.1023825435.5394.python-list at python.org>...
> > From: nealj [mailto:mwarden22 at hotmail.com] 
> > 
> > Version 2.1 of the Hap python debugger (and ide) has just 
> > been released to sourceforge.  This version improves 
> > performance, stability and adds a number of useful features 
> > such as an interactive window, watch tooltips and more.
> > 
> > We've been using it in house on our next product for six 
> > months now and everyone is very happy with it.
> > 
> > Try it out at http://sourceforge.net/projects/hapdebugger
> 
> This looks really good (I know a couple of people who're going to be
> pleased :-)
> 
> A couple of minor issues I noticed:
> 
>  - the color scheme is awful (the Idle/PythonWin scheme is
>    much less jarring on the eyes :-)
>  - I like the MSVC key bindings, but I was surprised that
>    Shift-F11 wasn't bound to "step out of".
>  - I couldn't step over a list comprehension (instead I was
>    taken through each iteration of the for 'loop').
> 
> All in all, great work however!
> 
> -- bjorn


From peter at engcorp.com  Mon Jun 10 14:06:14 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 10 Jun 2002 14:06:14 -0400
Subject: activating a webpage's click command from python
References: <mailman.1023727993.12908.python-list@python.org>
Message-ID: <3D04EA96.B3518A37@engcorp.com>

"Grimes, Jessica C" wrote:
> 
> I want to create a  program that will ask the user what stock he/she 
> wants to look up.  The program will then extract financial data from 
> yahoo and import it into a spreadsheet.
> 
> My problem is that I don't know how to make the application take the 
> user's input (the stock's ticker symbol) and place the user's input 
> from python into the space where yahoo accepts input for a ticker 
> symbol and activate yahoo's "Look up" click command.

You should investigate HTML forms.  The web page contains a form
which has several fields and a target URL.  If you view the source
for the page (in Netscape, hit Ctrl-U; in IE use the View->Source
menu) you can find buried within it the form and the names of the
fields.  Forms use the <form> tag and contain <input> fields.

Looking at finance.yahoo.com I can see the form calls "/q" when
executed, and has a field named "s" for the stock symbol.  You
can easily take such forms and compose a URL which retrieves the
data.  In this case this:

   http://finance.yahoo.com/q?s=nt.to

does the job for Nortel on the TSE (disclaimer: anyone even 
thinking of buying Nortel should probably just keep thinking :-).
The page returned contains the data, but it's embedded deep
in the HTML and you _really_ don't want to waste much of your
time trying to figure this out.  There's got to be a better 
approach, but I have never spent the time to research it.

Anyway, the following script does the job in a Pythonically
pragmatic way (in this case, that means "just barely")

import urllib
import re
import sys

URL = 'http://finance.yahoo.com/q?s=%s'
pattern = '%s</a>.*?<b>([0-9.]*)</b>'
symbol = sys.argv[1].upper()

data = urllib.urlopen(URL % symbol).read()
matches = re.findall(pattern % symbol, data)

if len(matches) == 1:
    print '%s last trade was %s' % (symbol, matches[0])
else:
    print 'Undefined result: %s' % (matches,)

C:\>python tick.py nt.to
NT.TO last trade was 2.69
C:\>python tick.py ibm
IBM last trade was 78.03
C:\>python tick.py engc
Undefined result: []

Hmm... Engenuity Corporation doesn't seem to be listed yet. :-)

-Peter


From mgerrans at mindspring.com  Mon Jun 17 00:06:29 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Sun, 16 Jun 2002 21:06:29 -0700
Subject: what *is* a class?
References: <MPG.17774343cc5a589098968a@news.rz.uni-karlsruhe.de>
Message-ID: <aejn8r$pnh$1@nntp9.atl.mindspring.net>

> g = Group()

Now g is a reference to an instance, or object, of the class Group.

> what actually does Group() return and is there any way to get hold of
> that again?

It creates and returns an instance (or object, if you prefer) of the class
Group.   g is referring to that instance, so you haven't lost it, unless you
assign g to something else.   Since you have a reference to the object you
created (g), you don't need to "get ahold of it again" -- you still have a
hold of (on?) it.

> i want to write a method which returns the instance, so that f.e. the
> following is possible:
>
> g = Group()
> g2 = g.foobar()
> g2 is g
>
> so that g2 is identical to g.
> any idea?

All you need to do is this:

g2 = g

But if you insist on having a method to do it, you could define it like so:

class Group:
   # ... other code.
   def getRef(self):
      return self

Then, instead of "g2 = g", you could do this:

g2 = g.getRef()

I can't think of any reason why you would want to do this though, unless you
just need the extra typing practice and want to confuse someone who might be
reading the code...

Perhaps I misunderstand your question and what you want is a factory.   That
would be a method that creates and returns objects.   Of course, the
constructor does this, but you might want a factory if you will instantiate
different types of objects based on some runtime consideration.   If this is
what you are after, then reply to that effect for more elaboration.

- mfg




From mhammond at skippinet.com.au  Wed Jun  5 18:59:51 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 05 Jun 2002 22:59:51 GMT
Subject: constructive critisism on pythonwin IDE
References: <6bd9f01b.0206050736.f469082@posting.google.com>
Message-ID: <3CFE9861.5090200@skippinet.com.au>

sameer wrote:
> First of all, let me start of by saying that Pythonwin is a great IDE.
>  Now for the constructive critisism.  The thing I hate about PythonWin
> is, that despite the fact that it's a development environment, there
> is a lot of caching done of modules, and the cached copy is not
> checked against a potentialy modified copy.  Whenever I change the
> contents of a file and try to run a module that depends on that file,
> I always end up with the old copy.  I have to then import and then
> reload the changed module in the interactive window for it to refresh.
>  It would greatly increase my productivity, if I can specify which
> modules can be cached.  I understand that there is a speed and
> development time issue, but I would like a little more fine grained
> control on how caching works on PythonWin, as the way it's currently
> implemented now is not acceptable.

All patches gratefully accepted.  Of course, if that is "not acceptable" 
I will cheerfully refund every cent you paid for it <wink>

Mark.



From eric.brunel at pragmadev.com  Thu Jun  6 14:40:24 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Thu, 6 Jun 2002 18:40:24 +0000
Subject: for , ordering
References: <mailman.1023377649.29116.python-list@python.org>
Message-ID: <ado2ur$fep$1@wanadoo.fr>

gabor wrote:

> hi,
> 
> i have a list and i want to create a copy of it with a bit different
> objects... by now i have something like this:
> 
> objects1 = []
> objects2 = []
> .
> .
> .
> 
> for obj1 in objects1:
> objects2.append(obj2(obj1.x))
> 
> 
> 1. isn't there a more elegant way to do this?

Yep. Do:
objects2 = [obj2(obj1.x) for obj1 in objects1]
It's called a "list comprehension" and is the exact equivalent of your code 
above, just shorter. And it's almost more legible than the 'for' loop :-)

> 2. is that guaranteed that the for cycle iterates over the elemenst of
> the list in the correct order ( 0,1,2,3,4,....)?

Yep. Lists are ordered, so the 'for' loop or the list comprehension above 
are guaranteed to browse the list in order. Just to mention it, it wouldn't 
be the case with dictionaries, that reorder their entries for optimization 
purposes.

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From greg at cosc.canterbury.ac.nz  Thu Jun 20 02:07:15 2002
From: greg at cosc.canterbury.ac.nz (greg)
Date: Thu, 20 Jun 2002 18:07:15 +1200
Subject: ANN: Pyrex 0.3.1 (the REAL one!)
References: <mailman.1024325000.4172.clpa-moderators@python.org>
Message-ID: <mailman.1024573203.7652.clpa-moderators@python.org>

It seems that what I advertised before as Pyrex 0.3.1
was the wrong version. The correct one should be there
now!

> Pyrex 0.3.1 is now available:
> 
>    http://www.cosc.canterbury.ac.nz/~greg/python/Pyrex/
> 
> Some fairly major problems with the new extension
> types facility have been fixed.
> 
> There are also some new features:
> 
>   -  Default argument values, * and ** arguments are
>      now supported in Pyrex-defined Python functions.
> 
>   -  External C functions with variable argument
>      lists can be declared and called.
> 
>   -  Full C type declaration syntax now supported,
>      including pointer-to-function types.
> 
>   -  More special methods for extension types:
>      __delitem__, __delslice__, __getattr__,
>      __setattr__, __delattr__
> 
>   -  ctypedef statement.
> 
>   -  String literal handling improved (triple
>      quoted strings, escape sequences in strings
>      should now work).
> 
>   -  Hex, octal and imaginary literals.
> 
>   -  Docstrings.
> 
> What is Pyrex?
> --------------
> 
> Pyrex is a new language for writing Python extension modules.
> It lets you freely mix operations on Python and C data, with
> all Python reference counting and error checking handled
> automatically.




From travis at enthought.com  Mon Jun 17 22:16:40 2002
From: travis at enthought.com (Travis N. Vaught)
Date: Mon, 17 Jun 2002 21:16:40 -0500
Subject: [ANN] SciPy '02 - Python for Scientific Computing Workshop
Message-ID: <KPEIKEPEIBGNGNKHDILIKENDDGAA.travis@enthought.com>

----------------------------------------
Python for Scientific Computing Workshop
----------------------------------------
CalTech, Pasadena, CA
Septemer 5-6, 2002

http://www.scipy.org/site_content/scipy02

This workshop provides a unique opportunity to learn and affect what is
happening in the realm of scientific computing with Python. Attendees will
have the opportunity to review the available tools and how they apply to
specific problems. By providing a forum for developers to share their Python
expertise with the wider industrial, academic, and research communities,
this workshop will foster collaboration and facilitate the sharing of
software components, techniques and a vision for high level language use in
scientific computing.

The two-day workshop will be a mix of invited talks and training sessions in
the morning. The afternoons will be breakout sessions with the intent of
getting standardization of tools and interfaces.

The cost of the workshop is $50.00 and includes 2 breakfasts and 2 lunches
on Sept. 5th and 6th, one dinner on Sept. 5th, and snacks during breaks.

There is a limit of 50 attendees.  Should we exceed the limit of 50
registrants, the 50 persons selected to attend will be invited individually
by the organizers.

Discussion about the conference may be directed to the SciPy-user mailing
list:

mailto:scipy-user at scipy.org

http://www.scipy.org/MailList


-------------
Co-Hosted By:
-------------

The National Biomedical Computation Resource (NBCR, SDSC, San Diego, CA)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
http://nbcr.sdsc.edu
The mission of the National Biomedical Computation Resource at the San Diego
Supercomputer Center is to conduct, catalyze, and enable biomedical research
by harnessing advanced computational technology.


The Center for Advanced Computing Research (CACR, CalTech, Pasadena, CA)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
http://nbcr.sdsc.edu
CACR is dedicated to the pursuit of excellence in the field of
high-performance computing, communication, and data engineering. Major
activities include carrying out large-scale scientific and engineering
applications on parallel supercomputers and coordinating collaborative
research projects on high-speed network technologies, distributed computing
and database methodologies, and related topics. Our goal is to help further
the state of the art in scientific computing.


Enthought, Inc. (Austin, TX)
^^^^^^^^^^^^^^^
http://enthought.com
Enthought, Inc. provides business and scientific computing solutions through
software development, consulting and training.  Enthought also fosters the
development of SciPy, an open source library of scientific tools for Python.





From whisper at oz.net  Thu Jun 27 15:02:45 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 27 Jun 2002 12:02:45 -0700
Subject: How to find out operating system
In-Reply-To: <uwuska8ny.fsf@online.no>
Message-ID: <GCEDKONBLEFPPADDJCOECEIGDMAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Syver Enstad
> Sent: Thursday, June 27, 2002 11:18
> To: python-list at python.org
> Subject: Re: How to find out operating system
> 
> 
> Peter Hansen <peter at engcorp.com> writes:
> 
> > David LeBlanc wrote:
> > > 
> > > You might try os.environ:
> > > 
> > > os.environ['WINOS'] -> WIN2000
> > > 
> > > I don't know if WINOS is defined on '98 or ME, but it should be good
> > for NT,
> > 
> > > 2K and XP
> > 
> > It is not.
> 
> Neither is it defined on win2k pro:
> 
> KeyError: WINOS
> 
> -- 
> 
> Vennlig hilsen 
> 
> Syver Enstad
> -- 

C:\>python
Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import os
>>> os.environ.has_key('WINOS')
1
>>> os.environ['WINOS']
'WIN2000'
>>>




From peter at engcorp.com  Fri Jun  7 02:07:45 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 07 Jun 2002 02:07:45 -0400
Subject: Method call from string
References: <znu-39AB22.01311007062002@news.fu-berlin.de>
Message-ID: <3D004DB1.62BEB68A@engcorp.com>

ZnU wrote:
> 
> Say I've got a variable with the value 'do_something', and I have an an
> object that happens to have a do_something method. How can I call that
> method based on the value of the variable? Preferably without using
> eval() or similar.

>>> var = 'do_something'
>>> class A:
...   def do_something(self):
...      print 'ooga booga'
...
>>> a = A()
>>> getattr(a, var)
<bound method A.do_something of <__main__.A instance at 0x007B0110>>
>>> f = getattr(a, var)
>>> f()
ooga booga


-Peter


From kosh at aesaeion.com  Fri Jun 21 03:55:05 2002
From: kosh at aesaeion.com (kosh at aesaeion.com)
Date: Fri, 21 Jun 2002 01:55:05 -0600 (MDT)
Subject: compile python witout tkinter
In-Reply-To: <20020621075938.GA896@lilith.my-fqdn.de>
Message-ID: <Pine.LNX.4.44.0206210150440.585-100000@aesaeion.com>

On Fri, 21 Jun 2002, Gerhard [iso-8859-15] H?ring wrote:

> * kosh at aesaeion.com <kosh at aesaeion.com> [2002-06-21 01:40 -0600]:
> > On Fri, 21 Jun 2002, Gerhard [iso-8859-15] H?ring wrote:
> >
> > > * kosh at aesaeion.com <kosh at aesaeion.com> [2002-06-21 01:26 -0600]:
> > > > On a machine I need to build python on I need to build it without tkinter.
> > >
> > > Why? Even if Tkinter gets compiled, it doesn't change anything for the
> > > rest. You could even delete the tkinter modules afterwards, if you're
> > > short on disk (quota).
> >
> > Because the libraries it depends on are not installed on that machine and
> > there is no point installing them since they are not needed.
>
> No problem, then. It won't get built if the dependencies are not met.
> As seen in Modules/Setup.

The problem is that it is trying to build it ayways. All the tkinter lines
in Modules/Setup that I can find are commented out but it doesn't seem to
have done any good.






From mwh at python.net  Fri Jun 14 05:38:08 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 14 Jun 2002 09:38:08 GMT
Subject: Zip makes sense! was: Re: why not "'in' in 'in'"?
References: <mailman.1024043660.31139.python-list@python.org>
Message-ID: <lk8z5imc9r.fsf@pc150.maths.bris.ac.uk>

"David LeBlanc" <whisper at oz.net> writes:

> Haskell - that's a mountain range in New York state? :p)

No, it's the first name of a pioneering logician.

Cheers,
M.

-- 
  I really hope there's a catastrophic bug insome future e-mail
  program where if you try and send an attachment it cancels your
  ISP account, deletes your harddrive, and pisses in your coffee
                                                         -- Adam Rixey


From peter at engcorp.com  Fri Jun 21 01:42:34 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 21 Jun 2002 01:42:34 -0400
Subject: Inhibiting duplicate apps in Windows?
References: <3D11F9E8.DFFFC8E4@tds.net>
		<Xns9233F02DB217Ccliechtigmxnet@62.2.16.82>  <3D1283EC.A28A45CB@engcorp.com> <mailman.1024636228.21073.python-list@python.org>
Message-ID: <3D12BCCA.3A0F7544@engcorp.com>

Cliff Wells wrote:
> 
[on locks 'n such]
> A fairly fool-proof/portable method is to open a socket. If the app is
> run a second time it will fail on the attempt.  Even if the app crashes,
> the socket will be closed by the OS (although it may take a couple of
> seconds on occasion).

Interesting, although I take it you mean "bind a socket to a port".
Are there any complications with this?  I created one and it's true
another instance could not be opened ("address already in us") but
the socket is listed with "netstat" as being LISTENING even though
I have not called listen() or accept().  Any problems with doing 
that, if something connects to the socket?  Could it cause the
socket to be held open for more than a couple of seconds after
closing?  Sometimes sockets can take minutes to be cleaned up.

-Peter


From jslowery at hotmail.com  Wed Jun 19 23:30:15 2002
From: jslowery at hotmail.com (Jeremy S Lowery)
Date: Wed, 19 Jun 2002 22:30:15 -0500
Subject: Pickling meta instances
Message-ID: <FPbQ8.18206$jT.4640794@e3500-atl1.usenetserver.com>

With the new features of 2.2, would there be any way to pickle class
definitions (meta-instances).






From garrett at bgb.cc  Sun Jun 16 00:08:26 2002
From: garrett at bgb.cc (Don Garrett)
Date: Sun, 16 Jun 2002 04:08:26 GMT
Subject: What If.....    Strong Types
Message-ID: <3D0C0F0D.4030900@bgb.cc>

   I'm not suggesting any real changes to any, only proposing a thought 
experiment.

   What if Python had been built from the start with strong types? And I mean 
strong types, not strong typing. The distinction I am trying to draw is that 
variables would not be typed, but that all types would have rigid interfaces.

   Primarily, what if classes always had rigidly defined interfaces. I mean 
that public members had to be declared to exist, and that methods couldn't be 
modified on an instance. Private members could exist but would be really private.

   All types would be classes (including ints and such). Amoung other things, 
I would add the concepts of Interfaces, instanceof operators.

   Local variables would not require declaration, and would work just like 
today. Introspection would work, but be read-only.

   My belief is that almost all the convenience of Python would be maintained, 
but that compilation, optimization, and automated code analysis would be simpler.

   I'm just wondering if the change would be good or bad. Would it really 
break anything important in Python today? Would it prevent any common errors? 
Would it help with automated code analysis (compilation and optimization 
included) as much as I think it would?

-- 
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc



From cliechti at gmx.net  Sun Jun 23 17:20:22 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 23 Jun 2002 23:20:22 +0200
Subject: sys.exc_info() into string
References: <KZoR8.44805$n4.10504231@newsc.telia.net> <Xns9236D90D4E2FAcliechtigmxnet@62.2.16.82> <moqR8.44823$n4.10512288@newsc.telia.net>
Message-ID: <Xns9236EDDB865A4cliechtigmxnet@62.2.16.82>

Magnus <spammers.do.not.bother at void.bogus> wrote in 
news:moqR8.44823$n4.10512288 at newsc.telia.net:

> Chris Liechti wrote:
> 
>> Magnus <spammers.do.not.bother at void.bogus> wrote in
>> news:KZoR8.44805$n4.10504231 at newsc.telia.net:
>>> in my program I am catching exceptions that I might have missed by
>>> using something like:
>>> 
>>> try:
>>>         doMyStuff()
>>> except:
>>>         type = sys.exc_info()[0]
>>>         syslog.syslog(syslog.LOG_EMERG, "My program: unexpected error"
>>>         + type)
>>> 
>>> The thing is that I can not figure out how to make "type" into a
>>> string above so I can use it for e.g. syslog(). Is there an easy way
>>> of doing this?
>> 
>> yes, the traceback module has functions to format an exception as string
>> or to print.
>> 
>> format_exc is a candidate for you. it return a list of lines from which
>> you can select to mosr informative or just join it the one big string.
>> 
>> chris
>> 
> 
> I think following is working...
> 
> except:
>         syslog.syslog(syslog.LOG_EMERG, "My program: unexpected er
> ror" + str(sys.exc_info()[0]))
> 
> ...thanks for the suggestion.

i encourage you to include the linenumber and file in the message too. when 
you want to fix the error, you don't have to try to reproduce it, which 
could be very tricky.
and it will be frustrating for you to see that your programm has errors, 
but you can't fix them because you don't know where they occured...

>>> try:
... 	e
... except:
... 	print ''.join(traceback.format_exception(
...                *sys.exc_info())[-2:]).strip().replace('\n',': ')
... 
File "<interactive input>", line 2, in ?: NameError: name 'e' is not 
defined

strip and replace are used to make one line of it. of course you could also 
print the entire traceback to the log.

chris

-- 
Chris <cliechti at gmx.net>



From huaiyu at gauss.almadan.ibm.com  Thu Jun 20 16:11:25 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Thu, 20 Jun 2002 20:11:25 +0000 (UTC)
Subject: How to represent the infinite ?
References: <3d11ee9b$1@news.swissonline.ch> <20020620193230.53d58f8f.christophe.delord@free.fr> <aet4p6$33kk$1@nntp6.u.washington.edu> <20020620202732.68e67873.christophe.delord@free.fr>
Message-ID: <slrnah4dnd.8m7.huaiyu@gauss.almadan.ibm.com>

Christophe Delord <christophe.delord at free.fr> wrote:
[regarding 1e1000 giving inf]
>With Python 2.2 and Linux it works. The example I gave is a copy-paste from IDLE. It also works with my Python 1.5.2 and 2.3 (I'm under Linux on an Intel processor)

Does 1e200**2 also work for you out of the box?  I had to get rid of the
artificial check for overflow in the source to really use inf in
computations.  I don't remember if 1e1000 works on my box without patch.

Huaiyu


From hesterloli at hotmail.com  Tue Jun 18 06:46:51 2002
From: hesterloli at hotmail.com (George Hester)
Date: Tue, 18 Jun 2002 10:46:51 GMT
Subject: python version?
References: <mailman.1024389443.4623.python-list@python.org>
Message-ID: <vcEP8.44423$GY.13969095@twister.nyroc.rr.com>

jeezushchrist

--
George Hester
_________________________________
"Troels Therkelsen" <troels at 2-10.org> wrote in message
news:mailman.1024389443.4623.python-list at python.org...
> > I remember once I tried to explain to my students the theory
> > behind the limit in
> > calculus.  To 35 students.  Maybe 3 understood it in such a way
> > that they could
> > present a solution back to me.  But most just didn't want to
> > know.  I was wrong.
> > I was not teaching Mathematicians but Buisness students that just
> > needed the
> > credit.  Now I am not someone who necessarily wants to code using
> > Python in ASP.
> > I just want some code to find my version of Python using ASP.
> > I'm sorry if I
> > have not lived up to your expectations.
> >
> > --
> > George Hester
>
> The document on this URL might help explain why you feel you have not
> gotten the answer to the question you thought you asked :-)
>
> http://www.tuxedo.org/~esr/faqs/smart-questions.html
>
> Regards,
>
> Troels Therkelsen
>
>
>




From cjensen at bioeng.ucsd.edu  Mon Jun 24 13:32:15 2002
From: cjensen at bioeng.ucsd.edu (Curtis Jensen)
Date: Mon, 24 Jun 2002 10:32:15 -0700
Subject: problem recursively removing directories on an NFS mount
Message-ID: <3D17579F.9030905@bioeng.ucsd.edu>

I have a function to recursively remove directories:
def removeDir( dirName ):
   """recursively remove a directory """

   for name in os.listdir( dirName ):
     full_name = os.path.join(dirName, name)

     if os.path.isdir( full_name ):
       removeDir( full_name )
     else:
       os.remove( full_name )
   #end for name

   os.rmdir( dirName )


Using python1.5.2 on an IRIX 6.5 system, this function usually works 
fine.  However, sometimes nfs files of the form ".nfs07D7761" get 
created.  This causes the os.rmdir( full_name ) command to throw an 
exception because the directory is not empty.  It appears that the 
os.remove() function causes NFS to generate these files.  Is this a bug 
in os.remove or nfs or my function?  What would be the best way to fix 
this problem?
Thanks.

-- 
Curtis Jensen
cjensen at bioeng.ucsd.edu
http://www-bioeng.ucsd.edu/~cjensen/
FAX (425) 740-1451



From peter at engcorp.com  Sat Jun 29 22:33:23 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 29 Jun 2002 22:33:23 -0400
Subject: Intercepting a process shutdown command under windows?
References: <d7fcf8b4.0206291804.3146614b@posting.google.com>
Message-ID: <3D1E6DF3.F15A76D2@engcorp.com>

Samuel Schulenburg wrote:
> 
> Is their a way to keep another process from shutdown your python
> application before you can perform some cleanup activities?
> 
> Sam Schulenburg

This may depend on which platform you're on, and on exactly what
kind of "shutdown" command you're faced with.

There will *always* be something that can terminate your app
without you getting any say.  Depending on the precise type
of signal, however, you might be able to intercept it.

-Peter


From irmen at NOSPAMREMOVETHISxs4all.nl  Mon Jun 10 17:04:11 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Mon, 10 Jun 2002 23:04:11 +0200
Subject: How convert all list/tuple...
References: <523aguoljk3qceh4kkki7emn6283d6sqpt@4ax.com>
Message-ID: <3D05144B.8010305@NOSPAMREMOVETHISxs4all.nl>

> I wish to translate something like:
>  [1,2,"string"]
> into:
>  ['1','2','string']

This is an easy one, either use map or a list comprehension:

map(str,[1,2,"string"])
   or,
[str(x) for x in [1,2,"string"]]

Irmen



From bokr at oz.net  Tue Jun 25 15:14:22 2002
From: bokr at oz.net (Bengt Richter)
Date: 25 Jun 2002 19:14:22 GMT
Subject: Suggestions for good programming practices?
References: <mailman.1024951532.5616.python-list@python.org> <af8n9i$3at$1@panix1.panix.com>
Message-ID: <afafee$fc$0@216.39.172.122>

On 24 Jun 2002 23:16:02 -0400, aahz at pythoncraft.com (Aahz) wrote:

>In article <mailman.1024951532.5616.python-list at python.org>,
>David LeBlanc <whisper at oz.net> wrote:
>>Mark:
>>>
>>> When evaluating whether a variable is None, don't compare it like this:
>>>
>>> 	if x == None:
>>>
>>> instead, use:
>>>
>>> 	if x:
>>>
>>> or:
>>>
>>> 	if not x:
>>>
>>> The encourages a polymorphic approach.  If you must compare to
>>> None, use the identify operator rather than equality:
>>>
>>> 	if x is None:
>>>
>>> or:
>>>
>>> 	if x is not None:
>>
>>I'm curious to know how this encourages polymorphism, and how making a
>>conditional test implicit rather then explicit is good practice?
>
>Depends on what you're testing for.  One of the big features of Python's
>object model is that any "empty" object tests as false: None, 0, "", (),
>[], and {}.  Testing for a specific false value (such as 0 or None)
>prevents you from easily changing the object being used.  Rely on the
>object to tell you whether it's true or false, and you'll be fine.
>
>That's polymorphic.

This kind of usage seems common:

 >>> def foo(x=None):
 ...     if x: print "Aahz says we'll be fine with %s" % `x`
 ...     else: print "But what about %s ?" % `x`
 ...     if x is None: print "Here we know we were passed %s (maybe by default)" % `x`
 ...
 >>> foo('something')
 Aahz says we'll be fine with 'something'
 >>> foo(0)
 But what about 0 ?
 >>> foo()
 But what about None ?
 Here we know we were passed None (maybe by default)

Do you recommend using def foo(*args) and testing len(args) instead?

Regards,
Bengt Richter


From piet at cs.uu.nl  Mon Jun 24 07:51:00 2002
From: piet at cs.uu.nl (Piet van Oostrum)
Date: 24 Jun 2002 13:51:00 +0200
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <aer2fn$frc$1@peabody.colorado.edu> <3D1214B6.E9BB77D2@noaa.gov> <3D125B79.6090903@skippinet.com.au> <af0lll$a7ntf$1@ID-99513.news.dfncis.de>
Message-ID: <wzlm94dhej.fsf@nono.cs.uu.nl>

>>>>> "P. Alejandro Lopez-Valencia" <dradul at yahoo.com> (PAL) writes:

PAL> "Mark Hammond" <mhammond at skippinet.com.au> escribi? en el mensaje
PAL> news:3D125B79.6090903 at skippinet.com.au...
>> Chris Barker wrote:
>> 
>> > On that note: does anyone know if there is a way to turn a Python
PAL> script
>> > into something that acts like an application on Windows...without
>> > resorting to Py2exe and the like. What I want is something just
PAL> like:
>> 
>> Add the .py extension to your PATHEXT environment variable, and you
PAL> can
>> execute "foo.py" at the command prompt.  Like much of Windows, this is
>> based on the extension rather than file attributes.

PAL> Ehem! :) This only works in WINNT class OSs that use cmd.exe as their
PAL> default command line processor. Lowly DOS32 OSs such as the Win98
PAL> mentioned by the OP, that only have command.com, just can't.

In that case use a program like PythonLauncher, which allows you to make a
small .exe file for each Python script (just copy the generic program to
script.exe where script is your script name (minus the .py extension).
This has the additional advantage of properly redirecting I/O from the
command line, which isn't the case with the other suggestions.
-- 
Piet van Oostrum <piet at cs.uu.nl>
URL: http://www.cs.uu.nl/~piet [PGP]
Private email: P.van.Oostrum at hccnet.nl


From tim at vegeta.ath.cx  Sun Jun  2 01:54:15 2002
From: tim at vegeta.ath.cx (Tim Hammerquist)
Date: Sun, 02 Jun 2002 05:54:15 GMT
Subject: What does sys.exit(1) means?
References: <adcb22$100q9d$1@ID-49758.news.dfncis.de>
Message-ID: <slrnafjd0r.t12.tim@vegeta.ath.cx>

Ken graced us by uttering:
> What does sys.exit(1) means?

Simply, shut down the current (Python) process and return a status of 1
to the shell.

But from the Python docstrings:

$ python
Python 2.2 (#1, Jan 17 2002, 22:02:17)
[GCC 2.95.3 20010315 (SuSE)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import sys
>>> print sys.exit.__doc__
exit([status])

Exit the interpreter by raising SystemExit(status).
If the status is omitted or None, it defaults to zero (i.e., success).
If the status numeric, it will be used as the system exit status.
If it is another kind of object, it will be printed and the system
exit status will be one (i.e., failure).
>>>

HTH
Tim Hammerquist
-- 
"Did someone open up your skull and carefully
replace your brain with rat droppings?   [Say yes!]"
    -- Tom Servo, MST3K


From Adndrew at novaresponse.com.au  Wed Jun  5 01:40:01 2002
From: Adndrew at novaresponse.com.au (Adndrew at novaresponse.com.au)
Date: Tuesday, 04 Jun 2002 23:40:01 -0600
Subject: Urgent News Flash
Message-ID: <04060223.4001@novaresponse.com.au>

URGENT NEWSFLASH

The government has installed BLACK BOXES into ISPs to watch your
Internet surfing.

When you are targeted, the first you will know is the KNOCK AT YOUR
DOOR and the JAIL SENTENCE.

If you think you have cleaned your computer - you are WRONG.

They can recover from your PC every single picture or item you have
ever watched on the Internet.

You need protection. You need an Evidence Eliminator. 

Click Here.

http://www.geocities.com/ee773636




From salmonia.nospam at cardiff.ac.uk  Mon Jun 10 06:02:44 2002
From: salmonia.nospam at cardiff.ac.uk (Alan James Salmoni)
Date: Mon, 10 Jun 2002 11:02:44 +0100
Subject: Confidence intervals, stats modules?
References: <ug2r3aac4etnef@corp.supernews.com> <adt0hs$3bl$1@news7.svr.pol.co.uk> <mailman.1023574153.24062.python-list@python.org>
Message-ID: <3D047944.7070305@cardiff.ac.uk>

Problem is that I don't think stats.py has confidence intervals (it does 
have mean, stddev etc). I am working on it myself for my own stats 
package, but it might be a few weeks as I am concentrating upon non 
parametric stuff.

Alan.

Tim Churches wrote:
> Duncan Smith wrote:
> 
>>"Stephen Boulet" <stephen at theboulets.net> wrote in message
>>news:ug2r3aac4etnef at corp.supernews.com...
>>
>>>Are there any stats modules to calculate things like confidence intervals,
>>>mean, standard deviation? Thanks.
>>>
>>>-- Stephen
>>
>>http://www.nmr.mgh.harvard.edu/Neural_Systems_Group/gary/python.html
>>
> 
> 
> Gary's stats module is excellent and is probably sufficient for the uses
> you
> describe. There are also some elementary statistical functions in Konrad
> Hinsen's
> equally excellent Scientific Python package - see
> http://starship.python.net/crew/hinsen/scientific.html
> 
> In fact, you could just use NumPy (Numerical Python) - have a look at
> the MLab 
> module in NumPy (which is at http://www.pfdubois.com/numpy/ ).
> 
> However, if you are interested in more advanced statistical analysis
> and/or in 
> drawing statistical graphs, then have a look at RPy, by Walter Moreira, 
> at http://rpy.sourceforge.net
> 
> RPy embeds the R stats environment within Python. R is good for most
> things
> statistical, from the elementary to the advanced and experimental, and
> it does
> truly beautiful graphics. It is a mature project backed by a number of
> eminent 
> statisticians in five continents.
> 
> Note that R is multi-platform (see http://www.r-project.org ) but RPy
> only
> works under Linux and Unix (and Mac Os X, perhaps) at this stage. I
> believe 
> a Windows version is in the works.
> 
> Tim C
> 
> 




From wurmy at earthlink.net  Sat Jun 29 19:33:49 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Sat, 29 Jun 2002 23:33:49 GMT
Subject: private
References: <pan.2002.06.29.17.38.51.789527.440@myself.com>
Message-ID: <3D1E439D.9020300@earthlink.net>

Rhymes wrote:
> Is there any possibility to build up a _real_ private attribute?
> Such as C++ private data members...
> I know that with double-underscore prefix (and the name mangling related)
> I can " hide " data members but I think it's not a so crafty trick...
> Is there any PEP about it? Will Guido insert a private keyword
> in a future release of Python ?

IMHO, the lack of "real" private members is a feature.

You might want to look at 2.2's properties. However, using them adds quite a 
bit of extra code, and soon you'll be writing ugly C++ or Java-like code with a 
getter and setter for every attribute. A high price to pay for paranoia... ;-)

There are cases where some control over attributes is useful, but there are 
other ways to solve that...

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/



From cliechti at gmx.net  Wed Jun 12 18:49:11 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 13 Jun 2002 00:49:11 +0200
Subject: why not "'in' in 'in'"?
References: <ae8dg70217g@drn.newsguy.com>
Message-ID: <Xns922C8BE1DC69cliechtigmxnet@62.2.16.82>

Grant Griffin <not.this at seebelow.org> wrote in
news:ae8dg70217g at drn.newsguy.com: 
> 
> One of the things I like about Python is that it works as expected. 
> However, I sometimes accidently expect it to allow one to use "in" to
> determine if one string is "in" another, e.g.:
> 
>>>> 'in' in 'in'
>
> which yields a traceback:
> 
>    Traceback (most recent call last):
>      File "<interactive input>", line 1, in ?
>    TypeError: 'in <string>' requires character as left operand
> 
> Evidently one is invited to use "count" or "find" instead:
> 
>>>> 'in'.count('in') > 0
> 1
>>>> 'in'.find('in') >= 0
> 1
> 
> I prefer "find" over "count" because it (presumably) punts after the
> first occurrence.  However, neither one reads nearly as well as the
> (illegal) "in" form.
> 
> So why is the "in" operator limited to single characters when used as
> a membership operator?  This peculiar lack of generality seems a bit
> (dare I say it?) "non-Pythonic".  (That said, "Pythonic" seems to be a
> bit like "enlightenment": those who claim to have mastered it wisely
> avoid spelling out any concrete criteria by which others can judge
> their mastery <wink>.) 

because in works with elements of a sequence.

here's a 'home made' "in":

>>> def _in(ele, seq):
... 	for x in seq:
... 		if x == ele: return 1
... 	else:
... 		return 0
... 	
>>> _in(7, range(3))
0
>>> _in(7, range(9))
1

the implementation of 'in' just counts up and calls __gettattr__ until it 
ge the requested element or the sequence has no more elements. the elements 
of a string are characters - and here's your limitation. of course its 
possible for an object to override the in operator to change this, see 
below.

> Perhaps its somehow due to the overloading of "in", which is used both
> as a membership operator and as a separator gizmo inside "for" loops. 
> ("It's a floor wax/No, it's a desert topping/No it's both!")  But is
> that any good reason to exclude the handy, sensible, and downright
> intuitive use of "in" as a substring detector?

it's certainly possible to do:

>>> class mstr(str):
... 	def __contains__(self, other):
... 		if len(other) > 1:
... 			return self.find(other) >= 0
... 		else:
... 			return str.__conatins__(self, other)
... 		
>>> mstr("Hello how are you?")
'Hello how are you?'
>>> s = mstr("Hello how are you?")
>>> 'how' in s
1

> Or maybe support for multiple characters would penalize the
> performance of the single-character "in".

that would only apply to strings and checking for length > 1 in the C-
source shouldn't be that time consuming (especialy as currently a check 
for != 1 is already in there).

don't know why 'in' doesn't work that way for strings - it bites be too 
from time to time...

maybe if someone did a patch of the C sources and place it on SF....

chris

-- 
Chris <cliechti at gmx.net>



From imbosol at vt.edu  Mon Jun 24 14:32:43 2002
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 24 Jun 2002 14:32:43 -0400
Subject: Suggestions for good programming practices?
References: <WEuR8.2074$5l4.75657@ozemail.com.au>
Message-ID: <bko7fa.9t.ln@grey.aerojockey.localdomain>

Dianne van Dulken wrote:
> Hi,
> 
> I'm fairly new to python, coming from a perl background, and was wondering
> if anyone has a list of things that they consider as good programming
> practice in Python (I thought this might be an interesting topic for
> discussion, anyway)
> 
> For example, in perl you obviously are always encouraged to use strict, and
> we always use eval around our main code to pick up any unexpected errors.  I
> presume the equivalent in python is the try statement.  Would you suggest
> that this should be used, or would that be redundant here?  Any other tips
> that you would suggest a new python user should be always doing as a good
> programming practice?


Python has relatively few of these.  I've been racking my brain to
think of any, but could only think of one really serious issue we face
in Python.  Python is a cautious language, and if you overlook
something there's a good chance you'll see a stack trace.

Anyways, the really important issue is:

* Do not mix spaces and tabs for indentation.  Choose one or the other
  and stick with it.  It helps to have an editor that indents
  intelligently and does nothing funny with the indents.


There's other stuff, but the same ideas apply to languages other than
Python.

* Avoid exec, execfile, eval, and input.

* Use local variables whenever possible.  Important in Python because
  local variables are much faster than global.  In practice, this
  means that people will often put their main code in a function, and
  call it with the "if __name__=='__main__': main()" cliche.

* Learn how references and the object system work as soon as possible.
  You should not be surprised when you do this:

  a = [1,2,3]
  b = a
  b[0] = 100
  print a

  And it prints "[100,2,3]"


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From gerhard at bigfoot.de  Sat Jun 15 19:07:37 2002
From: gerhard at bigfoot.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Sun, 16 Jun 2002 01:07:37 +0200
Subject: python version?
In-Reply-To: <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com>
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com>
Message-ID: <20020615230737.GC3672@lilith.my-fqdn.de>

* George Hester <hesterloli at hotmail.com> [2002-06-15 22:58 +0000]:
> Python ActiveX Scripting Engine error '80020009'
> 
> Traceback (innermost last): File "<Script Block >", line 5, in ?
> Response.Write(sys.version) NameError: name 'sys' is not defined
> 
> /python.asp, line 14
> 
> Response.Write(sys.version)
> 
> I think I need a script to do this.

You did not import the sys module.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From dswegen at software.plasmon.com  Thu Jun 27 07:53:57 2002
From: dswegen at software.plasmon.com (Dave Swegen)
Date: Thu, 27 Jun 2002 12:53:57 +0100
Subject: Difference between HTMLParser and htmllib.HTMLParser
Message-ID: <20020627115357.GA15813@software.plasmon>

I've just started playing with these, and I'm curious which is the
better of the two (I noticed the HTMLParser module only came in in 2.2).
And is there any reason why the names are so confusingly similiar? I
tried searching in google for info on this, but the daft naming made it
rather hard ;)

Cheers
    Dave




From jwilhelm at outsourcefinancial.com  Thu Jun 20 14:22:11 2002
From: jwilhelm at outsourcefinancial.com (Joseph Wilhelm)
Date: 20 Jun 2002 11:22:11 -0700
Subject: Proper number alignment
In-Reply-To: <aet5hu$9leln$1@ID-98166.news.dfncis.de>
References: <aet5hu$9leln$1@ID-98166.news.dfncis.de>
Message-ID: <1024597332.7665.2.camel@jwilhelm.ofsloans.com>

On Thu, 2002-06-20 at 11:05, mdk wrote:
> I am using \t\t (tabs) to align these numbers (disregard the math):
> 
> Total RAM                2,141,056
> Available RAM            12,934,032
> Used RAM                 1,024
> 
> But they are left justified and I need them right justified.
> 
> Any ideas?
> 
> Thanks.
> -- 
> http://mail.python.org/mailman/listinfo/python-list

Is this what you're looking for?

>>> print "%-20s %20s\n%-20s %20s\n%-20s %20s" % ( "Total RAM",
"2,141,056", "Available RAM", "12,934,032", "Used RAM", "1,024" )
Total RAM                       2,141,056
Available RAM                  12,934,032
Used RAM                            1,024


--Joseph Wilhelm





From eric.brunel at pragmadev.com  Wed Jun 19 09:33:48 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Wed, 19 Jun 2002 13:33:48 +0000
Subject: tkinter menu (newbie question)
References: <Xns9232778FCD335stefninnospamyahoofr@213.228.0.138>
Message-ID: <aepps0$al$1@wanadoo.fr>

Stephane Ninin wrote:

>   Hi all,
> 
> I am completely new to kinter and quite new too to Python,
> so ... (sorry if my question has already been asked many times here...)
> 
> 
> here is a simple program I am trying to make
> 
> <--->
> 
> #! /Python22/python
> 
> from Tkinter import *
> 
> class AppWin(Frame):
> 
>     def __init__(self,master=None):
>       Frame.__init__(self,master)
>       self.grid()
>         
>       self.menubar = Menu(self)
> 
>  self.filemenu = Menu(self.menubar)
> 
>  self.filemenu.add_command(label="New")
>  self.filemenu.add_command(label="Open...")
>  self.filemenu.add_command(label="Close")
>  self.filemenu.add_separator()
>  self.filemenu.add_command(label="Quit", command=self.nothing)
> 
> self.editmenu = Menu(self.menubar)
> 
>       self.helpmenu = Menu(self.menubar, name='help')
> 
>       self.helpmenu.add_command(label="About...")
> 
> self.menubar.add_cascade(label="File", menu=self.filemenu)
> self.menubar.add_cascade(label="Edit", menu=self.editmenu)
> self.menubar.add_cascade(label="Help", menu=self.helpmenu)
[snip end of code]

This doesn't suffice. With all that code, you just created a frame that has 
an attribute named menubar with the menu in it, but you didn't do anything 
to actually display the menu (there's no "magical" attributes in Python: if 
you do not do something with an attribute, it is only just that, an 
attribute...).

What you should do is to make MyApp a sub-class of Tk instead of Frame 
(this is more pratical this way), and add at the end of the constructor the 
line:

self.configure(menu=self.menubar)

And that should do the work. Oh: and don't forget to add parameter "self" 
to your "nothing" method, or you'll get a nice traceback when you do 
File->Quit...

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From loewis at informatik.hu-berlin.de  Fri Jun  7 15:44:45 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 07 Jun 2002 21:44:45 +0200
Subject: Python 2.2.1 vs. gcc 3.1?
References: <yu99vg8v6orb.fsf@europa.research.att.com> <15616.57963.519636.730290@12-248-41-177.client.attbi.com> <200206071747.g57HlbH01057@europa.research.att.com> <mailman.1023472897.6311.python-list@python.org> <yu99fzzy7uqj.fsf@europa.research.att.com>
Message-ID: <j4k7paj2ky.fsf@informatik.hu-berlin.de>

Andrew Koenig <ark at research.att.com> writes:

> ImportError: ld.so.1: ./python: fatal: relocation error: file /tmp/build-python/Python-2.2.1/build/lib.solaris-2.8-sun4u-2.2/struct.so: symbol PyString_Type: referenced symbol not found
> 
> So apparently the shared libraries are not being built properly.
> 
> Any more ideas?

The typical cause of this problem is that you are using GNU binutils,
in which case the compiler needs the option -Wl,--export-dynamic when
linking the Python executable. 

Now, there is machinery in configure.in to determine whether this flag
is needed. Traditionally (up to Python 2.2.1), this machinery invokes
'gcc -Xlinker -V' and expects to find 'BFD' in its output if it is GNU
ld, such as

GNU ld version 2.11.2 (with BFD 2.11.2)
  Supported emulations:
   elf32_sparc
   elf64_sparc

Starting with binutils 2.12, GNU ld stops printing BFD in its -V
output, so that this detection mechanism fails.

This has been fixed in the CVS, by checking presence of the
--export-dynamic linker option directly (configure.in 2.288.6.5, for
the 2.2 branch).

Regards,
Martin


From mgilfix at eecs.tufts.edu  Tue Jun 11 13:20:03 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Tue, 11 Jun 2002 13:20:03 -0400
Subject: Source formatting for long expressions.
In-Reply-To: <878z5lx4vl.fsf@twistedmatrix.com>; from radix@twistedmatrix.com on Tue, Jun 11, 2002 at 10:32:46AM -0400
References: <mailman.1023570972.27430.python-list@python.org> <m3it4srhq5.fsf@mira.informatik.hu-berlin.de> <874rgcxy99.fsf@twistedmatrix.com> <mailman.1023668414.25769.python-list@python.org> <6bd9f01b.0206110543.168b0f83@posting.google.com> <878z5lx4vl.fsf@twistedmatrix.com>
Message-ID: <20020611132003.I5683@eecs.tufts.edu>

On Tue, Jun 11 @ 10:32, Christopher Armstrong wrote:
> XML is a terrible format for object persistence, due to the lack of robust
> mappings. Twisted's XML uses attributes in some cases, but that is only
> possible when you have string:string mappings. In *any* other case, you
> have to resort to children nodes, with such ugly syntax as::
> 
> <instance class="twisted.some.Class">
> <dictionary>
>   <string role="key" value="someAttr" />
>   <list>
>     <string value="foo" />
>   </list>
> </dictionary>
> 
> Compare that with the equivalent AOT format::
> 
> Instance("twisted.some.Class", 
>   someAttr=["foo"],
> )
> 
> *That* is why I don't just use XML.

  I'm not disagreeing with you because I don't think you're wrong in
this case and well, XML isn't for everything. I think XML can work
here just as well but it's just more work. I'm not sure about the
lack of robustness of the mappings - you just have to define them and
can use a DTD to verify that your mappings are intact. But you have
to write extra code for the mappings that isn't quite worth it in
such a domain specific case. XML costs extra for being general.  Er,
you don't have any between version/backwards-compat issues though do
you? In which case, have you thought those out thoroughly?  Forgive me
if it isn't relavent. Not sure exactly where these will be used.

                    -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From cliechti at gmx.net  Sun Jun  2 14:13:53 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 2 Jun 2002 20:13:53 +0200
Subject: Pop a list from beginning ? and memory saving...
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de> <3CFA0676.68F80DE6@engcorp.com> <add289$vud42$1@ID-146704.news.dfncis.de>
Message-ID: <Xns9221CE2AF76D0cliechtigmxnet@62.2.16.82>

"Shagshag13" <shagshag13 at yahoo.fr> wrote in
news:add289$vud42$1 at ID-146704.news.dfncis.de: 
> and yes i'm working with large amount of data : i'm trying to index
> something like 1 go of plain text by using conventionnal text

1go thats 1GB of data for those who don't know french :-)

> retrieval structures (i mean without compression tricks, because i
> *don't* have time to implement theses, but with stoplist and so on) 
> and wishing to do it in one pass without relying on hdd temp files
> (i'm starting to think that i were dreaming...)
> 
> the most i think i could simply do is this : i use integer each time
> as possible, i free everything i don't need anymore with = None, and
> what could i do more ? Advices are ***greatly*** welcome !!!

- use "del var" instead of "= None".
- try to work in junks, do you realy hae 1GB in memory? maybe you can work 
in blocks of 1MB of 64KB or so.
- if the result is big and not immediatly needed in the program, then write 
it to the disk.
- maybe two-pass isn't that bad - especialy if your first step algorithm 
reduces the amount of data by some factors.

chris

-- 
Chris <cliechti at gmx.net>



From hesterloli at hotmail.com  Thu Jun 20 14:06:42 2002
From: hesterloli at hotmail.com (George Hester)
Date: Thu, 20 Jun 2002 18:06:42 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <g71Q8.33179$uk2.13779733@twister.nyroc.rr.com>
Message-ID: <SQoQ8.37979$uk2.15427856@twister.nyroc.rr.com>

If some kind soul had given me this when I asked for it ida gone away peacfully.
Maybe someone did up in the dreck.

--
George Hester
_________________________________
"George Hester" <hesterloli at hotmail.com> wrote in message
news:g71Q8.33179$uk2.13779733 at twister.nyroc.rr.com...
> Welp:
>
> <%@LANGUAGE=Python%>
> <HTML>
> <head></head>
> <body>
> <h1>Python Test</h1>
>
> <%
> #do some python stuff here
> import sys
> Response.Write('Python Test<br>')
> Response.write('<h3>Smaller heading</hr>')
> Response.Write(sys.version)
> %>
>
> </body>
> </html>
>
> Thank you Mark McEaheren.  He didn't give me this but pointed me in the right
> direction.  I suppose others tried to do the same thing but he gave a nudge
> using the command prompt and that is all it took.  Since my knowledge was <0
in
> this language all that I read looked Greek to me.
>
> --
> George Hester
> _________________________________
> "George Hester" <hesterloli at hotmail.com> wrote in message
> news:K7PO8.16057$uk2.8406537 at twister.nyroc.rr.com...
> > What is a simple script that will tell me the version of python I am
currently
> > running?  IIS 5.0 Windows 2000.
> >
> > --
> > George Hester
> > _________________________________
> >
> >
>
>




From fredrik at pythonware.com  Thu Jun 27 12:50:32 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 27 Jun 2002 16:50:32 GMT
Subject: How to find out operating system
References: <mailman.1025159195.9033.python-list@python.org>
Message-ID: <snHS8.45297$n4.10730843@newsc.telia.net>

"A" wrote:

> What is the best way of  finding out the kind operating system?
> I can use os.name but if the system is Windows I would like also
> know if the system is Windows98 or Windows ME or W2K or
> Windows XP.

some variation of

    import os
    if os.name == "nt":
        version = os.popen("ver").read()
        if version.find("Windows 2000") >= 0:
            ... windows 2000
        elif ...
        ...
    else:
        version = os.uname()
        version = version[0], version[1]
        ...

might do the trick.

or use the PythonWin extensions (see lawrence's post)

</F>




From pinard at iro.umontreal.ca  Tue Jun 25 16:36:58 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 25 Jun 2002 16:36:58 -0400
Subject: Parsing strings (\n and \\)
In-Reply-To: <lt0S8.45025$n4.10623651@newsc.telia.net>
References: <3D188006.7050202@thomas-guettler.de>
	<mailman.1025017712.20580.python-list@python.org>
	<lt0S8.45025$n4.10623651@newsc.telia.net>
Message-ID: <oqofdzxfh1.fsf@titan.progiciels-bpi.ca>

[Fredrik Lundh]

> > A simple avenue is to get Python itself to evaluate the string
> > as a constant (you ensure this by removing evaluation context).
> > Something like this:
> >
> >     PARSED = eval(UNPARSED, {}, {})

> >>> UNPARSED = "__import__('os').system('echo dream on!')"
> >>> PARSED = eval(UNPARSED, {}, {})
> dream on!

:-) :-)

In practice, I always matched UNPARSE with a regexp which is convenient
for a string.  I remove evaluation context because none is needed.  The the
danger is probably limited or does not exist in my actual cases.

But it did not occur to me how fragile the above writing may be.  Yet another
good reason to avoid some Python constructs, like `eval', whenever possible.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From peter at engcorp.com  Fri Jun  7 19:20:17 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 07 Jun 2002 19:20:17 -0400
Subject: Efficient python programming...
References: <mailman.1023452786.27923.python-list@python.org>
Message-ID: <3D013FB1.3D7477EA@engcorp.com>

Roman Suzi wrote:
> 
> On Fri, 7 Jun 2002, Peter Hansen wrote:
> > You forgot the even more important first thing for a beginner:
> > get it correct!  The only good way to do this is to write unit
> > tests that prove it.  If you don't have tests, you don't have
> > working code (or how can you prove it?).
> 
> To prove that code is correct is very difficult (practically impossible).
> Unittests are there to prove that you have errors, not the other way
> around.
> 
> Of course, they can increase your confidence in that optimized code is
> equivalent to what you had before.

I disagree, if the tests are written first (and run, proving that
they will fail when the code is wrong or missing).

The tests are executable requirement specifications.  Running
them "proves" (yes, not 100%... nothing could) that the code
meets the specs.  The specs could be wrong of course, but that
doesn't mean the code is broken...

-Peter


From nashden2002 at yahoo.com  Mon Jun 10 11:07:43 2002
From: nashden2002 at yahoo.com (Dennis Nash)
Date: 10 Jun 2002 08:07:43 -0700
Subject: Is python suitable for embedded programming?
Message-ID: <aa104658.0206100707.75889ccf@posting.google.com>

I am considering using Python to control embedded hardware, and would
like to know whether other people have done this, what their experiences
were, and whether they recommend it.

The system in question is in three parts: Industrial hardware talks to a
control system (a Linux or Windows2000 box). The control system outputs
the data from the hardware onto a web server. A 3rd computer contains a 
web browser which will interact with the web server, allowing an operator
to view what's happening with the hardware and change parameters.

I am thinking of using Python for the middle part, i.e. the control 
system which polls the hardware continually to get its state, and updates
the web pages accordingly.

Some specific questions:

* how good is serial port access in Python?

* the program will be running forever, in a big loop. Is it liable to 
gradually use all the memory up and then fall over?

* output to web pages: how good is Python Server Pages? is this the best
way to do it?


From shagshag13 at yahoo.fr  Tue Jun  4 12:56:57 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Tue, 4 Jun 2002 18:56:57 +0200
Subject: int, float and string...
References: <adhut1$10rtkl$1@ID-146704.news.dfncis.de> <mailman.1023206585.1785.python-list@python.org>
Message-ID: <adirg9$11hv44$1@ID-146704.news.dfncis.de>

"Mike C. Fletcher" <mcfletch at rogers.com> a ?crit dans le message de news: mailman.1023206585.1785.python-list at python.org...
> When you cast a float to an int you potentially lose information (the
> decimal fraction).  The conversion is also by way of truncation (for
> historic reasons), so 1.99999 -> 1 (not 2), which likely isn't what you
> wanted, (use round( num, 0) instead).  Floats (C doubles) are larger
> than ints (C longs) at the machine level (8 vs 4 bytes on 32-bit
> machines if I recall correctly), but the overhead of Python objects is
> huge compared to the size of the actual data, so don't worry about it
> unless you're using Numeric Python and slogging around huge arrays of
> numbers (where worrying about it might actually save you time).

Thanks again for your replies !

Where could i go to see and understand what you call "overhead of
Python objects" ?

thanks,

s13.





From aahz at pythoncraft.com  Tue Jun  4 10:02:17 2002
From: aahz at pythoncraft.com (Aahz)
Date: 4 Jun 2002 10:02:17 -0400
Subject: Need advice on multithreading problem
References: <slrnafpeei.p8.gerhard@developer2.opus-gmbh.net>
Message-ID: <adih99$rkm$1@panix1.panix.com>

In article <slrnafpeei.p8.gerhard at developer2.opus-gmbh.net>,
Gerhard =?iso-8859-15?Q?H=E4ring?=  <gerhard at bigfoot.de> wrote:
>
>The problem is this: I have a C library with a function of the following
>interface:
>
>    def sqlite_exec(conn, sql, callback, arg1):
>
>The only important parameter here is 'callback', it is a callback
>function that you need to give to sqlite_exec. The callback function
>will be called with the items sqlite_exec returns. If the callback
>functions returns a value other than zero, sqlite_exec will stop its
>processing.
>
>On the Python side, I wanted to force this C library call into a
>generator.  Which I did using a thread that's spawned for a new
>sqlite_exec call and Queue.Queue.

Is SQLite thread-safe?  Do you provide a thread-local conn?  Does
sqlite_exec() return a brand-new cursor?  If the answer to any question
is "no", your code is probably broken (haven't actually read it).
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"In the end, outside of spy agencies, people are far too trusting and
willing to help."  --Ira Winkler


From logiplexsoftware at earthlink.net  Tue Jun 11 20:20:59 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 11 Jun 2002 17:20:59 -0700
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
In-Reply-To: <20020611171553.077d5ebd.logiplexsoftware@earthlink.net>
References: <Xns9229EE722467Fgustafl@195.100.94.182>
	<DafN8.125526$4i.12087144@bin2.nnrp.aus1.giganews.com>
	<sNpN8.132081$4i.12440184@bin2.nnrp.aus1.giganews.com>
	<20020611171553.077d5ebd.logiplexsoftware@earthlink.net>
Message-ID: <20020611172059.741bcbaf.logiplexsoftware@earthlink.net>

On Tue, 11 Jun 2002 17:15:53 -0700
Cliff Wells wrote:
[snip]

> # ------------
> ranges = createRanges(10000)
> 
> print "Initialize Range:",
> t1 = time()
> r = Ranges(ranges)
> print time() - t1
> 
> t2 = time()
> print "Range():",
> for i in range(0, 100, 3):
>     r.has(i)
> t = time()
> print t - t2, "total", t - t1
> 
> t = time()
> print "isin():",
> for i in range(0, 100, 3):
>     isin(i, ranges)
> print time() - t
> 
> 
> ####################################
> Initialize Range: 0.331060051918
> Range(): 0.00136804580688 total 0.33601808548
> isin(): 4.98800003529

Oops.  This is a better test as it covers the entire range (the first test was
actually weighted towards isin()):

ranges = createRanges(10000)

print "Initialize Range:",
t1 = time()
r = Ranges(ranges)
print time() - t1

t2 = time()
print "Range():",
for i in range(0, 10000, 100):
    r.has(i)
t = time()
print t - t2, "total", t - t1

t = time()
print "isin():",
for i in range(0, 10000, 100):
    isin(i, ranges)
print time() - t

#####################
Initialize Range: 0.323083996773
Range(): 0.00372695922852 total 0.331498980522
isin(): 12.9388070107

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From mgilfix at eecs.tufts.edu  Tue Jun 11 14:26:03 2002
From: mgilfix at eecs.tufts.edu (Michael Gilfix)
Date: Tue, 11 Jun 2002 14:26:03 -0400
Subject: XML/Source Persistence (was: Source formatting for long expressions.)
In-Reply-To: <874rg9wvi7.fsf_-_@twistedmatrix.com>; from radix@twistedmatrix.com on Tue, Jun 11, 2002 at 01:55:12PM -0400
References: <mailman.1023570972.27430.python-list@python.org> <m3it4srhq5.fsf@mira.informatik.hu-berlin.de> <874rgcxy99.fsf@twistedmatrix.com> <mailman.1023668414.25769.python-list@python.org> <6bd9f01b.0206110543.168b0f83@posting.google.com> <878z5lx4vl.fsf@twistedmatrix.com> <20020611132003.I5683@eecs.tufts.edu> <874rg9wvi7.fsf_-_@twistedmatrix.com>
Message-ID: <20020611142602.L5683@eecs.tufts.edu>

On Tue, Jun 11 @ 13:55, Christopher Armstrong wrote:
> I meant the mappings that attributes provide. Creating mappings with multiple
> tags is overbearingly verbose, no matter if you formalize it or not. (Am I
> understanding you correctly?) Perhaps I should've said "flexible" rather
> than "robust". BTW, ONX (http://www.seairth.com/web/onx/onx.html) provides
> a better type of mapping: string:arbitrary object. It's not
> arbitrary object:arbitrary object, but it's a bit better.

  I agree that's very verbose and cumbersome. That's why I supported your
non-use of XML. ONX is interesting though. Gonna have to take a closer look.

> Versioning the persistent data is pretty much irrelevant wrt the format used --
> and yes, I have thought about it. Marmalade right now uses it's own versioned
> state-getters/setters (jellyToDOM_X where X is the version number), but AOT
> doesn't have its own interface -- it just uses __getstate__/__setstate__ like
> pickle. I'm probably going to write a mixin class for versioned state that
> supports methods like __getstate_X__ where X is the version of the persisted 
> state.

  If anything, I think that's where the XML format is nice - in versioning.
Usually, you get to do considerably more code-sharing with XML formats and
the update functions are very simplistic.

> Note that Twisted has a class for versioning state that's orthogonal to
> persisted format: twisted.persisted.styles.Versioned. This allows you
> to define 'upgradeToVersionX' methods that will be called in order to
> upgrade an object to the current version of the code. The only reason
> to use versioned state getters/setters as opposed to this is if you want
> the format of the persistent state to change for cosmectic reasons.

  This sounds prety cool. Will have to check it out.

                     -- Mike

-- 
Michael Gilfix
mgilfix at eecs.tufts.edu

For my gpg public key:
http://www.eecs.tufts.edu/~mgilfix/contact.html




From gerhard at bigfoot.de  Mon Jun 17 20:06:21 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Tue, 18 Jun 2002 02:06:21 +0200
Subject: Python Executable Code
In-Reply-To: <1024357771.569c0ff8rchopra@myrealbox.com>
References: <1024357771.569c0ff8rchopra@myrealbox.com>
Message-ID: <20020618000621.GA3119@lilith.my-fqdn.de>

* Rajat Chopra <rchopra at myrealbox.com> [2002-06-17 19:49 -0400]:
> Does anybody know how to convert a Python file (*.py) to a Windows
> Executable format (*.exe) ?
 
http://www.mcmillan-inc.com/install1.html
http://py2exe.sourceforge.net/

Ger'URL-bot'hard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From peter at engcorp.com  Sat Jun  1 17:49:14 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 01 Jun 2002 17:49:14 -0400
Subject: File Iterator - Print - Double New Lines
References: <mailman.1022957167.8339.python-list@python.org>
Message-ID: <3CF9415A.250B1260@engcorp.com>

Mike Brenner wrote:
> 
> print "QUESTION: Why did it jump from NO LINE SPACING to DOUBLE SPACING?"
> print "QUESTION: How to get SINGLE SPACING?"
> print
> print "Using Python 2.2.1, it does the same under win32all build 142 and "
> print "      running it from a DOS box without the windows extensions."

[snip code with troubles]

Mike, the print statement adds newlines after every line it prints.
If you want to see the real contents of the file, use sys.stdout.write().
Alternatively, you might want to strip the newlines from each line
as you read them in.  Depends what you are trying to do.

-Peter


From gerhard at bigfoot.de  Fri Jun 21 10:00:33 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 21 Jun 2002 14:00:33 GMT
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <Xns923496C17397Frcamesz@amesz.demon.nl>
Message-ID: <slrnah6c7h.154.gerhard@lilith.my-fqdn.de>

In article <Xns923496C17397Frcamesz at amesz.demon.nl>, Robert Amesz wrote:
> The most goto-like contructs in Python are 'return' (which exits from a 
> function/method before the end is reached) and 'break' (which exits a 
> loop, continuing with statement followin that loop).

And 'continue', which immediately goes to the beginning of the loop.

Speaking of which, unlike in some other languages, 'break' and 'continue'
cannot be used with named loops, so you need exceptions in this case.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From peter at engcorp.com  Thu Jun 20 21:07:28 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 20 Jun 2002 21:07:28 -0400
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com> <3D1010C3.C321F8B0@engcorp.com> <raXP8.47926$GY.15260255@twister.nyroc.rr.com> <3D107948.B3FEE4BB@engcorp.com> <slrnah4spv.557.quinn@retch.ugcs.caltech.edu>
Message-ID: <3D127C50.E4720CBA@engcorp.com>

Quinn Dunkan wrote:
> 
> On Wed, 19 Jun 2002 08:30:00 -0400, Peter Hansen <peter at engcorp.com> wrote:
> >George Hester wrote:
> >>
> >> Yes my take on it was that you said "I needed to crawl before I walk."
> Please
> >> don't hold my inference against me.
> >
> >That's precisely what I said.  Nothing derogatory about it, except
> 
> ... maybe if he's Balinese?
> 
> The name doesn't sound like it though.

I thought that was just plain weird of you, but Google knows all,
even when you expect to get zero results:

  http://www.google.ca/search?q=balinese+crawl+walk

Go figure. :)

-Peter


From gh_pythonlist at gmx.de  Sat Jun  1 13:51:31 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 1 Jun 2002 19:51:31 +0200
Subject: Copying a database
In-Reply-To: <FL6K8.771$NG1.217@newsread2.prod.itd.earthlink.net>
References: <FL6K8.771$NG1.217@newsread2.prod.itd.earthlink.net>
Message-ID: <20020601175131.GB14567@lilith.my-fqdn.de>

* Billy Ng <kwokng at earthlink.net> [2002-06-01 16:37 +0000]:
> I need you guys help on mysql again.  I need to push my python program to
> production now, but I don't want to recreate the mysql database on the
> production box.  I want to copy the database from the development box to
> production box.. Would anybody please tell me how to do it, thanks!

Use mysqldump to export a dump. Then use mysql to import the dump to the
remote database. Like this:

mysqldump {your options here} >dumpfile   # redirect to dumpfile
mysqldump {your options here} -h remotehost <dumpfile   # read from dump

If the database is not very big, this approach works fine. For
multi-gigabyte databases, I'm sure there is a more efficient approach
;-)

The MySQL manual should describe these commands in detail.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 20.0 ?C      Wind: 2.4 m/s




From narnett at mccmedia.com  Fri Jun 14 12:15:49 2002
From: narnett at mccmedia.com (Nick Arnett)
Date: Fri, 14 Jun 2002 09:15:49 -0700
Subject: executemany & UPDATE (MySQL & Python)
Message-ID: <NDBBLAEMOMEHKALAMKKJCEIGKPAA.narnett@mccmedia.com>

Can anyone confirm that it is *not* possible to pass a list to the Python
MySQLdb module in UPDATE operations?  In other words, this kind of thing
works:

self.dbh.execute("INSERT INTO Foo (blah, blorg,snork) VALUES
(%s,%s,%s)",myList)

... But this kind doesn't appear to work for me:

self.dbh.execute("UPDATE Foo SET blah=%s, blorg=%s WHERE snork=%s",myList)

Doing the latter with a loop in Python is much slower than I'd expect that
doing it with a list would be.

Nick

--
narnett at mccmedia.com
(408) 904-7198





From whisper at oz.net  Sun Jun  9 16:08:15 2002
From: whisper at oz.net (David LeBlanc)
Date: Sun, 9 Jun 2002 13:08:15 -0700
Subject: Newbie -  Python21.dll not found 
In-Reply-To: <_KNM8.11799$Q43.303520@nnrp1.ptd.net>
Message-ID: <GCEDKONBLEFPPADDJCOEEELFDHAA.whisper@oz.net>

It should be in c:\winnt\system32 and this directory should already be in
your path.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Stan
> Sent: Sunday, June 09, 2002 12:11
> To: python-list at python.org
> Subject: Newbie - Python21.dll not found
>
>
> Hi,
>         When I try to start up python a message "The dynamic link library
> python2.dll could not be found in the specified path". The app
> was installed
> along with other apps when  I have installed the Foxserv 2.1 on
> W2k system.
>
> Thanks in advance,
>
> Stan
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From zayats at red.seas.upenn.edu  Tue Jun  4 11:32:11 2002
From: zayats at red.seas.upenn.edu (Salim Zayat)
Date: 4 Jun 2002 15:32:11 GMT
Subject: Canvas.items
Message-ID: <adimhr$8v0$1@netnews.upenn.edu>

Hey there.  I noticed that the Canvas widget, depending on which version 
of Python or Tkinter you are using, may/may not have the Canvas.items 
attribute.  Is there another way to reference the actual CanvasItem if you 
had the id of the tag?  For example, if I had a CanvasText item with the 
id 21, I want to do something like :

item = canvas.items[21]

Thanks!

Salim


From sholden at holdenweb.com  Mon Jun 24 18:09:08 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 24 Jun 2002 18:09:08 -0400
Subject: Help with MySQLdb insert with data extracted from odbc <-- Oracle
References: <3d13d5fb.651829078@netnews.attbi.com> <mailman.1024712368.16609.python-list@python.org> <3d147e64.694941703@netnews.attbi.com>
Message-ID: <IKMR8.5776$mF.3188@atlpnn01.usenetserver.com>

<candiazoo at attbi.com> wrote ...
> Ah!  Thank you!
>
> Mike J.
>
> On Sat, 22 Jun 2002 04:17:58 +0200, Gerhard =?iso-8859-15?Q?H=E4ring?=
> <gerhard.haering at gmx.de> wrote:
>
> >* candiazoo at attbi.com <candiazoo at attbi.com> [2002-06-22 01:43 +0000]:
> >> Can someone tell me what this means?  I mean, I guess I *know* what it
means,
> >> but why?
> >>
> >> The insert statement into a mysql database is...
> >>
> >> 'insert into works values(?, ?, ?, ?, ? ,?)'
> >
> >Wrong. paramstyle = 'format' in MySQLdb, not 'qmark'. Look into the
> >DB-API specs (PEP 0249) for details about the various paramstyles that
> >are possible.
> >
> >> The invocation of the dynamic sql is...
> >>
> >>     status = cur_mysql.execute(sql,
> >>         (wrk_inst, ZERO, trs_inst, aas_inst, title, NADA))
> >
> >Simply use %s instead of ?. And read the fine MySQLdb docs that are
> >included :-)
> >

Please also note that you are potentially asking for trouble by assuming a
particular order of the fields in your table definition. It would be *much*
safer (though a little more tedious) to write

cursor.execute("""INSERT INTO
        wrk_inst, sinker, trs_inst, aas_inst, title, pub_name)
        VALUES (%s, %s, %s, %s, %s, %s)""")

That way, when someone else appears in a year or two and jiggles around with
the table definition it's less likely your code will blow up because you are
trying to insert 'N/A' into a numeric field.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From mwh at python.net  Mon Jun 17 05:34:12 2002
From: mwh at python.net (Michael Hudson)
Date: Mon, 17 Jun 2002 09:34:12 GMT
Subject: [development doc updates]
References: <mailman.1024057266.26182.python-list@python.org> <ugk341aivqpl27@news.supernews.com>
Message-ID: <lku1o2l062.fsf@pc150.maths.bris.ac.uk>

"John Roth" <johnroth at ameritech.net> writes:

> "Fred L. Drake" <fdrake at acm.org> wrote in message
> news:mailman.1024057266.26182.python-list at python.org...
> > The development version of the documentation has been updated:
> >
> >     http://www.python.org/dev/doc/devel/
> >
> > Updated to reflect recent changes.
> 
> I'm not certain I understand item 5. Extended Slices.

Where specifically?  In the whatsnew document?  That section isn't
finished yet.

> Does this mean that the standard sequence types now support
> the stride operand, 

Yes.  Well, at least lists, tuples, strings (both flavours), arrays.

> or that the stride operand is being depreciated and will eventually
> vanish?

No.  Assuming you mean "deprecated" -- a common typo!.

Cheers,
M.

-- 
93. When someone says "I want a programming language in which I
    need only say what I wish done," give him a lollipop.
  -- Alan Perlis, http://www.cs.yale.edu/homes/perlis-alan/quotes.html


From kasey at baac.net  Wed Jun 26 23:09:31 2002
From: kasey at baac.net (Kasey)
Date: 26 Jun 2002 20:09:31 -0700
Subject: text ads infrastructure in python?
Message-ID: <fcb222df.0206261909.7d3506be@posting.google.com>

Does anyone know of a text-based ads system available in python? 
Looking for some infrastructure that I could tie in to some mailing
list technology and a web app.  Ideally it would tie in with a payment
system like PayPal and would do inventory management and reporting.

Looking to do something similar to the Google text ads or
www.textads.biz ...

Thanks,
Kasey


From sbalu at nyx.net  Thu Jun 27 21:53:39 2002
From: sbalu at nyx.net (sarayu balu)
Date: 27 Jun 2002 19:53:39 -0600
Subject: Newbie: GUI
References: <3D1B9687.8060606@linuxfreemail.com> <i%MS8.13738$p56.4384836@newsb.telia.net>
Message-ID: <1025229219.198257@irys.nyx.net>

I have had fun with Jython where the whole java Swing library is accessible
-sb
Fredrik Lundh <fredrik at pythonware.com> wrote:
: "coh" <leaf00 at linuxfreemail.com>:

:> Is there a native gui, widget set, for python, or any in the works?

: http://www.python.org/doc/FAQ.html#4.13

: according to many, tkinter is still the de-facto standard:
:     http://www.python.org/topics/tkinter/

: if you want to write tkinter-compatible widgets in pure
: python, use the Tkinter 3K WCK:
:     http://tkinter.effbot.org

: search google (or google groups) for more information,
: and endless discussions of why toolkit A is better than
: toolkit B, or not.

: </F>



From urnerk at qwest.net  Mon Jun 17 20:09:57 2002
From: urnerk at qwest.net (Kirby Urner)
Date: Mon, 17 Jun 2002 20:09:57 -0400
Subject: Please help with Python Script/MS-Access (DAO)
References: <aell87$3uv$1@netnews.upenn.edu>
Message-ID: <pan.2002.06.18.00.09.54.672123.11168@qwest.net>

The script below uses the sgml parser, which is tag
aware, and so ignores them in the output data.  I 
call the below with a web page as a parameter, assuming
the hard-coded base url, e.g.

$ ./testurl.py Plant1.html

and data.txt looks like this (... = skipped lines):


 OrderID,OrderDate,PlantID,ProductID,OrderQty
 19034,4/1/02,1,7,1732
 19035,4/1/02,1,9,1888
 19036,4/1/02,1,4,1048
 19037,4/1/02,1,5,1708
 19038,4/1/02,1,6,876
 ...
 19411,4/27/02,1,10,1288
 19412,4/27/02,1,8,1732
 19413,4/27/02,1,1,732
 19414,4/27/02,1,2,236
 19415,4/27/02,1,9,1596

Now it's up to you in another function/module to read
this downloaded file and knock off the header.  If you 
don't trust the 4-lines pattern, you could trigger off
getting a numeric (digit) as the first non-blank or
whatever.  At least you don't have the <p> and other
tags to mess with, thanks to the sgml parser stripping
'em already.

Kirby


=======================

#!/usr/bin/python

# with thanks to Fredrik Lundh, Python Standard Library (O'Reilly)

import urllib,sys
import sgmllib

class FoundEnd(Exception):
    pass

class Extract(sgmllib.SGMLParser):

    def __init__(self,verbose=0):
        sgmllib.SGMLParser.__init__(self,verbose)
        self.data = []

    def handle_data(self,data):
        self.data.append(data)

    def start_body(self,attr):
        print "Body Start"

    def end_body(self):
        print "Body End"
        raise FoundEnd

def getwebdata(wp):
    p  = Extract()
    n = 0
    try:
        while 1:
            s = wp.read(512)
            if not s:
                break
            p.feed(s)
        p.close()
    except FoundEnd:
        return p.data
    return None

if __name__ == '__main__':
    webpage = sys.argv[1]
    baseurl = "http://opim.wharton.upenn.edu/~opim101/spring02/"
    fp = urllib.urlopen(baseurl + webpage)
    output = open("data.txt","w")
    results = getwebdata(fp)
    fp.close()

    if results:
        for i in results:
            output.write(i)
    output.close()



From jepler at unpythonic.net  Tue Jun  4 12:56:52 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 4 Jun 2002 11:56:52 -0500
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
In-Reply-To: <019301c20be7$933acb50$7201a8c0@holdenweb.com>
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <9A5L8.232032$%u2.84870@atlpnn01.usenetserver.com> <20020604112824.B24006@unpythonic.net> <019301c20be7$933acb50$7201a8c0@holdenweb.com>
Message-ID: <20020604115652.E24006@unpythonic.net>

On Tue, Jun 04, 2002 at 12:47:54PM -0400, Steve Holden wrote:
> Interesting. But there was no need to call me a sucker.

That wasn't intended to be addressed to you in particular, just anybody
who was about to be surprised at the way += works..

Jeff




From gcordova at hebmex.com  Thu Jun 27 15:18:25 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 27 Jun 2002 14:18:25 -0500
Subject: Pythoniac: Thoughts on a hardware Python processor
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233015DE661@mail.hebmex.com>

> 
> Christopher Saunter <Christopher.Saunter at durham.ac.uk> wrote:
> > Dear All,
> > 
> > I have seen the occasional post about creating a specific 
> > hardware 'Python Processor', that could then be built and
> > used in programmable logic or the like (Valves, relays,
> > core memory - Pythoniac ;-) - in other words, create
> > a processor that executes Python bytecode directly.  There would be
> > various different reasons for doing this (the 'neat' factor etc...)
> 
> The only route that has any chance of success is 
>     Python -> Forth -> stack engine
> which leverage works that has already gone into Forth area.  
> But, my days of hand assembling 68000 or programming HP calculators
> are long gone... :-)
> 

Ahh! Thanks for posting this; I was thinking of the same thing,
about Chuck Moore's forth chips and all that, but I didn't want
to post anything because I've already posted some Forth comments
and received no replies.

:-)

-gus

> -- 
> William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
> 8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 




From ykingma at accessforall.nl  Fri Jun  7 15:25:48 2002
From: ykingma at accessforall.nl (Ype Kingma)
Date: Fri, 07 Jun 2002 20:25:48 +0100
Subject: array, list, performance...
References: <mailman.1023391748.930.python-list@python.org>
Message-ID: <3D0108B7.3C206C2E@accessforall.nl>


Tim Peters wrote:
> 
> [Michael Chermside]
> >     li = [x0, x1, x2, ...]
> > ...
> >     li.append(x)   appending is O(1)
> >                    (actually, it SOMETIMES takes O(len(li)) as it
> >                     resizes the list. But if you grow a list by
> >                     continually appending, the amortized time is
> >                     linear (ie, constant amortized time per element)
> 
> [Ype Kingma]
> > Append is often enough linear in the length of the list
> > that growing by appending is O(len(li) * len(li)).
> 
> For Python 2.2 Michael is correct.  For Python 2.1 and earlier, you're
> correct "in theory", although in practice it was often linear-time anyway,
> depending on internal details of the platform realloc.

Good news.
Multiplying the list's size by a constant factor when necessary would 
give O(log(len(li)) * len(li)) behaviour, so how was this made
O(len(li)) in 2.2?
I checked 'What's new in 2.2' but found nothing about this.

Regards,
Ype


From detlev at die-offenbachs.de  Sat Jun 15 11:55:58 2002
From: detlev at die-offenbachs.de (Detlev Offenbach)
Date: Sat, 15 Jun 2002 17:55:58 +0200
Subject: IDE recommendations??
References: <b9veea.l85.ln@majestix.gallien.de> <3d0b06bd$0$52043$e4fe514c@dreader3.news.xs4all.nl>
Message-ID: <e2ofea.f75.ln@majestix.gallien.de>

Boudewijn Rempt wrote:

> Detlev Offenbach wrote:
> 
>> Hi all,
>> 
>> I have googled this list for IDE recommendations, but have found none
>> that satisfied me. I am looking for an IDE that allows development of
>> Qt/KDE applications using Python. So far I have found BlackAdder.
>> However, it looks as if development has stalled. I could use
>> individual applications (e.g. qt-designer for gui, scite as an editor,
>> pyuic, ...), but this lacks some comfort. I would appreciate any
>> recommendation for a suitable IDE
>> 
> 
> I've tried to use most IDE's out there -- from Pythonwin to
> Pythonworks. None was really comfortable, except for Wing IDE, which is
> really powerful. It doesn't include integration of a GUI designer,
> though, which makes it little more than a glorified editor/debugger
> (but _very_ glorified). At one point there was a version of KDevelop
> that supported Python development, but that has disappeared, it seems.
> 
> I use bash & XEmacs & Designer nowadays, and that gives me a fair
> productivity.
> 
Hi,

I just received an email from theKompany saying, that development on 
BlackAdder is going on. They hope to release a new beta supporting Qt 3 
and the new PyQt next week.

Detlev
-- 
Detlev Offenbach
detlev at die-offenbachs.de


From uwe at rocksport.de  Mon Jun  3 16:20:07 2002
From: uwe at rocksport.de (Uwe Schmitt)
Date: 3 Jun 2002 20:20:07 GMT
Subject: different gui-toolkits pros/cons...
References: <adfsoo$dmik4$1@hades.rz.uni-sb.de> <_rLK8.12$fd2.979@news.ecrc.de> <adg4e0$dq6mb$3@hades.rz.uni-sb.de> <3cfbc67b$0$52024$e4fe514c@dreader3.news.xs4all.nl>
Message-ID: <adgj1n$11edg$1@hades.rz.uni-sb.de>

Boudewijn Rempt <boud at valdyas.org> wrote:
| Uwe Schmitt wrote:

|> Markus Jais <mjais at web.de> wrote:
|> | with pyQT there might be problems with the license
|> | if you want to make your programm freely available
|> | (but I am not a lawyer, so I might be wrong)
|> 
|> That's a very important point: what do I have to pay if I used pyQt
|> for development ??? I did not find any explicit prices at the
|> trolltec-page...

| Trolltech doesn't sell PyQt: you need to go to http://www.thekompany.com
| for BlackAdder. 

O.K. But what if I only want pyQt ??? I suppose I have to pay Qt...

but: how much ???

Greetings, Uwe

-- 
Dr. rer. nat. Uwe Schmitt      ICQ# 159647634   Uwe.Schmitt at num.uni-sb.de


From bobnotbob at byu.edu  Thu Jun  6 11:16:05 2002
From: bobnotbob at byu.edu (Bob Roberts)
Date: 6 Jun 2002 08:16:05 -0700
Subject: programatically changing windows desktop background
Message-ID: <c4e6b17d.0206060716.6afae229@posting.google.com>

I want to write a python program that periodically takes an image
file, writes some useful infomation on it (like the Calendar, etc.),
saves the file, and then sets it as the current windows background
image.

Most of it is pretty straightforward, but does anybody know how I can
change the windows background imagage from a python program?


From DLNXPEGFQVEB at spammotel.com  Thu Jun 27 08:16:46 2002
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Thu, 27 Jun 2002 15:16:46 +0300
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com> <afdrlv$r83$0@216.39.172.122>
Message-ID: <nd0mhu8omept0dkl1deeac7pp4u2c17ufh@4ax.com>

On 27 Jun 2002 02:01:35 GMT, rumours say that bokr at oz.net (Bengt
Richter) might have written:

>Try accumulating your chunks in a list,
>like sl=[] followed by sl.append('chunk') instead, and then do f.write(''.join(sl))
>instead of the f.write(s) you would have done.

A minor correction, if I may add:

f.writelines(sl)
might be quite more efficient than
f.write(''.join(sl))

AFAIK the output is the same, as long as one remembers to insert the
necessary '\n' for line breaks where needed :)
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')


From aahz at pythoncraft.com  Mon Jun 24 23:16:02 2002
From: aahz at pythoncraft.com (Aahz)
Date: 24 Jun 2002 23:16:02 -0400
Subject: Suggestions for good programming practices?
References: <mailman.1024951532.5616.python-list@python.org>
Message-ID: <af8n9i$3at$1@panix1.panix.com>

In article <mailman.1024951532.5616.python-list at python.org>,
David LeBlanc <whisper at oz.net> wrote:
>Mark:
>>
>> When evaluating whether a variable is None, don't compare it like this:
>>
>> 	if x == None:
>>
>> instead, use:
>>
>> 	if x:
>>
>> or:
>>
>> 	if not x:
>>
>> The encourages a polymorphic approach.  If you must compare to
>> None, use the identify operator rather than equality:
>>
>> 	if x is None:
>>
>> or:
>>
>> 	if x is not None:
>
>I'm curious to know how this encourages polymorphism, and how making a
>conditional test implicit rather then explicit is good practice?

Depends on what you're testing for.  One of the big features of Python's
object model is that any "empty" object tests as false: None, 0, "", (),
[], and {}.  Testing for a specific false value (such as 0 or None)
prevents you from easily changing the object being used.  Rely on the
object to tell you whether it's true or false, and you'll be fine.

That's polymorphic.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From akakakhel at attbi.com  Wed Jun  5 18:44:04 2002
From: akakakhel at attbi.com (Ali K)
Date: Wed, 05 Jun 2002 22:44:04 GMT
Subject: Internat
References: <69QJ8.68711$ux5.82517@rwcrnsc51.ops.asp.att.net> <ad9tkq$g2n$1@agate.berkeley.edu>
Message-ID: <UuwL8.118689$ux5.177700@rwcrnsc51.ops.asp.att.net>

Thank you!!!




From larooy at xtar.co.nz  Sat Jun  1 03:43:50 2002
From: larooy at xtar.co.nz (John La Rooy)
Date: Sat, 1 Jun 2002 19:43:50 +1200
Subject: Why no open(f, "w").write()?
References: <mailman.1022702940.7544.python-list@python.org>
	<slrnafagek.64g.huaiyu@gauss.almadan.ibm.com>
	<slrnafcsor.12v.msoulier@pmerd071.ca.nortel.com>
	<20020531225504.248c92ce.larooy@xtar.co.nz>
	<m2lm9zh39j.fsf@mycroft.actrix.gen.nz>
Message-ID: <20020601194350.6366b7be.larooy@xtar.co.nz>

On 01 Jun 2002 19:22:32 +1200
Paul Foley <see at below> wrote:

> On Fri, 31 May 2002 22:55:04 +1200, John La Rooy wrote:
> 
> >> Well, as one of the credos of Python is that explicit is better than
> >> implicit, why are we relying on the implicit nature of file objects to close
> >> themselves? It's not nearly as easy to read.
> 
> > Do you explicitly del all your objects too, or wait for them to disappear when they go out of scope?
> 
> Neither.  You can't delete objects in Python ("del" doesn't do
> anything to objects; it does something to variables), and in theory
> they persist forever; GC is just an optimization to allow reuse of
> memory used by objects that the program will never access again,
> because you don't really have infinite memory.
> 
> Correct optimizations don't change the behaviour of correct code.
> 
> [Therefore, if the behaviour of your code does change depending on
> whether or not the optimization is done, either the optimization is
> in error or your code is.  Thus, relying on GC to close files (or do
> anything else) is Wrong.  Python really ought to print a warning when
> it GCs an open file]
> 
> -- 
> You don't have to agree with me; you can be wrong if you want.
> 
> (setq reply-to
>   (concatenate 'string "Paul Foley " "<mycroft" '(#\@) "actrix.gen.nz>"))

Good point. If I had infinite memory and was allowed an infinite number of
open files, I'd never need to close or flush my file objects, because the 
optimisation between my writing to a file and the actual event taking place
will have no effect on the behaviour of my code. But it DOES have an effect 
doesn't it? Buffering is an optimisation, isn't it?

John


From jbublitzNO at SPAMnwinternet.com  Sat Jun 29 20:55:33 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Sun, 30 Jun 2002 00:55:33 GMT
Subject: Newbie that don't understand
References: <AIlT8.13914$p56.4478653@newsb.telia.net> <3d1e262f.3755054@news.clara.net> <3D1E4F80.D8EFBB3E@engcorp.com>
Message-ID: <3D1E56EC.7080203@SPAMnwinternet.com>

Peter Hansen wrote:
> Just a dude wrote:

>>My son is an electronics
>>design engineer, and I still can't get my head around negative
>>voltages, how the hell do you get a -12 V supply... to him it's basic
>>but to me..............

> With a 12V supply, the power supply is trying to push electrons
> out the negative terminal and into the positive terminal.  If you
> have a -12V supply, it tries to push electrons out the positive
> terminal and into the negative terminal.

While this is way off topic and way pedantic, electrons
always leave and enter the same terminals (think electro-
chemistry, like a battery). Electrons leave the negative
terminal and flow towards the positive terminal (blame
this on whoever decided protons were positive and electrons
negative).  The electro-motive force (emf, which is what
voltage  measures) that causes this to happen is measured
as a difference in potential. The + terminal is 12v above
the - terminal.

If you connect the + terminal of a 12v battery to ground
(which is by definition a zero potential or 0v reference
point), the - terminal will be 12v below ground (the 12v
difference and polarity of difference are unchanged). IOW
12v volts below ground (0v) will be -12 volts. Some British
sports cars from the 50's and 60's have the positive
terminal connected to the chassis (ground for a car),
so they have a -12v (-6v?) electrical system (or "positive
ground"). There is no difference, however, between the
battery in an American car and these British cars - just
a difference in reference point. As Einstein said, it's
all relative..


Jim




From gerhard at bigfoot.de  Fri Jun 14 15:17:51 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 14 Jun 2002 21:17:51 +0200
Subject: executemany & UPDATE (MySQL & Python)
In-Reply-To: <NDBBLAEMOMEHKALAMKKJKEJJKPAA.narnett@mccmedia.com>
References: <slrnagk7tu.114.gerhard@lilith.my-fqdn.de> <NDBBLAEMOMEHKALAMKKJKEJJKPAA.narnett@mccmedia.com>
Message-ID: <20020614191751.GA1520@lilith.my-fqdn.de>

(Cc-ing Andy Dustman)

* Nick Arnett <narnett at mccmedia.com> [2002-06-14 11:02 -0700]:
> > Gerhard wrote:
> > > Nick Arnett wrote:
> > > > Can anyone confirm that it is *not* possible to pass a list to
> > > > the Python MySQLdb module in UPDATE operations?

...

> > For database modules that don't support prepared statements, the db module
> > implements executemany with multiple calls to execute, anyway.
> 
> Well, since MySQLdb is much faster at inserting that way, I'm assuming that
> if it'll do it on an UPDATE, that will also be much faster.

I mistakenly assumed that MySQLdb's executemany was implemented the same
way as pyPgSQL's. That assumption was wrong. Looking into the source
reveals that MySQLdb's implementation is really clever, but
unfortunately hardcoded to only work with INSERT statements.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 24.3 ?C      Wind: 1.9 m/s




From sandskyfly at hotmail.com  Sun Jun  9 07:59:13 2002
From: sandskyfly at hotmail.com (Sandy Norton)
Date: 9 Jun 2002 04:59:13 -0700
Subject: Comparison: HTTP with Java, Perl or Python
References: <3D009911.9000904@gmx.de> <4abd9ce7.0206080941.564ef212@posting.google.com>
Message-ID: <b03e80d.0206090359.541660cc@posting.google.com>

> import urllib
> print urllib.open("http://www.uni-dortmund.de").read()

python rules!


From tim.one at comcast.net  Tue Jun 11 19:32:25 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 11 Jun 2002 19:32:25 -0400
Subject: lists and sequences
In-Reply-To: <20020611221037.36421.qmail@web13303.mail.yahoo.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCKEDOPMAA.tim.one@comcast.net>

[Collin Monahan]
> What is a python list? Is it a linked list or a random access structure
> in contiguous memory?

A contiguous vector of pointers to objects.

> What type of situations should be avoided when using it?

Depends on the app and your judgment of appropriate tradeoffs.

> E.g. when does n^2 time happen using it?

alist.insert(0, newthing) physically moves len(alist) elements.  Ditto
alist.pop(0).

> What sorting algorithm is used by its sort member function?

A hybrid of samplesort and binary insertion sort.

> Are the answers to these questions the same for Jython?

Don't know.

grateful-for-the-easy-ones<wink>-ly y'rs  - tim





From whisper at oz.net  Sat Jun 29 01:35:40 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 28 Jun 2002 22:35:40 -0700
Subject: Python needs better error reporting
In-Reply-To: <LNBBLJKPBEHFEDALKOLCGEODAAAB.tim.one@comcast.net>
Message-ID: <GCEDKONBLEFPPADDJCOEOEPADMAA.whisper@oz.net>

Cute anthropomorphism is lots of fun, but doesn't _quite_ obscure the fact
that computers don't really "tell" anybody anything, programmers do. IIRC,
Python throws this "Syntax Error: invalid syntax" in a number of places. In
the place that I used as an example, looking at the language reference, an
appropriate error would be more like:
	Syntax Error: expected "and", "or", "xor" or ":"

This would at least narrow down what should be looked at to resolve the
problem.

Vague error reporting doesn't help new user's self confidance or acceptance
of the language.

I really DO have a point!

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Tim Peters
> Sent: Friday, June 28, 2002 21:22
> To: Python-List at Python. Org
> Subject: RE: Python needs better error reporting
>
>
> >>> "Syntax Error: invalid syntax" isn't very informative, yet it's
> >>> tossed out frequently. HOW is the syntax invalid?  For example:
> >>>  if s == ' '
>
> >> That is supposed to start a new block so it should be
> >> if s == ' ':
>
> > That's not the point - I know that.
>
> Or, IOW, SyntaxError is all the clue you actually needed.  If the parser
> were to tell you what *it* thinks is wrong, it would be along the lines of
> "the only arcs leading out of this state are 23, 167, and 528".
> Then you'd
> really have a point <wink>.
>
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From max at alcyone.com  Wed Jun 26 14:13:02 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 26 Jun 2002 11:13:02 -0700
Subject: Reminder of a python Project.
References: <2218.941T781T12224051threeseas@earthlink.net> <afc2mn$2kn$1@verence.demon.co.uk>
Message-ID: <3D1A042E.F8B2FBDA@alcyone.com>

Paul Wright wrote:

>                       ______
>                      /      \
>                    .' PLEASE `.
>                    |  DO NOT  |      _____
>                    | FEED THE |    ,'.....`.
>                    `. TROLLS ,'  ,'........ )
>                      \_    _/   |........ ,'

I think technically that's a crank/kook, not a troll.  :-)

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ See the son in your bad day / Smell the flowers in the valley
\__/ Chante Moore
    Bosskey.net: Aliens vs. Predator 2 / http://www.bosskey.net/avp2/
 A personal guide to Aliens vs. Predator 2.


From ulrich.pfisterer at rmb.co.za  Fri Jun  7 05:36:22 2002
From: ulrich.pfisterer at rmb.co.za (Ulrich Pfisterer)
Date: Fri, 07 Jun 2002 11:36:22 +0200
Subject: question on pattern
Message-ID: <3D007E96.3010505@rmb.co.za>

I want to do a pattern match on a string that spans multiple lines (c++ 
style comments, re.match('(/\*.*?\*/)', filestring)
If I use the re.findall(pattern, string) method it only finds those 
matches that are on a single line. If I use re.match(pattern, string, 
DOTALL) it spans multiple lines but only finds the first match.
Any help on this one would be greatly appreciated.



From jjl at pobox.com  Tue Jun  4 17:26:10 2002
From: jjl at pobox.com (John J. Lee)
Date: Tue, 4 Jun 2002 22:26:10 +0100
Subject: CGI questions
In-Reply-To: <mailman.1023176296.4832.python-list@python.org>
References: <mailman.1023176296.4832.python-list@python.org>
Message-ID: <Pine.SOL.4.30.0206042221180.10073-100000@mimosa.csv.warwick.ac.uk>

On Wed, 5 Jun 2002, Michael Hall wrote:

> 1.
> Does Python have the same or a similar problem that PHP had until recently
> with regard to automatic global registration of GET, POST and COOKIE
> variables within a script's namespace? My guess is that it doesn't, but
[...]

Of course not -- Python itself knows nothing about CGI.  I don't think the
cgi module does any tricky implicit stuff along those lines, either.

[...]
> alphanumeric string I'd expect from PHP. Most of it is definitely not
> [a-zA-Z0-9], and it is around 10 characters long. How can I get a PHP
> style md5 result?

Use the hexdigest method.

> 3.
> What is the decimal bit on the end of time.time()'s output? Hundredths of
> a second?

The seconds element of the time tuples that the standard Python library
function uses is a decimal number, in seconds.

> 4.
> Is the following Python the (best) equivalent to the PHP?
>
> PHP				Python
> ereg("bit","bigbits")		re.compile("bit").search("bigbits",1)

No idea, but Python uses Perl-style regular expressions, which aren't
exactly the same as 'extended' regexps.

All this stuff is in the std. library docs.


John



From leonw at gmx.net  Mon Jun 17 13:19:57 2002
From: leonw at gmx.net (Leon Widdershoven)
Date: Mon, 17 Jun 2002 19:19:57 +0200
Subject: Python GUI
References: <9ddd154e.0206122021.1b06f89a@posting.google.com>
Message-ID: <3D0E1A3D.ED47A96A@gmx.net>

Hi,

I'm quit satisfied with PyQt. It has loads of widgets, good docs,
a graphical interface designer (Qt's), and there is a python
interface compiler to create Python code.

It has at least 2 major drawbacks:
- The free version for windows does not read JPEG's
- It's not small at all

Greetz,
Leon



"H. Safarzadeh" wrote:
> 
> Hi all,
> I know that this subject is discussed in many other places, but I don't
> think it is harmful to talk about it more!
> 
>  I want to what library I should use to make GUI for my Python apps?
>  I have examined many of them, but I couldn't find what I need. Here is
> the result of some of my examinations:
> 
> -Tkinter: A nice simple lib, but really simple! It dosn't have enough
> widgets, and I do not like its look in addition!
> -wxPython: Really perfect. It has many widgets, and do many works(not
> only in GUI). But it is toooo big!(both on disk and in memory). I need a
> smaller one.
> -pyGTK and pyQt: I do not know enough about them. But I know that I do
> not like GTK look on Windows and Qt for Windows has some problems with
> its lisence(and it is tooo big, too!)
> -Jython/Swing: ....it's Jython, after all!
> -Win32 extensions: It is not portable!!!
> 
> Any comments? Other possibilities?
> 
> Thanks.


From imbosol at vt.edu  Mon Jun 24 16:11:52 2002
From: imbosol at vt.edu (Carl Banks)
Date: Mon, 24 Jun 2002 16:11:52 -0400
Subject: Suggestions for good programming practices?
References: <WEuR8.2074$5l4.75657@ozemail.com.au> <bko7fa.9t.ln@grey.aerojockey.localdomain> <af7sch$dcp$1@panix1.panix.com>
Message-ID: <8eu7fa.401.ln@grey.aerojockey.localdomain>

Aahz wrote:
> In article <bko7fa.9t.ln at grey.aerojockey.localdomain>,
> Carl Banks  <imbosol at vt.edu> wrote:
>>
>>* Learn how references and the object system work as soon as possible.
> 
> ObBrokenRecord: "bindings", not "references".

Of course, you are free to call it whatever you want.


-- 
CARL BANKS                                http://www.aerojockey.com
"Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."


From steph_lar at videotron.ca  Tue Jun 11 10:18:42 2002
From: steph_lar at videotron.ca (Stephane Larouche)
Date: 11 Jun 2002 07:18:42 -0700
Subject: Problem with wxPython
References: <mailman.1023629412.29376.python-list@python.org>
Message-ID: <b7226ba.0206110618.272c3a98@posting.google.com>

"Roman Yakovenko" <romany at actimize.com> wrote in message news:<mailman.1023629412.29376.python-list at python.org>...
> Could you post small piece of code. I have some computer configuration
> and it works fine to me. Also try Boa IDE. It is very helpful.
> 
> Roman

As I said in my previous message, I do run wxPython on the same
configuration on 3 other computers. But, on one of my computers, the
fastest one, wxPython crashes on wxFrame.__init__ if the style
includes wxMAXIMIZE_BOX, wxRESIZE_BOX, wxSYSTEM_MENU or
wxDEFAULT_FRAME_STYLE. For exemple, the wxPython demo crashes on

wxFrame.__init__(self, parent, -1, title, size = (800, 600),
style=wxDEFAULT_FRAME_STYLE|wxNO_FULL_REPAINT_ON_RESIZE)

in the file Main.py. This demo works perfectly well on the other
computers. Compiled software (with py2exe) also crashes.

Maybe i'll try Boa, but I don't think it will solve the problem. The
problem isn't the code. The code runs on every other computer I tried.
The problem is probably coming from an incompatibility of wxPython
with something else I have on my computer, I just can't figure out
what! Anybody's help is welcome, I'm close to reformat my hard-drive
and verify between every software installation if wxPython still works
in order to identify the conflict.

Thank you for your help!

Stephane Larouche
steph_lar at videotron.ca


From mutsuura at mutsuura.com  Sun Jun  9 12:30:19 2002
From: mutsuura at mutsuura.com (Attila Horvath)
Date: Sun, 9 Jun 2002 12:30:19 -0400 (EDT)
Subject: (no subject)
Message-ID: <Pine.LNX.4.44.0206091224010.1320-100000@adams.patriot.net>

All,

Am new to PYTHON and am looking for simple/sample PY-code to:

	- provide [generic] CGI functionality
	and if possible...
	- generate dynamic (DHTML) responses to CGI input parameters

Any suggestions? 

Thx
Attila

PS: Have already looked at 'Dive Into Python'.






From candiazoo at attbi.com  Thu Jun 27 14:14:55 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Thu, 27 Jun 2002 18:14:55 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com>
Message-ID: <3d1b55af.308978578@netnews.attbi.com>

trs_file.write("T%32d%-251s%-29s%-13s%-9s%6.2f%6.2f%-16s\n" % \
	(wrk_inst, title[0:251], pub_name[0:29], \
         std_num[0:13], date_range, base_fee, \
         page_fee, source_id))

This is how I am writing the record[s] now.  Not much gain but the other thing
is I need to handle nulls returned from the database (none values in pub_name)
and I need to limit the size (as you can see) of some of the fields because the
database field is much larger than the text field allows...

Mike


From emile at fenx.com  Sun Jun 23 11:54:59 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 23 Jun 2002 15:54:59 GMT
Subject: (Newbie)  Counting Instances ("Hits") with Regular Expressions
References: <3d15e215.43448155@news.texas.net>
Message-ID: <nbmR8.300036$cQ3.15150@sccrnsc01>

"Ben Fairbank" <baf at texas.net> wrote in message
news:3d15e215.43448155 at news.texas.net...
> I am new both to Python and to regular expressions, which may account
> for my difficulty.  I must count the frequenies of certain words in
> files of moderate length (about150 k bytes).  I have been reading
> files and then using count(s,sub), which is fast and easy.  I now have
> to allow for punctuation and eliminate words within words, etc, and so
> am trying to use regular expressions instead of simple words as
> targets.  I do not, however, find a similarly easy to use count
> function in the re module.  Yet this is such common operation it must
> be there, or easy to implement.  What is the usual way of simply
> counting "hits" in the re module?  (And what have I missed in the
> documentation; where is this to be found?  I have looked through Lutz
> and Ascher)
>

>>> s = 'spam and eggs and ham and eggs and spam and eggs'
>>> s.count('am')
3
>>>

--

Emile van Sebille
emile at fenx.com

---------



From m.faassen at vet.uu.nl  Thu Jun 20 19:34:43 2002
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 20 Jun 2002 23:34:43 GMT
Subject: GOTO w/ Python?
References: <3d10ed95@xpl-sdf1-sec1.> <uh1s1qdcl5b1b0@news.supernews.com> <aetbqv$kgq$1@newshost.accu.uu.nl> <uh4km1ffrb400c@news.supernews.com>
Message-ID: <aetoqj$i0g$1@newshost.accu.uu.nl>

John Roth <johnroth at ameritech.net> wrote:
[snip Djikstra being an evil clone of Dijkstra]
> Given the Professor's reputation as somewhat of a
> curmudgeon, wouldn't that be an identity function?

It's a possibility. Did you know Dijkstra originated in the
Netherlands but now lives in the US, just like Guido? Not implying
in any way that Guido is somewhat of a curmudgeon. :)

> Sorry about the misspelling.

I was just amused, not upset.

I can see where a non-Dutch speaker's brain would say 'Dijkstra' has *five*
consonants in a row, that can't be real, I'll make it 'Djikstra' instead.
But, in Dutch 'ij' is actually a vowel sound. My name for instance is
pronounced approximately like 'Mar-tine', where 'tine' rhymes with 'line'
and such. But only approximately.

Of course Guido has a hard Dutch 'g' in his name, the most horrible
'throat disease' sound in the Dutch language according to many foreigners
('ch' is another variety of our hard 'g' which may even be worse, though
'sch' also probably ranks highly). However, I grew up in
the south of the Netherlands and the accents there employ a much softer
'g', so I don't even pronounce Guido's name in the way he does.

Anyway, did you know that Dutch is the official natural language for
Python programmers? It's true! Here is some of the evidence:

http://groups.google.com/groups?hl=en&lr=&th=8fc168631bc6097a
(follow the thread)

Regards,

Martijn
-- 
History of the 20th Century: WW1, WW2, WW3?
No, WWW -- Could we be going in the right direction?


From padhia at yahoo.com  Sat Jun 29 10:52:35 2002
From: padhia at yahoo.com (P Adhia)
Date: 29 Jun 2002 07:52:35 -0700
Subject: Python COM installation and NT Administrative rights.
Message-ID: <3281a460.0206290652.51b800c9@posting.google.com>

Hi,

I'd like to install Python COM on my work PC - on which I don't have
administrative rights. Is there a Python COM installation available
that does not require administrative rights on NT ? (I don't know if
administrative rights is must for Python COM to install/work; but I
have installed some other softwares that detect current privilege
levels and install regitry entries reserved for per user instead of
system. I am hoping that Python COM can be made to do something
similar.)

TIA

P. Adhia


From tatebll at aol.com  Sun Jun  9 10:01:12 2002
From: tatebll at aol.com (Bill Tate)
Date: 9 Jun 2002 07:01:12 -0700
Subject: Creating Dynamic Web Pages (ZOPE)
References: <mailman.1023528013.13067.python-list@python.org>
Message-ID: <cb4ba455.0206090601.73b91f88@posting.google.com>

Michael Hall <olc at ninti.com> wrote in message news:<mailman.1023528013.13067.python-list at python.org>...
> I notice no-one has suggested good ol' CGI. For a relatively simple job
> such as this, a few basic Python CGI scripts, some templates and MySQLdb
> could also be used ... oh, just noticed the MS-SQL database. There is
> probably a module around to cope with MS-SQL, no?
> 
> Mick
> 
> On 7 Jun 2002, Bill Tate wrote ....
[snip]

It isn't that cgi wouldn't have been fine here.  From my perspective,
its a matter of what architecture is likely to work for me over the
long haul? While cgi is perfectly fine in many cases, I don't think I
would want to implement all the "housekeeping" infrastructure that
would already have been implemented in tools like zope, webware,
skunkweb, twisted, etc...


From rjones at ekit-inc.com  Wed Jun 12 19:24:20 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Thu, 13 Jun 2002 09:24:20 +1000
Subject: Wanted:  Potential Python Authors
In-Reply-To: <3d07d4a8@news.eol.ca>
References: <ae7gke$mhb$1@ccis12.baylor.edu> <db3fgu40sbnt3pno5v1i2h9lnmpcoti2bk@4ax.com> <3d07d4a8@news.eol.ca>
Message-ID: <200206130924.20528.rjones@ekit-inc.com>

On Thu, 13 Jun 2002 09:09, William Park wrote:
> Lothar Scholz <llothar at web.de> wrote:
> >>- Author of articles in the various trade magazines
> >
> > You really mean "trade" magazines ?
> > Normally they are written by marketing peoples who don't
> > have the competence to write more then a 4 page overview.
> > Hands off !!!!
>
> Bull.  I agree about trade magazines above, though there isn't any magazine
> covering Python.

Ehem.   http://www.pyzine.com/


    Richard





From janto_d at hotmail.com  Fri Jun 28 19:50:23 2002
From: janto_d at hotmail.com (Janto Dreijer)
Date: Sat, 29 Jun 2002 01:50:23 +0200
Subject: doctest for java
References: <afhv70$15mk$1@news.adamastor.ac.za>
Message-ID: <afisev$1f6e$1@news.adamastor.ac.za>

Tim Peters <tim.one at comcast.net> wrote in message
news:<mailman.1025281419.16432.python-list at python.org>...
> [Janto Dreijer]
> > Are there any attempts to port the "doctest" module to Java? (All
> > I can find using google is some testing class for the JavaDoc program
> > calling itself "DocTest".)
> >
> > At a glance it looks like a worthwhile effort...extracting Jython
> > "examples"
> > from /**...*/ comments also seems simple to implement.
>
> Last I heard, Python's std doctest.py worked fine under Jython.  If
anything
> else simple is needed, you're elected <wink>.

I in no way meant your code is trivial. Although I'm starting to think you
did that "import __future__" trick with the sole reason to confuse :-)

I'm actually thinking of running doctests on *.java files and not on *.py
files, so the current doctest.py wouldn't be 100% usable.

class Obj {
    /**
        "Three" is the number it should print. And the number of the
printing should be "three":
        This is a usage example written in Jython...maybe.

        >>> Obj.f()
        3

    */
    public static void f() {
        System.out.println(3);
    }
}

That's it. It can't be helped. I'm spending my weekend on this :-)




From paul at boddie.net  Fri Jun  7 07:41:35 2002
From: paul at boddie.net (Paul Boddie)
Date: 7 Jun 2002 04:41:35 -0700
Subject: Extracting data from HTML
References: <82096df4.0205311152.5891a17b@posting.google.com> <1022883953.6150.5802.camel@localhost> <mailman.1022908262.30070.python-list@python.org> <82096df4.0206010201.736c4b2c@posting.google.com> <23891c90.0206050321.d6a64c3@posting.google.com> <4slvfucpesg4ikfs3dfrjo8deio5de24ph@4ax.com>
Message-ID: <23891c90.0206070341.3b2a30e8@posting.google.com>

Giulio Cespuglio <giulio.agostini.remove.this at libero.it> wrote in message news:<4slvfucpesg4ikfs3dfrjo8deio5de24ph at 4ax.com>...
> >  http://www.boddie.org.uk/python/HTML.html
> 
> Thanks a lot for this resource. To be honest, I wonder how you could
> work out how to use htmllib, IMHO the documentation is very poor.
> Actually, I am happily using the flexibility of regular expressions to
> do these things ATM, but I'm willing to give this library a try.

I think that each technique has its advantages and disadvantages:

  Regular expressions:         good for mining for data without caring
                               about document structure;
                               bad for detecting and reasoning about
                               the document structure
  sgmllib
  (and SAX-like technologies): good for mining for data whilst
                               "binding" that data to specific
                               elements;
                               bad for dealing with complicated
                               document structures

  DOM-like technologies:       good for insisting on particular
                               document structures and for keeping
                               these structures intact;
                               bad for casual mining of data - effort
                               is required to find data before it can
                               be extracted

So, if regular expressions aren't giving you the control you require,
you may want to consider one of the other technologies.

Paul


From jjb5 at cornell.edu  Wed Jun 12 14:23:53 2002
From: jjb5 at cornell.edu (Joel Bender)
Date: Wed, 12 Jun 2002 14:23:53 -0400
Subject: dict = urllib.urldecode(string) ?
Message-ID: <jjb5-743571.14235312062002@newsstand.cit.cornell.edu>

I'm processing GET and POST data and I would like the functional 
inverse of urllib.urlencode().  That is, split the string apart by
the '&' chars, split each of those by '=' and build a dict of the 
results.  The key and the value should be passed to unquote_plus().

Here is my solution, comments?

>>> def urldecode(s):
...     rslt = {}
...     for item in s.split('&'):
...         keyValue = item.split('=')
...         rslt[ urllib.unquote_plus(keyValue[0]) ] = 
urllib.unquote_plus(keyValue[1])
...     return rslt
... 

>>> urldecode('a=b+c&d%26e=%20f')
{'a': 'b c', 'd&e': ' f'}
>>>


From cliechti at gmx.net  Sun Jun 23 08:50:09 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 23 Jun 2002 14:50:09 +0200
Subject: ConfigParser & converting strings to lists
References: <3D152A28.2177B054@tds.net> <3D159F06.2154D3F3@tds.net>
Message-ID: <Xns9236975A47F67cliechtigmxnet@62.2.16.82>

"Edward K. Ream" <edream at tds.net> wrote in news:3D159F06.2154D3F3 at tds.net:

> To complete this topic, let me restate the problem and JohnJacob's
> elegant solution.
> 
> Suppose files is a list of strings, and we have written a configuration
> section like this:
> 
>   config = ConfigParser.ConfigParser()
>   config.set("recent files", "recentFiles, files)
> 
> Then the elegant way of retrieving files _as a list_ is:
> 
>   files = eval(config.get("recent files", "recentFiles"))

it isn't that elegant i tend to call that sloppy ;-) the problem is that 
you execute arbitrary python expressions and you hope to get a list. but 
that is not sure, the user could have edited the config file.
exec and eval are always a risk in your code. e.g they could rise any 
exception because you don't know what they realy execute.

the split version is much more secure. you could also try a pickle. witch 
pickle.dumps you can get a string and reterive the old object using loads() 
but that isn't as secure as split.

if thats all too complicated, use your version with individual names 
"file1", "file2", ... it's much saver for a progrom that you release in the 
wild.

chris

PS you mentioned you see double backslashes. maybe you looked at the repr 
of a string:
>>> s = "he\tllo"
>>> repr(s)
"'he\\tllo'"
>>> str(s)
'he\tllo'
>>> print s
he	llo
>>> 

as you can see there is not realy a double backslash...

-- 
Chris <cliechti at gmx.net>



From Uwe.Mayer at ifib.uni-karlsruhe.de  Mon Jun 17 06:39:29 2002
From: Uwe.Mayer at ifib.uni-karlsruhe.de (Uwe Mayer)
Date: Mon, 17 Jun 2002 12:39:29 +0200
Subject: what *is* a class?
References: <MPG.17774343cc5a589098968a@news.rz.uni-karlsruhe.de> <Gxtt3F.upA@beaver.cs.washington.edu>
Message-ID: <3D0DBC61.9DDE4520@ifib.uni-karlsruhe.de>

Douglas Zongker wrote:
> "Group()" creates a new object which is an instance of the class
> Group.  It calls the object's initializer ("__init__" method), then
> returns the newly created object.
> 
> : i want to write a method which returns the instance, so that f.e. the
> : following is possible:
> :   g = Group()
> :   g2 = g.foobar()
> :   g2 is g
> : so that g2 is identical to g.
> : any idea?
> 
>     class Group:
>         def __init__( self ):
>             print 'creating instance'
> 
>         def foo( self ):
>             print 'in the foo method'
>             return self
> 
>     >>> g = Group()
>     creating instance
>     >>> g2 = g.foo()
>     in the foo method
>     >>> g2 is g
>     1
>     >>>
> 
> Note that there's usually no reason for a method (such as "foo") to
> return the self object, because the caller should already have the
> object to call the method in the first place.  I could leave "return
> self" out of the definition of "foo", and just write:
>
>     g = Group()
>     g.foo()
>     g2 = g
> 
> The result would be the same.

I am very sorry. Of course I tried that, returning self, but perhaps I
called: g2 = g.foo   instead of g2 = g.foo()
:-(((

The reason why I wanted that was for lazyness. I am writing classes
which should provide random access to arbitrary binary file formats with
a special focus on RIFF.
I wanted to create an object model of the file (without actually reading
all of it - especially skipping the large data chunks).
I had to distinguish between creating the object model and actually
reading the file in. Thus I decided to use the constructor for creating
the object model and provide a parse() method. I didn't like writing:

f = file('dummy.avi','rb')
avi = RIFF(<specify format>)
avi.parse(f)

instead I wanted to write:

avi = RIFF(<specify format>).parse(f)

but then of course parse() must return an instance object.
On the one side all parse() methods must explicitly "return self" (which
is not nice). On the other side from the semantics I thought it was ok
for parse to return an instance object of RIFF.

What do you think?
Thanks for your help!
Ciao
Uwe


From DLNXPEGFQVEB at spammotel.com  Tue Jun 11 09:41:37 2002
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Tue, 11 Jun 2002 16:41:37 +0300
Subject: Getting Tkinter to accept cyrillic text
References: <83034362.0206110052.351cb526@posting.google.com>
Message-ID: <mfvbguo95g1k5ot1dvm6467118hq8gge13@4ax.com>

On 11 Jun 2002 01:52:56 -0700, rumours say that ypoi at spray.se (DR) might
have written:

>I need Tkinter entry and label widgets to support cyrillic letters
>(koi8-r). I read in some post that you should change
>sys.setdefaultencoding(), but that method doesn't exist in my Python
>(I use version 2.2.1)

sys.setdefaultencoding gets deleted after the site.py module runs.

>I really would appreciate some kind of checklist, like: Do this, then
>change that, then evereything works! Or a patch or something (Win95)


- open http://groups.google.com
- copy the next line
group:comp.lang.python percolator.py fixtk.py recipe
- and paste it into the search box
Follow the instructions there.  Don't worry, it's not 2.1 only.  Also
there can be some minor improvements to the recipe (eg ability to print
*any* unicode character even if you don't have the means to enter them
through the keyboard apart from u"\uxxxx" sequences), but I need to
clean things a little in order to submit a patch (I think the info from
the post I directed you to already exists as a pending patch).

Good luck!
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')


From martin at strakt.com  Thu Jun 13 03:35:48 2002
From: martin at strakt.com (Martin =?iso-8859-1?Q?Sj=F6gren?=)
Date: Thu, 13 Jun 2002 09:35:48 +0200
Subject: Extension objects in Python 1.5 and 2.2
In-Reply-To: <15624.19045.700125.215173@12-248-41-177.client.attbi.com>
References: <20020613071813.GA16764@strakt.com> <15624.19045.700125.215173@12-248-41-177.client.attbi.com>
Message-ID: <20020613073548.GA16909@strakt.com>

On Thu, Jun 13, 2002 at 02:31:49AM -0500, Skip Montanaro wrote:
> 
>     Martin> How do I write extension objects so that they work in both
>     Martin> Python 1.5 and the 2.x series? 
> 
> Michael Hudson has been working on a new include file named pymemcompat.h.
> Since it's only used for backward compatibility it's in the Misc directory.
> You can grab it from CVS:
> 
>     http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/dist/src/Misc/pymemcompat.h
> 
> The idea is that you should be able to include it and program to the 2.3
> memory api in code that runs under 1.5.2 or newer versions of the
> interpreter.

Cool. Thanks a lot!


Martin

-- 
Martin Sj?gren
  martin at strakt.com              ICQ : 41245059
  Phone: +46 (0)31 7710870       Cell: +46 (0)739 169191
  GPG key: http://www.strakt.com/~martin/gpg.html




From peter at engcorp.com  Mon Jun 10 15:26:40 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 10 Jun 2002 15:26:40 -0400
Subject: activating a webpage's click command from python
References: <mailman.1023727993.12908.python-list@python.org> <3D04EA96.B3518A37@engcorp.com> <mailman.1023733034.32199.python-list@python.org>
Message-ID: <3D04FD70.B1AF7BD0@engcorp.com>

Andy McKay wrote:
> 
> > does the job for Nortel on the TSE (disclaimer: anyone even
> > thinking of buying Nortel should probably just keep thinking :-).
> 
> Oh come on its a bargain, a fraction of what it was, good buying
> opportunity ;)

And the same theory held true when it had dropped from C$140 to
$73, then to $22.15, then $9.85, then $5.65, then $4.40, then
$2.85, and ... where is it now?

Sure looks like the rate of dropping is slowing down, too, until
you start to look at it as a logarithmic curve!  :-)

-Peter

Disclaimer of total stupidity: I bought at only several of those
prices. :-)


From jon+usenet at unequivocal.co.uk  Sat Jun 22 16:13:23 2002
From: jon+usenet at unequivocal.co.uk (Jon Ribbens)
Date: Sat, 22 Jun 2002 20:13:23 -0000
Subject: Web templating/db tool with best designer/coder separation?
References: <af2gh6$2c2$0@216.39.172.122>
Message-ID: <slrnah9mj3.rdk.jon+usenet@snowy.squish.net>

In article <af2gh6$2c2$0 at 216.39.172.122>, Bengt Richter wrote:
> E.g., if you were hoping for an improvement over current practice,
> which do you consider to be the tool(s) to beat, and what would
> the killer improvement(s) be? TIA.

Take a look at http://jonpy.sourceforge.net/ . It almost completely
separates code and HTML so that the designer and the coder do not
interfere with each other.


From kosh at aesaeion.com  Fri Jun 28 21:48:48 2002
From: kosh at aesaeion.com (kosh at aesaeion.com)
Date: Fri, 28 Jun 2002 19:48:48 -0600 (MDT)
Subject: Python needs better error reporting
In-Reply-To: <GCEDKONBLEFPPADDJCOEEEOLDMAA.whisper@oz.net>
Message-ID: <Pine.LNX.4.44.0206281948130.17292-100000@aesaeion.com>

On Fri, 28 Jun 2002, David LeBlanc wrote:

> "Syntax Error: invalid syntax" isn't very informative, yet it's tossed out
> frequently. HOW is the syntax invalid? For example:
> if s == ' '
>

That is supposed to start a new block so it should be
if s == ' ':






From dgallion1 at yahoo.com  Tue Jun  4 12:33:08 2002
From: dgallion1 at yahoo.com (Darrell)
Date: 4 Jun 2002 09:33:08 -0700
Subject: Looking for a list subclassing example...
References: <adi4dv$118tko$1@ID-146704.news.dfncis.de>
Message-ID: <b0c5da3a.0206040833.456d184e@posting.google.com>

class Seq(list):
    """ list with key words """
    def __init__(self, *v, **kw):
        list.extend(self, v)
        for k,v in kw.items():
            setattr(self, k, v)

class LimitList(list):
    """ list limited in size """
    def __init__(self, limit):
        self.limit=limit
    def append(self, item):
        list.append(self, item)
        diff =  len(self) - self.limit
        if diff == 1:
            del self[0]
        else:
            self = self[-self.limit:]

--Darrell
"Shagshag13" <shagshag13 at yahoo.fr> wrote in message news:<adi4dv$118tko$1 at ID-146704.news.dfncis.de>...
> Hello,
> 
> I'm looking for a list subclassing example, and i can't find it...
> 
> Do you have one to show ?
> 
> Thanks,
> 
> s13.


From not.this at seebelow.org  Thu Jun 13 16:28:28 2002
From: not.this at seebelow.org (Grant Griffin)
Date: 13 Jun 2002 13:28:28 -0700
Subject: why not "'in' in 'in'"?
References: <mailman.1023977529.27368.python-list@python.org> <aeac0l013k@drn.newsguy.com> <aeakd1$2lg$1@bob.news.rcn.net>
Message-ID: <aeav9c01igi@drn.newsguy.com>

In article <aeakd1$2lg$1 at bob.news.rcn.net>, "Raymond says...
>
>"Grant Griffin" <not.this at seebelow.org> wrote in message
>news:aeac0l013k at drn.newsguy.com...
>> Yes, guilty as charged!  Here's an example of my poor taste.  Since "find"
>is
>> sort of a boolean concept ("I found/didn't find my tennis shoes"), and
>since the
>> "find" function really returns the index (position) of a substring, I
>probably
>> would have called it "position" or maybe just "pos" (in admiration for
>BASIC).
>
>Where did you find your tennis shoes?  When you set out to find your tennis
>shoes did you want to know where they were or just whether they were
>findable?  IMHO, it would be strange to find your shoes and then not know
>where they are.

Clearly you've never met my shoes <wink>.  (BTW, are you under age 30?)

Of course, it makes perfectly good sense to report the position along with the
find, if you wish, e.g., "I found my tennis shoes under the driveway."  But as
my sentence at top exemplifies, reporting a find brings with it no obligation to
report the location where found.  Unless you're thinking of the "location"
function.

i-shouldda-been-a-lawyer-ly y'rs,

=g2

_________________________________________________________________________

Grant R. Griffin                                           g2 at dspguru.com
Publisher of dspGuru                               http://www.dspguru.com
Iowegian International Corporation                http://www.iowegian.com



From gerhard at bigfoot.de  Sun Jun  9 07:28:51 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sun, 9 Jun 2002 13:28:51 +0200
Subject: MySQLdb warnings problem
In-Reply-To: <20020609112209.GA438@lilith.my-fqdn.de>
References: <3D022ACE.A545BF98@anansispaceworks.com> <20020609112209.GA438@lilith.my-fqdn.de>
Message-ID: <20020609112851.GB483@lilith.my-fqdn.de>

* Gerhard H?ring <gh_pythonlist at gmx.de> [2002-06-09 13:22 +0200]:
> Ok, so let's rewrite this to the appropriate form:
> 
>     sqlcmd= """UPDATE user SET nposts=%d, nrecent=%d, keytopic='%s'
>                WHERE username='%s'"""

     sqlcmd= """UPDATE user SET nposts=%d, nrecent=%d, keytopic=%s
                WHERE username=%s"""

_Without_ the single quotes, of course :)

>     dbcursor.execute(sqlcmd, (n_posts, n_recent, keytopic, user[0]))

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From fgeiger at datec.at  Fri Jun 21 16:02:25 2002
From: fgeiger at datec.at (F. GEIGER)
Date: Fri, 21 Jun 2002 22:02:25 +0200
Subject: Newbie-Question: Oberserving an directory
References: <3d1352a7@news.piro.net>
Message-ID: <af00q1$n8e$1@newsreaderm1.core.theplanet.net>

GIven that you are on NT/2k, a Google group search showed up a thread, where
Darrell postet this link:

http://www.dorb.com/darrell/win32WorkSvr/makeThumbsDG.py

in response to Jeff's question on how to use
win32file.FindFirstChangeNotification().

Maybe, this is what you are looking for.

Cheers
Franz

P.S.: The thread's contents are:

Von:Jeff Kunce (kuncej at mail.conservation.state.mo.us)
Betrifft:Monitoring file changes in NT
Newsgroups:comp.lang.python
      View this article only
Datum:2000/01/18

I'd like my python program to know when
directories and files have changed on
an NT system.

Could anyone share an example of how to use
win32file.FindFirstChangeNotification()
and its relatives?

Thanks.

  --Jeff
            Message 2 in thread
      Von:Darrell (news at dorb.com)
      Betrifft:Re: Monitoring file changes in NT
      Newsgroups:comp.lang.python
            View this article only
      Datum:2000/01/18


http://www.dorb.com/darrell/win32WorkSvr/makeThumbsDG.py

--
--Darrell
"Jeff Kunce" <kuncej at mail.conservation.state.mo.us> wrote in message
news:3h2h4.362$B5.8315 at news.more.net...
> I'd like my python program to know when
> directories and files have changed on
> an NT system.
>
> Could anyone share an example of how to use
> win32file.FindFirstChangeNotification()
> and its relatives?
>
> Thanks.
>
>   --Jeff
>




"Marcus Klein" <marcus1 at marcus1.de> schrieb im Newsbeitrag
news:3d1352a7 at news.piro.net...
> Hi there,
>
> I want to observe an directory to check if a new file has been put in, if
> that happens I want to start an event.
>
> Does anyone have a good idea how to start, I've tried it with signal, but
> either I am too stupid or it is the complete wrong way :-/
>
> Thx in advance, marcus
>
> --
> sick nature ?




From sjmachin at lexicon.net  Tue Jun 11 10:01:10 2002
From: sjmachin at lexicon.net (John Machin)
Date: 11 Jun 2002 07:01:10 -0700
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
References: <Xns9229EE722467Fgustafl@195.100.94.182>
Message-ID: <c76ff6fc.0206110601.6680e899@posting.google.com>

Gustaf Liljegren <gustafl at algonet.se> wrote in message news:<Xns9229EE722467Fgustafl at 195.100.94.182>...
> I'm making a function that takes one integer and returns a boolean that 
> says if the integer matches a number in a certain group of numbers. This 
> group consists of single numbers aswell as ranges spanning thousands of 
> numbers. I'm looking for an elegant way to write this function.
> 

Here's another approach:

=== code =============================

# assume non-negative integers ...

ranges = ((1,2), (4,4), (6,8), (12, 28), (30,30), (32,132))

def load_map(r):
   """Load the bitmap"""
   bigi = 0L
   for lo, hi in r:
      nbits = hi + 1 - lo
      mask = (1L << nbits) - 1
      bigi |= mask << lo
   return bigi

def gustaf(x, bitmap):
   """Return 1 if x is in the bitmap, otherwise 0"""
   return (bitmap & (1L << x)) != 0L
   
def try_it():
   bmap = load_map(ranges)
   for lo, hi in ranges:
      print "testing interval", lo, hi
      previ = -1
      for i in (lo-1, lo, lo+1, (lo+hi)//2, hi-1, hi, hi+1):
         if i > previ:
            previ = i
            print gustaf(i, bmap), i
         
if __name__ == "__main__":
   try_it()

=== output ================================
testing interval 1 2
0 0
1 1
1 2
0 3
testing interval 4 4
0 3
1 4
0 5
testing interval 6 8
0 5
1 6
1 7
1 8
0 9
testing interval 12 28
0 11
1 12
1 13
1 20
1 27
1 28
0 29
testing interval 30 30
0 29
1 30
0 31
testing interval 32 132
0 31
1 32
1 33
1 82
1 131
1 132
0 133
=== bot bait ================================
bignum, long, fast, speed, optimize snot out of

Is the function elegant enough?

Cheers,
John


From rganesan at myrealbox.com  Thu Jun 20 00:34:52 2002
From: rganesan at myrealbox.com (Ganesan R)
Date: 20 Jun 2002 10:04:52 +0530
Subject: Variable Interpolation - status of PEP 215
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <3D11558A.14451441@engcorp.com>
Message-ID: <ouhejypnyr.fsf@andlx-anamika.cisco.com>

>>>>> "Peter" == Peter Hansen <peter at engcorp.com> writes:

> Evading the question: are you aware of the ability of the Python 
> formatting operator to retrieve items from a dictionary rather 
> than just by order in a sequence?

I remember reading about this ..

>>>> foo = "trick"
>>>> bar = 95.5
>>>> print "This %(foo)s is loved by %(bar)s%% of programmers." % locals()
> This trick is loved by 95.5% of programmers.

> You can use any dictionary, but locals() or globals() can be handy
> in many cases.

but didn't know this particular trick. Pretty cool! Thanks. At least I am
glad that Python has the % formatting operator. I have to program in Java at
work and don't even ask me about the pain of printing messages in that
language :-(. And yes, I do know about Jython and love it :-).

Ganesan

-- 
Ganesan R


From shagshag13 at yahoo.fr  Mon Jun  3 06:12:20 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Mon, 3 Jun 2002 12:12:20 +0200
Subject: Efficient posting-list
Message-ID: <adffdk$10mep9$1@ID-146704.news.dfncis.de>

A posting list is something like that (hope i'm not too wrong) :

[key_0] -> ...
...
[key_i] -> [id_j, informations] -> [id_k, informations] -> ...
...
[key_n] -> ...

where in information retrieval :
- key are often words,
- id are document id (so we have key_i in document id_j and in document id_k and ...)
- informations are often in document frequency, but might be more...

I'm looking for an efficient way of implementing this in full python, by now i use a dict
for keys and a python list containing node objects for [id_j, informations].
But i have two troubles this is very slow to populate, and too memory consuming do you
have an idea for optimizing this ?

(keep in mind that : there are more than 500,000 text keys ; id are numbered from 1 to
150,000 ; length of a posting list might be from 1 to 500,000...)

s13.

ps : i could send my code to anyone asking it by mail...
ps2 : sorry again for my bad english...
psx : and thanks again to the guys which helps me before...




From uwe.schmitt at procoders.net  Fri Jun 28 09:43:21 2002
From: uwe.schmitt at procoders.net (Uwe Schmitt)
Date: 28 Jun 2002 13:43:21 GMT
Subject: distutils/boost question
Message-ID: <afhp5p$fqf0d$1@hades.rz.uni-sb.de>

Hi,

I used the Boost.Python library to write some python-extension.
How can I distribute the generated (binary) .so-file with distutils ?
I don't want to distribute the  sources, only the binaries.
In detail: I have to distribute the libbpl.so from boost and
my own .so-file. One problem is, that I have to add the
path of libbpl.so to $LD_LIBRARY_PATH. When I only copy it
to the extension-folder of python, it does not work...

I use Activestate Python2.2 and RedHat 7.2

Greetings, Uwe.

-- 
Dr. rer. nat. Uwe Schmitt    python, php, c++, codes, tips, tutorials at
uwe.schmitt at procoders.net    http://www.procoders.net


From whisper at oz.net  Tue Jun 18 03:53:57 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 18 Jun 2002 00:53:57 -0700
Subject: Listing all available (not just loaded) modules
In-Reply-To: <aemlao$85r2t$1@ID-140151.news.dfncis.de>
Message-ID: <GCEDKONBLEFPPADDJCOECENEDJAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Michael Steuer
> Sent: Monday, June 17, 2002 23:53
> To: python-list at python.org
> Subject: Re: Listing all available (not just loaded) modules
>
>
> > Idle, PythonWin and Boa (at least) do this, with PythonWin
> being the most
> > detailed - it shows classes and methods within the modules. The
> version of
> > Pythonwin I have does not show the contents of the
> site-packages tree, but
> > that would be easily fixable.
>
> They "do this" and how? Don't tell me you mean the dir() command ...
>
> Cheers,
>
> Mik.e
>

In Idle, look on the File menu for "Path Browser". It brings up a tree
window containing Python's idea of what it has installed.

In Pythonwin, it's on the Tools menu: "Browse Pythonpath". Does the same as
the above and also includes the pythonwin directories/modules. It also
drills down to the class/method/function level of everything it can.

In Boa, it's a tab called "Explorer" in their Editor window that has a tree
that offers varios chunks of information, including sys.path. Nice
integration.

How exactly do they do it? That's left as an exercise for the reader :). All
three of these apps have source available and all three of them accomplish
their tasks somewhat differently i'm sure. Good Hunting! (I'd start with
Idle :) .)

Regards,

Dave LeBlanc
Seattle, WA USA





From aahz at pythoncraft.com  Thu Jun 13 15:49:55 2002
From: aahz at pythoncraft.com (Aahz)
Date: 13 Jun 2002 15:49:55 -0400
Subject: Problems with importing; must do it twice?
References: <fgphgus0pv2c7hp8l27uaenjlerr97tdta@4ax.com>
Message-ID: <aeat13$pli$1@panix1.panix.com>

In article <fgphgus0pv2c7hp8l27uaenjlerr97tdta at 4ax.com>,
Stew LG  <stew at ten_INVALID_box.com> wrote:
>	
>My web host just switched hardware, and in the process I've been moved
>to Solaris and Python 2.1.1.

Your web host failed to compile Python correctly; talk with them.

>This seemed insane to me, so I tried Python from the command line.
>Same thing, but when I import it *again* it works:

No, it didn't.  Python caches import names even when the import fails
(protects against circular references); try using the cgi module for
anything that requires a socket and it'll die.  It might work for the
other functions in the cgi module, depending on when cgi tried to import
the socket code.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From plugin at supercable.es  Tue Jun 25 16:32:49 2002
From: plugin at supercable.es (Ugo García)
Date: Tue, 25 Jun 2002 22:32:49 +0200
Subject: Threads on embedded Python
References: <1024952585.111238@seux119> <Xns9237EECAA9CC5cliechtigmxnet@62.2.16.82> <1024957605.771843@seux119> <Xns9238944DA0FEcliechtigmxnet@62.2.16.82> <1025028733.290520@seux119> <Xns9238D6F8D5BBAcliechtigmxnet@62.2.16.82>
Message-ID: <1025037223.199076@seux119>

Thanks again Chris. I've just discovered how Py_BEGIN_ALLOW_THREADS works
and it works fine for me. So at this moment I'm going to leave the main loop
in C; but the article you give me has really makes me think.... it would be
more efficient to work with threads this way and to have the main loop in
Python. I would try this when I have a little more free time... or when I
publish the engine's web with the source code, let's see if someone wants to
try this for me!! :-)

Thanks very much, Chris
--ugo

"Chris Liechti" <cliechti at gmx.net> escribi? en el mensaje
news:Xns9238D6F8D5BBAcliechtigmxnet at 62.2.16.82...
> "Ugo Garc?a" <plugin at supercable.es> wrote in
> news:1025028733.290520 at seux119:
> > But, one question: you say that I don't have to have a main loop in
> > order to have threads running using Py_BEGIN_ALLOW_THREADS, but don't
> > known exactly how to use it. I explain: from the main loop in my C
> > application I launch threads in python using PyRun_SimpleString. But
> > these threads doesn't run unless I call Python to do something. How I
> > have to use Py_BEGIN_ALLOW_THREADS in order not to call Python
> > continuosly?
>
> (im realy not sure if it works that way, but im hoping)
>
> normaly Py_BEGIN_ALLOW_THREADS is used when python calls a C function in
an
> extension module. while the C function is executed, all python activities
> are stopped. to reactivate python, while still beeing in the C func. you
> can call that macro. this allows long running C funcs, without blocking
> python. (note that you must call Py_END_ALLOW_THREADS before leaving the C
> func).
>
> what you want is somewhat the other way round... the macros mentioned
above
> won't work directly, but if you look what they do, they are
> saving/restoring of thread state. it could possibly work if you use those
> functions to release the global interpreter lock. but i haven't done such
> things (i've started a native thread in an extension that called back to
> python on events, but the main loop is in python)
>
> i think its much easier to lauch a loop in python and let it call back to
c
> functions...
>
> http://python.org/doc/current/api/threads.html
>
> chris
>
> --
> Chris <cliechti at gmx.net>
>




From bl at netgroup.dk  Wed Jun 12 04:23:59 2002
From: bl at netgroup.dk (Bo Lorentsen)
Date: 12 Jun 2002 10:23:59 +0200
Subject: Embedding and threads ?
In-Reply-To: <83ofeiq071.fsf@panacea.canonical.org>
References: <mailman.1022667286.17372.python-list@python.org>
	<j44rgrnkfm.fsf@informatik.hu-berlin.de>
	<mailman.1022740196.32586.python-list@python.org>
	<j4d6vdwvyk.fsf@informatik.hu-berlin.de>
	<1022771667.689.136.camel@netgroup>
	<j4znyhzguh.fsf@informatik.hu-berlin.de> <1022830475.893.17.camel@netgroup>
	<j4znygd7iv.fsf@informatik.hu-berlin.de> <1022839266.823.66.camel@netgroup>
	<j4d6vac9b7.fsf@informatik.hu-berlin.de>
	<mailman.1023091388.3109.python-list@python.org>
	<j4wutg8shk.fsf@informatik.hu-berlin.de>
	<mailman.1023108789.25792.python-list@python.org> 
	<83ofeiq071.fsf@panacea.canonical.org>
Message-ID: <1023870239.692.63.camel@netgroup>

On Mon, 2002-06-10 at 23:43, Kragen Sitaker wrote:

> > I will use Python for this project as long as what I do only need a
> > single thread model, and when I need more than that I'll se what to do
> > about it :-)
> 
> It sounds like you want processes, not threads.  Assuming you're not
> on Microsoft Windows, os.fork() will give you one Python interpreter
> state per process.  Different processes have to make special
> arrangements to communicate, though, through explicitly shared memory
> or explicit channels.
Yes, I use thread somehow like processes, but I like to be able to share
recources between them, more easily (cache, db, sockets etc.).

A process gives me a total copy of myself, where a thread only make a
new stack, but share data. I know I can transfere file (and socket)
handlers while using processes, but the thread model is much more
elegant.

/BL





From ak at silmarill.org  Fri Jun  7 15:22:33 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Fri, 07 Jun 2002 19:22:33 GMT
Subject: How could i get execution times ?
References: <adkjro$12183k$1@ID-146704.news.dfncis.de>
Message-ID: <slrnag21vg.5c0.ak@ak.silmarill.org>

In article <adkjro$12183k$1 at ID-146704.news.dfncis.de>, Shagshag13 wrote:
> 
> I'm looking for a way to get execution time at different points of a program, by now i use the
> code below, but i've seen in some post that time.clock() could be a better way to measure this,
> 
> So my question, is how you pythoners do it ??? and to understand _why_ ?
> 
> Thanks
>
profile:

>>> import profile
>>> def f():
...  for i in range(10000):
...   5*350 

>>> profile.run('f')
         2 function calls in 0.020 CPU seconds

   Ordered by: standard name

   ncalls  tottime  percall  cumtime  percall filename:lineno(function)
        1    0.000    0.000    0.000    0.000 <string>:1(?)
        1    0.020    0.020    0.020    0.020 profile:0(f)
        0    0.000             0.000          profile:0(profiler)


 - Andrei


-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From shalehperry at attbi.com  Fri Jun 28 21:40:38 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 28 Jun 2002 18:40:38 -0700 (PDT)
Subject: I'm an idiot
In-Reply-To: <Xns923BBCF03699Adavidgriswold1netsca@204.127.202.16>
Message-ID: <XFMail.20020628184038.shalehperry@attbi.com>

On 29-Jun-2002 David Griswold wrote:
> Thanks.  I actually figured out the missing parentheses, and the typo on 
> the split/strip..  I will blame Active Python for being TOO helpful on 
> the former and a brainfart on the latter.
> 
> Now one more thing is happening I can't figure out.  When readline gets 
> to the end of the file (the file is only 34k) it just keeps on reading 
> blank lines.  Am I missing something?  I even changed my error I am 
> checking for to EOFError.
> 

sorry fell asleep at the wheel ....

the python idiom is usually either:

while 1:
  line = f.readline()
  if not line: break
  
  # process line

or

for line in f:
  # process line




From phlip_cpp at yahoo.com  Sat Jun  1 02:01:31 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 01 Jun 2002 06:01:31 GMT
Subject: ANN: PyUnitTestBrowser 003
References: <63604d2.0205310937.15d47972@posting.google.com> <uit54nem7.fsf@online.no> <ZSSJ8.10245$Om3.446676066@newssvr21.news.prodigy.com>
Message-ID: <ad9nvr$hav@dispatch.concentric.net>

>> Have you any hints on how to hack browser.py so that it run
>> unittest.TestCase derived classes in a file without a suite method
>> that generates it?
> 
> Someone sent me a patch that does this, so it's going in the next
> "release".

This tweak and a few other cosmetic issues are in the bag.

Also db() works.

-- 
  Phlip
   http://www.greencheese.org/InMildDefenseOfTheGnats
  --  Now collecting votes for a new group:
      news:alt.flame.moderated  --



From pyth at devel.trillke.net  Fri Jun  7 09:45:38 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 7 Jun 2002 15:45:38 +0200
Subject: SLP bindings?
Message-ID: <20020607154538.B6609@prim.han.de>

Hello,

does anyone know of *any pythonic efforts* to implement
a 'Service Locator Protocol' (SLP) as defined by

    http://www.openslp.org/doc/rfc/rfc2614.txt

where openslp.org provides a C/java implementation already?

AFAIK dynamic platform independent registration and locating 
via SLP seems to be very promising. I know of some LAN-instances 
where all services (nfs, cups, smb, ...) are located via this protocol.

thanks,

    holger




From eager at eagercon.com  Sun Jun  9 16:37:05 2002
From: eager at eagercon.com (Michael Eager)
Date: Sun, 09 Jun 2002 13:37:05 -0700
Subject: import problem
Message-ID: <2se7guc3kkmm3bgooj776j55igpkmgpst4@4ax.com>

I'm trying to use Python 2.2 and PyQt on a RedHat-7.2 system.

I get an error when I execute "from qt import *" -- there is a failure
in qt.py when it executes "import libsup".  The error is "No module
named libsup".

libsup.so.9 is a link to libsup.so.9.0.3 and is in
/usr/lib/python2.2/site-packages.  If I enter "import libsup", I get
the same error message.

Any suggestions?

-- Mike Eager, eager at eagercon.com


From arosado at softhome.net  Tue Jun  4 20:12:24 2002
From: arosado at softhome.net (Andres Rosado)
Date: Tue, 04 Jun 2002 20:12:24 -0400
Subject: manipulating child processes under Windows 
In-Reply-To: <200206041603.g54G3Ad19721@gaffa.mit.edu>
References: <Your message of Tue, 04 Jun 2002 06:42:57 -0600. <courier.3CFCB5D1.000044EC@softhome.net>
Message-ID: <5.1.0.14.0.20020604200622.00c1c870@mail.softhome.net>

At 12:03 PM 6/4/2002 -0400, Douglas Alan wrote:
> > I managed to do the contrary with one of the spawnl() function. Pass as a
> > first parameter the os.P_WAIT to wait for the program to finish.
>
>But this will only wait for the child process to finish.  In my case,
>the child process is just a stub program that starts up a grandchild
>process and then terminates.  I need my Python program to wait for the
>grandchild process to terminate.  spawnl() doesn't provide a way of
>doing this.

In that case, you need the process information of the grandchild and a call 
for the funcion os.wait(). Sorry I can't help anymore, but I don't know how 
to get that information in Windows. :(


-----------------------------------
Andres Rosado
Email: andresr at despammed.com
ICQ: 66750646
Homepage: http://andres980.tripod.com/

Impartial, adj.:
         Unable to perceive any promise of personal advantage from
espousing either side of a controversy or adopting either of two
conflicting opinions.
                 -- Ambrose Bierce, "The Devil's Dictionary"





From ftobin at neverending.org  Sun Jun 30 16:45:43 2002
From: ftobin at neverending.org (Frank Tobin)
Date: Sun, 30 Jun 2002 16:45:43 -0400
Subject: How to get rid the new line
In-Reply-To: <4phT8.45578$n4.10822996@newsc.telia.net>
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org>
 <3D1AD534.A74EC9F4@ipm.fhg.de> <3D1B1A76.563E5394@engcorp.com>
 <pan.2002.06.28.00.14.48.789371.16627@linuxmail.org> <afg513$2426$1@news.idiom.com>
 <3D1BA535.1EE16646@engcorp.com> <4phT8.45578$n4.10822996@newsc.telia.net>
Message-ID: <Pine.LNX.4.44.0206301641200.7942-100000@palanthas.neverending.org>

Fredrik Lundh, on 2002-06-29, wrote:

> or add the single character needed to make chomp a
> bit more robust:
> 
>     def chomp(line):
>         if line[-1:] == '\n':
>             line = line[:-1]
>         return line

I think the fact that the first few tries here were *wrong* about how to
implement a chomp indicates that it really should part of Python.  Just
because it's easy to implement doesn't mean it can't be part of the, say,
string methods.  I mean, look at str.{upper,lower,isdigit}, etc.  A chomp
method would be popular; rstrip or implementing your own doesn't cut it.

-- 
Frank Tobin			http://www.neverending.org/~ftobin/



From ark at research.att.com  Fri Jun 14 13:50:09 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 14 Jun 2002 17:50:09 GMT
Subject: a possibly foolish question about slices
Message-ID: <yu99sn3prbqm.fsf@europa.research.att.com>

Suppose x is a string or list, and I want to refer to
a subsequence of x with length n starting at position p.
under the assumption that this subsequence is entirely

To do so, I can use x[p:p+n] in all circumstances but one,
namely when p < 0 and p == -n.  Is there a similarly concise
expression that works even in this special case?

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From maxm at mxm.dk  Thu Jun 20 02:51:11 2002
From: maxm at mxm.dk (Max M)
Date: Thu, 20 Jun 2002 08:51:11 +0200
Subject: Python+ASP->import problem
References: <slrnah0pat.fi1.jarvin24@lehtori.cc.tut.fi>
Message-ID: <3D117B5F.4030206@mxm.dk>

J?rvinen Petri wrote:


> When importing my modules, I always get AttributeError for attributes that 
> surely exist on imported module.
> 
> <clip>
> Response.Write(enactHTML.defineCSS('Enact Main'))
> AttributeError: 'module' object has no attribute 'defineCSS'
> </clip>


I'll betcha that that your module has a bug, or that it imports another 
module which has a bug.

If the module encounters errors during import of other modules, it 
registers it's own name in the module list but is only an empty "zombie" 
module. As far as I understand. So your module seems to be imported 
allright in your asp page, but it really is not.

Then this rather cryptic message can appear, as your page cannot find 
the attribute in the "zombie" module.

Hmm ... hope this was clear enough.

regards Max M



From occeanlinux at linuxmail.org  Thu Jun 27 11:32:28 2002
From: occeanlinux at linuxmail.org (SiverFish)
Date: Thu, 27 Jun 2002 15:32:28 +0000
Subject: Questiion on list
References: <pan.2002.06.27.14.36.03.151659.5728@linuxmail.org> <slrnahl7j8.fd3.quinn@upchuck.ugcs.caltech.edu>
Message-ID: <pan.2002.06.27.15.32.28.404968.5728@linuxmail.org>

On Thu, 27 Jun 2002 05:11:06 +0000, Quinn Dunkan wrote:

> On Thu, 27 Jun 2002 14:36:03 +0000, SiverFish
> <occeanlinux at linuxmail.org> wrote:
>>I got the list of words now i want to calculate how many word in that
>>list which is the same word (insensitive case),can anyone show me how to
>>do it for example ls = ['abc','cd','abc','adf',abc','dwc','cd'] it will
>>show abc = 3 ,cd = 2, adf = 1, dwc=1
> 
> import string
> d = {}
> for e in map(string.lower, ls):
>     d[e] = d.get(e, 0) + 1
> freqs = [ (count, w) for w, count in d.items() ] freqs.sort()
> freqs.reverse()
> print ', '.join([ '%s = %d' %(count, w) for count, w in freqs ])

Could you  do it in python 1.5.2 and explain clearly to me please 
How about we have to do the same problem here when read in the multiple
files,Thank a lot


From jadestar at idiom.com  Mon Jun 17 23:12:08 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 18 Jun 2002 03:12:08 GMT
Subject: Pop a list from beginning ? and memory saving...
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de>
Message-ID: <aem8e8$4kq$1@news.idiom.com>

Shagshag13 <shagshag13 at yahoo.fr> wrote:

> I had a huge list containing consuming data, as i only need a one pass in a
> for statement,
> i'm wondering if by doing this with a pop i could save memory ? (as i think
> that python
> free it after the pop...)
> and is there a way to pop from the beginning  of the list ? (i think i could
> use reverse but
> i don't know if this memory consuming)

 The problem with l.pop(0) is that it will run at O(n) time
 (linearly proportional to the list's size).

 From what I gather the underlying C structure of a list allows
 things to be appended and popped (from the end) in constant time.
 So the obvious answer would be to build the list in reverse order
 and then pop() them all off the end with a simple:

	while len(l):
		i=l.pop()
		# ...

 If that's not feasible (you can't build the list in reverse order)
 then you might try profiling list.reverse() (it might be implemented
 as an inplace reversal in the underlying C or Java.  If so the reverse()
 method will be O(n) and the processing will be O(1) for each item
 for a total of 2 * O(n).  That may not be ideal but it's much better
 than ~ O(n) * 1/2 * 0(n) (which is the roughly the result of the 
 original approach where each iteration takes O(n)).

 So, the critical question is:

	Is the list's reverse() method done in place and does it
	run in linear time (or better?  if so, how)?
 


From whisper at oz.net  Sun Jun  9 02:38:08 2002
From: whisper at oz.net (David LeBlanc)
Date: Sat, 8 Jun 2002 23:38:08 -0700
Subject: Determination of screen resolution?
In-Reply-To: <3D02F4C8.C2DF2EE0@bellatlantic.net>
Message-ID: <GCEDKONBLEFPPADDJCOEAEKCDHAA.whisper@oz.net>

On what platform - it will differ between linux/unix, windows and mac?

David LeBlanc
Seattle, WA USA 

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of David Lees
> Sent: Saturday, June 08, 2002 23:25
> To: python-list at python.org
> Subject: Determination of screen resolution?
> 
> 
> I would like to use tkinter and have the canvas object automatically
> scaled fit the screen.  Is it possible to read the screen resolution
> (800x600, 1024,768...) from somewhere.
> 
> David Lees
> -- 
> http://mail.python.org/mailman/listinfo/python-list




From johnroth at ameritech.net  Wed Jun  5 07:41:47 2002
From: johnroth at ameritech.net (John Roth)
Date: Wed, 5 Jun 2002 07:41:47 -0400
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <slrnafps1c.4ea.huaiyu@gauss.almadan.ibm.com> <ufqm0voskv86cc@news.supernews.com> <slrnafqppg.4ea.huaiyu@gauss.almadan.ibm.com> <MCeL8.114006$%y.11961913@bin4.nnrp.aus1.giganews.com>
Message-ID: <ufru76lis4k867@news.supernews.com>

"Terry Reedy" <tjreedy at udel.edu> wrote in message
news:MCeL8.114006$%y.11961913 at bin4.nnrp.aus1.giganews.com...
>
> "Huaiyu Zhu" <huaiyu at gauss.almadan.ibm.com> wrote in message
> news:slrnafqppg.4ea.huaiyu at gauss.almadan.ibm.com...
> > Objects are referenced by their ids (or equivalent).  Their memory
> location
> > may or may not change when it is modified, but if it is the same
> object,
> > every name that was refering to it will remain referencing to it
> after the
> > modification, regardless of its memory location.
>
> Conceptually, PyObjects do not even *have* a memory location.  In the
> current CPython implementation for computers, they have a *fixed*
> location, which happens to be used for the unique integer id.  A
> (fixed-location) list object contains a pointer to an auxiliary array
> which may be moved when resized.  Same for dicts.
>
> Terry J. Reedy

I see your point, which explains why the list in the first
example always updates. The issue then seems to be that
in cases where the '+=' operation returns the same object
it started with, the generated code still tries to update
the container object, causing an exception if the
container object happens to be immutable.

John Roth

>
>
>




From kragen at pobox.com  Sat Jun  8 22:07:30 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 08 Jun 2002 22:07:30 -0400
Subject: constructive critisism on pythonwin IDE
References: <6bd9f01b.0206050736.f469082@posting.google.com> <3CFE9861.5090200@skippinet.com.au> <3D00E261.CA641C7E@noaa.gov>
Message-ID: <83vg8t4331.fsf@panacea.canonical.org>

Chris Barker <Chris.Barker at noaa.gov> writes:
> > sameer wrote:
> > > The thing I hate about PythonWin
> > > is, that despite the fact that it's a development environment, there
> > > is a lot of caching done of modules, and the cached copy is not
> > > checked against a potentialy modified copy.  Whenever I change the
> > > contents of a file and try to run a module that depends on that file,
> > > I always end up with the old copy.  I have to then import and then
> > > reload the changed module in the interactive window for it to refresh.
> 
> This is how Python works, it's not anything specific to PythonWin. It
> works this way because a given module might be imported dozens of times
> in a given Python program, and you would want that to be done fresh each
> time.

                                    VVV
I think you might mean, "you would *not* want that to be done fresh".
                                    ^^^
Is that right?



From gumuz at looze.net  Mon Jun 24 04:13:35 2002
From: gumuz at looze.net (Guyon Morée)
Date: Mon, 24 Jun 2002 10:13:35 +0200
Subject: Socket Disconnection
References: <3d1304d0$0$224$4d4ebb8e@news.nl.uu.net> <3d16817e$1@news.nz.asiaonline.net> <7934d084.0206232354.52c2163d@posting.google.com>
Message-ID: <3d16d3c7$0$229$4d4ebb8e@news.nl.uu.net>

ok, thanx a lot!

what do you mean by an empty datastring? does it, before disconnecting, send
a "" or none?

and for the case of abrupt(hardware) disconnecting, should i use a 'try:' or
does this not have any effect?

thanx,

Guyon

"Andrae Muys" <amuys at shortech.com.au> wrote in message
news:7934d084.0206232354.52c2163d at posting.google.com...
> "Colin Brown" <cbrown at metservice.com> wrote in message
news:<3d16817e$1 at news.nz.asiaonline.net>...
> > Hi Guyon
> >
> > I assume here that you are talking TCP/IP.
> >
> > If you are waiting on a receive [ data = connection.recv() ] then you
will
> > have an empty data string returned when the client disconnects.
> >
> > If you are sending you will get an error telling you the link was shut
down.
> >
>
> Also note that the only reliable way to detect a client disconnect is
> the latter.  If you want to demonstrate the failure of the recv()
> approach, just yank out the ethernet cable on the client machine, and
> notice that although the client is now disconnected, the server has no
> idea.
>
> Might I suggest taking a look at TCP/IP Illustrated Vol 1, by
> W.Richard Stevens.  Definately a book to read if you are doing any
> TCP/IP network programming.
>
> Andrae Muys




From chris.gonnerman at newcenturycomputers.net  Sun Jun 23 22:52:21 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sun, 23 Jun 2002 21:52:21 -0500
Subject: Newbie question (anyone programming with curses?)
References: <af4ovd$1bt3$1@pingwin.acn.pl> <ummR8.290538$352.27434@sccrnsc02>
Message-ID: <004901c21b2a$31254e20$0101010a@local>

----- Original Message ----- 
From: "Emile van Sebille" <emile at fenx.com>


> hellboy
> > I'm quite new to Python, but I'm positive that version 2.2 (for
> Windows)
> > comes without curses module, though documentation states it is
> included.
> > import curses gives me "Import Error: no module named curses" message.
> > Anyone knows where to download curses module (for Windows)?
> >
> 
> from http://py-howto.sourceforge.net/curses/node2.html
> 
> No one has made a Windows port of the curses module. On a Windows
> platform, try the Console module written by Fredrik Lundh. The Console
> module provides cursor-addressable text output, plus full support for
> mouse and keyboard input, and is available from
> http://effbot.org/efflib/console.
> 
> Hope this is current and helps,

I found the effbot's version a bit unstable on Windows 9x (a lot actually)
and I wound up writing WConio, based on the TCCONIO.C sources for MinGW32.

    http://newcenturycomputers.net/projects/wconio.html

but it's still not curses.  I have considered many times creating a GConio
(Generic Console I/O) module wrapping both WConio on Win32 and curses on 
Unixoid OS's.  Haven't done much other than think about it though.






From kragen at pobox.com  Mon Jun 10 03:38:10 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 03:38:10 -0400
Subject: Where is Berkeley DB on your system? Can you live without DB 1.85?
References: <mailman.1023559228.8459.python-list@python.org>
Message-ID: <83lm9n4m8t.fsf@panacea.canonical.org>

"David LeBlanc" <whisper at oz.net> writes:
> AFAIK, this is true for any version of Sleepycat > 1.85, and has been true
> for years. Including a later version of Sleepycat BSDDB imposes a
> constraint, not on Python itself, which is free, but on someone who would
> like to distribute Python along with their _commercial_for_sale_ app (think
> Py2Exe for example).

I don't think anybody has proposed including a modern version of BSD
DB in Python, only dropping support for an old version, which is also
not included in Python.

"commercial for-sale" is misleading.  The current license of BSD DB
allows distribution in commercial for-sale apps with no further hassle
unless said commercial for-sale apps are, approximately, free
software, and it forbids inclusion in noncommercial not-for-sale apps
unless said not-for-sale apps are, approximately, free software.

> If you want to change Python's market dynamics and require _all_ Python apps
> to be free and open source, then you can include any version of Sleepycat
> BSDDB you like. I doubt it will have a positive impact on Python's future
> prospects. It didn't work for Perl in it's early years and that's why the
> Artistic License was change to allow "for fee" distributions that included
> Perl.

I thought the AL allowed this from the beginning.  Certainly
Sleepycat's license doesn't forbid it.

In any case, your irresponsible, alarmist accusation is a red herring,
as you will no doubt agree once you understand the situation properly.
Dropping support for db 1.85 will not require all Python apps to be
free software or open source; it won't even require all distributions
of Python to be free software or open source.  It will only require
distributions of Python with bsddb support to either link to an
external BSD DB library (like the one included with glibc) or be free
software.

If your application doesn't use BSD DB, it won't be affected by this,
even if you ship binary distributions of it, made with e.g. McMillan
Installer.  If your application uses BSD DB and you are shipping db
1.85, you are screwing your users, probably without knowing it (unless
you know enough about db 1.85's bugs to work around them.).  This
change will allow you to continue to screw your users in the same way,
but you will be required to know that was what you were doing.

At present, there are irresponsible vendors shipping proprietary
distributions of Python including BSD DB 1.85.  This course of action
subjects the entire Python community to pressure to drop or deprecate
support for BSD DB, because BSD DB 1.85 is badly broken.  Those of us
who use sane versions of BSD DB are hurt by this, because BSD DB is by
far the most featureful, performant, and reliable of the other
dbm-style databases Python supports.

I assume these vendors are being irresponsible because they don't know
they are imperiling their users' data by using this buggy old
software.  This change will let them know that the problem exists, and
they can either stop shipping proprietary distributions, stop
supporting BSD DB, or start maintaining BSD DB 1.x.

If it's not clear, I'm in favor of this change.  It helps everyone
everywhere.



From plugin at supercable.es  Mon Jun 24 17:02:10 2002
From: plugin at supercable.es (Ugo García)
Date: Mon, 24 Jun 2002 23:02:10 +0200
Subject: Threads on embedded Python
Message-ID: <1024952585.111238@seux119>

Hi all. I have an applicattion who uses Python to control some processes.
I've made a little script library to work with threads and each proccess is
a Python thread. The main application make calls to Python in order to
create these proccesses. The problem is that the proccess doesn't run unless
I call Python to do something. Now in the main loop I have a line like this:
    PyRun_SimpleString("dummy=1");

This make Python the threads to run. The problem is that this is a little
slow 'cause if I call Python from tha main application too many times the
proccesses run well but not the main application, and if a make a big pause
between calls the proccesses doesn't work at a good speed. Is posible to
have some threads running in an embedded Python without making callings to
it continuosly? It would be a better idea to have another thread who's only
task is to call Python (a line that the one written before) in order to have
the Python proccesses running?

Thanks in advance,
--ugo





From bielen at stafa.nl  Tue Jun 25 05:53:40 2002
From: bielen at stafa.nl (Patrick Bielen)
Date: Tue, 25 Jun 2002 11:53:40 +0200
Subject: Python & zipfiles
Message-ID: <uhgfcopr269e0@corp.supernews.com>

Hello all,

I use a module to work with zipfiles.
This one is only for unpacking files
from the zip. But i only want to know
if i have to much code to do the trick.

I'm barely new to python, that's why i ask.

--- begin script

import os
from zipfile import ZipFile

zipdir = '/as400/Beheer/Databases/'
os.chdir('/tmp')
zipfiles = ZipFile(zipdir + 'dbs.zip').namelist()
ZipFile(zipdir + 'dbs.zip','r')
for file in zipfiles:
    print file
    source = ZipFile(zipdir + 'dbs.zip').read(file)
    open(file, 'w').write(source)
ZipFile(zipdir + 'dbs.zip').close()

--- end of script ---

The problem is, that i don't know if i use the
ZipFile-command too much... because the manual
told me i have to close it afterwards, but it's
not clear to me if i have to expecitely open it.

Any help or code cleaning would be really appre-
ciated.

Greetings

Patrick


From dsavitsk at e-coli.net  Sun Jun 23 01:49:44 2002
From: dsavitsk at e-coli.net (dsavitsk)
Date: Sun, 23 Jun 2002 05:49:44 GMT
Subject: connect to running internet explorer (IE) via shdocvw.dll
References: <3D106564.4040602@pretection.com>
Message-ID: <YjdR8.1715$Kt4.681644@newssrv26.news.prodigy.com>

if it is for your own use, you can place the VB in a public method of a
class and control that via python and COM.

-d


"Marc Spoorendonk" <marc.spoorendonk at pretection.com> wrote in message
news:3D106564.4040602 at pretection.com...
> This is how my problem can be handled in VB:
>
> Dim SWs As New SHDocVw.ShellWindows
> Dim IE As SHDocVw.InternetExplorer
>
> For Each IE In SWs
>        Set Doc = IE.document
>        If TypeOf Doc Is HTMLDocument Then
>           List3.AddItem Doc.url
>        End If
> Next
>
> The question is simple: How can I accomplish this same thing in Python?
>
> Mark Hammond answered me this by email:
> "I think you will find SHDocVw is a vtable interface, and therefore not
> supported by Python."
>
> I've read up about vtable, COM etc. but I am fairly new to this. (I was
> born as a Unix programmer.)  I've played with the interfaces and
> makepy.py. Should I write a python-plugin in C++ or can it be done by
> native pytoncode?
>
> Regards,
>
> Marc
>




From ethanw at rediffmail.com  Thu Jun 13 18:55:09 2002
From: ethanw at rediffmail.com (Ethan)
Date: 13 Jun 2002 15:55:09 -0700
Subject: PythonWin Problem!
Message-ID: <1a52c7f1.0206131455.6df50cf8@posting.google.com>

I have been using ActiveState's port of Python for a while. PythonWin
has been very helpful. Since a couple of days PythonWin developed a
problem. When I start it, it appears in the taskbar and tray but does
not appear on the screen.

My Win98 did recently develop a problem which prevents me from getting
the right click menu on the taskbar entries. This was secondary to
some registry cleaning I did. I am not sure if this is related to
Pythonwin problem but all others apps are working fine. I do get the
menu for the tray icon and the "Close Program" window does not say
"not responding"

I uninstalled Python - deleted the Python21 folder and reinstalled. It
did not help. Any suggestions? I really prefer PythonWin over IDLE.


From skip at pobox.com  Fri Jun  7 14:02:14 2002
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 7 Jun 2002 13:02:14 -0500
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <200206071756.g57HulT01147@europa.research.att.com>
References: <yu99vg8v6orb.fsf@europa.research.att.com>
        <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
        <200206071649.g57GnmC00708@europa.research.att.com>
        <15616.62131.521803.383860@12-248-41-177.client.attbi.com>
        <200206071756.g57HulT01147@europa.research.att.com>
Message-ID: <15616.62758.469825.704384@12-248-41-177.client.attbi.com>

    Andrew> No -- LD_LIBRARY_PATH tells the runtime system ...  LD_RUN_PATH
    Andrew> tells the linker ...

Like I said, it's been ages... ;-)

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html





From emile at fenx.com  Tue Jun 11 10:21:53 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 11 Jun 2002 14:21:53 GMT
Subject: Defining a method final
References: <6glN8.35376$86.884025@twister1.libero.it> <mailman.1023796095.1427.python-list@python.org> <phmN8.36164$K4.893500@twister2.libero.it> <KwmN8.153269$352.7754@sccrnsc02> <ae4tvv$rvd$1@wanadoo.fr>
Message-ID: <5InN8.160528$cQ3.4621@sccrnsc01>

"Eric Brunel" <eric.brunel at pragmadev.com> wrote in message
news:ae4tvv$rvd$1 at wanadoo.fr...
> Python now have an "official" means to make attributes private by
prefixing
> them with a double-underscore. And it *does* prevent their use or
> overloading:
>

I don't know why I thought name mangling only happened at the module
level... But note that even here it's still usable, and that the user
has broken the rule of 'leading underscore == leave it alone'.

>>> class Test:
...     def __private(self): pass
...
>>> dir(Test)
['_Test__private', '__doc__', '__module__']
>>> dir(Test)
['_Test__private', '__doc__', '__module__']
>>> Test._Test__private
<unbound method Test.__private>
>>>

--

Emile van Sebille
emile at fenx.com

---------



From fredrik at pythonware.com  Sat Jun 29 08:06:23 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 29 Jun 2002 12:06:23 GMT
Subject: incorrect upper()/lower() of UTF-8
References: <afiecp$oss$1@web1.cup.hp.com>
Message-ID: <3phT8.45577$n4.10823047@newsc.telia.net>

Jim Henry wrote

> It appears that python does not uppercase/lowercase
> utf-8 strings properly:

It's not supposed to: the 8-bit string type can hold either
text strings or binary buffers.  The encode method takes
a text string and returns a binary buffer.

For reliable processing of Unicode data, use Unicode strings.

</F>




From chris.gonnerman at newcenturycomputers.net  Sat Jun  8 10:45:53 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sat, 8 Jun 2002 09:45:53 -0500
Subject: Removing ^M
References: <Pine.LNX.4.21.0206091005170.7346-100000@localhost.localdomain>
Message-ID: <004501c20efb$386ad1e0$0101010a@local>

----- Original Message ----- 
From: "Michael Hall" <olc at ninti.com>


> I am trying remove ^M characters (some kind of newline character) from an
> HTML file. I've tried all sorts of string.replace and sed possibilities
> but the
> things just won't go away. Does anyone have a way of removing such
> characters?

Unixoid OS, right? 

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

import sys

data = sys.stdin.read()

for ch in data:
    if ch != "\r":
        sys.stdout.write(ch)

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

Run it like this:

    python cleaner.py <source.html >destination.html

Yeah, it could have a bunch more nice stuff...  This is the Q&D 
version.

If the file is really large this won't be real efficient, but for 
small files it's probably the best way.


Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net





From stuart at stu.org.uk  Thu Jun 13 10:35:32 2002
From: stuart at stu.org.uk (Stuart)
Date: Thu, 13 Jun 2002 15:35:32 +0100
Subject: Embedded systems
Message-ID: <aeaajv$5db01$1@ID-17980.news.dfncis.de>

Is Python practical for embedded systems?   I am working on a project at the
moment that uses a GSM/GPRS modem and a GPS receiver.  I have prototyped the
whole thing in Visual Basic, and now I need to port it over to a suitable
processor (possibly on a 68000 device using Linux).

I don't really want  to get into C for this project, and was wondering if
Python was suitable?  Are memory requirements an issue?

Stuart
www.stu.org.uk




From Joseph.Dornisch at raytheon.com  Wed Jun 19 15:53:55 2002
From: Joseph.Dornisch at raytheon.com (Joe Dornisch)
Date: Wed, 19 Jun 2002 15:53:55 -0400
Subject: outside email
Message-ID: <3D10E153.1010608@raytheon.com>

Hey Jon,

What's your outside email address. I got a pretty funny email yesterday 
that I forwarded
to a bunch of people, but you shouldn't open it up at work.  So what's 
your home email address.

-Joe

-- 
--------------------------------
Joseph A Dornisch 
Joseph.Dornisch at raytheon.com
Phone: (814) 278-3130

Raytheon State College
Science Park Road
State College PA, 16801



From martin at v.loewis.de  Thu Jun 20 03:53:57 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 20 Jun 2002 09:53:57 +0200
Subject: Unicode hexadecimal characters
References: <20020620015059.11518.52464.Mailman@mail.python.org> <mailman.1024552525.29935.python-list@python.org>
Message-ID: <m3d6umjsh6.fsf@mira.informatik.hu-berlin.de>

Terry Hancock <hancock at anansispaceworks.com> writes:

> Makes me wonder what the procedure is for suggesting a new
> code block or if it's already been proposed and rejected
> for some reason.  I suppose a search on the site above would
> probably reveal this.

There is actually a link for this on the Consortium's home page, to

http://www.unicode.org/pending/proposals.html

It is recommended that you post your idea on the Unicode mailing list
first.

Regards,
Martin



From ktf at nospamathotmail.com  Sat Jun 15 01:34:57 2002
From: ktf at nospamathotmail.com (KTF)
Date: Sat, 15 Jun 2002 17:34:57 +1200
Subject: VBA to Python
Message-ID: <HjAO8.4199$iR2.398698@news.xtra.co.nz>

In moving from using Access front end forms and code connected to an Access
back end I am looking at changing to Java front end to SQL back end with a
business rules engine using Python.

These are busy forms and there is a lot of data validation (checking data
input for correct format, string length etc), if this do that's, and
directing of drop-down list content based on the entry or selection of the
previous field (if this make, populate the next list with the models of the
chosen make etc).

Is Python a good choice to manage and run the rules and is there a way of
transcribing / translating the VBA code across to Python other than
literally doing it
line by line?

Thanks,
Keith






From dietmar at bronko.fh-wedel.de  Sun Jun 23 06:37:30 2002
From: dietmar at bronko.fh-wedel.de (Dietmar Lang)
Date: 23 Jun 2002 10:37:30 GMT
Subject: Parsing
References: <3D14CEBC.C6655F86@vip.fi>
Message-ID: <slrnahb9gi.bo.dietmar@bronko.fh-wedel.de>

Hi,

Pekka Niiranen wrote:
> Has anybody used python's parser -module to parse user defined
> text structure? It is possible ?

No, but it seems that the Toy Parser Generator[1] might be just what you
need. The parser module is for parsing Python code. If configurations
are specifically what you want to store and retrieve, the ConfigParser
module from the standard library will be easier to use.

Greetings,
Dietmar

[1] http://christophe.delord.free.fr/en/tpg/index.html

-- 
I haven't lost my mind -- it's backed up on tape somewhere.


From sholden at holdenweb.com  Thu Jun  6 10:22:06 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 6 Jun 2002 10:22:06 -0400
Subject: Using mxDateTime
References: <I9KL8.125444$ux5.179714@rwcrnsc51.ops.asp.att.net>
Message-ID: <MiKL8.314322$%u2.224305@atlpnn01.usenetserver.com>

"LJD" <ljd at nospam.com> wrote ...
> Hi:
>
> I'm new to python and I have a need to use time/date functions.
>
> I am running redhat 7.3 with the default python 1.52.  I just downloaded
the
> appropriate mxDateTime RPM and installed it.
>
> How do I now make those mxDateTime functions available to me in my python
> program?  Since this is an extension module rather than a built in
library,
> I assume I can't simply...
>
> import mxDateTime
>

Almost. You actually need to

    import mx.DateTime

because the mx extensions are nowadays structured as components of the mx
package.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From krissepu at vip.fi  Sat Jun 22 15:23:40 2002
From: krissepu at vip.fi (Pekka Niiranen)
Date: Sat, 22 Jun 2002 22:23:40 +0300
Subject: Parsing
Message-ID: <3D14CEBC.C6655F86@vip.fi>

Has anybody used python's parser -module to parse user defined
text structure? It is possible ?

-pekka-



From pinard at iro.umontreal.ca  Thu Jun 20 11:58:46 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 20 Jun 2002 11:58:46 -0400
Subject: Fill a file with some text.
In-Reply-To: <Xns9233739A9FCB8mdkmdkmdkmdk@130.133.1.4>
References: <Xns9233739A9FCB8mdkmdkmdkmdk@130.133.1.4>
Message-ID: <oqr8j2kkll.fsf@titan.progiciels-bpi.ca>

[mdk]

> I need to fill a file with differnt amounts of some text, either just
> the letter x or random, doesn't matter.  But I need the size to be, for
> example, 6070784 bites.  What is the best and or fastest way to do this?

The simplest may be:    open(FILE, 'w').write('x' * 6070784)

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From richie at entrian.com  Thu Jun 27 08:59:56 2002
From: richie at entrian.com (richie at entrian.com)
Date: Thu, 27 Jun 2002 13:59:56 +0100
Subject: HTML Error from Python CGI Script
References: <79b54d28.0206241618.6c78dbde@posting.google.com>
Message-ID: <il2mhuou82avfd348m7u3tkuou4p2h6633@4ax.com>

Chris,

> syntax error at line 2: `newline or ;' unexpected

Are you writing your temporary HTML file into your cgi-bin directory?  Is
the server configured to run files in that area as scripts rather than to
serve them as documents?  If so, you should either write the file into the
'pages' area of your web server (or whatever it's called on your system)
rather than cgi-bin, or reconfigure your web server to only run certain
extensions as CGI scripts.

-- 
Richie Hindle
richie at entrian.com



From stefnin.nospam at yahoo.fr  Wed Jun 19 05:53:13 2002
From: stefnin.nospam at yahoo.fr (Stephane Ninin)
Date: 19 Jun 2002 09:53:13 GMT
Subject: tkinter menu (newbie question)
Message-ID: <Xns9232778FCD335stefninnospamyahoofr@213.228.0.138>

  Hi all,

I am completely new to kinter and quite new too to Python,
so ... (sorry if my question has already been asked many times here...)


here is a simple program I am trying to make

<--->

#! /Python22/python

from Tkinter import *

class AppWin(Frame):

    def __init__(self,master=None):
      Frame.__init__(self,master)
      self.grid()
        
      self.menubar = Menu(self)

 	self.filemenu = Menu(self.menubar)
	
 	self.filemenu.add_command(label="New")
 	self.filemenu.add_command(label="Open...")
 	self.filemenu.add_command(label="Close")
 	self.filemenu.add_separator()
 	self.filemenu.add_command(label="Quit", command=self.nothing)

	self.editmenu = Menu(self.menubar)

      self.helpmenu = Menu(self.menubar, name='help')

      self.helpmenu.add_command(label="About...")

	self.menubar.add_cascade(label="File", menu=self.filemenu)
	self.menubar.add_cascade(label="Edit", menu=self.editmenu)
	self.menubar.add_cascade(label="Help", menu=self.helpmenu)


    def nothing():
        pass
    
   
# main
        
appwin = AppWin()
appwin.mainloop()

<--->

The main problem is that except a small window (Tk), 
there is no menu on it...
Does anybody have an idea why ? and how to change that ?
(that's probably quite a stupid mistake... :) )
 

   Thanks in advance.

   Regards,


   Stephane


-- 
  Stephane Ninin
stefnin.nospam at yahoo.fr


From BPettersen at NAREX.com  Sat Jun 29 11:50:25 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Sat, 29 Jun 2002 09:50:25 -0600
Subject: private
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192F1A9F3@admin56.narex.com>

> From: Rhymes [mailto:rhymes at myself.com] 
> 
> Is there any possibility to build up a _real_ private 
> attribute? Such as C++ private data members... I know that 
> with double-underscore prefix (and the name mangling related) 
> I can " hide " data members but I think it's not a so crafty 
> trick... Is there any PEP about it? Will Guido insert a 
> private keyword in a future release of Python ?

Ok, here's a challenge for you... Write a C++ program with private data
members, then find at least two trivial ways of circumventing it.

Since private is just an indicator that you need to do some more work if
you want to access the member in C++, I don't see why Python should
bother to offer something more. If you can't expect your programmers to
act like grown-up's you really have more serious problems...

-- bjorn




From mhammond at skippinet.com.au  Mon Jun 17 18:46:37 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 17 Jun 2002 22:46:37 GMT
Subject: MS Word -- finding text
References: <mailman.1024247060.12127.python-list@python.org> <3D0D1862.5050708@skippinet.com.au> <ugsc48godgaja8@news.supernews.com>
Message-ID: <3D0E6767.9070907@skippinet.com.au>

Mike MacDonald wrote:
> Hi Mark !
> Thanks for this and also for the superb win32 contribution.
> My copy of the book is close to worn out !
> Best regards

Thanks!  The copy I use is not yet worn out, but does have some nasty 
red wine stains :)

I've actually been amazed (and happy) how often I have had to refer to 
my own book :)

Mark.



From mickey at tm.informatik.uni-frankfurt.de  Sat Jun 15 10:31:50 2002
From: mickey at tm.informatik.uni-frankfurt.de (Michael 'Mickey' Lauer)
Date: 15 Jun 2002 16:31:50 +0200
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net>
Message-ID: <3d0b4fd6@nntp.server.uni-frankfurt.de>

David Abrahams <david.abrahams at rcn.com> wrote:
> > "Michael 'Mickey' Lauer" <mickey at tm.informatik.uni-frankfurt.de> wrote in
> message news:3d0a8b6e at nntp.server.uni-frankfurt.de...
> 
> 
>> I had a quick look at the other wrappers (SILOON, GRAD, CXX, etc.)
>> but most look either outdated or not appropriate for this task.
>> Do you know an (automatic) wrapper generator which has this feature?
>>
>> I know sip can do this for c++ libraries - can sip also be used
>> to wrap plain c-libs without "handwriting" code?
> 
> On most platforms, Boost.Python would work for you. The determining issue is
> whether or not the C++ compiler treats "C" and "C++" linkage function
> pointers the same way. If the appended program will compile with your C++
> compiler, you're probably OK:

The attached program compiles fine here. However, regarding boost.python I
thought this was more of a c++ library for handcrafting tightly integrated
bindings rather than an automatic wrapper generator? Is this understanding wrong?

Yours,

:M:



From jason at tishler.net  Mon Jun 24 12:54:45 2002
From: jason at tishler.net (Jason Tishler)
Date: Mon, 24 Jun 2002 12:54:45 -0400
Subject: Python+Cygwin+XML SAX build failure of PyXML 0.7.1
In-Reply-To: <59d8a5f6.0206221501.232d4b3b@posting.google.com>
References: <59d8a5f6.0206210804.25353aa3@posting.google.com>
 <mailman.1024711848.1696.python-list@python.org>
 <59d8a5f6.0206221501.232d4b3b@posting.google.com>
Message-ID: <20020624165445.GB1508@tishler.net>

Dmitri,

On Sat, Jun 22, 2002 at 04:01:27PM -0700, Dmitri wrote:
> > Would you like to contribute a pre-built Cygwin PyXML to the Cygwin
> > distribution?
> 
> That's possible, as long as I will be satisfied by the results.

I'm sure you will be "satisfied by the results" of contributing a
pre-built Cygwin PyXML to the Cygwin community. :,)

Or, were you referring to my suggested solution?

Jason




From whisper at oz.net  Fri Jun 21 00:27:57 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 20 Jun 2002 21:27:57 -0700
Subject: Dispatching a Win32 COM in the background?
In-Reply-To: <3D127FE0.74D47795@engcorp.com>
Message-ID: <GCEDKONBLEFPPADDJCOEKEKADKAA.whisper@oz.net>

Hmmm... actually I don't think either the queue.empty or the sleep are
useful. I believe if you profile this, it's spending all it's time blocked
on the get().

Would this do as well?:

#done gets set by the other thread.
while not done:
	# block here until queue has content
	text = self.text_queue.get()
	speech.Speak(text)

#done became true
#do clean up/exit stuff

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Peter Hansen
> Sent: Thursday, June 20, 2002 18:23
> To: python-list at python.org
> Subject: Re: Dispatching a Win32 COM in the background?
>
>
> "Steven M. Castellotti" wrote:
> >
> > Thanks a lot guys, your suggestions have worked perfectly!
> >
> >         Just for the record, here's what the full
> implementation looks like.
>
> Beautiful... thanks for the followup Steven.  I especially like
> your use of deferred imports to speed up initial program loading.
>
> [...]
> >       if not(self.text_queue.empty()):
> >          text = self.text_queue.get()
> >          speech.Speak(text)
> >          time.sleep(.001)
>
> Note that the time.sleep() as shown above is doing nothing
> useful for you.  I suspect you had it unindented but your
> posting probably used tabs and spaces which makes it hard
> to see properly on some readers.
>
> The reason it does nothing useful is that Python will
> automatically switch between different threads every so
> often anyway, so the concept of giving up the CPU for
> what amounts to a tiny fraction of a second is somewhat
> unnecessary.
>
> -Peter
> --
> http://mail.python.org/mailman/listinfo/python-list





From Chris.Barker at noaa.gov  Thu Jun 20 17:34:41 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Thu, 20 Jun 2002 14:34:41 -0700
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <aer2fn$frc$1@peabody.colorado.edu> <3D122E79.A189C72E@astro.cornell.edu> <aetapt$kld$1@peabody.colorado.edu> <3D124575.2A1393FF@astro.cornell.edu>
Message-ID: <3D124A3D.3858EE2A@noaa.gov>

Tom Loredo wrote:

> Well, there is now a recommendation each for WinXP and Win2k in this
> thread.  If anyone has tried both and can offer a comparison/recommendation,
> I'm all ears.  If I'm going to upgrade, I'd rather do it sooner than later.

This is getting kind of OT, but I'm seriously unthrilled with the
licencing/privacy/registration issues with XP, so I havn't tried it. I
also see no reason to, the only improvements I've heard about are better
integration of various (mostly MS) multi-media stuff with the desktop,
which is no use to me. I'm working on a largish Python project with two
other progammers. I'm on linux, one of them on win2k the other on XP, I
havn't see much functional difference.

If I had to use windows more, I would definately give cygwin a good
look-see. There is also at least one other package of *nix command line
tools for windows that looks like it's less of a complete environment,
but rather ports of the unix tools to windows.

-Chris



-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From eric.brunel at pragmadev.com  Mon Jun 24 06:26:18 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Mon, 24 Jun 2002 10:26:18 +0000
Subject: Newbie to HTMLgen
References: <3d16132a$0$27325$afc38c87@news.easynet.co.uk>
Message-ID: <af6knq$kbl$1@wanadoo.fr>

chris lyon wrote:

> I'm porting an app from Cold Fusion to Python, for various reasons.
> 
> And am looking at the possibility of using HTMLgen
> 
> I am trying to place the Submit button of a form into a table element
> using Tablelite
> since the Submit button is generated by the Form Class is there any way to
> put it into the table header tag,
> of a table that has rows of table data and inputs.
> 
> I can't seem to find a way to place the submit button into a table header
> cell where the client likes it.
[snip HTML code]

I used HTMLgen a few time ago and I almost didn't ever use the "auto-submit 
button" feature of the Form class. This is annoying, and as you pointed it 
out, you can't control where the button will be placed (not to mention some 
forms do not even have a submit button, since form submitting can be made 
via Javascript...).

So my advice would be: do not follow the rules in the manual ("Do not 
manually append your submit button. HTMLgen will append it for you.") and 
do:

myForm = Form(..., submit='')

then create your own submit button via the Input class with the type 
"submit". This is much easier this way...

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From peter at engcorp.com  Fri Jun 21 09:32:12 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 21 Jun 2002 09:32:12 -0400
Subject: RE strings (was: Variable Interpolation - status of PEP 215)
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <4PfQ8.44245$n4.10307683@newsc.telia.net> <KhiQ8.44269$n4.10312432@newsc.telia.net> <29e28c51.0206201156.7e61a7f5@posting.google.com> <3D127B89.EE1C17F4@engcorp.com> <aeuprv$ek$1@news1.wdf.sap-ag.de>
Message-ID: <3D132ADC.D04B52EA@engcorp.com>

Daniel Dittmar wrote:
> 
> > Cimarron Taylor wrote:
> >>   m = re'^EMP:([^,]*),([^,]*),([^,]*),([^,]*)$'.match(line)
> >>   if m:
> >>         ...
> >>
> >> is much clearer and more maintainable than
> Peter Hansen replied:
> > Given that the only real difference is you skip an import, and
> > at the cost of modifying the language definition in an unusual
> > manner, is there any point to this at all?
> 
> The implementation could choose to cache the result of re.compile at module
> level, thus
> - speeding up the 'naive' re usage
> - freeing the programmer from doing the caching in code

Any reason the re.match() method could not do the same, internally?

> I'm still against it, the road to Perl is paved with syntax extensions.

:-)

-Peter


From cjw at sympatico.ca  Fri Jun 14 09:41:31 2002
From: cjw at sympatico.ca (Colin J. Williams)
Date: Fri, 14 Jun 2002 09:41:31 -0400
Subject: PythonWin - an operational question
Message-ID: <3D09F28B.19AB5929@sympatico.ca>

An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020614/517bbcc7/attachment.html>

From llothar at web.de  Tue Jun 25 09:46:06 2002
From: llothar at web.de (Lothar Scholz)
Date: Tue, 25 Jun 2002 15:46:06 +0200
Subject: Python and Eclipse
References: <3D17FAF4.35CC9F1B@earthlink.net> <81CC01E58BF2A99B.DFCD5EC6E4654907.32D731380951B9CA@lp.airnews.net>
Message-ID: <3esghukikkd91beijpk977d8vgijctmqrn@4ax.com>

On 25 Jun 2002 07:33:14 -0500, claird at starbase.neosoft.com (Cameron
Laird) wrote:

>In article <3D17FAF4.35CC9F1B at earthlink.net>,
>Joseph A Knapka  <jknapka at earthlink.net> wrote:
>>Hi folks,
>>
>>I am interested in having Python support in
>>Eclipse <URL: http://www.eclipse.org> - interested
>>enough to contribute code, or to tackle the
>>task myself, if need be. Of course, I'd like
>>the Python support to be at least as good as the
>>Java support, which is a pretty tall order,
>>'cause the Java support in Eclipse is awfully
>>damn good --- it nearly allows me to *enjoy*
>>writing Java code :-)
>>
>>Anyway, I know some other folks have expressed
>>interest in this as well, and I'd like to hook
>>up with them. If you're working on Python+Eclipse
>>and you've got code in any state of completion,
>>I'm willing to alpha- or beta-test, fix stuff,
>>add new stuff, etc.
>			.
>			.
>			.
>Can you say a few more words about Eclipse?  It ... well, I'm
>wary.  I understand its marketing more than its technology,
>and on hostile days, I think it's *only* marketing.  Maybe you
>can make it more appealing.  "[It] nearly allows me to *enjoy*
>writing Java ..." is certainly a good start.

Yes and thats why i'm not sure about the future of Eclipse. At 
the moment its nothing else then IBM's way to force (enterprise)
developers using there Websphere Application Server - and 
in a second step sell there consultants. If they  not succeed with
this they will drop support as fast as you can  shake your head. 

I worked to much with IBM and there breakthrough technologies 
like  DSOM, OpenDoc or OS/2 to see that IBM can decide from
one day to another that a technologie is dead. 

And you can't maintain a system like Eclipse with volunteers only.
All larger OS projects showed this.

And by the way i can't see a real difference between Eclipse, JBuilder
IntelliJ, CodeGuide or Netbeans for a person who is not developing
Enterprise Java Beans. They only prove one thing about globalization:
It makes the few products that survive look like twins or better
clones.



From markus.vonehr at ipm.fhg.de  Mon Jun  3 03:03:21 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Mon, 03 Jun 2002 09:03:21 +0200
Subject: socket module and broadcasts ?
References: <3CF780D6.DFEDCFB7@ipm.fhg.de> <hrLJ8.80417$%u2.72437@atlpnn01.usenetserver.com>
Message-ID: <3CFB14B9.CF8FAA4B@ipm.fhg.de>

Hi Steve,


Steve Holden wrote:

> Do you really want your server to have top
> listen for broadcasts from potential clients?

that is exactly what I want to do.

> What if the server and the
> client are of different IP networks?

I assume that they are both in the same network, but the server is
on an embedded system and initially it can be that its IP address is
unknown, so I want to query for the server and assign an IP address.

Markus



From stefan.kirchberg at web.de  Thu Jun 20 12:29:21 2002
From: stefan.kirchberg at web.de (Stefan Kirchberg)
Date: Thu, 20 Jun 2002 18:29:21 +0200
Subject: Grep in Python
References: <f562ef59.0205210754.7b2c6737@posting.google.com> <83it5giuy2.fsf@panacea.canonical.org> <f562ef59.0205220752.5dc839de@posting.google.com>
Message-ID: <3D1202E1.2060108@web.de>

coolslife wrote:

>>>I saw that lots of people have implemented their own version of grep
>>>and most have called it "pygrep". (Am I right about that?). Since this
>>>is a resource consuming activity,  does anyone have any idea what
>>>might be a optimum way to implement this?


>>See "NR-grep: A Fast and Flexible Pattern Matching Tool", by Gonzalo
>>Navarro, at http://www.dcc.uchile.cl/~gnavarro/ps/spe01.ps.gz for how
>>to search text fast.
> 
> Thanx for your replies.

> are there any other resources I can look up?

Since grep uses Regular Expressions, you might want to have a look at 
the standard module 're'. One could try something like the following 
example. It's quite simple (because it's Python ;-)

 >>>>>>>>>>>>>>>>snip
import re, sys

# your search expression here, maybe
pat = sys.argv[1]

# your input text filename, maybe
fname = sys.argv[2]

# now scan the textfile
for line in open("C:/temp/test.txt").readlines():
   if re.search(pat, line):
     print line,

# that's it.
<<<<<<<<<<<<<<<<snap

Hope this helps. Sure it's not perfect... ;-)

   Stefan



From gh_pythonlist at gmx.de  Sun Jun  2 23:09:43 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Mon, 3 Jun 2002 05:09:43 +0200
Subject: cursors, etc
In-Reply-To: <20020602221459.05355.00002083@mb-mm.aol.com>
References: <20020602221459.05355.00002083@mb-mm.aol.com>
Message-ID: <20020603050943.A65224@gargamel.hqd-internal>

* BCBOOKER <bcbooker at aol.com> [2002-06-03 02:14 +0000]:
> Is there any way for a user to actively move a cursor or some other marker on a
> plane?  That is, if there is a 4x4 square with the cursor at (1,1), can the
> user press up, n, 1 or whatever and make the x increase?  I tried making
> something along the lines of this, but i had to go type in every situation -
> that is, make the map for (2, 3) or (1, 2).
> 
> I'm a newbie to programming as a whole, so please excuse any nievete that
> this message may exude...

You don't need to hard-code all possibilities. I'd suggest an approach like the
following Python pseudo-code:

def valid_position(coordinates):
    """Coordinates is a tuple of (x,y) coordinates.
    This function returns 1 if these are valid coordinates, 0 otherwise."""
    # Fill in code here ...

coordinates = (0, 0)
while 1:
    # Assume we have a function get_key() to get the a key from keyboard
    # and constants for the various cursor keys.
    x, y = coordinates
    key = get_key()
    if key == CURSOR_UP:
        y -= 1
    elif key == CURSOR_DOWN:
        y += 1
    elif key == CURSOR_LEFT:
        x -= 1
    elif key == CURSOR_RIGHT:
        x += 1
    else:
        # handle other keys here
    
    new_coordinates = (x, y)
    if valid_position(new_coordinates):
        coordinates = new_coordinates
    else:
        # Beep, display error message or something similar.


HTH,

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 10.8 ?C      Wind: 2.1 m/s




From gerhard at bigfoot.de  Fri Jun 21 10:30:17 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 21 Jun 2002 14:30:17 GMT
Subject: I'd give up Perl tomorrow if only...
References: <mailman.1024667909.336.python-list@python.org>
Message-ID: <slrnah6dv9.154.gerhard@lilith.my-fqdn.de>

Michael Chermside wrote:
> Where can I find the pyperl module?

Google - Enter "pyperl" - then hit "I feel lucky" :-)

This will bring you to
http://www.cpan.org/modules/by-module/LWP/GAAS/pyperl-1.0.readme which looks to
a non-Perl guy like me that its available as a CPAN module.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From pyth at devel.trillke.net  Sat Jun 29 17:11:30 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 29 Jun 2002 23:11:30 +0200
Subject: Synchrinized chat?
In-Reply-To: <3D16E40D.500A5DD4@gol.ge>; from gleki@gol.ge on Mon, Jun 24, 2002 at 11:19:09AM +0200
References: <3D16E40D.500A5DD4@gol.ge>
Message-ID: <20020629231130.A20310@prim.han.de>

Giorgi Lekishvili wrote:
> Hi all!
> 
> Is somebody making a synchronized chat in Python? I don't need the
> script of the chat itself, rather a learning example of a real-world
> synchronized (multi)threading.

Would you mind to rephrase your questions? What is a "synchronized
chat", a "script of the chat" and what do you think you need
multithreading for? 

For a start you might simply describe in non-programming terms 
what you want to achieve.

    holger




From charter-news at portzer0.virtualave.net  Wed Jun 19 18:58:33 2002
From: charter-news at portzer0.virtualave.net (Locke)
Date: Wed, 19 Jun 2002 18:58:33 -0400
Subject: using os.chmod in windows
Message-ID: <3D110C99.5090607@portzer0.virtualave.net>

I never really learned the number codes for file permissions in UNIX. I 
always just did "chmod +r" because it is much easier to remember. I need 
to change a file to be 'hidden' in windows. In UNIX, this is done by 
renaming it with a dot in front of the name. But the python os.chmod 
command says you need to specify the number as the argument. As far as I 
know, there is no number for 'hidden' since it is windows only. So how 
do I do this?

Thanks.



From Chris.Barker at noaa.gov  Mon Jun 17 16:55:43 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Mon, 17 Jun 2002 13:55:43 -0700
Subject: NEWbie asks "is Python a good tool for automated file/folder 
 generation?"
References: <6k6kgukhuk8pfgos97a38h0d4r5jr0u3sr@4ax.com>
Message-ID: <3D0E4B81.2B81D9B7@noaa.gov>

Unconditional Grace wrote:

> What I would like to do is largely string-based file and folder
> manipulation:

you've already gotten a couple of suggestions for some partsw of your project.

> 3.  Within each folder, create one HTML page consisting of a table
> with a link to each image, preferably with a thumbnail (can be
> original image resized by HTML since this is all local = no bandwidth
> issues).

I can do you one better. Following is a scripit I wrote that does
something similar. It takes a set of directories with jpegs in them, and
for each one creates an html page with a set of thumbnails of the
images, and links to the images. I used the Python Imaging Library (PIL)
to make the thumbnails, and HTMLGen to help generate the HTML. You can
find both of these packages by doing a google search on the web. There
are probably a lot of other options for creating htm thatn HTMLgen, I
havn't looked recently. You will find that one of the stregths of Python
is the available modules, both the "standard" modules, and a huge
assortment of others. Use them. If you are doing something that seems
like a common task, there is probably a module out there to help you.
 
> 4.  Create a top level HTML page that links to all 100 image HTML
> pages created in step 3.  Can be text only.

The top level page links to the thumbnail pages, but it could just as
easily have a full set of links.
 
> 5.  Optional:  in step 3, make each image part of an HTML page with
> "previous image" and "next image" links, rather than just being called
> by the browser as a jpg image.

Exercise left for the reader.

> I've never used Python, and I'm wondering if you, collectively, think
> that Python is a good language to do this.

It's a fabulous tool for this sort of thing.

> While it seems like Python
> is an incredibly useful tool, I'm not sure I have time right now to
> invest in learning it if it can't do this.

It wouldn't be such a useful tool if it couldn't do this!! It is well
worth taking the time to learn.

> It seems I frequently want
> to do repetitive Windows file/folder chores like this, so I want to
> get smart-like.

Good idea. In this case, you are manipulating files/folder, building
html, and doing image manipulation (if you want the thumbnails) Python
is an excellent tool for all these tasks.

> I'm not asking anyone to write such code for me, heh heh,

sorry, I already did...

good luck,

-Chris

The Code: note that it was written in Python 1.5, a couple of years ago.
If I were to re-write it now, I would make some changes. It needs PIL
and HTMgen

#!/usr/bin/python

# This is a Python script that builds a bunch of web pages out of a set
# of directories with pictures in them

import Image, glob, os
import HTMLgen
HT = HTMLgen

main_name = "index.html"

num_columns = 4 # number of columns for thumbnail table

maindir = os.getcwd()
maindirname = os.path.split(maindir)[1]

maindoc = HT.SimpleDocument(title=maindirname)
maindoc.append(HT.Heading(1,maindirname))

dirlist = filter(lambda name: os.path.isdir(name),os.listdir(os.getcwd()))

for dirname in dirlist:
        os.chdir(os.path.join(maindir,dirname))
        doc = HT.SimpleDocument(title=dirname)
        doc.append(HT.Heading(1,dirname))
        
        table = HT.Table('Click on the small images to see them full size')
        table.width = ''
        doc.append(table)
        
        table.body = []
        
        count = 0
        picts = glob.glob("*.jpg")
        picts = filter(lambda name:(not (name[-10:] == '.thumb.jpg')),picts)
        while count < len(picts):
                row = ['']*num_columns
                for i in range(num_columns):
                        count = count+1
                        if count > len(picts):break
                        pict = picts[count-1]
                        im = Image.open(pict)
                        im.thumbnail((128,128))
                        im.save((pict[:-3]+'thumb.jpg'),'JPEG')
                        row[i] = (HT.Href(pict,HT.Image(pict[:-3]+'thumb.jpg')))
                table.body.append(row)
                                  
        doc.append(HT.BR())

        doc.append(HT.Href(os.path.join(os.pardir,"index.html"),"Back to Main
Index"))
        
        doc.write("index.html")

        maindoc.append(HT.BR()) 
        maindoc.append(HT.Heading(2,HT.Href(os.path.join(dirname,"index.html"),dirname)))

os.chdir(maindir)
maindoc.append(HT.BR()) 
maindoc.append(HT.Href("http://www.python.org",HT.Image("PythonPoweredAnim.gif")))
maindoc.write('index.html')



-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From fredrik at pythonware.com  Thu Jun 27 17:47:49 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 27 Jun 2002 21:47:49 GMT
Subject: How to find out operating system
References: <mailman.1025210498.607.python-list@python.org>
Message-ID: <9KLS8.45365$n4.10750643@newsc.telia.net>

David LeBlanc wrote

> > > > You might try os.environ:
> > > > os.environ['WINOS'] -> WIN2000
> > > >
> > > > I don't know if WINOS is defined on '98 or ME, but
> > > > it should be good for NT, 2K and XP
> > >
> > > It is not.
> >
> > Neither is it defined on win2k pro:
> > KeyError: WINOS

> >>> import os
> >>> os.environ['WINOS']
> 'WIN2000'

on the other hand:

c:\> ver
Microsoft Windows 2000 [Version 5.00.2195]

(swedish version, all available service packs and
security patches)

c:\> python
Python 2.1.1 (#20, Aug 23 2001, 11:27:17) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
>>> import os
>>> os.environ["WINOS"]
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "C:\py21\Lib\os.py", line 372, in __getitem__
    return self.data[key.upper()]
KeyError: WINOS

(fwiw, googling for WINOS brought up a whole bunch of
pages telling me that if I want to build mozilla, I need to
set the WINOS variable.  are you sure your computer
doesn't spend the nights building browsers, when you're
not looking?)

</F>




From sandysj at juno.com  Fri Jun 21 12:34:55 2002
From: sandysj at juno.com (Jeff Sandys)
Date: Fri, 21 Jun 2002 16:34:55 GMT
Subject: newbie string question
References: <aeu40s$9jg2s$1@ID-145503.news.dfncis.de>
Message-ID: <3D1355AF.901DAE68@juno.com>

Other good replies already.
My metaphor: Fence posts and panels.
(A 50 foot fence using 10 foot panels has 6 fence posts and 5 panels)
The fence post problem is common in computational math.

Don Low wrote:
> 
> There's probably a very simple explanation, 
> but I can't see it for now, so bear with me.
> 
> >>>pystr = 'Python'
> >>>pystr = [5]
> 'n'
> >>>pystr [2:5]
> 'tho'
> 
> The question is, why isn't pystr [2:5] 'thon' instead of 'tho'. 
> Could somebody explain?


From dyoo at hkn.eecs.berkeley.edu  Fri Jun 28 20:38:00 2002
From: dyoo at hkn.eecs.berkeley.edu (Daniel Yoo)
Date: Sat, 29 Jun 2002 00:38:00 +0000 (UTC)
Subject: Why the colon?
References: <mailman.1025306860.24026.python-list@python.org>
Message-ID: <afivh8$1t35$1@agate.berkeley.edu>

Erv Young <res04ft9 at gte.net> wrote:

: 1. def ask_ok(prompt, retries=4, complaint='Yes or no, please!'):
: 2.     while 1:
: 3.         ok = raw_input(prompt)
: 4.         if ok in ('y', 'ye', 'yes'): return 1
: 5.         if ok in ('n', 'no', 'nop', 'nope'): return 0
: 6.         retries = retries - 1
: 7.         if retries < 0: raise IOError, 'refusenik user'
: 8.         print complaint

: where I have added line numbers for the sake of the discussion.  I
: understand the colons in the middle of lines 4, 5, and 7 as statement
: separators.  Makes perfect sense.

Hello!

Colons are really used for the beginning of blocks.  In lines 4, 5,
and 7, the example is using a shortcut --- if the block is just one
line long, it can be placed on the same line for brevity.  The code
could have been reformatted as:

###
def ask_ok(prompt, retries=4, complaint='Yes or no, please!'):
     while 1:
         ok = raw_input(prompt)
         if ok in ('y', 'ye', 'yes'):
             return 1
         if ok in ('n', 'no', 'nop', 'nope'):
             return 0
         retries = retries - 1
         if retries < 0:
             raise IOError, 'refusenik user'
         print complaint
###

which should be treated equivalently.


Note that the one-liner shortcut doesn't work for more than one level
of block nesting, to avoid language abuse.  That is, something like:

###
>>> try: if 42 == 42: print "ok"
  File "<stdin>", line 1
    try: if 42 == 42: print "ok"
          ^
SyntaxError: invalid syntax
###

is immediately flagged as a SyntaxError.  Thank goodness.  *grin*


If you'd like to delve into the reference material about this, you can
look at:

    http://www.python.org/doc/ref/compound.html


Hope this helps!



From fredrik at pythonware.com  Thu Jun 13 06:37:57 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 13 Jun 2002 10:37:57 GMT
Subject: newbie question - python blocking on int()?
References: <4LaN8.2488$Vr2.586019@news20.bellglobal.com> <Xns922A149F19723cliechtigmxnet@62.2.16.82> <pan.2002.06.11.03.00.52.273636.25803@fuckspam.hetdigitalegat.nl>
Message-ID: <9C_N8.11931$p56.3660777@newsb.telia.net>

"Wladimir" wrote:

> Parsing the output of ftp.dir in any way is never comphrehensively safe
> because it it platform dependent according to the FTP RFC. Spaces in
> usernames are not possible though AFAIK, and most servers use a unix ls
> like semantic for dirs so it's *pretty safe*. (although I've seen ugly DOS
> like listings as well)

Dan Bernstein's ftpparse library can parse the output from
most FTP servers.  you can find a wrapper here:

    http://effbot.org/downloads/

(look for ftpparse)

</F>




From whisper at oz.net  Mon Jun 17 17:35:57 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 17 Jun 2002 14:35:57 -0700
Subject: How to distribute py-scripts on windows without requiring the full ActivaState package
In-Reply-To: <2e0f5ffe.0206171151.4305c306@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEEELJDJAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of DV
> Sent: Monday, June 17, 2002 12:51
> To: python-list at python.org
> Subject: How to distribute py-scripts on windows without requiring the
> full ActivaState package
>
>
> Hello,
>
> A couple of years ago, I experimented with Python on Linux, but
> somehow got out of touch.
>
> Just recently I discovered, that there is a pretty comprehensive
> Windows-Version for it that even supports pure windows-stuff such as
> COM or Win32API.
>
> Yet, I found out that the ActiveState package is very large to
> download and install, so this leads me to the question(s):
>
> 1)Is there some way to distribute a script foo.py by maybe bundling it
> with python.exe and python.dll, delivering these three files ONLY (and
> of course any .py library files that foo.py calls) and not requiring
> the recipient to download and install the whole ActiveState package?

The "standard" Python distribution lives at www.python.org, and doesn't
include the PythonWin Win32 extensions. It's smaller.

There are 3rd party packages like Py2Exe that bundle python, required
libraries and scripts into a single .exe file. You can find out a lot more
about distributing Python apps at http://www.mcmillan-inc.com/install1.html

> 2)Are there different versions of python.exe and python.dll? What I
> would find useful was, if there was ONE version of
> python.exe&python.dll that only could be used for console stuff, so
> you wouldn?t have such big files to distribute along with foo.py, and
> another one for true GUI-stuff

Python.exe is used for console mode programs and Pythonw.exe is used for GUI
applications that don't need a console window. Python.exe can also be used
for GUI programs, but then you get a console window in addition to your GUI
window(s).

> 3)Concerning COM and other heftier Win32 dependend stuff, or anything
> that python needs a third-party dll for to use: Can the necessary dll
> be bundled along with python.exe, python.dll and foo.py, just as
> mentioned in 1)?

Yes

> Mind, I would?t want to distribute stuff as a single EXE. I could live
> with having to distribute more than one file, but I would very much
> like it, if an end user receiving my script would?t have to install
> the whole ActiveState package.

If you don't mind distributing a .zip file containing your app, distutils
(part of the Python distribution) can help you create installation scripts.
There are other alternatives like the Wise Installer ($) and InoSetup
(free).

> I could imagine that one weak point about this is, that i have read
> somewhere that dlls must be registered before they are usable, but I
> also read that this is possible by using rundll.exe or some other
> Windows system program manually. If so, the registering process could
> be done by a .BAT file.

Only dlls containing COM components need to be pre-registered. A regular dll
only needs to be found in the path or PYTHONPATH, or in the directory
containing the .exe file that's trying to use it.

> As for the harddisk-space penalty of havin more than one python.exe
> and python.dll (one for each distributed script), this is the reason
> why I asked, whether a console version exists: If it did exist, you
> could do the GUI by designing a HTML-page that drives it.

Python is independent of it's GUI components. If you don't use them, you
don't have to distribute them. If you do, then you'll have a larger .zip or
.exe file. If you know you're going to use several scripts, then it makes
sense to just download and install a standard distribution of Python that
all the scripts can share.

> Sorry, if this posting appears twice under different topics, if so, it
> is most likely a temporary fault of deja.com, through which I posted
> this.
>
> I would appreciate any info on this matter.
>
> Thank you in advance,
> Bye.
> Franz

Dave LeBlanc
Seattle, WA USA





From martin at strakt.com  Thu Jun  6 02:37:59 2002
From: martin at strakt.com (Martin =?iso-8859-1?Q?Sj=F6gren?=)
Date: Thu, 6 Jun 2002 08:37:59 +0200
Subject: prototyping good OOdesign in Python?
In-Reply-To: <20020605203051.GC1455@isis.gerrietts.net>
References: <XFMail.20020601131915.shalehperry@attbi.com> <Pine.LNX.4.33.0206030725420.5660-100000@rnd.onego.ru> <20020605203051.GC1455@isis.gerrietts.net>
Message-ID: <20020606063759.GA28596@strakt.com>

On Wed, Jun 05, 2002 at 01:30:51PM -0700, Geoff Gerrietts wrote:

> Since Python doesn't care about most of this stuff, modelling it is
> only useful if you expect to port the code later. And I prefer not to
> plan for an unfortunate event. ;)

I still it is good to think in terms of 'public' and 'private', even
though the names are badly chosen. I prefer 'interface' and
'implementation'. If you don't decide which of your class members are to
be part of the interface and which are implementation-specific, I feel
that is a good way to shoot yourself in the foot.


Martin

-- 
Martin Sj?gren
  martin at strakt.com              ICQ : 41245059
  Phone: +46 (0)31 7710870       Cell: +46 (0)739 169191
  GPG key: http://www.strakt.com/~martin/gpg.html




From mcherm at destiny.com  Fri Jun 21 10:00:24 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Fri, 21 Jun 2002 10:00:24 -0400
Subject: I'd give up Perl tomorrow if only...
Message-ID: <3D133178.5000105@destiny.com>

Where can I find the pyperl module?

-- Michael Chermside





From bokr at oz.net  Mon Jun 24 13:24:51 2002
From: bokr at oz.net (Bengt Richter)
Date: 24 Jun 2002 17:24:51 GMT
Subject: Subclass of array
References: <3d1714c3_7@news.newsgroups.com>
Message-ID: <af7kl3$743$0@216.39.172.122>

On Mon, 24 Jun 2002 14:54:21 +0200, JB <jblazi at hotmail.com> wrote:

>Can I derive an own class from array? (I use 2.2.)

Good thought ;-) It doesn't look like it, but if it can be done, I'd like
to know too. You can make a class that wraps instead of inherits, though.
There are lots of ways to do that. See one limited way in another reply
of mine to you in the other thread (which I did because subclassing the
obvious way didn't work ;-).

Regards,
Bengt Richter


From mario at kyb291.biologie.uni-bielefeld.de  Mon Jun  3 10:19:33 2002
From: mario at kyb291.biologie.uni-bielefeld.de (Mario Wehbrink)
Date: 3 Jun 2002 10:19:33 MET
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org>
Message-ID: <3cfb2695$1_1@news.uni-bielefeld.de>

Am  Sat, 01 Jun 2002 14:38:46 GMT  schrieb/wrote/  
Andrei Kulakov  < ak at silmarill.org > in comp.lang.python : 
 
> Hello,
> 
>     "hello world" type program from wxPython tutorial takes 6
> seconds to start on my cel 366 128mb system (running Debian). Is
> this typical? Why is it so slow? I'm looking for a graphic
> toolkit that'd be fairly quick. I don't want to use Tk because of

Hi,

try pygtk. I also switched to it, after i found wxPython too slow for my
PPro-200.

http://www.daa.com.au/~james/pygtk/


Mario

-- 
http://www.porcupinetree.com


From radix at twistedmatrix.com  Wed Jun 12 09:57:04 2002
From: radix at twistedmatrix.com (Christopher Armstrong)
Date: 12 Jun 2002 09:57:04 -0400
Subject: Bits and bytes?
In-Reply-To: <3d074e6d$0$224$4d4ebb8e@news.nl.uu.net>
References: <3d074e6d$0$224$4d4ebb8e@news.nl.uu.net>
Message-ID: <87elfcvbv3.fsf@twistedmatrix.com>

>>>>> "Guyon" == Guyon Mor?e <gumuz at looze.net> writes:

    Guyon> ok, this might be a strange question.... but i was wondering: how
    Guyon> can i read or write 'bits'?

    Guyon> if i have a string, how can i convert to it's basic 0's and 1's?

    Guyon> and offcourse, the other way around.... let's say i want to store
    Guyon> some bit-sequence(0011101110) to a file.


    Guyon> how do i do this?

http://twistedmatrix.com/users/moshez/binary.py

Moshe Zadka wrote this because he was annoyed at all the people in #python 
asking how to do this ;-)


-- 
                                Chris Armstrong
                         << radix at twistedmatrix.com >>
                http://twistedmatrix.com/users/carmstro.twistd/





From paul at fxtech.com  Mon Jun 24 12:55:04 2002
From: paul at fxtech.com (Paul Miller)
Date: Mon, 24 Jun 2002 16:55:04 GMT
Subject: embedded Python - optimization for creating/destroying many small 
 objects
Message-ID: <3D174EEA.EFAE79C@fxtech.com>

I have Python embedded and have some code that create thousands of small
Python extension objects. But my usage for these is basically to pass
some data to a Python callback. Essentially, I have a loop, and I create
my Python object (with data from a C object), call the Python function
(the callback), then decref/destroy the object.

I would like to simply create ONE of these objects and reuse it for the
entire loop, so I am not paying for the memory allocation/deallocation
overhead.

Is there a prescribed way of doing stuff like this for extension
objects?


From phr-n2002b at NOSPAMnightsong.com  Wed Jun 26 21:48:53 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 26 Jun 2002 18:48:53 -0700
Subject: Pythoniac: Thoughts on a hardware Python processor
References: <afc52t$2pj$1@sirius.dur.ac.uk>
Message-ID: <7xeletjxtm.fsf@ruckus.brouhaha.com>

I don't think it makes sense to stick with existing Python bytecodes
for hardware execution.  Take a look at the work done on Lisp machines
in the 70's and 80's instead.  


From maxm at mxm.dk  Sun Jun 23 06:22:43 2002
From: maxm at mxm.dk (Max M)
Date: Sun, 23 Jun 2002 12:22:43 +0200
Subject: bloody freshman question regarding strings
References: <mailman.1024821149.18077.python-list@python.org>
Message-ID: <3D15A173.9010904@mxm.dk>

Christian Guenther wrote:

> My biggest question at the moment is why can't I do something like this:
> 
> def mailmsg(severity message opts target recipient):
> 	# first create the message body
> 	msgbody="computer: " + COMPUTER + "\n
> 	         system:   " + SYSTEMTYPE + "\n
> 		 cpu:      " + CPUTYPE + "\n
> 		 kernel:   " + KERNELVER + "\n
> 		 date:     " + DATE + "\n
> 		 \n
> 		 severity: " + severity + "\n
> 		 message:  " + message + "\n\n"
> 
> 
> I want to generate messages - to concole and to mail-recipients - that
> show the status of user-processing. This message should include all the
> information you see above - they are stored in global vars - a severity
> and of course a message itself. Let alone, it just doesn't work and I
> know I do this in a nodo way, I just don't have alot of time left ;-)


A few approaches:

def mailmsg(severity, message, opts, target, recipient):
     # first create the message body
     msgbody = "computer: %s\n" % COMPUTER)
     "system:   %s\n" % SYSTEMTYPE)
     "cpu:      %s\n" % CPUTYPE)
     "kernel:   %s\n" % KERNELVER)
     "date:     %s\n\n" % DATE)

     "severity: %s\n" % severity)
     "message:  %s\n\n" % message


def mailmsg(severity, message, opts, target, recipient):
     # first create the message body
     bodyParts = []
     a = body.append
     a("computer: %s\n" % COMPUTER)
     a("system:   %s\n" % SYSTEMTYPE)
     a("cpu:      %s\n" % CPUTYPE)
     a("kernel:   %s\n" % KERNELVER)
     a("date:     %s\n\n" % DATE)
     a("severity: %s\n" % severity)
     a("message:  %s\n\n" % message
     msgbody = ''.join(bodyParts)


def mailmsg(severity, message, opts, target, recipient):
     # first create the message body
     msgbody = """computer: %(COMPUTER)s
system:   %(SYSTEMTYPE)s
cpu:      %(CPUTYPE)s
kernel:   %(KERNELVER)s
date:     %(DATE)s

severity: %(severity)s
message:  %(message)s

""" % vars()

regards Max M



From rjones at ekit-inc.com  Sun Jun 30 21:13:08 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Mon, 1 Jul 2002 11:13:08 +1000
Subject: I'd give up Perl tomorrow if only...
In-Reply-To: <afo9rh$jlb$1@panix1.panix.com>
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au> <Pine.LNX.4.44.0206301756130.22298-100000@palanthas.neverending.org> <afo9rh$jlb$1@panix1.panix.com>
Message-ID: <200207011113.08851.rjones@ekit-inc.com>

On Mon, 1 Jul 2002 11:04, Aahz wrote:
> Frank Tobin  <ftobin at neverending.org> wrote:
> >I suggest that you be a Perl programmer for a while

Gah, never again! :)


> But somehow one doesn't hear it very often from people who've been using
> Python for a long time -- and certainly nobody has truly stepped up to
> the plate to put all the finishing details into place.  Sean
> Reifschneider came pretty close, then dropped the ball.  Where's the
> itch-scratching that induces people to do the work?

I recently stepped up to the plate, but got zero response from catalog-sig. 
I'll be contacting authors of the various half-finished systems directly if I 
get no response soon.


> You might be able to get some traction if you concentrate on just the
> testing issue (and possibly the README issue).  That's the kind of thing
> that gets at least public verbal support from Guido.

Yep. Generally, it requires someone to just actually _write_ the PEP, submit 
it, put up with the flamage long enough to get a consensus (or something very 
close to) and finally the Python community will have a spec they can call 
their own. I won't mind being that person - I just need to determine the 
current state of the three (or so) efforts out there...


   Richard





From egbert at bork.demon.nl  Fri Jun 14 09:36:40 2002
From: egbert at bork.demon.nl (Egbert Bouwman)
Date: Fri, 14 Jun 2002 15:36:40 +0200
Subject: Python GUI
In-Reply-To: <9ddd154e.0206122021.1b06f89a@posting.google.com>; from safarzadeh@hotsheet.com on Wed, Jun 12, 2002 at 09:21:09PM -0700
References: <9ddd154e.0206122021.1b06f89a@posting.google.com>
Message-ID: <20020614153640.A1863@bork.demon.nl>

On Wed, Jun 12, 2002 at 09:21:09PM -0700, H. Safarzadeh wrote:

> -pyGTK and pyQt: I do not know enough about them. But I know that I do
> not like GTK look on Windows [...]

Does PtGTK have any other drawbacks ?
egbert
-- 
Egbert Bouwman - Keizersgracht 197 II - 1016 DS  Amsterdam - 020 6257991
========================================================================




From noone at here.invalid  Thu Jun 20 23:04:49 2002
From: noone at here.invalid (JohnJacob)
Date: Thu, 20 Jun 2002 22:04:49 -0500
Subject: newbie string question
References: <aeu40s$9jg2s$1@ID-145503.news.dfncis.de>
Message-ID: <aeu54i0lp3@enews1.newsguy.com>

"Don Low" <mt at open2web.com> wrote in message
news:aeu40s$9jg2s$1 at ID-145503.news.dfncis.de...
> There's probably a very simple explanation, but I can't see it for now, so
> bear with me.
>
> >>>pystr = 'Python'
> >>>pystr = [5]
> 'n'
> >>>pystr [2:5]
> 'tho'
>
> The question is, why isn't pystr [2:5] 'thon' instead of 'tho'. Could
> somebody explain?
>
> --
> Thanks,
>
> Mark

It's sort of analogous to time: from 2:00 to 5:00 is 3 hours. And the 5th
(zero based) hour of the day is from 5:00 to 6:00

0   1   2   3   4   5   6   7      o'clock (slicing)
| 0 | 1 | 2 | 3 | 4 | 5 | 6 |      counting the hours (indexing)
  P   y   t   h   o   n

greg




From iwk_spamisanono at nospam_xs4all.nl  Tue Jun 11 17:16:10 2002
From: iwk_spamisanono at nospam_xs4all.nl (iwk)
Date: Tue, 11 Jun 2002 23:16:10 +0200
Subject: Installing PyQt on Mandrake 8.2
Message-ID: <3D06689A.9040807@nospam_xs4all.nl>

Hi there,

I need to install PyQt on Mandrake 8.2 but I've been running into some 
trouble. PyQt doesn't seem be included in the PowerPack edition which I 
bought so I downloaded the RPM from the contributions directory on a 
mirrorsite but that one failed because it was linked to an older version 
of SIP. Then I tried to install from source but ./configure failed 
because it couldn find the Python SIP modules. Apparently these ones 
aren 't included with Mdk8.2 either.

Has anyone been more succesfull? What should be the proper steps to get 
PyQt installed on Mdk 8.2?

Regards,

Iwan



From cliechti at gmx.net  Mon Jun 24 17:25:51 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 24 Jun 2002 23:25:51 +0200
Subject: Threads on embedded Python
References: <1024952585.111238@seux119>
Message-ID: <Xns9237EECAA9CC5cliechtigmxnet@62.2.16.82>

"Ugo Garc?a" <plugin at supercable.es> wrote in
news:1024952585.111238 at seux119: 

> Hi all. I have an applicattion who uses Python to control some
> processes. I've made a little script library to work with threads and
> each proccess is a Python thread. The main application make calls to
> Python in order to create these proccesses. The problem is that the
> proccess doesn't run unless I call Python to do something. Now in the
> main loop I have a line like this: 
>     PyRun_SimpleString("dummy=1");
> 
> This make Python the threads to run. The problem is that this is a
> little slow 'cause if I call Python from tha main application too many
> times the proccesses run well but not the main application, and if a
> make a big pause between calls the proccesses doesn't work at a good
> speed. Is posible to have some threads running in an embedded Python
> without making callings to it continuosly? It would be a better idea
> to have another thread who's only task is to call Python (a line that
> the one written before) in order to have the Python proccesses
> running? 

why not run python in a separate thread (created by your app) and doing the 
loop in python, like PyRun_SimpleString("while 1: mainloop()") or so?
you do have a main loop in python, don't you? otherwise you wouldn't need 
python threads, right? :-) 

you could possibly also release the GIL, but don't forget to grab it again 
when accessing the python interpreter. i'm not sure about this one, so i 
would try the above.

chris

-- 
Chris <cliechti at gmx.net>



From rotundo52 at hotmail.com  Wed Jun 19 23:13:44 2002
From: rotundo52 at hotmail.com (Jon J. Morin)
Date: Thu, 20 Jun 2002 03:13:44 GMT
Subject: working with secure shell
Message-ID: <ILbQ8.1749$CZ4.555208@news1.news.adelphia.net>

One question: Is there a python library or module for working with the 
secure shell (ssh)?

Jon J. Morin


From markus.vonehr at ipm.fhg.de  Thu Jun 20 07:35:51 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Thu, 20 Jun 2002 13:35:51 +0200
Subject: Advice sought on ps module
References: <6e990e29.0206200325.355f0413@posting.google.com>
Message-ID: <3D11BE17.C6AE67A@ipm.fhg.de>

Hi,

do you know the PSDraw Module in the PIL?
http://www.pythonware.com/library/pil/handbook/psdraw.htm
Maybe this is a solution for you.

Markus


Hans-Joachim Widmaier schrieb:

> About 2 years ago I wrote a primitive PostScript module to help me
> create various things like CD inlays or diagrams. It served that
> purpose quite well, despite its primitiveness.
> Since long I wanted to rewrite the whole thing from scratch and do it
> (at least kind of) "right." But I'm still not so sure as which way to
> go. At one time I thought I might employ ghostscript, which would have
> enabled me to, e. g. get the string width for some text, so I can make
> adjustments (not only to the text). Apart from that I couldn't get a
> sensible communication with gs through popen2, the PostScript output
> is created somewhat nonlinear, making it harder to ensure that the gs
> context is correct (in that query case).
> I'm convinced that's the wrong way, but the alternative doesn't look
> too good either. If you want to do some optimizations, you have to
> double some functionality of the ps interpreter.
>
> Examples:
>   - To calculate the with of a string, you have to read the AFM file,
> add the character widths, aplly kerning, transform it with the CTM and
> the fonts TM.
>   - To know whether a font must be reset or is already the current
> font you have to keep track of all the save/restore, gsave/grestore
> and the like.
>
> I do not want to write a "do all, end all" PostScript module. I do not
> even want to give it away (I'd be simply too embarrassed).
>
> Still I hope that some of you might give me some good advice on how to
> best do it.
>
> Thanks for any suggestion!
> Hans-Joachim



From omega at temple-baptist.com  Mon Jun  3 00:28:24 2002
From: omega at temple-baptist.com (Erik Walthinsen)
Date: Sun, 02 Jun 2002 21:28:24 -0700
Subject: Problems with exec scoping (solved)
References: <adep6r01ntg@enews1.newsguy.com>
Message-ID: <aderf60vsg@enews2.newsguy.com>

On Sun, 02 Jun 2002 20:49:54 -0700, Erik Walthinsen wrote:


> The problem I'm running into now is that if I print out the repr() of
> the 'objects' variable from within the python in the HTML page, I get a
> dict with the right keys, but the values of each of the keys are
> non-existant. This obviously puts a bit of a crimp in my plans, since
> the whole point of making the dict available is to use the objects
> contained.

Well, it turns out that I was being an idiot and not checking the source
of the returned HTML page.  When repr() is run on an object, it returns a
string such as:

<captiveclient.CaptiveClient instance at 0x8189264>

...which conveniently doesn't get displayed as text in a browser, looking
as much like HTML as it does.  I discovered this as I wrote a __repr__()
function for the class and had large chunks of the page disappear because
I didnt stick a trailing ">" on the returned string.  Switching to "[]"
solved the problem.

I will now go and see about writing a HTML-textizer for use in such
situations.

      Erik Walthinsen <omega at temple-baptist.com> - System Administrator
        __
       /  \                GStreamer - The only way to stream!
      |    | M E G A        ***** http://gstreamer.net/ *****
      _\  /_


From calves at coelce.com.br  Mon Jun 17 13:37:07 2002
From: calves at coelce.com.br (Alves, Carlos Alberto - Coelce)
Date: Mon, 17 Jun 2002 14:37:07 -0300
Subject: C or gcc lists
Message-ID: <29A97D00F387D411AC7900902770E148058700E1@lcoeexc01.coelce.net>

Does anybody can indicate me lists of C or gcc discusions?!

Carlos Alberto
COELCE/DPRON-Departamento de Projetos e Obras Norte
Fone: 677- 2228
e-mail: calves at coelce.com.br
\|||/
(o o)
--ooo0-(_)-0ooo--

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020617/c978fda2/attachment.html>

From fperez528 at yahoo.com  Thu Jun 13 22:02:59 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 13 Jun 2002 20:02:59 -0600
Subject: What language - platform use to data acquisiton/numerical work?
References: <1c849ea9.0206130130.6cdaa36f@posting.google.com>
Message-ID: <aebism$rht$1@peabody.colorado.edu>

Matteo wrote:

> The data analisys numerical procedures have to be done in matlab due
> to their complexity and specialization, and I fear even the plotting
> will have to be done in Matlab.

Absolutely false. Python will be just fine for this, coupled with anything 
from good ole gnuplot all the way to something like mayavi/vtk (with grace, 
dislin and many other options in between). Here are some examples:

http://www-hep.colorado.edu/~fperez/pub/topo_2412.b585m008.36.1000.slices/
http://www-hep.colorado.edu/~fperez/pub/topo_viz_sample.jpg
http://www-hep.colorado.edu/~fperez/pub/topo_vol_ray.jpg

Those are 3-d slices of 4-d datasets, with either multiple isosurfaces and cut 
planes with contour colormaps computed, or full-blown raytraced volume 
renderings. Those can be either sent automatically to html pages in 10 lines 
of code (try that in Matlab!) or can be manipulated interactively with mouse 
control of rotation, position and zoom level.

Believe me, unless you need to rely on a large amount of code which is 
_already_ written in matlab, you'll do just fine with python/Numeric/Scipy. 
Using weave (from scipy) inlining C for speed is a trivial exercise too. You 
can find some material on that topic here:
http://www-hep.colorado.edu/~fperez/python/python-c/

I know people who are doing multi-electron atom structure calculations in 
supercomputers using python, so I'm sure your problem can also be handled in 
python.

Good luck,

f.


From sameer_ at email.com  Fri Jun  7 11:44:25 2002
From: sameer_ at email.com (sameer)
Date: 7 Jun 2002 08:44:25 -0700
Subject: email.Message question.  including email attachments
Message-ID: <6bd9f01b.0206070744.64cb30c0@posting.google.com>

I am trying to use the new (since Python 2.2) email module to send
email with a pdf attachment.  I have tried various ways to send the
pdf file, but always end up corrupting the file.  Since the whole
thing only requires a few lines of Python code, can anyone here be
gracious enough to show me how to include a pdf attachment in a
Message object from the Message class defined in the email module.

Thanks!


From support at internetdiscovery.com  Fri Jun 14 01:36:20 2002
From: support at internetdiscovery.com (Mike Clarkson)
Date: Fri, 14 Jun 2002 05:36:20 GMT
Subject: Detecting OS and Window Manager
References: <mailman.1023304388.9659.python-list@python.org>
Message-ID: <3d0980b9.6504256@nntp.ix.netcom.com>

On Wed, 05 Jun 2002 12:12:06 -0700 (PDT), "Sean 'Shaleh' Perry"
<shalehperry at attbi.com> wrote:

>
>On 05-Jun-2002 Andrew Wilkinson wrote:
>> Hi,
>> 
>> I'm writing a program that integrates with a number of different OSes and
>> Window Managers (under Linux), but I need to check what the user is running.
>> I know that...
>> 
>> import os
>> if os.name == 'nt':
>> 
>> will tell me they're running NT. But under Linux I'd like to know if they
>> are running KDE or Gnome, does anyone know a (preferably) simple test for
>> this?
 
This is quite a hard problem because of the possible DeskTop and
window manager combinations. For a stab at a solution (in part using X
atoms), take a look at the logic in the WmDefault style of tix 8.1.3
on http://tix.sourceforge.net/. The logic is in Tcl, but it is used by
Tix in Python.

>the only fool proof way to detect KDE, GNOME, etc is to check for X atoms.  The
>bigger question is why do you care?  As long as they can display a GUI you
>should not be bothered about which one it is.  Especially under the X Window
>System it was designed so that random applications DIDN'T know what they were
>running under.

There can be a number of reasons you'd care. For example, if you are
using TkInter, you may want it to pick up the user's default settings
for that environment for things like colors and styles.  If you are
using CDE/KDE/GNOME, then Tix.Tk() should(*) pick these up using
the WmDefault style, whereas Tkinter.Tk() won't, except under Windows.

>Also remember KDE is not a window manager, it is a desktop environment.  It is
>perfectly reasonable to run KDE and use fvwm (or even twm) as the window
>manager.  Same applies to GNOME.  Also, what about someone who only runs say
>Window Maker and never uses GNOME or KDE?

Each case has to be checked and as you point out,  X atoms are the
safest.

Mike.


From skip at pobox.com  Thu Jun 20 21:16:21 2002
From: skip at pobox.com (Skip Montanaro)
Date: Thu, 20 Jun 2002 20:16:21 -0500
Subject: RE strings (was: Variable Interpolation - status of PEP 215)
In-Reply-To: <3D127B89.EE1C17F4@engcorp.com>
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com>
        <4PfQ8.44245$n4.10307683@newsc.telia.net>
        <KhiQ8.44269$n4.10312432@newsc.telia.net>
        <29e28c51.0206201156.7e61a7f5@posting.google.com>
        <3D127B89.EE1C17F4@engcorp.com>
Message-ID: <15634.32357.874081.785019@gargle.gargle.HOWL>

    Peter> Given that the only real difference is you skip an import, and at
    Peter> the cost of modifying the language definition in an unusual
    Peter> manner, is there any point to this at all?

Yeah...  Makes it more like Perl (more magic under the covers instead of out
where you can see it). ;-)

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html




From killes at geneseo.net  Wed Jun 26 09:48:54 2002
From: killes at geneseo.net (SteveK)
Date: Wed, 26 Jun 2002 08:48:54 -0500
Subject: getting python source
Message-ID: <afcgv0$aie$1@news.geneseo.net>

Is the CVS repository the only way to get the source for Python 2.2?  Is
there a .zip file somewhere I can just download?





From james.kew at btinternet.com  Fri Jun  7 16:27:53 2002
From: james.kew at btinternet.com (James Kew)
Date: Fri, 7 Jun 2002 21:27:53 +0100
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <20020606215349.GA13474@node1.opengeometry.net> <mailman.1023401650.18434.python-list@python.org>
Message-ID: <adr50a$1n43o$1@ID-71831.news.dfncis.de>

"gabor" <gabor at realtime.sk> wrote in message
news:mailman.1023401650.18434.python-list at python.org...
>
> for file in directory:
> file.printsize()
>
>
> ran = range(len(directory))
> for r in ran:
> directory[r].printsize()
>
> [file.printsize() for file in directory]
>
> they're the same, and everyone would recommend you the last one..

Well, as a relative newbie, I'd recommend the first as much more
comprehensible and maintainable by whoever inherits the code.

The explicit loop clearly describes, in the code itself, the intent of the
code. Both the alternatives exercise darker corners of the language and are
less immediately descriptive, at least to my inexperienced eyes.

I'm a little uneasy about the list comprehension: it feels like it almost
abuses the spirit of list comprehensions by using an expression with
side-effects and then discarding the result of the comprehension.

Finally, "printsize" as a method name suggests to me that the loop is more
likely to be IO-bound than CPU-bound, so attempts to optimise the loop
overhead may buy very little benefit.

If it were me, I'd go with the first option for clarity; leave it unless and
until a performance problem emerges and profiling points at this loop;
consider algorithmic changes first (possibly merge the multiple IO
operations into a single larger one?); and only then start sacrificing
clarity for speed.

James





From skim at adobe.com  Wed Jun  5 13:21:44 2002
From: skim at adobe.com (Sung H. Kim)
Date: Wed, 5 Jun 2002 10:21:44 -0700
Subject: Freeze questions
Message-ID: <p04320404b923f8b99b71@[153.32.58.245]>

Can the Freeze tool be used with C libraries? i.e. extensions?

Will it create a truly standalone app (on Windows)?

Does it support tk/tkinter apps yet?

Finally, is there a good how-to on the Windows platform?

Thank you!

--Sung H. Kim
   skim at adobe.com




From mwh at python.net  Tue Jun 11 11:24:09 2002
From: mwh at python.net (Michael Hudson)
Date: Tue, 11 Jun 2002 15:24:09 GMT
Subject: unknown opcode Exception then AV
References: <ae546s$bg5$1@ctb-nnrp2.saix.net>
Message-ID: <lkk7p5yh2w.fsf@pc150.maths.bris.ac.uk>

"Geoff Bien" <gbien at intekom.co.za> writes:

> I have one script out of many (python 2.1 activestate on win32)
> which gives an "invalid opcode" exception and then also
> sometimes causes an AV.
> 
> This script  uses intebase with the gvib or kinterbasdb
> extension modules. since using gvib or kinterbasedb
> makes no difference to the problem they are probably
> not the cause
> 
> this is the only script I have problems with. it is too long
> to post here.
> 
> Traceback (most recent call last):
>   File "<string>", line 452, in ?
>   File "<string>", line 446, in go
>   File "<string>", line 425, in doPeriod
>   File "<string>", line 341, in doMonth
> SystemError: unknown opcode
> 
> the interpreter seems quite confused with this script,
> in other places simple statements like x=x+1  are giving
> the wrong results

Hard to give meaningful advice on this information!  Check your RAM?
Blow away .pycs?  Use dis do disassemble the bytecode of the function
that blows up?

Cheers,
M.

-- 
  > Look I don't know.  Thankyou everyone for arguing me round in
  > circles.
  No need for thanks, ma'am; that's what we're here for.
                                    -- LNR & Michael M Mason, cam.misc


From whisper at oz.net  Thu Jun 13 22:18:57 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 13 Jun 2002 19:18:57 -0700
Subject: why not "'in' in 'in'"?
In-Reply-To: <LNBBLJKPBEHFEDALKOLCIEPBPMAA.tim.one@comcast.net>
Message-ID: <GCEDKONBLEFPPADDJCOEGEAIDJAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Tim Peters
> Sent: Thursday, June 13, 2002 18:57
> To: python-list at python.org
> Cc: Grant Griffin
> Subject: RE: why not "'in' in 'in'"?
>
>
> [Grant Griffin]
> > One of the things I like about Python is that it works as expected.
> > However, I sometimes accidently expect it to allow one to use "in"
> > to determine if one string is "in" another, e.g.:
> >
> > >>> 'in' in 'in'
> >
> > which yields a traceback:
> >
> >    Traceback (most recent call last):
> >      File "<interactive input>", line 1, in ?
> >    TypeError: 'in <string>' requires character as left operand
> >
> > ...
> >
> > So why is the "in" operator limited to single characters when used as a
> > membership operator?
>
> In The Beginning, "x in y" was implemented as if via:
>
> def like_in(x, y):
>     for element in y:
>         if cmp(x, element) == 0:
>             return 1
>     return 0
>
> That's still close(*) to what it does if the class of y doesn't implement
> the newer-fangled __contains__ method.  Ponder what like_in(x, y)
> does when
> x and y are strings:  it can't possibly return 1 unless x is a 1-character
> string, because each value element takes on is a 1-character string.
> Therefore "str1 in str2" raised an exception when len(str1) != 1, as you
> almost certainly didn't intend to use this as an expensive and obscure way
> to spell 0.  If that's what you really wanted, fine, but then Guido wanted
> to help you achieve that goal by forcing you to use some way even more
> expensive and obscure <wink>.
>
> But that was then.  Python does allow defining __contains__ methods now,
> precisely so that a type can define the meaning of "x in y" in
> whatever way
> makes most sense for type(y).  The meaning of "in" is no longer
> *defined* by
> like_in, and consequently Guido agrees that the string behavior is much
> harder to justify now.
>
> So if somebody wants this change enough to submit a patch (code,
> doc, + test
> suite changes), it's likely to be accepted for 2.3.  As code changes go,
> this one should be particularly easy.  We have an internal bet going as to
> whether this invitation is enough to kill the idea <wink>.
>
> some-parts-of-this-were-actually-true-ly y'rs  - tim
>
>
> (*) The default implementation today uses "x == element" instead
> of cmp, so
> that a type implementing only __eq__ works fine.  Other than that, same
> thing.
>

Hmmm... How about calling it "contains" instead of "in"? "if 'feefifofum'
contains 'fi'" seems nicer then "if 'fi' in 'feefifofum'" maybe?

Of course maybe this makes too much sense for a language where zip has zip
to do with compression and yield nothing to do with threads.

Dave LeBlanc
Seattle, WA USA





From DLNXPEGFQVEB at spammotel.com  Tue Jun 11 11:45:26 2002
From: DLNXPEGFQVEB at spammotel.com (Christos TZOTZIOY Georgiou)
Date: Tue, 11 Jun 2002 18:45:26 +0300
Subject: [OT] Antispamming the email address
References: <3D05FF31.4060701@araknos.it> <mailman.1023804795.22623.python-list@python.org>
Message-ID: <9c6cgu425qgap7hfbkfnpqnq1582g4604d@4ax.com>

On Tue, 11 Jun 2002 07:12:06 -0700, rumours say that "David LeBlanc"
<whisper at oz.net> might have written:

[Nicola Larosa commenting on my base64 encrypting of my real email addr]

[David:]
>I think both of them are too easy to programatically detect and unwind. Even
>"fred_nospam_ at some.org is only going to have a limited deterrent lifetime.

I used to have it rot13 encrypted to no avail, as it seemed.  It was
here on the python list that I saw sb else encoding their email address
with base64, can't remember who it was though, so it was not my idea but
I liked it, and I think that, although easy to decode once you get the
idea, it would need human intervention, so I'll drink to the
spamming-list-collector that will bother, and let them spam me once (I
administrate my company's network, and I try to improvise a lot to
reject spam during their delivery (before the message is received :) )
before they become inexistant, e-mail-wise.

Phew! a long sentence, if I've seen one.  No wonder I always liked
'Catch-22'.  At least I was on topic (re the subject, which is OT)
throughout the sentence!
-- 
TZOTZIOY, I speak England very best,
Real email address: 'dHpvdEBzaWwtdGVjLmdy\n'.decode('base64')


From eppstein at ics.uci.edu  Fri Jun 14 15:08:25 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Fri, 14 Jun 2002 12:08:25 -0700
Subject: a possibly foolish question about slices
References: <yu99sn3prbqm.fsf@europa.research.att.com>
Message-ID: <eppstein-FCC2A4.12082514062002@news.service.uci.edu>

In article <yu99sn3prbqm.fsf at europa.research.att.com>,
 Andrew Koenig <ark at research.att.com> wrote:

> Suppose x is a string or list, and I want to refer to
> a subsequence of x with length n starting at position p.
> under the assumption that this subsequence is entirely
> 
> To do so, I can use x[p:p+n] in all circumstances but one,
> namely when p < 0 and p == -n.  Is there a similarly concise
> expression that works even in this special case?

x[p:][:n]

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From mcherm at destiny.com  Fri Jun  7 10:30:03 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Fri, 07 Jun 2002 10:30:03 -0400
Subject: Medium to Large Scale Python Deployments
Message-ID: <3D00C36B.10405@destiny.com>

Folks, I think this is a pretty useful thing to collect.

In fact, it might be really nice to collect a list of large (successful) 
Python projects and post it somewhere so the Python community could use 
it for advocacy.

So do me a favor... if you have (or know of) a large Python project 
which you think our community ought to be bragging about, please either 
post it to this thread or email it to me. I'll collect the list, and 
then see if I can get it posted somewhere appropriate. Please run 
pycount (http://starship.python.net/crew/gherman/playground/pycount/ ) 
as "pycount.py -R *.py" on the top of your source tree, (or modify that 
appropriately for your project!).

The exact definition of "large" is fairly vague... I'd say 30K loc is 
borderline and 50K loc is definitely big.

Please note that the project does NOT need to be open-source or freely 
available to qualify!!!

If I get at least 8 responses, I'll assume there's sufficient interest, 
and arrange to get it posted.

Alternatively, if such a list already exists, let me know where, and 
I'll volunteer to help update and promote it.

Here's what I have so far:

Project      Size(Kloc)  URL
-------        ----      ---------------------
Zope            ???      http://www.zope.org/
twisted         >50      http://www.twistedmatrix.com
ekit            ~50      http://www.ekit.com/
Domenic's
   ERP system   >100      --
Something
   by aahz      30-40     ???


Sources:
   http://groups.google.com/groups?th=b7ec40fc188a2a52&rnum=1





From spammers.do.not.bother at void.bogus  Fri Jun  7 16:59:57 2002
From: spammers.do.not.bother at void.bogus (Magnus)
Date: Fri, 07 Jun 2002 20:59:57 GMT
Subject: Comparison: HTTP with Java, Perl or Python
References: <3D009911.9000904@gmx.de>
Message-ID: <h99M8.11172$p56.3361428@newsb.telia.net>

Ingo Linkweiler wrote:

> Hello,
> 
> for a comparison of languages I need a complete "translation" of this
> short skript into Perl, Tcl and Java:
> 
>  myurl = "http://www.uni-dortmund.de"
>  import urllib
>  file  = urllib.urlopen(myurl)
>  text  = file.read()
>  file.close()
>  print text
> 
> Can you send me some solutions?
> The skripts must be able to run, please do not post "partial" solutions.
> 
> Ingo


Try following...

Regards,
Magnus
--

import java.io.*;
import java.net.*;

public final class UrlGet {

    final public static void main(String args[]) {
 
        try {
            String theContent = new String();
            URL theUrl = new URL("http://www.uni-dortmund.de");
            URLConnection theUrlConnection = theUrl.openConnection();
            BufferedReader theReader =
                new Buffe
redReader(new InputStreamReader(theUrlConnection.getInputStream()));
            String theInputLine;
            while ((theInputLine = theReader.readLine()) != null) {
                theContent += theInputLine;
            }
            theReader.close();
            System.out.println(theContent);
        }
        catch (Exception e){
            System.err.println(e.getMessage());
            System.err.println(e.toString());
            e.printStackTrace();
        }
    }
}



From jadestar at idiom.com  Thu Jun 20 03:51:10 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 20 Jun 2002 07:51:10 GMT
Subject: AF_UNIX + SOCK_DGRAM
Message-ID: <aes1he$gv2$1@news.idiom.com>

 I was playing with the socket module and Unix domain sockets,
 trying to make it work with SOCK_DGRAM.  I could get messages
 recvfrom() but any attempt to send responses was failing, 
 complaining about the return address.

 Now I think I understand the problem.  The recvfrom() tuple
 does contain two parts (which WOULD BE address and data in UDP)
 but I'm guessing that the address is meaningless for Unix 
 domain sockets.  It occurs to me that the "return" address 
 must be the same socket (filesystem node of type 's') as the 
 one that I did a .bind() to.  

 That suggests that using a Unix domain socket as a bidirectional
 DATAGRAM communications channel requires that we implement a flow 
 control protocol to force the other end to do its recvfrom().
 I honestly know now what happens when more than two processes 
 get involved and I'm curious what the value being returned as
 u.recvfrom()[1] actually means.
 
 This is academic curiosity, and I'll read through the C sources
 to see what I can glean therefrom, eventually.  I'd like to know
 how I can tell what process has opened a connection to my 
 AF_UNIX socket and I'd love to know how to pass an open file
 descriptor through an AF_UNIX socket (in Python).  I've read that
 this is possible in C and that it is one of the few features of 
 AF_UNIX sockets which is NOT available via other address/protocol 
 families.  I've also heard that this can be used as a way for a
 client to request read-only or append-only access to a file
 --- since the UNIX APIs (in C) apparently offer ways for the server
 to obtain credential information about the requesting process as
 well as passing open file descriptors (which are then roughly akin
 to "capabilities" since their access mode can't be changed)
 



From mal at lemburg.com  Tue Jun 18 03:28:02 2002
From: mal at lemburg.com (M.-A. Lemburg)
Date: Tue, 18 Jun 2002 09:28:02 +0200
Subject: unicode strings and strings mix
References: <Pine.LNX.4.33.0206180848090.28845-100000@rnd.onego.ru>
Message-ID: <3D0EE102.5050905@lemburg.com>

Roman Suzi wrote:
> Is there any way to define "default" encoding so the following error
> will not happen?
> 
> 
>>>>unichr(0x3345) + "?????"
>>>
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> UnicodeError: ASCII decoding error: ordinal not in range(128)
> 
>>>>unichr(0x3345) + "babab"
>>>
> u'\u3345babab'

There is, but it's better not to rely on it since it's a site
specific setting.

You should use unicode("?????", "koi8-r") instead (or
define a wrapper for this like

def n(text):
     return unicode(text, "koi8-r")

-- 
Marc-Andre Lemburg
CEO eGenix.com Software GmbH
______________________________________________________________________
Company & Consulting:                           http://www.egenix.com/
Python Software:                   http://www.egenix.com/files/python/
Meet us at EuroPython 2002:                 http://www.europython.org/





From whisper at oz.net  Wed Jun 19 20:43:17 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 19 Jun 2002 17:43:17 -0700
Subject: using os.chmod in windows
In-Reply-To: <3D110C99.5090607@portzer0.virtualave.net>
Message-ID: <GCEDKONBLEFPPADDJCOEGEFCDKAA.whisper@oz.net>

I didn't find any doc for this, so I tried the simple expedient of:

# Attrtest.py

import os

os.stat('testfile.txt')

raw_input('Change file attribute using File Explorer, then press return')

os.stat('testfile.txt')

When I did this, I got:

(At the start, normal r/w file)
>>> os.stat('testfile.txt')
(33206, 0L, 9, 1, 0, 0, 26L, 1024026085, 1023320564, 1023320564)
(Add hidden attribute using File Explorer)
>>> os.stat('testfile.txt')
(33206, 0L, 9, 1, 0, 0, 26L, 1024532993, 1023320564, 1023320564)
(Add readonly in addition to hidden)
>>> os.stat('testfile.txt')
(33060, 0L, 9, 1, 0, 0, 26L, 1024532997, 1023320564, 1023320564)

Notice that the last five digits (an int in decimal) of the 8th group have
changed. How you parlay that into something chmod() can use, I have no idea.

HTH,

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Locke
> Sent: Wednesday, June 19, 2002 15:59
> To: python-list at python.org
> Subject: using os.chmod in windows
>
>
> I never really learned the number codes for file permissions in UNIX. I
> always just did "chmod +r" because it is much easier to remember. I need
> to change a file to be 'hidden' in windows. In UNIX, this is done by
> renaming it with a dot in front of the name. But the python os.chmod
> command says you need to specify the number as the argument. As far as I
> know, there is no number for 'hidden' since it is windows only. So how
> do I do this?
>
> Thanks.
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From bokr at oz.net  Sat Jun 29 17:38:41 2002
From: bokr at oz.net (Bengt Richter)
Date: 29 Jun 2002 21:38:41 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com> <mailman.1025336262.12245.python-list@python.org> <3d1dc547.468619140@netnews.attbi.com> <afkv73$f2274$1@ID-139865.news.dfncis.de> <OxnT8.45616$n4.10835602@newsc.telia.net>
Message-ID: <afl9d1$5k0$0@216.39.172.122>

On Sat, 29 Jun 2002 19:05:18 GMT, "Fredrik Lundh" <fredrik at pythonware.com> wrote:

>Dave Kuhlman wrote:
>
[...]
>> Is finding things in globals slower than finding them in locals?
>
>yes.  Python assigns integer indices to local names, and
>stores the corresponding objects in an array.
>
Is this done for built-in keywords also? ISTM a standard list
of indices could be defined for a particular Python version,
and code generated to take advantage.

Maybe even a range of byte codes could be reserved for much-used
functions, as opposed to looking them up and then executing
a general function-call byte code to go there. It should be
possible to make this an efficient default, while still allowing
for straight-forward local shadowing bindings, IWT.

E.g., instead of the code for abs(x) being

           3 LOAD_NAME                0 (abs)
           6 LOAD_NAME                1 (x)
           9 CALL_FUNCTION            1

it might be
          6 LOAD_NAME                1 (x)
          9 CALL_BUILTIN_0           1

where e.g. CALL_BUILTIN_0 .. CALL_BUILTIN_31 might be the reserved range
for builtin functions in the special list.

However
          3 LOAD_NAME                0 (x)
          6 LOAD_CONST               0 (0)
          9 CALL_BUILTIN_X           1

would probably be about as fast, though not as compact.

If necessary to make it unambiguous and to make compilation easier, you could have
a "builtin" declaration analogous to "global." Or the usage might be popular enough
that you'd want a way to opt out instead of opt in. Since it's a bad idea to rebind
keywords anyway, maybe backwards compatibility wouldn't be too much of a problem if
it were default for some builtins functions in a future version?

Hm, seems like you could do static arg count checks too for this kind of builtin use.

Hm2, wonder if you could allow user-defined assignment of functions to dynamically
assigned byte codes within a user range using "builtin user_func" declarations,
so that compilation would make analogous code like that suggested for real builtins.

OTTOMH just now, not exactly PEP-ready ;-)

Regards,
Bengt Richter


From fredrik at pythonware.com  Thu Jun 20 05:28:00 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 20 Jun 2002 09:28:00 GMT
Subject: 'else' clause in 'for' and 'while'
References: <ad496f8.0206200107.291c5a40@posting.google.com>
Message-ID: <AehQ8.44259$n4.10310580@newsc.telia.net>

Ville Vainio wrote:
> How do you people use the else clause in for and while constructs? Are
> they used widely for some specific applications? Do they offer
> elegant, idiomatic solutions to some problems?

    for item in list:
        if item == my_target:
            print "found it"
            process(item)
            break
    else:
        print "not found"

    while more_items():
        item = get_next_item()
        if item == my_target:
            print "found it"
            process(item)
            break
    else:
        print "not found"

    try:
        item = lookup(database, key)
    except KeyError:
        print "not found"
    else:
        print "found"
        process(item)

    try:
        handler = getattr(object, "handler")
    except AttributeError:
        print "no handler"
    else:
        handler(event)

# (etc)

</F>




From skip at pobox.com  Fri Jun  7 14:11:58 2002
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 7 Jun 2002 13:11:58 -0500
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <200206071804.g57I4fp01212@europa.research.att.com>
References: <yu99vg8v6orb.fsf@europa.research.att.com>
        <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
        <200206071747.g57HlbH01057@europa.research.att.com>
        <15616.62634.635544.905055@12-248-41-177.client.attbi.com>
        <200206071804.g57I4fp01212@europa.research.att.com>
Message-ID: <15616.63342.559215.644086@12-248-41-177.client.attbi.com>

    Skip> Try running the just built python executable and then manually
    Skip> execute "import struct".  What error message(s) do you see?

    Andrew> $ python
    ...

Showing that you cd'd to your build directory and executed "./python" would
give me greater faith that you were executing the just-built interpreter and
importing the proper .so file.  Note that since the build process renamed
the file, you will probably have to go back and mv it back to struct.so
before trying this...

Skip




From Nixdisciplenews at hotmail.com  Thu Jun  6 23:07:25 2002
From: Nixdisciplenews at hotmail.com (Felix)
Date: Thu, 06 Jun 2002 23:07:25 -0400
Subject: Compiling Xchat to enable python scripting 
Message-ID: <3D00236D.5010000@hotmail.com>

I have tried looking for documentation on this particular problem but ot 
no avail I wasn't able to find anything on it. This is my problem, After 
I compiled and installed xchat for Red hat 7.3 I noticed that python 
scripting was disable is there anyway of enabling it from within xchat? 
Or do I have ot recompile the whole program and include some sort of 
enable commmand within the ./configure option? If that is the case 
please I would appreciate if someone would give me all the necessary 
commands needed to correct my problem please.




From loewis at informatik.hu-berlin.de  Fri Jun  7 11:07:35 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 07 Jun 2002 17:07:35 +0200
Subject: How do you read unicode files?
References: <adpd7h$6sv$1@slb3.atl.mindspring.net> <m3vg8vaa9u.fsf@mira.informatik.hu-berlin.de> <adpt72$dds$1@nntp9.atl.mindspring.net>
Message-ID: <j4660vjfew.fsf@informatik.hu-berlin.de>

"Matt Gerrans" <mgerrans at mindspring.com> writes:

> Thanks Mike, Francis and Martin for the quick replies!
> 
> My mistake was using readlines() to read the text file, instead of read().
> (I should have posted a snippet of code in the first place and probably you
> all would have noticed that immediately).

Yes, you cannot read-and-decode UTF-16 line-by-line. Unfortunately,
the Python UTF-16 stream reader also does not support .readline, even
though it should.

Regards,
Martin



From mboedick at mboedick.org  Thu Jun 13 18:11:09 2002
From: mboedick at mboedick.org (Matthew Boedicker)
Date: Thu, 13 Jun 2002 18:11:09 -0400
Subject: returning traceback strings
Message-ID: <20020613181109.A1019@gurthang.twcny.rr.com>

Hello,

I am looking for something exactly like traceback.print_exc(), but instead
of writing the traceback text to a filehandle, I want it to return it to
me as a string.  I was unable to find any way to do this using the traceback
module, since everything there seems to be in terms of file handles being
passed around.

The problem is I need to pass the entire multi-line traceback string to a
file-like object's write() method in one call, not line by line.

This is the best workaround I was able to come up with:

import cStringIO
import traceback

def print_exc_str():
    buf = cStringIO.StringIO()
    traceback.print_exc(file=buf)
    return buf.getvalue()

Any suggestions for a better way?  What about the traceback module using
strings internally, and wrapping another layer around that to write to files
and file-like objects?

-- 
Matthew Boedicker http://mboedick.org




From i.linkweiler at gmx.de  Mon Jun 24 04:55:09 2002
From: i.linkweiler at gmx.de (Ingo Linkweiler)
Date: Mon, 24 Jun 2002 10:55:09 +0200
Subject: PyQT craches when using webbroswer.open or os.system
Message-ID: <3D16DE6D.3040108@gmx.de>

Hello!

all my Python applications are crashing with Python-QT, if I try to open an html-document:

import webbrowser
webbrowser.open(myurl)

or 

os.startfile(myurl)

==> General protection fault in PYTHON.EXE

This crash only occures with Windows 2000 and Internet Explorer 6 as Default Browser, but can be seen on many systems. It does NOT crash with Netscape and not with Windows 98.

It does not CRASH, if the Internet Explorer is started before doing the calls.
os.startfile does not crash, when opening other documents like text files with notepad.

I can avoid this crashes, if I close all my QT-Windows before doing the calls, but this is not a good solution.

All tested systems have:

Windows 2000
Internet Explorer 6
Python 2.2 or older
QT 2.3 NC
PyQT 3.2.4 or older

Do you have any ideas to solve this problem?

Ingo




From mcfletch at rogers.com  Sun Jun  2 19:58:10 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Sun, 02 Jun 2002 19:58:10 -0400
Subject: List
References: <20020602205610.31575.qmail@linuxmail.org> <009b01c20a7e$12a5ad80$0101010a@local>
Message-ID: <3CFAB112.2040900@rogers.com>

I don't know if you'll consider this a more elegant way to divide up the 
lists, but I think it's a little more beautiful than the index-mucking 
approaches IMO:

def into( source, count=3 ):
     source = source[:]
     return intoDestructive( source, count )

def intoDestructive( source, count=3 ):
     if source and not count:
         raise ValueError( """Cannot divide into 0-length sub-elements""" )
     elif not source:
         return []
     results = []
     if count > 0:
         while source:
             results.append( source[:count])
             del source[:count]
     else:
         while source:
             results.append( source[count:])
             del source[count:]
     return results

It will be very slow for very large lists with positive slice values 
(something the index-mucking approach doesn't suffer from), but for most 
reasonably sized lists, should be fairly good.  It handles negative 
steps as well, so you can slice up the list from the right _or_ the left 
side :) .

Enjoy all,
Mike

Chris Gonnerman wrote:
...
> 
>     Lists = []
>     N = 2
> 
>     adj = 0
> 
>     if len(BigList)%N:
>         adj = 1
> 
>     for i in range(len(BigList)/N+adj):
>         Lists.append(BigList[i*N:(i+1)*N])
> 
> (If anyone knows a more elegant solution please post it,
> as I am getting headaches trying to think of one.)
...
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From pyth at devel.trillke.net  Wed Jun  5 12:30:55 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 5 Jun 2002 18:30:55 +0200
Subject: self
In-Reply-To: <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com>; from sholden@holdenweb.com on Wed, Jun 05, 2002 at 11:47:26AM -0400
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu> <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com>
Message-ID: <20020605183055.D30696@prim.han.de>

Steve Holden wrote:
> "Vojin Jovanovic" <vjovanov at stevens-tech.edu> wrote ...
> [...]
> >
> > > Well, good luck with your project -- and don't spring any more
> > > requirements on us, it's getting harder each time to solve your problem.
> >
> > Sorry, but it is not that my problem is hard.  It is that the concept of
> > having
> > to reference instance variables always through self. is making it hard and
> > inconvenient which was my point in my original post.
> > My opinion is that those points explained in FAQ are irrelevant.  The only
> > one relevant is .....
> >
> [...]
> 
> Could you maybe use self.__dict__ as a namespace argument to the eval()?

no his questions go deeper :-)

the problem is that the values in the dict have to be 'evaluated' if accessed. 
For every name in such a value you need to evaluate its value again.
IOW it's recursive. The obvious solution is to intercept the namevalue-lookups 
during evaluation of an equation. But this doesn't work for exec/eval:
They ignore the interception points. 

IOW you cannot successfully wrap or inherit a dictionary type and pass it 
to exec/eval :-(

The next best solution IMO is to preprend a namescope like 'self.' 
to every name in an expression. But you may have to parse the expression to 
do this. Maybe fixing exec/eval behaviour is a better idea?

regards,

    holger




From prema at prema.co.nz  Sat Jun  8 21:31:59 2002
From: prema at prema.co.nz (Prema)
Date: 8 Jun 2002 18:31:59 -0700
Subject: RE Module How to escape $
References: <ae58a20e.0206080509.44da4bcb@posting.google.com> <3D0223D5.4050900@earthlink.net>
Message-ID: <ae58a20e.0206081731.456e6200@posting.google.com>

Hi Hans!
Thanks so much for your post -- it works perfectly.
I had been head bashing for a while on that one for some reason.
Thanks for getting back so quickly
Best regards 
Mike

Hans Nowak <wurmy at earthlink.net> wrote in message news:<3D0223D5.4050900 at earthlink.net>...
> Prema wrote:
> > Hi People !
> > I'm trying to use the re.sub function to replace several instances of a
> > PHP Variable eg "$Amount". Could someone please help with an example
> > Thanks very much
> > Mike
> 
> Sure...
> 
>  >>> import re
>  >>> s = "I have $amount dollars"
>  >>> re.sub("\$\S+", "foo", s)
>  'I have foo dollars'
>  >>>
> 
> HTH,


From phr-n2002b at NOSPAMnightsong.com  Sat Jun 29 13:35:02 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Jun 2002 10:35:02 -0700
Subject: private
References: <pan.2002.06.29.17.38.51.789527.440@myself.com>
Message-ID: <7x8z4y9eex.fsf@ruckus.brouhaha.com>

Rhymes <rhymes at myself.com> writes:
> Is there any possibility to build up a _real_ private attribute?
> Such as C++ private data members...
> I know that with double-underscore prefix (and the name mangling related)
> I can " hide " data members but I think it's not a so crafty trick...
> Is there any PEP about it? Will Guido insert a private keyword
> in a future release of Python ?

Well, as several people have said, C++ private members aren't that
private either.  If you mean something like private members in Java
which try to really enforce privacy against hostile callers, no,
there's nothing like that in Python.  The only way to simulate it is
with a remote object scheme, putting the protected objects in a
separate process (or separate computer) from the caller, and letting
them run their own access policy.  This has more overhead than
ordinary method calls, but can be made very secure.



From bokr at oz.net  Wed Jun 12 19:45:01 2002
From: bokr at oz.net (Bengt Richter)
Date: 12 Jun 2002 23:45:01 GMT
Subject: map -> class instance
References: <m2r8jcp3pv.fsf@video.paradise.lost>
Message-ID: <ae8mdt$8qb$0@216.39.172.122>

On Wed, 12 Jun 2002 16:49:32 -0500, John Hunter <jdhunter at nitace.bsd.uchicago.edu> wrote:

>
>I want to convert dictionary instances to instances of a class that
>has the keys of the map as attributes.  I naively tried this:
>
>class Map2Class:
>    def __init__(self,m):
>        self.m = m
>
>    def __getattr__(self, key):
>        return self.m[key]
>
>    def __setattr__(self, key, val):
>        self.m[key] = val
>
>    def __delattr__(self, key):
>        if self.m.has_key(key):
>            del self.m[key]
>
>m = {'first' : 'John',
>     'last' : 'Hunter',
>     'age' : 34}
>
>c = Map2Class(m)
>print c.first, c.last, c.age
>
>But got an infinite recursion:
>
>~/python/test $ python map_to_class.py 
>Traceback (most recent call last):
>  File "map_to_class.py", line 20, in ?
>    c = Map2Class(m)
>  File "map_to_class.py", line 3, in __init__
>    self.m = m
>  File "map_to_class.py", line 8, in __setattr__
>    self.m[key] = val
>  File "map_to_class.py", line 5, in __getattr__
>    return self.m[key]
>  File "map_to_class.py", line 5, in __getattr__
>    return self.m[key]
>        [snip ... lots more line fivers ]
>
>Is there a better/right way to do what I want?
>
This seems to work. I'm using 2.2:

 >>> def makeM2C(m):
 ...     class PersInf:
 ...         def __repr__(self):
 ...             return '<%s: "%s" "%s" %s" instance at %s>' % (
 ...                 self.__class__.__name__,self.first, self.last, self.age, id(self)
 ...             )
 ...     PersInf.__dict__.update(m)
 ...     return PersInf()
 ...
 >>> makeM2C(m)
 <PersInf: "John" "Hunter" 34" instance at 8208688>

I assume you're only interested in a single instance here, but if
you want to define an __init__ for PersInf, you can, and then return
Persinf instead of the instance PersInf(). You could then store the
returned class like AClass = makeM2C(aDict), and then do anInst = AClass(initparams).

But here it's one instance, with a few variations on feeding the factory a dictionary:
(You can just put pass where I defined __repr__ if you don't care)

 >>> c = makeM2C(m)
 >>> print c.first, c.last, c.age
 John Hunter 34
 >>> c2 = makeM2C({'first':'Ratoncito','last':'Miguel','age':'??'})
 >>> c2
 <PersInf: "Ratoncito" "Miguel" ??" instance at 8208416>
 >>> print c2.first, c2.last, c2.age
 Ratoncito Miguel ??
 >>> oi = 'Kalle Anka (gammal)'.split()
 >>> c3 = makeM2C(dict(zip(['first','last','age'],oi))
 ... )
 >>> c3 = makeM2C(dict(zip(['first','last','age'],oi)))
 >>> c3
 <PersInf: "Kalle" "Anka" (gammal)" instance at 8192544>
 >>> print c3.first, c3.last, c3.age
 Kalle Anka (gammal)
 
HTH

Regards,
Bengt Richter


From mgerrans at mindspring.com  Thu Jun 27 12:17:00 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Thu, 27 Jun 2002 09:17:00 -0700
Subject: How to get rid the new line
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org> <3D1B1BE7.D482AE@millfilm.co.uk>
Message-ID: <affdtb$6n6$1@slb6.atl.mindspring.net>

> ## remove all the leading and tail white space
> line = string.strip(aline)
> if line:
>     # this is an emptyLine

I like this one best; a regular expression seems like overkill and assuming
that the line is just the one '\n' character (as in other posts) is not good
either.  A line with only whitespace and nothing else probably also should
be considered empty, and the above takes care of that.

However, I think there is a bug in the comment: it should be "this is NOT an
emptyLine."

- mfg




From gcordova at hebmex.com  Thu Jun 27 15:55:16 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 27 Jun 2002 14:55:16 -0500
Subject: self destruction :)
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233015DE6C2@mail.hebmex.com>

> 
> i've got a list of dynamicaly instantiated objects of a class 
> of my own. how can i let one of the objects delete itself?
> 
> i've tried: 'del self', but somehow it didn't work. 
> afterwards it should also _not_ be in the list anymore.
> 
> thanx,
> guyon
> 

In what list?

Can a snake eat it's tail?

An object cannot "forget" itself, it's illogical.

Objects deallocate automagically when the last reference
to it disappears.

-gus




From gerson.kurz at t-online.de  Fri Jun 14 13:15:25 2002
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Fri, 14 Jun 2002 17:15:25 GMT
Subject: I feel stoopid ... this code is to verbose
References: <3D0A1791.3060307@mxm.dk>
Message-ID: <3d0a240f.15785671@news.t-online.de>

Correct me if I'm wrong, but all of the previous suggestions only seem
to insert dots, not dots and commas; i.e. you want the output like
this:

>>> 42.034.539.823,94

not

>>> 4.203.453.982.394

so here is my one liner:

strfmt = lambda x:"".join(map(lambda c,l,y=len(x):(l==3) and c+"," or
(not (l-3)%3 and c+"." or c),list(x),range(len(x),0,-1)))
    
print strfmt("4203453982394")

>>> 42.034.539.823,94

Of course, the statement in the above code should be written in one
line.



From bjarne_christiansen at hotmail.com  Wed Jun 26 09:26:20 2002
From: bjarne_christiansen at hotmail.com (Bjarne Christiansen)
Date: Wed, 26 Jun 2002 13:26:20 +0000
Subject: Python file upload
Message-ID: <F68N9KAy4MoSn03t3Z4000002a8@hotmail.com>

Hi Fredrik,

> > #!c:\Python\python -d
>
>do you really need optional debug output from the parser?
>maybe you meant "-u"?

It works with the -u option. Thanks a lot - you are the best!!!
Well, I always use the -d option as default while in testing phase...

Best Regards,
~Bjarne

>
> > I have changed the script to the following:
> > ...
> >   item = form["filename"]
> >   if item.file:
> >     data = item.file.read()
> >     f = open("file1.jpg","wb")
> >     f.write(data)
> > ...
> >
> > But I still run into the same problem, any ideas?
>
>how much data is missing?  if it's a couple of kilobytes missing
>at the end, you might need to explicitly close the file:
>
>     f.close()
>
></F>
>
>
>--
>http://mail.python.org/mailman/listinfo/python-list


_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp.





From phd at phd.pp.ru  Wed Jun  5 16:27:08 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 6 Jun 2002 00:27:08 +0400
Subject: Urllib thread safe?
In-Reply-To: <just-9B7B3B.21511805062002@news1.xs4all.nl>; from just@xs4all.nl on Wed, Jun 05, 2002 at 09:51:18PM +0200
References: <mailman.1023280808.19305.python-list@python.org> <m3it4xobkb.fsf@mira.informatik.hu-berlin.de> <mailman.1023304328.9010.python-list@python.org> <just-9B7B3B.21511805062002@news1.xs4all.nl>
Message-ID: <20020606002708.A31521@phd.pp.ru>

On Wed, Jun 05, 2002 at 09:51:18PM +0200, Just wrote:
> > > Yes, that is thread-safe. urllib does not maintain any per-url state,
> > 
> >    Hah? It maintains global _urlopener...
> 
> But that's not per-url state, is it?

   I am afraid you would have problems mixing calls to the global
_urlopener from different threads.

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From pyth at devel.trillke.net  Fri Jun  7 16:38:51 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 7 Jun 2002 22:38:51 +0200
Subject: Help with Black Adder...
In-Reply-To: <3D00FB5F.5010101@verio.net>; from vlindberg@verio.net on Fri, Jun 07, 2002 at 12:28:47PM -0600
References: <adpc36$179jq$1@ID-139866.news.dfncis.de> <mailman.1023432372.7831.python-list@python.org> <3D00FB5F.5010101@verio.net>
Message-ID: <20020607223851.D6609@prim.han.de>

VanL wrote:
> The Book mentioned above (GUI programming with QT or some 
> such) is also available for free on the web.
> 
>     http://www.opendocspublishing.com/pyqt/index.lxp

can anyone point me to a reseller in germany which doesn't
insist on credit cards?

thanks,

    holger




From bokr at oz.net  Fri Jun 21 12:09:37 2002
From: bokr at oz.net (Bengt Richter)
Date: 21 Jun 2002 16:09:37 GMT
Subject: converting an array of chars to a string
References: <3d12dda2_6@news.newsgroups.com>
Message-ID: <aevj41$fg2$0@216.39.172.122>

On Fri, 21 Jun 2002 10:10:01 +0200, JB <jblazi at hotmail.com> wrote:

>How can I convert an array of chars to string?
>
Depends on what you mean by "an array of chars."

>The problem is this: I should like to perform 8 bit 
>calculations in a buffer. After this has been done, the
Sounds like you mean a "buffer" of 8-bit integers. How are
you representing this "buffer"?

Have a look at the array module. It can construct an array
of signed or unsigned character elements optionally initialized
with values from a list or string. You can operate on elements
of the array using indices, and you can convert the result to
string or list, or you can write it directly to a file. E.g.,
here is an example using unsigned chars:

Get the array module:
 >>> import array

Make an array of unsigned chars, initialized from a list:
 >>> nums = array.array('B',[65,66,67,0,1,2,3])
 >>> nums
 array('B', [65, 66, 67, 0, 1, 2, 3])

Look at it as a string:
 >>> nums.tostring()
 'ABC\x00\x01\x02\x03'

Convert to actual list:
 >>> nums.tolist()
 [65, 66, 67, 0, 1, 2, 3]

Now we'll mess with the [2] element (67) in the array:
 >>> nums[2]+=1
 >>> nums
 array('B', [65, 66, 68, 0, 1, 2, 3])

Result as string (note C->D):
 >>> nums.tostring()
 'ABD\x00\x01\x02\x03'

>buffer has to ba written to a binary file (and file.write 
>takes strings instead of lists).
>
Check doc string for .tofile() method:
 >>> print nums.tofile.__doc__
 tofile(f)

So:
 >>> f=file('abdetc.txt','wb') # write binary
 >>> nums.tofile(f)
 >>> f.close()

Now read it back into a string
 >>> f=file('abdetc.txt','rb') # read binary
 >>> numstr = f.read()
 >>> numstr
 'ABD\x00\x01\x02\x03'

This gets you the whole string without an exception,
whereas you can read an exact number of characters
using the .fromfile() method, if you know how many you want.

Use as string, and/or convert to another array:
 >>> nums2=array.array('B',numstr)
 >>> nums2
 array('B', [65, 66, 68, 0, 1, 2, 3])

We've gone full circle:
 >>> nums2.tostring()
 'ABD\x00\x01\x02\x03'

Or, you could do something else with the file, e.g.,
seek past two bytes in the file and get three bytes into an array
of unsigned chars:
 >>> f=file('abdetc.txt','rb') # read binary
 >>> f.seek(2)

Make empty array:
 >>> nums3=array.array('B')
 >>> nums3
 array('B')

Append elements from file:
 >>> nums3.fromfile(f,3)
 >>> nums3
 array('B', [68, 0, 1])

See as string again:
 >>> nums3.tostring()
 'D\x00\x01'

You can also do all this without the array module, with
lists of numbers, converting to chars and strings:

 >>> numlist = map(ord,'ABC\x00\x01\x02')
 >>> numlist
 [65, 66, 67, 0, 1, 2]
 >>> map(chr,numlist)
 ['A', 'B', 'C', '\x00', '\x01', '\x02']
 >>> ''.join(map(chr,numlist))
 'ABC\x00\x01\x02'

Regards,
Bengt Richter


From janeaustine50 at hotmail.com  Thu Jun 13 08:51:04 2002
From: janeaustine50 at hotmail.com (Jane Austine)
Date: 13 Jun 2002 05:51:04 -0700
Subject: Is the weakref.ref callback is called deterministically?
Message-ID: <ba1e306f.0206130451.795085d6@posting.google.com>

Dear Pythonistas,  

I'm studying weakref module recently to make a variation of WeakList.

In Andres Tuells' WeakList[1] and weakref module, I have found
something interesting. There are "while 1" loops in fetch methods.

[weakref module]
    def popitem(self):
        while 1:
            key, value = self.data.popitem()
            o = key()
            if o is not None:
                return o, value

[WeakList module]
	def __getitem__(self, i):
		while 1:
			result  = self.data[i]()
			if result is not None:
				return result

Even though the index i does not change, it loops till the
result is not None. Seemingly it depends on the weakref.ref
callback to remove dead weakrefs.

However, the while loop wouldn't be needed if the callback
was called right after the last reference to the object
was "del"ed.

Is it that we can't expect exactly when the weakref.ref callback
will be called?

With warmest regards,

 Jane Austine.
------------------------------------
[1] http://sourceforge.net/tracker/download.php?group_id=5470&atid=305470&file_id=14021&aid=487738


From peter at engcorp.com  Mon Jun 10 23:43:47 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 10 Jun 2002 23:43:47 -0400
Subject: Correct code/unit testing (was Re: Efficient python programming...)
References: <mailman.1023762854.21527.python-list@python.org>
Message-ID: <3D0571F3.ED5B7F@engcorp.com>

"Delaney, Timothy" wrote:
> 
> > From: Peter Hansen [mailto:peter at engcorp.com]
> > Perhaps, assuming you put a lot of bugs in the code, some
> > amount of staring will be necessary.
> >
> > I believe staring is of little or no use, however, having
> > tried it many times with my and others' code which had bugs and
> > not having found those bugs by staring.
> 
> Staring in and of itself is not of much use ... however, simply getting
> another set of eyes on the code is. I have found that in the majority of
> cases, one of two things will occur:
> 
> 1. The new set of eyes will immediately see the thing you've been failing to
> see for the past 2 hours ...
> 
> OR
> 
> 2. While explaining to the new set of eyes enough for them to understand the
> problem, *you* will see what you've been failing to see for the past 2 hours
> ...

Agreed!  Of course, in XP that's called "pair programming" and you do
it all the time on production code, not just after you've wasted two
fruitless, frustrating hours. :-)

-Peter


From johnroth at ameritech.net  Thu Jun 20 07:08:29 2002
From: johnroth at ameritech.net (John Roth)
Date: Thu, 20 Jun 2002 07:08:29 -0400
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu>
Message-ID: <uh3drhh14us0eb@news.supernews.com>

"Tom Loredo" <loredo at astro.cornell.edu> wrote in message
news:3D110B55.4D8AD7FA at astro.cornell.edu...
>
> Hi folks-
>
> I've been using Python heavily on a Solaris box and a Mac for a
> few years now.  For reasons I'll spare you from, I now am venturing
> into Windows territory (a Dell Inspiron 7500 laptop, if that matters).
> I see at least two Windows versions of Python---that at Python.org,
> and the ActiveState version.  The obvious "pro" for the ActiveState
> version is simplicity of the install (batteries included, etc.).
> But what about after that?  Are their substantial differences between
> the versions that I will bump into down the line?  In particular,
> I write many C extensions; can one use distutils and the Borland
> or gcc free compilers to build C extensions in both versions?
>
> Thanks for your patience and advice for this PythonWin newbie!
>
> -Tom Loredo

You should be able to use any compiler you want to - .dll's are
supposed to be language neutral, at least in the low level sense
of accessing them. As another poster pointed out, you might have
some problems with object files, since they aren't necessarily
compatible between different vendor's products.

John Roth




From shredwheat at attbi.com  Mon Jun  3 11:44:29 2002
From: shredwheat at attbi.com (Pete Shinners)
Date: Mon, 03 Jun 2002 15:44:29 GMT
Subject: solarwolf display ?
References: <slrnafl9gv.5h4.Taekyon@macchinetta.miadimora>
Message-ID: <3CFB8ED2.2070507@attbi.com>

Tae kyon wrote:
> On a Mandrake 8.2 system I compiled pygame 1.5 and solarwolf 1.5.1.
> I have a matrox millennium g200 8 Mb video ram graphics card with 3d support 
> by XFree86 4.2.0, and a SVGA non-interlaced video with HSync 30-54,
> VSync 50-90. I even get to play tuxracer at a good framerate ! But for
> some reason 800x600 fullscreen displays will not work (some loki games
> have to be played in a window, for instance, and nethack must be played
> on a 1024x768 display).
> 
> Unfortunately it seems that solarwolf uses this type of display.
> I know very little about python, nothing about pygame; anyway I tried
> setting the video mode by changing the relevant options in 
> pygame.display.set_mode(()) ... but nothing seems to work.

i know it's not documented, but try running solarwolf with the "-window" 
flag and it will run in a window. strange that your games will run at 1024x 
but no 800x. Pygame uses SDL, and if SDL cannot change to the desired video 
mode it will pick a bigger one and 'letterbox' the display.





From pyth at devel.trillke.net  Tue Jun 11 09:32:30 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 11 Jun 2002 15:32:30 +0200
Subject: Defining a method final
In-Reply-To: <phmN8.36164$K4.893500@twister2.libero.it>; from fcorneti@libero.it on Tue, Jun 11, 2002 at 12:45:09PM +0000
References: <6glN8.35376$86.884025@twister1.libero.it> <mailman.1023796095.1427.python-list@python.org> <phmN8.36164$K4.893500@twister2.libero.it>
Message-ID: <20020611153230.R6609@prim.han.de>

Hello Fabio,

[you explained]
> I would like to define a method for a class which has
> not to be overriden. Of course, since the developer of
> this project are two, this could be established by
> convention, but if the project will grow consistently 
> another developer could override the method by mistake,
> causing unpleasant side effects.

I see. But did you know that usually pythonic objects work like this:

class stream:
    def readline(self):
        return ..someline...

class filestream:   # no inheritance needed!
    def readline(self):
        return ..readfromfile...

and that a method wanting to have an instance which it can
call readline on, would not need to check types etc...

def tokenize(provider):
    l=provider.readline()

works with a stream or a filestream-instance, it even works
with a 'file' instance or a StringIO-instance :-)

I imagine that this might sound fragile to you. But you
spare yourself and other a *lot* of effort if you don't
insist on exact type checking but rather use the
*behaviour* of objects. After all, working with behaviour 
is the great promise of OOP and python has it. Java/C++
would require you to group all objects into an
inheritance-hierarchy which usually makes things *more* 
difficult and verbose. 

OTOH for big enough programs (say >50.000 LOC) you should:

- write many tests which thoroughly check for desired behaviour
  (ask Peter Hansen for more details:-)

- dig deeper into 'interface' concepts for python. It's 
  also discussed on python-lists from time to time...

not that someone will certainly point out, that you should
write tests for smaller programs, too :-)

regards,

    holger




From christophe.delord at free.fr  Thu Jun 20 13:21:46 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Thu, 20 Jun 2002 19:21:46 +0200
Subject: How to represent the infinite ?
References: <3d11ee9b$1@news.swissonline.ch>
Message-ID: <20020620192146.0487f7a0.christophe.delord@free.fr>

On Thu, 20 Jun 2002 17:02:33 +0200
"erreur" <dev at newdeal.ch> wrote:

You can try to redefine __cmp__ to handle all the cases.

For example :

Inf(+1) is +oo
Inf(-1) is -oo

class Inf:
	def __init__(self, sign=+1):
		self.sign = sign
	def __str__(self):
		return self.sign<0 and "-Inf" or "+Inf"
	def __cmp__(self, other):
		if isinstance(other, Inf):
			return cmp(self.sign, other.sign)
		else:
			return self.sign
	def __pos__(self): return Inf(self.sign)
	def __neg__(self): return Inf(-self.sign)

inf = Inf()
for t in ( \
	"+inf", "-inf",
	"10<-inf", "10>-inf", "10<+inf", "10>+inf",
	"-inf<10", "-inf>10", "+inf<10", "+inf>10",
	"-inf<-inf", "-inf==-inf", "-inf<+inf", "-inf==+inf",
	"-inf>=-inf", "-inf>=+inf"):
	print t, "=", eval(t)

This gives:

+inf = +Inf
-inf = -Inf
10<-inf = False
10>-inf = True
10<+inf = True
10>+inf = False
-inf<10 = True
-inf>10 = False
+inf<10 = False
+inf>10 = True
-inf<-inf = False
-inf==-inf = True
-inf<+inf = True
-inf==+inf = False
-inf>=-inf = True
-inf>=+inf = False

You may also redefine __add__, __mul__, ...


Best regards,
Christophe.

> Will somebody have an idea, to represent the infinite one?
> 
> I have variables to initialize with is +inf (or - inf).  To be sure that
> later, all will be smaller (or larger) than my variables.
> I tried to redefine the operators on an object.  It goes for Inf>10 but I do
> not arrive for 10>Inf (because that takes > of Int).
> 
> --------------------
> 
> def __gt__(self, val):
> 
>     return 1
> 
> --------------------
> 
> 
> Not need to say to me that I can use 99999999999999 or -99999999999999 or I
> do not know what.
> Nor from a MAX_INT of the machine... bus Python can go higher.
> 
> Thank you for you answers (thank you to also answer by email if possible,
> because I will not have any more access to the news this weekend).
> 
> PS: Sorry for my English....
> 
> 


-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From peter at engcorp.com  Mon Jun 24 22:32:54 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 24 Jun 2002 22:32:54 -0400
Subject: adding attributes to builtin functions
References: <Gy8Fr6.J7D@beaver.cs.washington.edu> <3D17CE49.E3D16ABE@engcorp.com> <Gy8oFx.qEw@beaver.cs.washington.edu>
Message-ID: <3D17D656.B1B667A7@engcorp.com>

Douglas Zongker wrote:
> 
> Well, that works, but what I *really* wanted to do was to set the
> attributes from within C.  I tried something like this:
> [...]
> but it gives the same exception as doing it in Python ("builtin
> function has read-only attributes").  I was hoping if someone could
> tell me how to do it in Python I could translate that back into C...

Well, since as it says the item has read-only attributes, I 
don't suppose you _can_ do it, even in Python.  Or you can
do what I showed, but from C.  

Maybe at this point you need to ask and answer the question "why?".  
There's often a better approach when you get stumped by something
like this.  What's your real goal in doing this?

(I've never had a use for function attributes and I'm not sure 
I ever will.  Maybe it's just me... I do tend to avoid anything
that's not just "vanilla" code though, to keep myself sane.)

-Peter


From CousinStanley at HotMail.Com  Wed Jun 12 07:48:17 2002
From: CousinStanley at HotMail.Com (Cousin Stanley)
Date: 12 Jun 2002 11:48:17 GMT
Subject: PIL _imaging
References: <Xns922AA829A8C5FHotMailcs@130.133.1.4> <3D06E694.4040103@mxm.dk>
Message-ID: <Xns922B3150A6B32HotMailcs@130.133.1.4>

> Cousin Stanley wrote:
> 
> 
>>   There are 2 files in K:\Python\Lib\PIL\DLLs
>>   but apparently these are not the ones that are needed ...
>> 
>>     _imaging.pyd
>>     _imagingtk.pyd 
> 
> 
> Put those two files into:
> K:\Python\DLLs
> 
> And you should be good to go.
> 
> regards Max M
> 
> 

Cousin Max M ... 

Thanks for the reply ... 

I copied the 2 files into the  K:\Python\DLLs  folder
and tried again, but the result is the same, leading to
the same error I originally posted ...  

   python Samples/viewer.py ButtonD.jpg

     File "K:\Python\Lib\PIL\PIL\Image.py", line 37, in __getattr__
        raise ImportError, "The _imaging C module is not installed"
     ImportError: The _imaging C module is not installed

I also tried adding  K:\Python\DLLs  
to the PYTHONPATH environment variable
in the autoexec.bat file ... no luck ... 

There is a file  K:\Python\Lib\PIL\PIL.pth  
that contains only the single 3 character entry PIL ... 

I've also tinkered with this file 
trying various path combinations
with no luck ...

Cousin Stanley


From tdelaney at avaya.com  Sun Jun  2 21:26:33 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 3 Jun 2002 11:26:33 +1000 
Subject: Why no open(f, "w").write()?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A33D@natasha.auslabs.avaya.com>

> From: Andreas Kostyrka [mailto:andreas at kostyrka.priv.at]
>
> > Me ...
> >
> > __files = []
> >
> > def open (path, mode):
> >     f = __open(path, mode)
> >     __files.append(f)
> >     return f
> >
> > or
> >
> > class file:
> >
> >     __files = []
> >
> >     def __init__(self, path, mode):
> >         __files.append(self)
> >
> > Whilst I know this is not the implementation, there is nothing that
> > prevents it from being like this. So you get back an object 
> that has one
> > more reference count than expected. It never gets collected.

> Well, then the implementation is broken. If you allow the 
> implementation
> to collect additional references at random places, you can prove quite
> easily that nothing is safe :)

Not at all. My implementation conforms 100% to the documented behaviour of
the function/type. I should be able to replace __builtin__.file with the
above implementation and expect everything to work (except that I may run
out of memory - but that *also* conforms 100% to documented behaviour).
Indeed, in 2.2.8 File Objects there is in fact no mention that the file
object will be closed when its destructor is called. I searched through much
of the rest of the docs and found no reference to a file object destructor.
Therefore, my above implementation of class file is completely correct in
that it does not have a destructor.

> > You should never rely on automatic freeing of *any* 
> resources unless it is
> > guaranteed. Always explicitly free resources when you have 
> finished with
> > them.
> Well, then you should del all your names too.
> Oops wait, how can we be sure that del works? What if del 
> also increments the 
> refcount instead of decrementing?

"""
__del__(self)

   Called when the instance is about to be destroyed. This is also
   called a destructor  If a base class has a __del__() method, the
   derived class's __del__() method must explicitly call it to ensure
   proper deletion of the base class part of the instance. Note that it
   is possible (though not recommended!) for the __del__() method to
   postpone destruction of the instance by creating a new reference to
   it. It may then be called at a later time when this new reference is
   deleted. It is not guaranteed that __del__() methods are called for
   objects that still exist when the interpreter exits. 

   Note: "del x" doesn't directly call x.__del__() -- the former
   decrements the reference count for x by one, and the latter is only
   called when its reference count reaches zero. Some common situations
   that may prevent the reference count of an object to go to zero
   include: circular references between objects (e.g., a doubly-linked
   list or a tree data structure with parent and child pointers); a
   reference to the object on the stack frame of a function that caught
   an exception (the traceback stored in sys.exc_traceback keeps the
   stack frame alive); or a reference to the object on the stack frame
   that raised an unhandled exception in interactive mode (the traceback
   stored in sys.last_traceback keeps the stack frame alive). The first
   situation can only be remedied by explicitly breaking the cycles; the
   latter two situations can be resolved by storing None in
   sys.exc_traceback or sys.last_traceback. Circular references which
   are garbage are detected when the option cycle detector is enabled
   (it's on by default), but can only be cleaned up if there are no
   Python-level __del__() methods involved. Refer to the documentation
   for the gc module for more information about how __del__() methods
   are handled by the cycle detector, particularly the description of
   the garbage value. 

   Warning: Due to the precarious circumstances under which __del__()
   methods are invoked, exceptions that occur during their execution are
   ignored, and a warning is printed to sys.stderr instead. Also, when
   __del__() is invoked in response to a module being deleted (e.g.,
   when execution of the program is done), other globals referenced by
   the __del__() method may already have been deleted. For this reason,
   __del__() methods should do the absolute minimum needed to maintain
   external invariants. Python 1.5 guarantees that globals whose name
   begins with a single underscore are deleted from their module before
   other globals are deleted; if no other references to such globals
   exist, this may help in assuring that imported modules are still
   available at the time when the __del__() method is called.
"""

An implementation of del which failed to decrement the refcount by one would
not conform to the documented behaviour, and hence would be broken.

Also, in reference to the rest of this thread, note the sentence "It is not
guaranteed that __del__() methods are called for objects that still exist
when the interpreter exits."

> Basically what you have proven is, that if you have a faulty open() 
> implementation, the open(f,t).write(d) idiom doesn't work.

So is the Jython implementation faulty? The open(f, t).write(d) idiom does
not work there (it may, if you get lucky).

> In respect to CPython, refcounting behaviour, or more general 
> retaining 
> references in any gc regime, is a quite important part of the 
> interface.

It is explicitly not a documented part of the interface - the documented
part is that it is not safe to rely on this behaviour.

When it is explicitly documented that this behaviour is guaranteed now and
forever, for all implementations of Python after a specified version, I will
*consider* using this behaviour. Until then it is not worthy of
consideration.

Tim Delaney




From rjones at ekit-inc.com  Thu Jun  6 19:40:34 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Fri, 7 Jun 2002 09:40:34 +1000
Subject: Medium to Large Scale Python Deployments
In-Reply-To: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com>
References: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com>
Message-ID: <200206070940.35245.rjones@ekit-inc.com>

On Fri, 7 Jun 2002 09:22, Sam Penrose wrote:
> [Domenic R. Merenda raises the topic of Python used in large contexts]
>
> I'm also curious about large amounts of Python, period. Who knows of
> ~50K Python lines-of-code (LOC) systems? 100K LOC? More?

I'm just taking a punt here, but I'd say we've got around 50k or more lines in 
action here (http://www.ekit.com/). It's hard to tell, 'cos there's stuff in 
Zope and it doesn't play well with "wc" :)


   Richard





From not.this at seebelow.org  Fri Jun 14 13:43:00 2002
From: not.this at seebelow.org (Grant Griffin)
Date: 14 Jun 2002 10:43:00 -0700
Subject: why not "'in' in 'in'"?
References: <mailman.1024019967.16051.python-list@python.org>
Message-ID: <aed9v40e68@drn.newsguy.com>

In article <mailman.1024019967.16051.python-list at python.org>, Tim says...
>
>[Grant Griffin]
>> One of the things I like about Python is that it works as expected.
>> However, I sometimes accidently expect it to allow one to use "in"
>> to determine if one string is "in" another, e.g.:
>>
>> >>> 'in' in 'in'
>>
>> which yields a traceback:
...
>So if somebody wants this change enough to submit a patch (code, doc, + test
>suite changes), it's likely to be accepted for 2.3.  As code changes go,
>this one should be particularly easy.  We have an internal bet going as to
>whether this invitation is enough to kill the idea <wink>.

Thanks for the info, Tim.  I'm not in a position to do this work myself, so if
somebody else wants to do it, I don't mind.  (Besides, I was just asking a
_question_ <wink>.)

But until this change magically appears, just knowing that there is no grand
philosophical reason why something as intuitive and practical as this must be
excluded from Python restores my faith.

faithful-ly y'rs,

=g2

_________________________________________________________________________

Grant R. Griffin                                           g2 at dspguru.com
Publisher of dspGuru                               http://www.dspguru.com
Iowegian International Corporation                http://www.iowegian.com



From jonathan at onegoodidea.com  Fri Jun 28 05:21:25 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Fri, 28 Jun 2002 10:21:25 +0100
Subject: Python 2.2 __slots__ Bug?
References: <mailman.1025236118.663.python-list@python.org>
Message-ID: <B941E925.D1AF%jonathan@onegoodidea.com>

On 28/6/2002 4:47, in article mailman.1025236118.663.python-list at python.org,
"Tim Peters" <tim.one at comcast.net> wrote:

>> ...
>> I can't see an easy way to fix this without changing the way that the
>> "delete later" list is constructed.
> 
> Sometimes we just agree to pretend that we never saw the bug report <wink>.

Actually I think I just figured out an easy way to fix the problem.

I was thinking that if there had been a liberal sprinkling of dicts in the
chain then it wouldn't overflow. But there's no way of doing that from the
Python side of things.

Then it occurred to me that it could be done from the C side of things:

Surround the slots deallocation in 'subtype_dealloc' with Py_TRASHCAN type
macros with the modification that if the recursion limit is hit, we create a
new tuple and set its items to be the contents of the slots, then DECREF it,
set the slots to NULL, and continue. The tuple deallocating code will see
that the recursion limit has been hit and add itself to the delete later
list.

This has the effect of inserting a container into the chain every few items
which can take part in the stack unwinding and restarting magic.

Simple and effective, no?

If I want to have a go at this, should I just upload a patch to Sourceforge,
or should I email it to someone?

Jonathan



From cliechti at gmx.net  Mon Jun  3 18:22:00 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 4 Jun 2002 00:22:00 +0200
Subject: Time
References: <3cfbdf0c_1@news.iprimus.com.au>
Message-ID: <Xns922341B4698Ccliechtigmxnet@62.2.16.82>

Gold Fish <occeanlinux at linuxmail.org> wrote in
news:3cfbdf0c_1 at news.iprimus.com.au: 

> Anyone know how to determine the time in python. If the mtime in the
> file filename.py is Mon Jun 3 03:25:57 2002 ,if i want to set the time
> forward or backwards in 5 days time, how can i do that. 

the "time" module has various functions to manipulate times in varous 
formats.

to add an offset, convert to seconds since epoch, manipulate and convert 
back:

>>> time.strftime("%a, %d %b %Y %H:%M:%S",
...    time.localtime(time.time()+24*3600*5))

chris

-- 
Chris <cliechti at gmx.net>



From ljd at nospam.com  Sat Jun  8 17:49:18 2002
From: ljd at nospam.com (LJD)
Date: Sat, 08 Jun 2002 21:49:18 GMT
Subject: Program mysteriously STOPS!
Message-ID: <yZuM8.146064$ux5.186314@rwcrnsc51.ops.asp.att.net>

Hi all:

The following code is designed (at least thats my intent) to run forever.  
The program should sit in a loop until it connects with the server.  It 
should just loop and loop and loop. 

It does, for a while, maybe five or ten minutes, but then the process just 
seems to vanish.  It's in the c1==0 loop when it quits. When I do a PS AUX, 
the process is just gone.  Obviously it quit, but more importantly WHY???, 
and what can I do so it doesn't?

Thanks,
Larry

------------------------------------------------------------------------------
(the code)

# Setup Parameters
HostIP = "192.168.0.3"
HostPort = 5020
HB_Interval = 1.0

############################################################################################################################

import socket
import time
import select

c1=0
c2=1

while c1 == 0:
        try:
                s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
                s.connect(HostIP, HostPort)
                c1=1
                c2=1

        except socket.error,e:
                print e
                
                fe=open('/share/details.txt','a')
                t=time.time()
                fe.write(strftime('%H',t)
                fe.close()
                
                print time.time()
                time.sleep(5.0)
                c1=0
                c2=0

        while c2 == 1:
                EL=time.time()

                try:
                        r,w,e = select.select([s], [s], [s])
                        if w:
                                s.send(chr(32) + chr(5) + chr(0) + chr(8) +
 chr(0) + chr(0) + chr(0) + chr(0))

                        if r:
                                data=s.recv(50000)
                                tuple = time.localtime(time.time())
                                YR = time.strftime('%Y',tuple)
                                MO = time.strftime('%m',tuple)
                                DA = time.strftime('%d',tuple)
                                HR = time.strftime('%H',tuple)

                                f=open('/share/F' + YR + '_' + MO + '_' + D
A + '_' + HR + '_data.raw', 'ab')
                                f.write(data)
                                f.close()

                except socket.error,e:
                        print e
                        c2=0
                        c1=0
                        s.close()

                DL=time.time()
                TIL=DL-EL
                TR=HB_Interval-TIL
                if TR > 0:
                        time.sleep(TR)



From ak at silmarill.org  Sat Jun  1 10:38:46 2002
From: ak at silmarill.org (Andrei Kulakov)
Date: Sat, 01 Jun 2002 14:38:46 GMT
Subject: wxPython performance
Message-ID: <slrnafhmoa.389.ak@ak.silmarill.org>

Hello,

    "hello world" type program from wxPython tutorial takes 6
seconds to start on my cel 366 128mb system (running Debian). Is
this typical? Why is it so slow? I'm looking for a graphic
toolkit that'd be fairly quick. I don't want to use Tk because of
the looks.

 - Andrei

-- 
Cymbaline: intelligent learning mp3 player - python, linux, console.
get it at: cy.silmarill.org


From nessus at mit.edu  Fri Jun 21 18:26:32 2002
From: nessus at mit.edu (Douglas Alan)
Date: 21 Jun 2002 18:26:32 -0400
Subject: Doing a ps in windows.. (Is application XYZZY running?)
References: <m38z8g74hw.fsf@localhost.localdomain> <nnmv9uc6gtkge4at1cfiv0rbijsol31oc4@4ax.com> <lchejw5lez.fsf@gaffa.mit.edu>
Message-ID: <lcd6uk5kvb.fsf@gaffa.mit.edu>

I wrote:

> It seems that the list of processes is being cached, and then
> EnumObjectItems just returns the value from the cache.  How do I
> flush the cache?

Nevermind -- I figured it out myself.  To refresh the cache, you do:

   win32pdh.EnumObjects(None, None, 0, 1)

|>oug


From jepler at unpythonic.net  Wed Jun 26 14:25:02 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 26 Jun 2002 13:25:02 -0500
Subject: Python 2.2 __slots__ Bug?
In-Reply-To: <3D1A057F.A81FF84@alcyone.com>
References: <mailman.1025089474.17780.python-list@python.org> <3D1A057F.A81FF84@alcyone.com>
Message-ID: <20020626182502.GB3710@unpythonic.net>

On Wed, Jun 26, 2002 at 11:18:39AM -0700, Erik Max Francis wrote:
> Glyph Lefkowitz wrote:
> 
> > I get "I'm fine" and then a segmentation fault.  This only happens
> > when
> > __slots__ is defined on a new-style class.  Seriously, i'm not making
> > those
> > numbers up -- reliably, 43551 objects will crash the interpreter;
> > fewer will
> > not.
> 
> Hmm, the script runs fine on my Slackware 8.0 system, even if I crank
> the number up to 100000.

When run in a thread, it dies on my redhat 7.3 machine.  When not run in
a thread, it seems to succeed even for very large values.

Non-main threads are presumably started with a smaller stack..

Jeff




From whisper at oz.net  Thu Jun 27 19:19:51 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 27 Jun 2002 16:19:51 -0700
Subject: How to find out operating system
In-Reply-To: <9KLS8.45365$n4.10750643@newsc.telia.net>
Message-ID: <GCEDKONBLEFPPADDJCOEAEJKDMAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Fredrik Lundh
> Sent: Thursday, June 27, 2002 14:48
> To: python-list at python.org
> Subject: Re: How to find out operating system
> 
> 
> David LeBlanc wrote
> 
> > > > > You might try os.environ:
> > > > > os.environ['WINOS'] -> WIN2000
> > > > >
> > > > > I don't know if WINOS is defined on '98 or ME, but
> > > > > it should be good for NT, 2K and XP
> > > >
> > > > It is not.
> > >
> > > Neither is it defined on win2k pro:
> > > KeyError: WINOS
> 
> > >>> import os
> > >>> os.environ['WINOS']
> > 'WIN2000'
> 
> on the other hand:
> 
> c:\> ver
> Microsoft Windows 2000 [Version 5.00.2195]
> 
> (swedish version, all available service packs and
> security patches)
> 
> c:\> python
> Python 2.1.1 (#20, Aug 23 2001, 11:27:17) [MSC 32 bit (Intel)] on win32
> Type "copyright", "credits" or "license" for more information.
> >>> import os
> >>> os.environ["WINOS"]
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "C:\py21\Lib\os.py", line 372, in __getitem__
>     return self.data[key.upper()]
> KeyError: WINOS
> 
> (fwiw, googling for WINOS brought up a whole bunch of
> pages telling me that if I want to build mozilla, I need to
> set the WINOS variable.  are you sure your computer
> doesn't spend the nights building browsers, when you're
> not looking?)
> 
> </F>

Ewps... I have been known to compile the lizard of browsers

David LeBlanc
Seattle, WA USA 





From zopestoller at thomas-guettler.de  Tue Jun 25 11:01:42 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Tue, 25 Jun 2002 17:01:42 +0200
Subject: Python and Eclipse
References: <3D17FAF4.35CC9F1B@earthlink.net>
Message-ID: <3D1885D6.2080802@thomas-guettler.de>

Joseph A Knapka wrote:

> Hi folks,
> 
> I am interested in having Python support in
> Eclipse <URL: http://www.eclipse.org> - interested
> enough to contribute code, or to tackle the
> task myself, if need be. Of course, I'd like
> the Python support to be at least as good as the
> Java support, which is a pretty tall order,
> 'cause the Java support in Eclipse is awfully
> damn good --- it nearly allows me to *enjoy*
> writing Java code :-)


Can someone tell me what eclipse is about?

Someone said it is IBM's intend spread the use of
WebSphere and someone else said it is an IDE.

The homepage does not say much, exept that a lot of
big company are in this consortium.

I wonder why SUN is not in the concortium.

  thomas







From op73418 at mail.telepac.pt  Mon Jun 24 08:51:36 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Mon, 24 Jun 2002 13:51:36 +0100
Subject: Generic Python
References: <3D16E465.F0D2509E@ifib.uni-karlsruhe.de>
Message-ID: <u85ehuo7mf7dccu1ghdsq7l6b7ujrlljns@4ax.com>

On Mon, 24 Jun 2002 11:20:37 +0200, Uwe Mayer
<Uwe.Mayer at ifib.uni-karlsruhe.de> wrote:

>Hi,
>
>is it possible to to write a class which, f.e. takes an argument in
>__init__() and that doesn't return an instance object but a new class
>object?
>
>The problem is that I've got a base class and there should be many
>subclasses of it. Each subclass just overwriting an output method and
>serving as a template: 
>
>class base1:
>    def __init__(self):
>        ...
>    def output(self):
>        ...
>
>class template1(base1):
>    def output(self):
>        ...
>
>class template2(base1):
>    def ouptut(self):
>        ...
>
>I need many, many of these "template" classes, so a sorter way would be
>to accustom the base class to take another argument which then outputs
>the right thing:
>
>
>class base2:
>    def __init__(self, text):
>       ...
>    def output(self):
>       print self.text
>
>template1 = base2('some text')
>template2 = base2('another example')
>...
>
>My problem with this is that this produces instance objects and in the
>above example I had class objects. I can hardly use the latter case
>because I would always modify the original.
>I also cannot create new templates as I need them, because I'm writing
>utility classes which are much more complex and take much more
>arguments.
>
>Is it somehow possible to have the "base2" class return a class instead
>of an instance?
>Is there another way to solve this problem?
>
>Thanks in advance
>Uwe

This leads into metaclass programming. Read GvR's descintro on Python
2.2, it has some material - it may be all you need. Roughly you need
something like

class Whatever(type):
    ...

type is the mother of all metaclasses, its instances are classes
themselves.

Hope it helped, all the best,
Gon?alo Rodrigues 



From cben at tx.technion.ac.il  Mon Jun 10 01:57:15 2002
From: cben at tx.technion.ac.il (Beni Cherniavksy)
Date: Mon, 10 Jun 2002 08:57:15 +0300
Subject: Does Python need a '>>>' operator?
In-Reply-To: <a9gnvs$42p$0@216.39.172.122>
References: <mailman.1018919132.7451.python-list@python.org>
 <a9gnvs$42p$0@216.39.172.122>
Message-ID: <Pine.GSO.4.33.0206100850550.12342-100000@techst02.technion.ac.il>

[OK, that thread is long dead but I din't notice how many miles of
massages I'm behind when I started reading ;-]

On 2002-04-16, Bengt Richter wrote:

> On Mon, 15 Apr 2002 21:03:45 -0400, Tim Peters <tim.one at comcast.net> wrote:
>
> >In practice, if I'm working with 300-bit bitsets represented as longs, I
> >never do hex(x), but hex(x & mask) instead, where mask = ~(-1L << 300) (a
> >string of 300 low-order bits).  Regardless of x's sign, that "shows the
> >bits" in a 1-means-on 0-means-off sense.  The key is that I *know* the width
> >I have in mind, much as hex(int) knows the width of a platform C long today,
> >so can pull the same trick for regular ints.
> >
> Sure. But you can also pick a width for minus_sign-less-but-not-sign_bit-less
> hex purposes by defining a normalized width that encompasses just enough bits
> to show a hex 0 or F as the first digit, knowing that all to the left are
> redundant duplication. (Credit to Greg Ewing for the suggestion of making sure
> it's always a reader-friendly leading 0 or F to indicate sign, vs. my ensuring
> just an msb of 0 or 1 within the ms hex digit).
>
I just got another idea: use 0x1234 for 0-filled numbers and 1xABCD for
1-filled ones.  That way you impose no restrictions on what follows the
prefix and keep backward compatibility.  0xFFFFFFFF stays a 2^n-1
_positive_ number, as it should be.  The look of 1x is weird at first but
it is very logical...

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>



From talbotdr at starband.net  Thu Jun 13 16:58:58 2002
From: talbotdr at starband.net (Dave)
Date: Thu, 13 Jun 2002 20:58:58 GMT
Subject: Python 2.2.1
Message-ID: <II7O8.622$e27.275217555@twister1.starband.net>

    Hi, I am 12 years old and would like to begin learning Python over the summer. I have Windows 98 SE. I have python 2.2.1 and all of its utilities on my computer. I am taking a tutorial from  http://softwaredev.earthweb.com/sdopen/article/0,,12382_626311,00.html  If you go to it and scroll down to the paragraph entitled Getting Started there is a paragraph that says, "You will need to cause the directory containing the file named python.exe to be listed in your system environment variable named path. If you already know how to set the path, go ahead and do it.  If you don't already know how, you may need to get some help." 


    How do I set the path and what should the name of the path be? All help and information would be appreciated. Thanks.:)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020613/89b1b133/attachment.html>

From Marcel.Opsteegh at nob.nl  Wed Jun  5 08:40:36 2002
From: Marcel.Opsteegh at nob.nl (Marcel Opsteegh)
Date: Wed, 5 Jun 2002 14:40:36 +0200 
Subject: Urllib thread safe?
Message-ID: <F1DB8BE87BC3D311A693009027DCE1C54EA3E9@dasboot.int.nob.nl>

Does somebody know whether the urllib is thread safe? 
I use the following call: urllib.URLopener().open(...)

Thanks,
Marcel.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020605/ced7d45f/attachment.html>

From whisper at oz.net  Fri Jun 28 21:56:12 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 28 Jun 2002 18:56:12 -0700
Subject: Python needs better error reporting
Message-ID: <GCEDKONBLEFPPADDJCOEEEOLDMAA.whisper@oz.net>

"Syntax Error: invalid syntax" isn't very informative, yet it's tossed out
frequently. HOW is the syntax invalid? For example:
if s == ' '

David LeBlanc
Seattle, WA USA





From shalehperry at attbi.com  Tue Jun  4 14:35:25 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Tue, 04 Jun 2002 11:35:25 -0700 (PDT)
Subject: directory
In-Reply-To: <3cfcfd4a_1@news.iprimus.com.au>
Message-ID: <XFMail.20020604113525.shalehperry@attbi.com>

On 04-Jun-2002 Gold Fish wrote:
> How to read the subdirectories in the directory
> Support i have the directory /home/ which have some subdirectories
> /home/sub1
> /home/sub2
> /home/sub3
> I am using os.listdir('/home/') to read in the list so it will be 
> ['sub1','sub2','sub3'] now i want to read inside the sub directory sub2
> once again i using os.listdir(os.listdir('/home/')[1]) but it came out with 
> an errors: not a file or directory. Can anyone tell me what did i wrong. 
> Anh how to fix it.

notice the values given by listdir('/home/') -> 'sub1', 'sub2'.  listdir()
expects full path to a file.  You need to do
listdir(os.path.join('/home', 'sub1')).  More likely a better option would be
to use os.path.walk().




From mcfletch at rogers.com  Fri Jun 21 00:19:52 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Fri, 21 Jun 2002 00:19:52 -0400
Subject: newbie string question
References: <aeu40s$9jg2s$1@ID-145503.news.dfncis.de> <3D129BB7.C7EA15D2@engcorp.com> <aeu76k$9brq1$1@ID-145503.news.dfncis.de>
Message-ID: <3D12A968.8040503@rogers.com>

pystr[3:]

is equivalent to:

pystr[3:len(pystr)]

and

pystr[:3]

is equivalent to:

pystr[0:3]

That is, when you leave off a slice index, the "end of the sequence in 
that direction" is used.  Thus:

pystr[ : ]

is:

pystr[0:len(pystr)]

HTH,
Mike

Don Low wrote:
...
> So if I understand, pystr[3] refers to the subscript index, but pystr[3:]
> refers to the slice index. pystr[2:5] means slice at slice index 2 and 5,
> not start at subscript index 2 and slice at slice index 5.
> 
> I just want to understand this once and for all.
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From shagshag13 at yahoo.fr  Sun Jun  2 06:07:36 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Sun, 2 Jun 2002 12:07:36 +0200
Subject: Pop a list from beginning ? and memory saving...
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de> <3CF9EB89.7CA7204B@doc.ic.ac.uk>
Message-ID: <adcqon$1012ko$1@ID-146704.news.dfncis.de>

> You could try a.pop(0), which pop the first instance of the list, like a FIFO.
>
> Thanks
>
> Ben

and is it a way to free memory ?

thanks,

s13.




From jepler at unpythonic.net  Wed Jun 12 18:10:12 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Wed, 12 Jun 2002 17:10:12 -0500
Subject: why not "'in' in 'in'"?
In-Reply-To: <ae8dg70217g@drn.newsguy.com>
References: <ae8dg70217g@drn.newsguy.com>
Message-ID: <20020612171011.A1215@unpythonic.net>

If
    'in' in 'in' == 1
then what about
    [1, 2] in [1, 2]
?  How do you tell that situation from
    [1, 2] in [[1, 2]]
?

The members of a string are characters, not arbitrary substrings.

Jeff




From peter at engcorp.com  Fri Jun 21 00:41:03 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 21 Jun 2002 00:41:03 -0400
Subject: newbie string question
References: <aeu40s$9jg2s$1@ID-145503.news.dfncis.de> <3D129BB7.C7EA15D2@engcorp.com> <aeu76k$9brq1$1@ID-145503.news.dfncis.de>
Message-ID: <3D12AE5F.26999DBB@engcorp.com>

Don Low wrote:
> 
> >    0   1   2   3   4   5    <-- subscript index
> >  +---+---+---+---+---+---+
> > "| P | y | t | h | o | n |"
> >  +---+---+---+---+---+---+
> >  0   1   2   3   4   5   6  <-- slice index
> >
> So if I understand, pystr[3] refers to the subscript index, but pystr[3:]
> refers to the slice index. pystr[2:5] means slice at slice index 2 and 5,
> not start at subscript index 2 and slice at slice index 5.
> 
> I just want to understand this once and for all.

Exactly right.  And as Fernando reminded us, negative indexes
can be used as well, but with similar interpretation:

   -6  -5  -4  -3  -2  -1    <-- negative subscript index
  +---+---+---+---+---+---+
 "| P | y | t | h | o | n |"
  +---+---+---+---+---+---+
 -6  -5  -4  -3  -2  -1   :  <-- negative slice index

so pystr[-5:-2] is "yth", pystr[-4:] is "thon",
and pystr[-5] is just "y".

You can start to confuse yourself again, maybe :-), by
combining positive and negative slice indices:

   pystr[-5:4]

The two diagrams above make it clear what happens tho. ;-)

-Peter


From not.this at seebelow.org  Thu Jun 13 01:15:14 2002
From: not.this at seebelow.org (Grant Griffin)
Date: Thu, 13 Jun 2002 00:15:14 -0500
Subject: why not "'in' in 'in'"?
References: <ae8dg70217g@drn.newsguy.com> <3D080482.7A7FD514@alcyone.com>
Message-ID: <3D082A62.CF800C75@seebelow.org>

Erik Max Francis wrote:
> 
> Grant Griffin wrote:
> 
> > So why is the "in" operator limited to single characters when used as
> > a
> > membership operator?  This peculiar lack of generality seems a bit
> > (dare I say
> > it?) "non-Pythonic".

> 
> On the contrary, it is exquisitely Pythonic.

(I sure hope my "enlightenment" light will go on one day.  Until then, I
guess I'll just have to bask in the warm glow of others <wink>.)

> You used the magic word
> yourself:  _membership_.

In that case, perhaps I should have thought of a better one <wink>.  In
any event, in the case of strings, I think one might stretch the concept
of "membership" to include substrings.  In fact, note that since Python
has no actual character type (because a character is adequately
represented as a string of length one) the "in" operator actually _does_
test membership of strings in strings.  But for some reason, it's
limited to testing strings of length one.  Go figure!

> A string is a sequence of characters, so when
> testing membership with the `in' operator, the left-hand side of the
> expression should be a character (really a string of length one).  `in'
> tests membership, not for the existence of subsequences.  Would you have
> expected
> 
>         [1, 2] in [1, 2, 3, 4]
> 
> to evaluate to true?

Honest answer: although I can't see much use for that construct, I can't
see much harm in it either.  So although it's probably not worth adding
to Python, I'm sure _somebody_ would find a use for it.

But anyway, let's "just say no" to the tuple case rather than confuse
ourselves; let's just make the string version of "in" a special case. 
Then, we can rest easy knowing that:

   7.  Readability counts. 
   8.  Special cases aren't special enough to break the rules. 
   9.  Although practicality beats purity.

(Emphasis on the last one.)

see-what-i-mean-about-"concrete"-principles?-<wink>-ly y'rs,

=g2
-- 
_____________________________________________________________________

Grant R. Griffin                                       g2 at dspguru.com
Publisher of dspGuru                           http://www.dspguru.com
Iowegian International Corporation            http://www.iowegian.com


From dfackrell at DELETETHIS.linuxmail.org  Fri Jun 28 10:38:16 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Fri, 28 Jun 2002 08:38:16 -0600
Subject: Dialing out on MODEM (Partially Resolved)
References: <mailman.1025269958.7325.python-list@python.org>
Message-ID: <3d1c74d9$1_2@hpb10302.boi.hp.com>

Were it me, I'd probably just add the additional data to the string with
enough ","s between to allow the paging system to answer.

"ATDT5555555\r\n" -> "ATDT5555555,,,,,12345\r\n"

The length of the pause created by "," can be set for some (most?) modems,
but for more information you should probably look to a reference on the AT
command set for modems.

--
Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.


"David McInnis" <david at dataovation.com> wrote in message
news:mailman.1025269958.7325.python-list at python.org...
> Peter,
>
> Thanks for the tip.  I forgot that I am on Windows and needed \r\n
> rather than just \n
>
> I get the modem to dial, now how do I get it to dial a number when the
> phone is picked up.  I am dialing a pager.  When it picks up I need to
> outdial a numeric message.
>
> David McInnis
>
> -----Original Message-----
> From: python-list-admin at python.org [mailto:python-list-admin at python.org]
> On Behalf Of Peter Hansen
> Sent: Friday, June 28, 2002 5:50 AM
> To: python-list at python.org
> Subject: Re: Dialing out on MODEM
>
> David McInnis wrote:
> >
> > I need to write a small script that opens com3 and writes an ATDT to
> > dial out, answer the line and send a numeric message.  Basically, I
> > paging utility.
> >
> > I am using pyserial and cannot seem to get my modem to dial with
> >
> > Modem.write("ATDT5555555")
>
> Is Modem an object from pyserial, or one of yours?  And does it
> automatically append the \n or CR or do you have to do that yourself?
>
> -Peter
> --
> http://mail.python.org/mailman/listinfo/python-list
>
>
>




From cerbarus at gmx.de  Wed Jun 19 16:56:18 2002
From: cerbarus at gmx.de (cerbarus)
Date: 19 Jun 2002 13:56:18 -0700
Subject: PyOpenGL 2.0.0.44
Message-ID: <9508151c.0206191256.2ec23088@posting.google.com>

Hallo,
I have download the "PyOpenGL 2.0.0.44" for Windows and installed it.
Also I have download the OpenGLContext from the PyOpenGL-Project.

If I try to run the examples, the programm cant find the modules
whitch should be imported. So I read the page once again but I cant
find where to get this modules. (e.g. OpenGL.GL)

Can be, I'm to stupid and overread where to get it, but can tell me
someone where to get it?

Thanks
Dirk


From vlindberg at verio.net  Fri Jun  7 15:15:39 2002
From: vlindberg at verio.net (VanL)
Date: Fri, 07 Jun 2002 13:15:39 -0600
Subject: Working with different word processing formats?
Message-ID: <3D01065B.7070505@verio.net>

Hello,

Does anyone know of a solution for exporting editing (and 
ideally, importing) common word processing formats?

I am trying to find any existing work in this area, pure 
python or a wrappable library, that I could use or extend. 
I don't even need to go straight to the target format -- I 
could output XML or LaTeX that could then be transformed.

I would like this to be as cross-platform as possible (which 
is why I'm just not using Pythonwin + com).  I'd like to hit 
Linux, Windows, and MacOS X.

If anyone knows of anything, please speak up!  I just cannot 
find *anything*...  The closest thing that google turned up 
was an apparently defunct Python + LaTeX scientific text 
editor, and that was only a mention, not any actual code.

Thanks,


VanL



From cliechti at gmx.net  Fri Jun  7 18:29:35 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 8 Jun 2002 00:29:35 +0200
Subject: Configuration recommendations
References: <adra81$qf9$1@nntp2-cm.news.eni.net>
Message-ID: <Xns9227567C15A9cliechtigmxnet@62.2.16.82>

"Stuart D. Gathman" <stuart at bmsi.com> wrote in news:adra81$qf9$1 at nntp2-
cm.news.eni.net:

> I have an application - of the background daemon variety - written in
> python.  It has some configuration data like the following:
> 
> scan_rfc822 = 1
> scan_html = 1           # can be CPU intensive
> block_chinese = 0
> hide_path = ()          # domains that block visible private nodes
> #hide_path = ( 'jcpenney.com' )
> block_forward = {}      # people that hate forwards
> #block_forward = { 'mycorp.com': ('egghead','busybee') }
> log_headers = 0
> blind_wiretap = 1
> wiretap_users = {}
> wiretap_dest = None
> #wiretap_users = { 'bigcorp.com': ('disloyal', 'bigmouth') }
> #wiretap_dest = '<spy at bigcorp.com>'
> check_user = {}
> #check_user = { 'mycorp.com': ('joe','mary','bill') }
> reject_virus_from = ()
> #reject_virus_from = ( 'mycorp.com' )
> 
> What is standard practice for incorporating such data?  I would like it
> to remain executable python code for maximum flexibility (plus I don't
> have to deal with parsing or XML). Should the application exec the config
> script?  Can this be done so that the module namespace is updated?
> Or should the config script be the entry point, setting up the data, then
> calling the main application?

have a look at the ConfigParser module. it uses ini-file style and it's 
easy to load the defaults from a file and update with the user preferences 
save settings etc. it also has sections so that you can group the options. 
dictionaries and list have to be inserted and read back as strings. you can 
use eval (i don't recomend this as it could have nasty sideeffects 
depending on your users...) or you can make some simple code to parse the 
data.
e.g. for a list:
inifile:
reject_virus_from = comp1, comp2

code:
s = config.get('section','reject_virus_from')
lst = [q.strip() for q in s.split(',')]

(or separate by whitespace instead of colon: lst = s.split())

or for a dict:
d = dict([(k.strip(),v.strip())
          for k,v in q.split(':')
          for q in s.split(',')
         ])

(add some appropriate exception handling to detect errors)
note that you don't need to write quotes around the strings, which might be 
user-frendlier to non programmers that use your program.

chris

-- 
Chris <cliechti at gmx.net>



From peter at engcorp.com  Sat Jun  8 00:07:34 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 08 Jun 2002 00:07:34 -0400
Subject: Efficient python programming...
References: <mailman.1023504191.26068.python-list@python.org>
Message-ID: <3D018306.D01A8C9B@engcorp.com>

Sean 'Shaleh' Perry wrote:
> 
> >
> > I'll step back from any claim that unit tests really prove anything,
> > but in a very Pythonic and pragmatic fashion they are Good Enough to
> > make them a /necessity/ for anyone trying to deliver reliable software.
> 
> my current project is a window manager for X Windows.  It is all X lib calls
> and UI events.  I can maybe test 5% of it.  Seems most of the programs I have
> worked on turn out that way.  I write as many tests as I can but usually it
> just doesn't work out for me.

There are approaches that could increase that significantly.

The key pattern is to create a layer which accepts the various calls
you make into the UI and library and turns the information into something
like strings which you can easily test.  Take simple code which makes
such calls, test your test code with it, and then freeze the test code.
Now you have a layer which can replace ("mock") the UI or library 
and allow you to intercept and check calls that would have been made to it.

Now all your UI stuff is testable at least in terms of whether the right
routines are called at the right times with the right parameters.  You 
aren't testing whether the right pixel turns on, but then you are using
a library for that: you don't need to test it.

Stepping back even farther: keep the untestable code in a very thin
layer, test it manually and thoroughly, then never change it.  You
don't have to test this bit of code that is "so simple it cannot fail", 
when the alternative is to skip testing of 95% of your application.

Automated testing of GUIs is a very underdeveloped area of software
development, but it's a frontier that almost nobody is bothering to
explore.  If more people check it out I think we'll make better 
progress developing it.

-Peter


From xscottgjunk at yahoo.com  Sat Jun  1 01:36:43 2002
From: xscottgjunk at yahoo.com (Scott Gilbert)
Date: 31 May 2002 22:36:43 -0700
Subject: Generating unique numbers?
References: <3CF7EBC8.4080802@verio.net>
Message-ID: <79b2b0bb.0205312136.5365df9c@posting.google.com>

VanL <vlindberg at verio.net> wrote in message news:<3CF7EBC8.4080802 at verio.net>...
> Is there a good algorithm for generating unique numbers 
> (ints, specifically) that can be used as object identifiers?
> 
> I am discounding the obvious newid += 1 because I will need 
> to merge different sets, and while I may have some number 
> overlap, using a simple count for each object would 
> guarantee that I have the maximum number of clashes.
> 
> Specifically, I will be using these numbers as dictionary keys.
> 

If you're on windows, you could use the pythoncom module CreateGuid
function.

Basically GUIDs are 128 bit numbers that are pretty much going to be
unique no matter where or when you call the function.


From deblnononospammmyyy at theworld.com  Sun Jun  9 21:01:59 2002
From: deblnononospammmyyy at theworld.com (David Lees)
Date: Mon, 10 Jun 2002 01:01:59 GMT
Subject: Screen resolution again
References: <3D039794.9549C54@bellatlantic.net> <mailman.1023652097.14513.python-list@python.org>
Message-ID: <3D03FA8A.BB874824@theworld.com>

Martin,

I just ran on my win98se box and your code works perfectly.  Thank you
kindly.

david lees


Martin Franklin wrote:
> 
> On Sunday 09 Jun 2002 5:59 pm, you wrote:
> > I am still trying to figure out how to automatically create a canvas
> > that is sized according to the screen resolution using tkinter on a
> > win95 box and python 2.2.1.  The hack below sort of does that, but I
> > would appreciate advice, suggestions or pointers to a url on how to do
> > it correctly.  Note that I create an extra canvas get the root height
> > and windth.
> >
> > david lees
> > -------------------
> >
> > from Tkinter import *
> > root = Tk()
> >
> > root.title('Drawing')
> > canvas = Canvas(root,width=20,height=20,bg='white')
> > canvas.pack()
> >
> > h = canvas.winfo_vrootheight()
> > w = canvas.winfo_vrootwidth()
> >
> >
> > print w,h
> > canvas2 = Canvas(root,width=w-30,height=h-30,bg='white')
> > canvas2.pack()
> >
> > root.mainloop()
> 
> Why not call winfo_vrootheight() on root, that at least gets rid of the first
> canvas.  I used this to test and it seemed to work but did not fill the
> entire screen I guess there is a bit of slack for the window decorations etc.
> however i'm running RedHat linux 7.2 with KDE What happens when you run this
> exactly?
> 
> import Tkinter
> root=Tkinter.Tk()
> root_height=root.winfo_vrootheight()
> root_width=root.winfo_vrootwidth()
> 
> print root_height, root_width # on my screen this prints 768 1024
> 
> canvas=Tkinter.Canvas(root, width=root_width, height=root_height)
> canvas.pack()
> root.mainloop()


From peter at engcorp.com  Wed Jun 19 01:04:03 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 19 Jun 2002 01:04:03 -0400
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com>
Message-ID: <3D1010C3.C321F8B0@engcorp.com>

George Hester wrote:
> 
> The whole issue started from here.  I did not deserve to be told to "crawl
> before I walk."  I may have responded to that inappropriately I confess.  But
> that does not excuse the implied derogatory.  Since that the lambasting
> crecscendoed.

Perhaps "inferred derogatory" is more appropriate.  "Implied"
suggests that I intended it to be so, which is not true.  "Inferred"
refers to how you interpreted it.

Anyway, maybe we'd best move on... no long-term harm done
to any party, I think.

-Peter


From rajarshi at presidency.com  Thu Jun 13 14:57:08 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Thu, 13 Jun 2002 14:57:08 -0400
Subject: a question on SOAP
Message-ID: <pan.2002.06.13.14.57.08.774939.27301@presidency.com>

Hi,
  I'm, using SOAP.py (0.9.7) and Python 2.2. I have a small python script
acting as a 'server' and another script calls functions on the server.
Currently both client and server are on the same machine. However when
the client calls the server I get the following errors:
'
Traceback (most recent call last):
  File "./cdbclient.py", line 84, in ?
    c =  s.getCAS(cas = config['cas'],theory = config['theory'])
  File "/home/rajarshi/src/py/modules/SOAP.py", line 3616, in __r_call
    self.__hd, self.__ma)
  File "/home/rajarshi/src/py/modules/SOAP.py", line 3534, in __call
    config = self.config)
  File "/home/rajarshi/src/py/modules/SOAP.py", line 3410, in call
    r.putrequest("POST", real_path)
  File "/usr/local//lib/python2.2/httplib.py", line 453, in putrequest
    self.send(str)
  File "/usr/local//lib/python2.2/httplib.py", line 395, in send
    self.connect()
  File "/usr/local//lib/python2.2/httplib.py", line 379, in connect
    raise socket.error, msg
socket.error: (111, 'Connection refused')

My firewall is off, nmap shows that the port 8080 is open, so why would
the server be refusing connections?

Any pointers would be appreciated.

TIA,


From jblazi at hotmail.com  Mon Jun 24 08:54:21 2002
From: jblazi at hotmail.com (JB)
Date: Mon, 24 Jun 2002 14:54:21 +0200
Subject: Subclass of array
Message-ID: <3d1714c3_7@news.newsgroups.com>

Can I derive an own class from array? (I use 2.2.)
-- 
Janos Blazi


-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From david.abrahams at rcn.com  Tue Jun 18 07:37:17 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Tue, 18 Jun 2002 07:37:17 -0400
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de> <aeftmj$m29$1@bob.news.rcn.net> <3d0c6d67@nntp.server.uni-frankfurt.de> <3D0CE698.1050806@SPAMnwinternet.com> <aeiu3a$4ff$1@bob.news.rcn.net> <3D0D3223.5070500@SPAMnwinternet.com> <aekrd1$4ao$1@bob.news.rcn.net> <3D0E2606.5050001@SPAMnwinternet.com>
Message-ID: <aen6ch$d5u$1@bob.news.rcn.net>

"Jim" <jbublitzNO at SPAMnwinternet.com> wrote in message
news:3D0E2606.5050001 at SPAMnwinternet.com...
> David Abrahams wrote:

> > Hmm, I wonder about that. The C++ type system is pretty sophisticated
(to
> > the point of being able to represent entire compile-time programs!) and
all
> > of the parser-based wrapping systems I've seen eventually fail to handle
> > some constructs I use, especially if I'm wrapping template-based code.
> > That's part of the reason I chose to use the compiler itself in
> > Boost.Python. The compiler (in theory) has no problem with C++ syntax
;-)
>
> > On the other hand, I realize there's a large category of simple C++
> > interfaces which can be parsed will less-than-complete machinery.
>
> Template-based code does require special treatment with sip.
> In most cases where I've wrapped template-based types or
> classes, I manually write C++ code for conversion C++ <-> Py
> (sip has some functions that make that easier) and implement
> it as a sip 'mapped type', which lets me refer to the same
> conversion code from any return value or arg.

That's a given with Boost.Python; with v2 conversions are automatically
registered in a global repository so they can be re-used automatically
across extension modules.

> I generally
> assign a new name (eg QList<int> -> IntList) but wouldn't
> have to (sip will parse QList<int> and look for a mapped
> type definition corresponding).
>
> sip won't handle all possible C++ syntax, and it sometimes
> takes some creativity to shoehorn something in, but something
> well over 80% of h file code translates directly (after
> removing variable names), and only 1 or 2% really takes
> major effort.

    .../your/ h file code...

> The real problem wouldn't seem to be C++ types as such, but
> finding a useful (to the Python programmer) representation
> via PyObject, and still maintaining the range of abilities
> built into the C++ code (esp things like subclassing).

That's built-in to Boost.Python. V1 had to implement a complicated Python
class look-alike itself; V2 uses Python's new-style classes so subclassing
"just works". Making virtual functions overridable in Python requires a
small amount of boilerplate on the part of the programmer. I often wish I
had a good C++ parser front-end which I could use to automate that part.

> >>The other features that are helpful on large projects are the
> >>built-in support for versioning
>
> > How so? Can you say more about this?
>
> sip uses directives: %If (A - B) ... %End to wrap a block
> of code for versioning over a range. It could also be
> %If (A - ) (>= A) or %If ( - B) (< B). Acceptable values
> for A and B are defined in a %Timeline declaration; actual
> values for A and B are passed to sip via the command line,
> usually  derived from a version #define in the C++ source
> code via configure (at least on Linux). Evaluation is by
> lexical ordering of the A and B values. A and B are
> strings like 'KDE_3_0_0'.
>
> Not very different from #ifdefs, but maybe a little
> easier.

OK. The C++ code you're wrapping still has to get by with #ifdef, I
suppose...

> > Do you mean docstring generation?
>
> No - similar, but sip provides %ExportDoc ... %End directives
> inside which you could place just about anything you want I
> suppose. For PyQt/PyKDE the directives wrap SGML marked-up text.
> You can construct an input file which along with the proper sip
> command line switch, will collect and concatenate all of the
> doc blocks distributed over the description files. The
> concatenated file can be processed (in the case of SGML) to
> provide HTML docs, or whatever you require.

Oh, there are plenty of automated tools for extracting doc from C++. Lots of
people use Doxygen with Boost.Python. No need to invent a new way to do it!

> > Boost.Python works on a similar range of platforms.
> > I'm not sure if this configuration stuff is an appropriate job for a
> > wrapping system to provide specific support for; what's the argument for
> > integrating it with the wrapping system?
>
> PyKDE consists of about 400 description files spread over 10-12
> libraries depending on version and not necessarily with the same
> description file <-> library correspondence in each version. PyQt
> is slightly smaller, but supports more platforms. The alternative
> to embedding version/platform/feature info in the description files
> is to maintain separate sets of files. In the case of PyKDE (depends
> on PyQt) for KDE < 3.0.0, it might build with 2.3.0 <= Qt <= 2.3.2
> depending on what the user has. If the user compiled Qt himself, he
> might have a different feature set than the official distributed Qt
> for his platform. The number of permutations is fairly large and that
> would be multiplied by the number of files.
>
> On the other hand, changes or bug fixes are usually the same across
> versions/platforms/features. So in addition to only supporting one
> fileset instead of many, it's only necessary to repair/edit one
> fileset instead of many. Similarly, it isn't possible to test every
> possible platform - the potential for simple typos crippling a
> particular version alone is pretty large, and most code is common.
>
> Again, #ifdefs probably provide similar functionality - this is
> just a little more formalized/standardized and supported by
> the tools (so error checking is possible, for instance).

I see.

> > Not sure about this; we'd need to compare specific functionality.
>
> My guess is that Boost does at least some of this a little more
> naturally. I'm not sure the extent that Boost can modify things
> like return values or args without resorting to C++

There is no "resorting to C++" with Boost.Python -- Boost.Python *embraces*
C++! Since you're coding the wrappers in C++ anyway, it's trivial to make a
"thin wrapper function" which calls the function being wrapped if you want
to change the interface somehow. Then you just wrap the thin wrapper.

> >>I suppose those are a series of small advantages, but the end
> >>result seems to me that you write less code
>
> > Are you counting the IDL file? That's code, IMO.
>
> Yes, but I don't write it - I just edit it mostly to correct
> mistakes my description file generator creates and to handwrite
> code where the generator has flagged that. Crude but effective (TM)
> Even without automation, the process is easiest if you start by
> saving the h file with a .sip extension and then (mostly) delete
> the things that sip doesn't want/need, so it's still mostly
> editing (sometimes A LOT of editing).

That seems to be the key difference. It's not clear wether it's really an
advantage, but I suppose that time will tell.

> >>Boost (IIRC) also has a big advantage in being able to bind much
> >>more selectively.
>
> > Can you explain what you mean by that?
>
> I should probably go back and review Boost. I had thought that
> with Boost it was possible to wrap as much or as little of a
> C++ library as you want to.

That's true.

> > Heh, yeah. Unfortunately, documentation is often harder than coding.
Even
> > the current official release of Boost.Python is not very
well-documented.
> > We're working hard to correct that for v2, which is in pre-release.
>
> Well, you have docs ... I found the web site to be pretty good at
> explaining how to use Boost (and perhaps haven't remembered it all
> correctly).

Yeah, it's a reasonable tutorial, but there are still a lot of undocumented
features in v1.

> >>Could you automate code generation to some extent?
>
> > Not sure what you have in mind here. In a sense, the metaprogramming
> > techniques used by Boost.Python already *do* code generation.
>
> Yeah - 'code generation' is ambiguous. sip itself is a code generator
> (it produces C++/Python files which are the actual bindings or Python
> lib). What I meant in this case is generating the description files
> directly from the h files. I currently take a set of h files, run them
> through a handwritten parser and generate *.sip files which sip then
> uses to generate the actual C++ code for the bindings.

Oh, sure, these guys are doing the same thing with Boost.Python:

http://mambo.peabody.jhu.edu/~karlmac/publications/gamera_python_2001/gamera
.html

> I had the
> impression SWIG was supposed to do most of that in a single step. Just
> wondering if I could implement something like that with Boost, or
> need to code a line manually for each method/class/whatever.

If you have something which can mostly parse your headers, you can generate
much of the wrapping code automatically.

> It's pretty amazing to find out who's using Python and things
> like Boost, pyGTK. PyQt, PyKDE, etc.

Yep. BTW, Boost is much more than just my Python binding library, so please,
for the sake of the other authors: "Boost.Python" when discussing my lib.

> I left your web site with the impression that I (personally) would
> find sip easier for large projects, but Boost would have advantages
> on smaller projects (unfortunately I remember the impression better
> than the reasons for it). That's especially interesting to me at
> the moment, because I have ideas for a lot of smaller projects. I expect
> I'll stick with sip, but will have to take another look at Boost, if
> only to steal good ideas :)

Not sure that you'll find much to steal; they're so different in approach...

Regards,
Dave





From loewis at informatik.hu-berlin.de  Wed Jun 19 12:20:34 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-15?q?L=F6wis?=)
Date: 19 Jun 2002 18:20:34 +0200
Subject: i18n/l10n and python modules
References: <mailman.1024498229.26760.python-list@python.org>
Message-ID: <j4vg8fb5pp.fsf@porthos.informatik.hu-berlin.de>

Marko Samastur <markos at elite.org> writes:

> I've read Python Library Reference, but I'm obviously too dumb to make it 
> work. Whatever I do, when I run the application, I get either app strings 
> translated or those in modules but never both.

Can you explain in a few words what it is that you are currently
doing? I assume you use gettext. Correct?

Do you use a single textual domain (.mo file) or many of those? Do you
have any bindtextdomain calls? Do you use gettext.gettext,
gettext.dgettext, or gettext.GNUTranslation?

In short, what you want to do sounds reasonable, and I cannot see any
reason why it won't work precisely the way you are doing it.

Regards,
Martin


From gerhard.haering at gmx.de  Fri Jun 21 06:17:13 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 21 Jun 2002 12:17:13 +0200
Subject: I'd give up Perl tomorrow if only...
In-Reply-To: <3D12F6FC.3000608@wedgetail.com>
Message-ID: <20020621101713.GA13168@lilith.my-fqdn.de>

* Derek Thomson <derek at wedgetail.com> [2002-06-21 09:50 +0000]:
> David Mitchell wrote:
> >.. there was something equivalent in scope and breadth to CPAN.
> 
> I (almost) agree, and sympathise :)
> 
> >
> >Is ANYONE working on this, or thinking about working on it, or knows
> >someone who's thinking about working on it, or dreaming about it, or ...?
> 
> Among other things, it's one of the reasons I'm tinkering with a Perl ORB.

Have you checked out PyPerl? I've seen questions about it on this
mailing list in the past, but can't find the URL at the moment.

> [...] What we *really* need to do is dynamically create CORBA
> interfaces for Perl modules, but that might be impossible given that
> Perl is like Python in that module and class interfaces aren't
> pre-declared :(

Why is that a problem? Why not just override __getattr__ or its
Perl-equivalent and switch to SOAP/XML-RPC instead of CORBA? You're
restricted to simple data types, then, but that would probably be the
same for CORBA, right?

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))





From fredrik at pythonware.com  Sat Jun 29 08:32:02 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 29 Jun 2002 12:32:02 GMT
Subject: M2Crypto: select() behaves weird on SSL sockets
References: <3D04ED17.5070606@NOSPAMREMOVETHISxs4all.nl> <ae2str$b0q$1@nntp6.u.washington.edu> <3D066E77.4070901@NOSPAMREMOVETHISxs4all.nl> <ae6bmb$frc$1@dahlia.singnet.com.sg> <3D0790E1.9080904@NOSPAMREMOVETHISxs4all.nl> <831ybavprx.fsf@panacea.canonical.org> <3D0919AF.7050903@NOSPAMREMOVETHISxs4all.nl> <83it45lxfb.fsf@panacea.canonical.org> <3D1B92A2.7030605@NOSPAMREMOVETHISxs4all.nl> <q0ZS8.106386$mF.41854@atlpnn01.usenetserver.com> <3D1C9DF8.3010607@NOSPAMREMOVETHISxs4all.nl>
Message-ID: <6NhT8.45585$n4.10824579@newsc.telia.net>

Irmen de Jong wrote:

> 3. able to handle very many concurrent connections (no thread overhead
>     per connection -- but is a thread really that heavy?)

lot's of information and links here:

    http://www.kegel.com/c10k.html

</F>




From zopestoller at thomas-guettler.de  Thu Jun 27 08:05:12 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 27 Jun 2002 14:05:12 +0200
Subject: Strings across lines
References: <3D1AF1CD.8010503@thomas-guettler.de> <3D1AF36B.7000202@wedgetail.com>
Message-ID: <3D1AFF78.8060704@thomas-guettler.de>

Derek Thomson wrote:

> Thomas Guettler wrote:


[cut]

> print ( 'foo ................. %s ' +
>         'bar ................. %s '   ) % (foo, bar)


That's better than "foo" + \ "bar". Thank you!

  thomas





From peter at engcorp.com  Thu Jun 20 21:39:56 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 20 Jun 2002 21:39:56 -0400
Subject: Inhibiting duplicate apps in Windows?
References: <3D11F9E8.DFFFC8E4@tds.net> <Xns9233F02DB217Ccliechtigmxnet@62.2.16.82>
Message-ID: <3D1283EC.A28A45CB@engcorp.com>

Chris Liechti wrote:
> 
> "Edward K. Ream" <edream at tds.net> wrote in news:3D11F9E8.DFFFC8E4 at tds.net:
> > What I would like is for the second copy of Leo to detect that another
> > copy of leo.py is already running, send a message to the first copy and
> > then exit. I wonder if anyone knows how this might be done?

> dumb solution: create a "lock file" and remove it on programm exit. the
> other problem is that you have to delete it manually when the app crashes
> (i deleted may lock files with older Netscapes ;-)

I wonder whether there's not a simple way to avoid manual deletion.
If one application creates the lock file, it can hold it open as 
long as it is running.  A second invocation can detect the first
by its failure to open the same file for writing.  If the file
already exists, it can try to delete it with os.remove(). That
should fail if the first application still has it open, but if 
the first application actually crashed, the OS should allow the
file to be removed, at which point the new invocation can safely
recreate it, knowing it now has the lock.

Haven't actually tried this myself... :)

-Peter


From mgerrans at mindspring.com  Mon Jun 24 01:23:35 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Sun, 23 Jun 2002 22:23:35 -0700
Subject: Python hits the spot
References: <mailman.1024772248.9399.python-list@python.org> <yppfsn3dlpco.fsf@hotlips.cs.chalmers.se> <3D1634ED.B9ACF9EC@engcorp.com> <af5d0r$nlt$1@slb4.atl.mindspring.net> <7xfzzdhf07.fsf@ruckus.brouhaha.com>
Message-ID: <af6ajm$scg$1@slb5.atl.mindspring.net>

> "Matt Gerrans" <mgerrans at mindspring.com> writes:
> > > > for i in range(50*1000*1000):
> > > >    pass
> >
> > You call that a bug?   Did you consider trying xrange()?
>
> Maybe not a bug, but a pretty serious misfeature IMO.

It is neither a "misfeature" nor a bug, it is simply a matter of
understanding how to use your tools, as well as understanding how they work
and applying that knowledge.

I remember when I first read Learning Python and it said that range()
creates and returns a list.   It immediately occurred to me that if I was to
do a huge range that I would instead use a while loop with a counter;
shortly thereafter I was pleased to see xrange() introduced as the way to do
this.   In fact, one of my favorite aspects Python was that when I was
learning Python, I would be thinking, "it would be cool if it did blah" only
to turn the page and see that it does indeed -- this happened over and over
and continues to, as I learn more and as new releases come out.

Being the fearless maverick that I am, I just tried running this loop:

   for i in xrange(1000*1000*1000):  # The original loop X 200.
      pass

Guess what?   No memory problems whatsoever.   It took a few minutes (about
six and a half), but not much memory.






From dgallion1 at yahoo.com  Fri Jun 21 08:19:27 2002
From: dgallion1 at yahoo.com (Darrell)
Date: 21 Jun 2002 05:19:27 -0700
Subject: Looking for Python tools for extracting information from PDF files.
References: <mailman.1024567408.12098.python-list@python.org>
Message-ID: <b0c5da3a.0206210419.404bb0a5@posting.google.com>

http://groups.google.com/groups?q=python+pdf+darrell&hl=en&lr=&ie=utf-8&oe=utf-8&selm=nFuB6.9537%24jh.1800126%40typhoon.nyroc.rr.com&rnum=2

Now I'm using Pyrex to parse PDF and it works very well.
What I have isn't ready for posting. But I must say Pyrex 
is nice.

--Darrell

Pearu Peterson <pearu at cens.ioc.ee> wrote in message news:<mailman.1024567408.12098.python-list at python.org>...
> Hi,
> 
> Can anyone suggest a (preferably Python) tool that can be used to extract
> various information from PDF files? In particularly, I am interested in
> extracting graphics from the PDF files.
> 
> I tried to search the web but all I found were tools to
> create/modify/transform PDF files but I could not find any tools that can
> extract graphics. Any hints are appreciated.
> 
> Thanks,
> 	Pearu


From whisper at oz.net  Mon Jun 24 22:18:55 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 24 Jun 2002 19:18:55 -0700
Subject: Suggestions for good programming practices?
In-Reply-To: <Xns923818184AEC3cliechtigmxnet@62.2.16.82>
Message-ID: <GCEDKONBLEFPPADDJCOEKEONDLAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Chris Liechti
> Sent: Monday, June 24, 2002 17:20
> To: python-list at python.org
> Subject: RE: Suggestions for good programming practices?
>
>
> "David LeBlanc" <whisper at oz.net> wrote in
> news:mailman.1024957832.5633.python-list at python.org:
>
> > <snip>
> >>
> >> * Avoid exec, execfile, eval, and input.
> >
> > Might one ask why? What do you have to know to use them successfully?
>
> where does the string you exec come from? from the user: is he a python
> programmer? could he make subtle errors that break your code and
> then blame
> you? could it be somebody that like to play tricks with others and insert
> harmful expression or satements (like os.system('rm -fR * /*')?
> How do you
> handle exceptions, as any exception could be raised by the unknown code
> (hint: 'try: ... except: pass' is bad too...)?
>
> if you don't want to deal with such problems, avoid exec* and eval. they
> are fine for a throw away script, and they're needed when you embedd a
> python interpreter in your app (for this case there is code.interact and
> even better reexec). for all the other cases they're a risk.
>
> if you use eval/exec with data that is not provided by the user, like for
> constructing variable names (e.g. eval("name%d"%i)), then it's bad coding
> style. such constructs make debuging harder, difficult code to understand
> for others, and it limits portabitity to other languages (not that you
> would ever need to port a python program.... but sometimes one needs to
> work with other languages to earn money etc...)
> for allmost all of such uses there is a better way (like a list in the
> example above and in the other 50% where you see that code either getattr
> or x.__dict__ is the solution ;-).

I suspect that exec, execfile and eval are useful programmatically for
executing external programs, especially when one is assembling the path/file
oneself. One might even find a use for creating and evaluating code on the
fly.

Good points about input though.

> > Does "input" imply that "input-raw" is also to be avoided?
>
> input(x) is like eval(raw_input(x)). this means that raw_input itself is
> save as it only returns strings. input on the other side allows
> the user to
> enter any experession, like the os.system on from above.
>
> e.g. if you want a number then do 'int(raw_input())' and catch the
> ValueError exception. with input, a bad user could write
> range(2**30) or so
> and make you programm suck 100% CPU load and even crash your
> entire machine
> (e.g. Linux, Kernel 2.2.x) (and yes the power or reset button will be the
> only thing that works in the later ;-)

More good points about input

> >> * Use local variables whenever possible.  Important in Python because
> >>   local variables are much faster than global.  In practice, this
> >>   means that people will often put their main code in a function, and
> >>   call it with the "if __name__=='__main__': main()" cliche.
> >
> > Does code run at global scope pay a global scope lookup cost?
> >
> > And for large loops in defs that use class instance data, copying the
> > data to a local variable to avoid the lookup cost each time through
> > the loop. (Does copying globals to local also give a speed gain?)
> >
> > class myclass():
> >      def __init__(self):
> >           self.min = 1
> >           self.max = 1000
> >      def loop(self):
> >           min = self.min
> >           max = self.max
> >           for i in range(min,max):
> >                print i
>
> here the copying to local scope doesn't speed up anything. note that you
> use min and max (which are builtin names by the way) only once.
> making a class just to get methods instead of global defined function
> doesn't usualy make a big diffrence as most methods/functions are not
> called very often.
>
>
> >>> def f(x): return x*1.1
> >>> def loopfu():
> ...    	q = f
> ...    	for i in range(10000):
> ...    	    	q()
>
> making f a local (as q) does give some speedup as the namelookup
> for q will
> be faster than for f and it has to be looked up 10000 times. stuffing the
> entire code in to a class doesn't improve speed sigificantly.

The example was contrived for the purpose of discussion. I'm sure willing to
bet that the numerous (well, several) times I've seen "self" variables
copied into methods where there for performance. Actually, I don't see a
huge difference between your example and mine wrt the idea of making "far"
things "near" to enhance performance. BTW, "min" and "max" are not builtins
in the example - they're instance variables of the class.

I would still be interested in knowing if global code paid a price for
accessing global variables since they're both running in the same scope?

> you can look here for more details, interesting anyway:
> http://manatee.mojam.com/~skip/python/fastpython.html
>
> chris
>
> --
> Chris <cliechti at gmx.net>

Dave LeBlanc
Seattle, WA USA





From christophe.delord at free.fr  Thu Jun 20 14:53:35 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Thu, 20 Jun 2002 20:53:35 +0200
Subject: Bug in Python 2.3 ?
Message-ID: <20020620205335.0c1d31b4.christophe.delord@free.fr>

It seems there's a bug in python 2.3 (available on CVS).

The following script works with python 2.2, not with 2.3

def flatten(L):
	for i in L: 
		if type(i) == list: 
			for j in flatten(i):
				yield j
		else:
			yield i 

def mklist(n):
	if n:
		return [ str(n), mklist(n-1), str(n) ] 
	else:
		return []
 
L = mklist(6) 
print "".join(flatten(L))
 
flatten seems to work but join makes a segmentation fault. 
 
It works fine with print "".join(list(flatten(L)))

I'm using Python 2.3 with Linux (Redhat)


-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From tdelaney at avaya.com  Tue Jun  4 19:50:37 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 5 Jun 2002 09:50:37 +1000 
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A351@natasha.auslabs.avaya.com>

> From: Jeff Epler [mailto:jepler at unpythonic.net]
> 
> On Tue, Jun 04, 2002 at 12:01:56PM -0400, Steve Holden wrote:
> > >>> a = b = [1,2,3]
> > >>> a += [4]
> > >>> a, id(a), b, id(b)
> > ([1, 2, 3, 4], 269472088, [1, 2, 3, 4], 269472088)
> > >>>
> > 
> > Where did the rebinding take place? ISTM that "a" is still 
> bound to the same
> > list, which has been modified in place.
> 
> Sure it rebinds a (just to the same object formerly bound to it) ..
> >>> class X:
> ...     def __setattr__(self, a, v):
> ...         print "rebinding %s, sucker" % a
> ...         self.__dict__[a] = v
> ... 
> >>> x = X() 
> >>> x.a = b = [1,2,3]
> rebinding a, sucker
> >>> x.a += [4]
> rebinding a, sucker
> >>> x.a is b
> True

To put it another way ... (for __iadd__, assume every inplace assignment
method).

__iadd__ must return an object (or None). The result of __iadd__ is rebound
to the original name.

For an object that modifies in place (such as a list), it returns self from
__iadd__.

So there definitely is rebinding going on with every augmented assignment.

The problem is that the rebinding occurs *after* __iadd__ is called, and I
can't see any way to avoid that. Yes, you can refuse to rebind, but any side
effects have already occurred.

Tim Delaney




From peter at engcorp.com  Sat Jun 22 22:26:37 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 22 Jun 2002 22:26:37 -0400
Subject: ConfigParser & converting strings to lists
References: <3D152A28.2177B054@tds.net> <3D152C69.CC41ADB0@tds.net>
Message-ID: <3D1531DD.ACF06E31@engcorp.com>

"Edward K. Ream" wrote:
> 
> the set method writes:
> 
> recentFiles = ['file1' 'file2' ...]
> 
> so it is strings of the form "['file1' 'file2' ...]" that must be
> converted to a list of strings. 

>>> s = "['file1' 'file2' '...']"
>>> [x[1:-1] for x in s[1:-1].split()]
['file1', 'file2', '...']

Does that do it?

-Peter


From jiri at baum.com.au  Sat Jun  1 23:32:43 2002
From: jiri at baum.com.au (Jiri Baum)
Date: Sun, 2 Jun 2002 13:32:43 +1000
Subject: Python and HTML frames
References: <ada6hf$vico3$1@ID-49758.news.dfncis.de> <r08ada.b7b.ln@10.0.0.1> <ada9ur$vcigh$1@ID-49758.news.dfncis.de>
Message-ID: <rk3cda.nbb.ln@10.0.0.1>

>> Ken:
>> > How do I output HTML onto frames? This code below doesn't work. Can
>> > someone help me?

Jiri:
>> This isn't really a python question, it's a HTML question (or possibly
>> HTTP; are you ouputting the right headers?).
...

Ken:
> Why you think it isn't python question? I want that HTML frame output
> from the python script. Can you tell me how?

1) Get your HTML right first, using static pages.

2) Output those same static pages from your python script.

3) Then worry about the dynamic stuff.


Jiri
-- 
Jiri Baum <jiri at baum.com.au>           http://www.csse.monash.edu.au/~jirib
  MAT LinuxPLC project --- http://mat.sf.net --- Machine Automation Tools


From sarmstrong13 at mac.com  Fri Jun 14 13:03:21 2002
From: sarmstrong13 at mac.com (SA)
Date: Fri, 14 Jun 2002 12:03:21 -0500
Subject: Translating Tcl/Tk code into Python/Tkinter help please...
In-Reply-To: <m3sn3qs0ba.fsf@fkogtsp1.bmc.uu.se>
References: <B92E4241.6F6B%sarmstrong13@mac.com> <m3sn3qs0ba.fsf@fkogtsp1.bmc.uu.se>
Message-ID: <B92F8C09.7161%sarmstrong13@mac.com>

On 6/14/02 3:59 AM, in article m3sn3qs0ba.fsf at fkogtsp1.bmc.uu.se, "Thomas
Sicheritz-Ponten" <thomas at cbs.dtu.dk> wrote:

> "try" and "exec" is used instead of "catch"
> but in this case you could use commands.getoutput
> 
> Untested code:
> 
> import commands
> from Tkinter import *
> 
> 
> def execute_python_code1():
>   global t1, t2
> 
>   txt = t1.get(0.0, END)
>   com = "/sw/bin/python -c %s" txt
>   
>   try: 
>       output = commands.getoutput(com)
>   except:
>       output = 'Error running command: %s' % com
> 
>   t2.delete(0.0, END)
>   t2.insert(END,output)
> 
> 
>   
> # getoutput returns errors as catch does, so there is no need to use
> try/except
> # IMHO, you should strip newlines from the command returned from the text
> widget
> 
> def execute_python_code2():
>   t2.delete(0.0, END)
>   output = commands.getoutput(com = "/sw/bin/python -c %s" % t1.get(0.0,
> END).strip())
>   t2.insert(END,output)
>       
>       
Thank You Everyone for your help.

I changed:

def execute_python_code2():
    t2.delete(0.0, END)
    output = commands.getoutput(com = "/sw/bin/python -c %s" % t1.get(0.0,
END).strip())
    t2.insert(END,output)
 to:

def execute_python_code2():
    t2.delete(0.0, END)
    output = commands.getoutput(t1.get(0.0,END)
    t2.insert(END,output)


This seems to do the trick. I can now enter a python command in t1 hit the
execute button and the result is printed in t2.

This seems to accomplish what I need. Once again, thank you everyone.

Thanks.
SA



From aahz at pythoncraft.com  Mon Jun 24 15:36:49 2002
From: aahz at pythoncraft.com (Aahz)
Date: 24 Jun 2002 15:36:49 -0400
Subject: Suggestions for good programming practices?
References: <WEuR8.2074$5l4.75657@ozemail.com.au> <bko7fa.9t.ln@grey.aerojockey.localdomain>
Message-ID: <af7sch$dcp$1@panix1.panix.com>

In article <bko7fa.9t.ln at grey.aerojockey.localdomain>,
Carl Banks  <imbosol at vt.edu> wrote:
>
>* Learn how references and the object system work as soon as possible.

ObBrokenRecord: "bindings", not "references".
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From kragen at pobox.com  Mon Jun 10 18:03:18 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 18:03:18 -0400
Subject: htmllib examples?
References: <ut1nfuo5skj0d5g0ff6dngo1icelpdfmg1@4ax.com> <mailman.1023308471.31469.python-list@python.org>
Message-ID: <83adq2pza1.fsf@panacea.canonical.org>

Geoff Gerrietts <geoff at gerrietts.net> writes:
> Quoting Giulio Cespuglio (giulio.agostini.remove.this at libero.it):
> > Hi there,
> > 
> > I'm trying to use htmllib, but I can't find any examples in the
> > documentation
> > (http://www.python.org/doc/current/lib/module-htmllib.html).
> > Is there a (very well hidden) example codebase in www.python.org?
> > Where else can I look?
> 
> I sent this out the other day -- it's probably not exactly what you're
> looking for, but it's a step in the direction:
> 
> http://groups.google.com/groups?hl=en&lr=&selm=mailman.1022875851.20681.python-list%40python.org

Here's another example:

http://groups.google.com/groups?selm=mailman.1023244388.32117.python-list%40python.org



From shagshag13 at yahoo.fr  Tue Jun  4 12:49:20 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Tue, 4 Jun 2002 18:49:20 +0200
Subject: Looking for a list subclassing example...
References: <adi4dv$118tko$1@ID-146704.news.dfncis.de> <mailman.1023206706.2689.python-list@python.org>
Message-ID: <adir20$11bumr$1@ID-146704.news.dfncis.de>

"Jeff Epler" <jepler at unpythonic.net>
"Mike C. Fletcher" <mcfletch at rogers.com>

Thanks to you !!! Your answers help me,

s13.




From bokr at oz.net  Sat Jun 29 14:13:52 2002
From: bokr at oz.net (Bengt Richter)
Date: 29 Jun 2002 18:13:52 GMT
Subject: Python needs better error reporting
References: <mailman.1025350057.14486.python-list@python.org>
Message-ID: <afktd0$5ng$0@216.39.172.122>

On Sat, 29 Jun 2002 07:25:58 -0400, Tim Peters <tim.one at comcast.net> wrote:

>[David LeBlanc]
>> Cute anthropomorphism is lots of fun, but doesn't _quite_ obscure the fact
>> that computers don't really "tell" anybody anything, programmers do.
>
ISTM that's a cute counter-anthropomorphism ;-) I.e., a computer is more than a
passive communication link.

>When we ship Guido with every compiler, the world will be a better place for
>sure.
>
More cuteness ;-)

>> IIRC, Python throws this "Syntax Error: invalid syntax" in a number of
>> places.
>
>Yes.  Study the code a little and you'll soon discover why.
>
>> In the place that I used as an example, looking at the language
>> reference, an appropriate error would be more like:
>> 	Syntax Error: expected "and", "or", "xor" or ":"
>
>Well, "xor" isn't legit there, but other tokens that are OK after
>
>    if s == ' '
>
>include
>
>    '
>    "
>    '''
>    """
>    ==
>    !=
>    <>
>    <
>    <=
>    >
>    >=
>    +
>    -
>    /
>    *
>    **
>and
>    \
>
>Call me prophetic, but you wouldn't be happy with a list of all
>possibilities either.  Collapse oodles of them into "binary operator" and
>you lose half the people you're trying to help.  Etc.
>
>> This would at least narrow down what should be looked at to resolve the
>> problem.
>
>Not really.  Such lists tend to baffle more than clarify.  I once used a
>Pascal compiler whose favorite trick was to point a caret at a trailing
>semicolon, and then complain "semicolon expected"; the message was always
>correct when it appeared, while appearing dead wrong, and was useless in
>determining the *real* problem regardless.  It did appear friendlier than
>"syntax error" the first 719 times I saw it, though <wink>.
>
>> Vague error reporting doesn't help new user's self confidance or
>> acceptance of the language.
>>
>> I really DO have a point!
>
>Error-reporting in parsers has been the subject of intense research.  Do a
>little digging and come back when you can make a case that your point can be
>addressed effectively.  Keep in mind that most consumer-grade computers
>still don't have telepathy chips <wink>.
>
What if you had a daemon watching you? It could notice what you did to correct
syntax errors. Then it could say, "Last time you got this kind of syntax
error, you just added a ':' to the end of the line and that took care of it."
I guess you could look at before and after parse structure to figure out the
role of the changed tokens ...

Ok, there's a lot of hand waving in that, but the idea is for it to recognize
actual erroneous patterns of usage and do statistics, as opposed to accounting
for all possible errors and reporting error instances as if all errors were
equally probable.

And then maybe add some AI hype and an animated GUI paper clip in the shape
of a dancing python with a policeman's cap ;-)

Regards,
Bengt Richter


From gleki at gol.ge  Wed Jun 12 07:41:46 2002
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Wed, 12 Jun 2002 13:41:46 +0200
Subject: AttributeError? Why?
Message-ID: <3D07337A.870E82E8@gol.ge>

Hi all!

I have been making a module, which, as I believed, was organized
correctly. It has __init__.py file, which is empty, and several modules.

Now, what happens: import goes fine. But when I try to create a class,
>>> s=SimAn.SimAn([],[],None,0,0)
Traceback (innermost last):
  File "<pyshell#18>", line 1, in ?
    s=SimAn.SimAn([],[],None,0,0)
  File "D:\Python20\SED\SimAn.py", line 18, in __init__
    self.Scenario=ARScenario.ARScenario(self.xdata)
AttributeError: ARScenario
=====
Of course, ARScenario is there:
>>> dir(SimAn)
['ARScenario', 'SimAn', '__builtins__', '__doc__', '__file__',
'__name__', 'copy', 'math']
=====
and, of course, the class ARScenario based on the class Scenario, has
__init__(self,..) method.


The question: where's my error?

Thank you.
Giorgi



From phr-n2002b at NOSPAMnightsong.com  Thu Jun 27 17:20:34 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 27 Jun 2002 14:20:34 -0700
Subject: self destruction :)
References: <f4KS8.1326$qq1.14832@castor.casema.net>
Message-ID: <7x6604ctb1.fsf@ruckus.brouhaha.com>

"Guyon Mor?e" <gumuz at looze.net> writes:
> i've got a list of dynamicaly instantiated objects of a class of my own. how
> can i let one of the objects delete itself?
> 
> i've tried: 'del self', but somehow it didn't work. afterwards it should
> also _not_ be in the list anymore.

You mean you have 

  x = [a, b, c]

and you want to remove b from the list?

Just say "del x[1]".


From pyth at devel.trillke.net  Wed Jun 19 14:33:25 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Wed, 19 Jun 2002 20:33:25 +0200
Subject: Pop a list from beginning ? and memory saving...
In-Reply-To: <3D10CC06.3010700@destiny.com>; from mcherm@destiny.com on Wed, Jun 19, 2002 at 02:23:02PM -0400
References: <3D10CC06.3010700@destiny.com>
Message-ID: <20020619203325.J15079@prim.han.de>

Michael Chermside wrote:
> As for the original question (is it a good idea), I'd say YES, but with 
> one caveat. I think that it is important that the "iterable" utilities 
> operate on ANY iterator ... and that isn't possible with the intelligent 
> implementation of reverse. So I think that if iterable.reverse() were to 
> check the type of it's argument and use the clever approach (counting 
> backwards) for types list and tuple, but fall back on creating a temp 
> list and reversing it for other types, then that'd be quite useful. 
> Optimize the common case, but don't lose the universality.

i completly agree. Iterators in an iterators module should work 
with any iterator.  And reverse iterators are not really a brand 
new invention, as anyone using the C++-STL stuff knows. 

i-am-sure-lisp-always-had-it-ly y'rs, holger




From sameer_ at email.com  Fri Jun 28 14:55:11 2002
From: sameer_ at email.com (sameer)
Date: 28 Jun 2002 11:55:11 -0700
Subject: wxpython threads help needed
Message-ID: <6bd9f01b.0206281055.2b01a2a6@posting.google.com>

Does anyone have an example of wxpython threads in action?  I am
trying to create  an application with several (tabbed) panels, each of
which is being updated on its on thread.  Basically, I don't want the
application to freeze up, or to have long startup times as all the
panels are being updated.  I want the startup to happen as soon as the
main panel is built.  I would appreciate it if someone can give an
example either from their own app or code, or a simple example cooked
up on the fly.

Thanks!


From dougz at cs.washington.edu  Fri Jun 14 22:09:27 2002
From: dougz at cs.washington.edu (Douglas Zongker)
Date: Sat, 15 Jun 2002 02:09:27 GMT
Subject: getgrouplist() ?
References: <mailman.1024100982.24519.python-list@python.org>
Message-ID: <Gxq5zr.LBs@beaver.cs.washington.edu>

Graham Guttocks <graham_guttocks at yahoo.co.nz> wrote:
: I need a way to get a back a list of supplemental group ids for a
: specified uid.  Something like getgrouplist() in my C standard lib.  I
: don't see a convenient method for this, have I missed it?

Use the "grp" and "pwd" modules.  I didn't see a convenient way to do
exactly what you want; I had to iterate over the list of all groups to
find the ones with a specific user:

   >>> import grp
   >>> def groups(uname):
   ...    return [i[0] for i in grp.getgrall() if uname in i[-1]]
   ...
   >>> groups('root')
   ['root', 'bin', 'daemon', 'sys', 'adm', 'disk', 'wheel']
   >>>

The 'pwd' module can be used to translate between username and numeric
uid.
   
dz



From synthespian at uol.com.br  Fri Jun 14 23:46:40 2002
From: synthespian at uol.com.br (synthespian)
Date: Sat, 15 Jun 2002 00:46:40 -0300
Subject: Numeric module won't load on Debian Woody
Message-ID: <pan.2002.04.21.20.58.43.196357.31994@uol.com.br>

Hello-

	Numeric Python o Woody does not work, even though I did apt-get install
python-numeric without any problem.Theoretically, it's installed.

	Has anyone experienced problems with the Numeric module on Debian Woody?
	What should I check? 
	Is there a workaround?

	I have Python2.2 installed woth no problem.


	TIA,
	Regs
	Henry
	synthespian at uol.com.br



From ftobin at neverending.org  Sun Jun 30 18:13:22 2002
From: ftobin at neverending.org (Frank Tobin)
Date: Sun, 30 Jun 2002 18:13:22 -0400
Subject: I'd give up Perl tomorrow if only...
In-Reply-To: <Xns923DF22CCC54Ccliechtigmxnet@62.2.16.82>
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au>
 <mailman.1024681648.27687.python-list@python.org>
 <Pine.LNX.4.44.0206301550120.7807-100000@palanthas.neverending.org>
 <Xns923DF22CCC54Ccliechtigmxnet@62.2.16.82>
Message-ID: <Pine.LNX.4.44.0206301756130.22298-100000@palanthas.neverending.org>

Chris Liechti, on 2002-06-30, wrote:

> well it has. the module unittest provides all thats needed. the module just 
> needs a class derrived from unittest.TestSuite and the unittest module can 
> find the tests.

No it doesn't, for two reasons.

1) unittest is merely a testing framework, and there is no standard for 
using it from an extracted archive.

2) unittests are only one way to test; it is generally best suited for
small-unit testing.  I personally use it in conjunction with whole-program
blackbox tests.

Perl also has some fairly standard testing suites, but there is a wide
variety you can choose from.  However, the *interface* from an extracted 
Perl module is *always* "make test".

> its also a convention to have a test() function, but the former is more 
> powerful. yes, it should be used more widely.
> 
> > 3) a standard installation mechanism (Perl and Python do have)
> > 4) a standard README system (Perl has, Python doesn't)
> 
> i don't know the perl way but with python every object has a __doc__ 
> attribute, every module, class, function and method etc. and with pydoc 
> there is a standard extraction tool right in the standard distro. ever 
> tried it?

I'm very familiar with docstrings.  Perl also has its POD.  But each of
these is API information.  README's are a different sort.  dostrings do
not substitute for what a README is for:

1) description of the module
2) describes dependencies
3) other information that is not API-oriented

> as metioned above, python has all it needs, now if you talk of an
> organization that runs test on non-standard extension modules, no we
> don't have that right now. the python distro has good tests for the
> standard lib, but as there is no organization that distributes
> extensions, everyone is responsible for testing its modules. (there are
> collections of modules, like The Vaults of Parnassus:
> http://www.vex.net/parnassus/ or ActiveStates Programmer's Package
> Manager (PPM) but thats not as strong or established as CPAN)

Yes, I'm talking about non-bundled extension modules.  That's all we're 
talking about here.  The standard lib is irrevelant to the discussion.

I suggest that you be a Perl programmer for a while, and see what having
CPAN accessible and authoring for it really is like.  You explicitly state
you aren't familiar with it.  I've been author of a couple CPAN modules
for about 4 years now, and an extensive user for just as long, and have
been a Python module writer for just over a year now.  Python is a great
language, but in terms of a reusable code repository, *nothing* comes
close to the jewel that is CPAN.  There is a reason you'll hear this over
and over from people coming to Python from Perl.

The Vaults and PPM just don't hold a candle to CPAN.  For all practical
purposes, they are just big FTP repositories.

-- 
Frank Tobin			http://www.neverending.org/~ftobin/



From gerhard at bigfoot.de  Fri Jun  7 19:27:09 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 8 Jun 2002 01:27:09 +0200
Subject: Python MTA
In-Reply-To: <GCEDKONBLEFPPADDJCOEAEHDDHAA.whisper@oz.net>
References: <GCEDKONBLEFPPADDJCOEAEHDDHAA.whisper@oz.net>
Message-ID: <20020607232709.GA12947@lilith.my-fqdn.de>

* David LeBlanc <whisper at oz.net> [2002-06-07 15:44 -0700]:
> Has anyone developed a simple/stupid Mail Transfer Agent using Python that
> does't rely on any external mail agents (and thus would be platform
> independent)?

Python 2.2.1 (#1, Apr 29 2002, 11:12:54)
[GCC 2.95.3 20010315 (release) [FreeBSD]] on freebsd4
Type "help", "copyright", "credits" or "license" for more information.
>>> import smtpd
>>>

Looks like so ;-)

Also check out TMDA.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From cliechti at gmx.net  Thu Jun 13 14:19:49 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 13 Jun 2002 20:19:49 +0200
Subject: What language - platform use to data acquisiton/numerical work?
References: <1c849ea9.0206130130.6cdaa36f@posting.google.com>
Message-ID: <Xns922CCF37357A2cliechtigmxnet@62.2.16.82>

mdtorre at freemail.it (Matteo) wrote in 
news:1c849ea9.0206130130.6cdaa36f at posting.google.com:

> I'm a programmer, I've worked with C, Java, Perl, Matlab, PHP, and a
> bit of almost every widespread language. I love OO programming. I have
> recently doscovered Python thanks to the great enthusuasm of its
> supporters. I love Java for its elegance and OO-ness, and Perl for its
> raw power. I don't like C++ for its long developng time and complexity
> and hate Visual C++ for being the worst (most complex, most slow, most
> ... ugly) develop environment on the earth.
> 
> I have to do (from scratch and alone) a work without any language
> restriction. The system will have to work under MSWin2000 and maybe
> that, for stability, we will port it to unix-likes.
> 
> The work involve:
> - a system of data acquisition (and control) from (of) a PLC via
> serial port. The data acquisition is in real time with low frequency
> (about 1 - 2 hz)
> - a big system of data analisys, pattern recognition, data
> visualization and heavy user interaction mostly for experiment with
> algorithms and for diagnostics
> - a end user system, mostly a wrapping of siplified GUI to the
> previous two subsystems.
> 
> I think Python would be suitable for my work. What do you think?

of course. python is suitable for almost anything

> I think I will handle serial port comm vith small C libraries, I don't
> know if Python is suited to the problem.

or you could use my lib: http://pyserial.sf.net it runs on windows and 
posix systems. i use it to program microprocessors.

> Having the base drivers should not be a problem ro realize the data
> acquisition system in Python Or not?

easy. from the serial port you get strings, working on binary data is easy 
as there is the struct package and the array module.

> The data analisys numerical procedures have to be done in matlab due
> to their complexity and specialization, and I fear even the plotting
> will have to be done in Matlab. I have seen there are some bridges
> from Phyton to Matlab, how they are?

there are some modules but there are better solutions. e.g. Numpy is very 
powerful for work with vectors and matrices (it has some sililaities with 
matlab). SciPy and ScientificPython give additional toolboxes. (you should 
find them easily using google.com)

> My problem here is where to start developing in Pyyhon and where to
> stop in ML. ML has a very simple and good sistem of building GUIs, how
> is Pithon?

you have many choices. Tk is supplied, wxWindows, pyQT and pyGTK are often 
used ones. (or try anygui if you don't want to decide :-)

> Finallly I hope to develop the wrapping GUI layer in Phyton. Again,
> has Python some simple and good GUI support? I've seen that the most
> widely used GUI module is something in tcl/tk (which I don't know at
> all). Considering that I don't know python nor tcl/tk, how is the
> learnig and developing curve?

getting started with tkinter should be easy. i often use wxWindows which is  
nastier to debug but good looking with nice license.
you might be able to find a visualisation component (SciPy and co deliver 
some tools i think) if you find such a component you might want to choose 
the toolkit the component uses :-)

chris


-- 
Chris <cliechti at gmx.net>



From johnroth at ameritech.net  Wed Jun  5 07:51:49 2002
From: johnroth at ameritech.net (John Roth)
Date: Wed, 5 Jun 2002 07:51:49 -0400
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023234674.21472.python-list@python.org>
Message-ID: <ufrupva3edm6fd@news.supernews.com>

"Delaney, Timothy" <tdelaney at avaya.com> wrote in message
news:mailman.1023234674.21472.python-list at python.org...
> > From: Jeff Epler [mailto:jepler at unpythonic.net]
> >
> > On Tue, Jun 04, 2002 at 12:01:56PM -0400, Steve Holden wrote:
> > > >>> a = b = [1,2,3]
> > > >>> a += [4]
> > > >>> a, id(a), b, id(b)
> > > ([1, 2, 3, 4], 269472088, [1, 2, 3, 4], 269472088)
> > > >>>
> > >
> > > Where did the rebinding take place? ISTM that "a" is still
> > bound to the same
> > > list, which has been modified in place.
> >
> > Sure it rebinds a (just to the same object formerly bound to it) ..
> > >>> class X:
> > ...     def __setattr__(self, a, v):
> > ...         print "rebinding %s, sucker" % a
> > ...         self.__dict__[a] = v
> > ...
> > >>> x = X()
> > >>> x.a = b = [1,2,3]
> > rebinding a, sucker
> > >>> x.a += [4]
> > rebinding a, sucker
> > >>> x.a is b
> > True
>
> To put it another way ... (for __iadd__, assume every inplace
assignment
> method).
>
> __iadd__ must return an object (or None). The result of __iadd__ is
rebound
> to the original name.
>
> For an object that modifies in place (such as a list), it returns self
from
> __iadd__.
>
> So there definitely is rebinding going on with every augmented
assignment.
>
> The problem is that the rebinding occurs *after* __iadd__ is called,
and I
> can't see any way to avoid that. Yes, you can refuse to rebind, but
any side
> effects have already occurred.
>
> Tim Delaney

It mostly depends on what we want to have happen in
this case. I see three distinct possibilities:

1. Forbid the operation when the container is immutable,
and the embedded object does not update in place.
2. Don't attempt to update an immutable container.
3. Don't attempt to update when the embedded operation
returns itself.

Each of these require additional overhead.

1 requires two checks, on the container and on
the contained object.

2 requires only one check, on the container object

3 requires a check on the object returned from the
embedded operation.

I don't see a compatability issue here, because
the current behavior raises an exception.

John Roth
>

>




From pyth at devel.trillke.net  Wed Jun 12 20:01:42 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 13 Jun 2002 02:01:42 +0200
Subject: methods to parse big foreign ascii-file?
In-Reply-To: <3d079dbc$1@sol.wh-hms.uni-ulm.de>; from igor.stroh@wohnheim.uni-ulm.de on Wed, Jun 12, 2002 at 09:15:06PM +0200
References: <mailman.1023899956.13534.python-list@python.org> <3d079dbc$1@sol.wh-hms.uni-ulm.de>
Message-ID: <20020613020142.A6609@prim.han.de>

Igor Stroh wrote:
> On Wed, 12 Jun 2002 18:05:51 +0200, "holger krekel"
> <pyth at devel.trillke.net> wrote:
> 
> > - come up with a grammar and let a parser do
> >   the hard work (which one?)
> 
> if the file is structured in a certain form, that is you can describe the
> data structure with a EBNF, then have a look at 'simpleparse' (I've lost
> the link, but google will certainly have some for you :) )...

thanks. This might be simple yet powerful enough for the task. I must
try hard to remember EBNF-grammars from my diploma exams, though :-)

> The idea behind optimistic concurrency control is surprisingly simple:
> just go ahead and do whatever you want to, without paying attention
> to what anybody else is doing. If there is a problem, worry about it
> later. Many politicians use this algorithm, too. (c) Andrew S. Tanenbaum

quite true though his Operating System oppinions were not :-)

    holger




From cliechti at gmx.net  Mon Jun 24 20:19:31 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 25 Jun 2002 02:19:31 +0200
Subject: Suggestions for good programming practices?
References: <mailman.1024957832.5633.python-list@python.org>
Message-ID: <Xns923818184AEC3cliechtigmxnet@62.2.16.82>

"David LeBlanc" <whisper at oz.net> wrote in
news:mailman.1024957832.5633.python-list at python.org: 

> <snip>
>>
>> * Avoid exec, execfile, eval, and input.
> 
> Might one ask why? What do you have to know to use them successfully?

where does the string you exec come from? from the user: is he a python 
programmer? could he make subtle errors that break your code and then blame 
you? could it be somebody that like to play tricks with others and insert 
harmful expression or satements (like os.system('rm -fR * /*')? How do you 
handle exceptions, as any exception could be raised by the unknown code 
(hint: 'try: ... except: pass' is bad too...)?

if you don't want to deal with such problems, avoid exec* and eval. they 
are fine for a throw away script, and they're needed when you embedd a 
python interpreter in your app (for this case there is code.interact and 
even better reexec). for all the other cases they're a risk.

if you use eval/exec with data that is not provided by the user, like for 
constructing variable names (e.g. eval("name%d"%i)), then it's bad coding 
style. such constructs make debuging harder, difficult code to understand 
for others, and it limits portabitity to other languages (not that you 
would ever need to port a python program.... but sometimes one needs to 
work with other languages to earn money etc...)
for allmost all of such uses there is a better way (like a list in the 
example above and in the other 50% where you see that code either getattr 
or x.__dict__ is the solution ;-).

> Does "input" imply that "input-raw" is also to be avoided?

input(x) is like eval(raw_input(x)). this means that raw_input itself is 
save as it only returns strings. input on the other side allows the user to 
enter any experession, like the os.system on from above.

e.g. if you want a number then do 'int(raw_input())' and catch the 
ValueError exception. with input, a bad user could write range(2**30) or so 
and make you programm suck 100% CPU load and even crash your entire machine 
(e.g. Linux, Kernel 2.2.x) (and yes the power or reset button will be the 
only thing that works in the later ;-)

>> * Use local variables whenever possible.  Important in Python because
>>   local variables are much faster than global.  In practice, this
>>   means that people will often put their main code in a function, and
>>   call it with the "if __name__=='__main__': main()" cliche.
> 
> Does code run at global scope pay a global scope lookup cost?
> 
> And for large loops in defs that use class instance data, copying the
> data to a local variable to avoid the lookup cost each time through
> the loop. (Does copying globals to local also give a speed gain?)
>
> class myclass():
>      def __init__(self):
>           self.min = 1
>           self.max = 1000
>      def loop(self):
>           min = self.min
>           max = self.max
>           for i in range(min,max):
>                print i

here the copying to local scope doesn't speed up anything. note that you 
use min and max (which are builtin names by the way) only once.
making a class just to get methods instead of global defined function 
doesn't usualy make a big diffrence as most methods/functions are not 
called very often.


>>> def f(x): return x*1.1
>>> def loopfu():
...    	q = f
...    	for i in range(10000):
...    	    	q()

making f a local (as q) does give some speedup as the namelookup for q will 
be faster than for f and it has to be looked up 10000 times. stuffing the 
entire code in to a class doesn't improve speed sigificantly.

you can look here for more details, interesting anyway:
http://manatee.mojam.com/~skip/python/fastpython.html

chris

-- 
Chris <cliechti at gmx.net>



From sholden at holdenweb.com  Wed Jun 12 23:05:31 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 12 Jun 2002 23:05:31 -0400
Subject: About writing Jython tutorial...
References: <c427d639.0206070612.4ea8ce0c@posting.google.com>
Message-ID: <aYTN8.528767$%u2.263854@atlpnn01.usenetserver.com>

I'd "Robert Hanlin" <findler_lambda at yahoo.com> wrote in message
news:c427d639.0206070612.4ea8ce0c at posting.google.com...
> Hi all,
>
> I'm indoctrinating friends into the dark, beautiful world of Jython.
> But I find there is no good online documentation, and I only have an
> easy time because I happen to own the O'Reilly Jython Essentials book.
>  (For example, when you use jythonc to compile classes, if something
> is missing on the classpath you're more likely to get a null pointer
> exception than an appropriate class not found exception.)
>
> If I happened to write a walkthrough, is there a website where it
> would be welcome?  If not, maybe I should make a weblog...
>
I'd be quite happy to host such content for you, assuming it were useful.
I'm sure I could find a directory on pydish...
regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From loewis at informatik.hu-berlin.de  Mon Jun  3 08:19:19 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 03 Jun 2002 14:19:19 +0200
Subject: Embedding and threads ?
References: <mailman.1022667286.17372.python-list@python.org> <j44rgrnkfm.fsf@informatik.hu-berlin.de> <mailman.1022740196.32586.python-list@python.org> <j4d6vdwvyk.fsf@informatik.hu-berlin.de> <1022771667.689.136.camel@netgroup> <j4znyhzguh.fsf@informatik.hu-berlin.de> <1022830475.893.17.camel@netgroup> <j4znygd7iv.fsf@informatik.hu-berlin.de> <1022839266.823.66.camel@netgroup> <j4d6vac9b7.fsf@informatik.hu-berlin.de> <mailman.1023091388.3109.python-list@python.org>
Message-ID: <j4wutg8shk.fsf@informatik.hu-berlin.de>

Bo Lorentsen <bl at netgroup.dk> writes:

> Pseudo, is because it can't really take advantage of SMP (or the system
> process scheduler), and its the software that controlles the
> "scheduling".

Yes, it can't make use of threads, but no, it is not (necessarily)
software scheduled. Python gives up the GIL when executing byte code
every so many instructions; it is the the choice of the operating
system to schedule then next thread to run. If not executing byte
code, each Python thread gives up the GIL when starting a long-running
operation (such as a blocking OS call).

So under normal conditions, on a single processor, with enough
runnable Python threads: the processor will always be busy, and all
threads will make progress. You can hardly ever tell the difference to
free threading.

Regards,
Martin



From candiazoo at attbi.com  Sat Jun 29 09:43:47 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Sat, 29 Jun 2002 13:43:47 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com> <mailman.1025336262.12245.python-list@python.org>
Message-ID: <3d1db962.465573593@netnews.attbi.com>

I don't really think you want to see ALL of it.  :)  A) it ain't purdy and B)
it's a few thousand lines of code...  but I'll post snippets a little later,
after my wife goes out.  :)

Mike


From aahz at pythoncraft.com  Wed Jun 19 19:46:04 2002
From: aahz at pythoncraft.com (Aahz)
Date: 19 Jun 2002 19:46:04 -0400
Subject: Flame fests (was Re: python version?)
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <3D1010C3.C321F8B0@engcorp.com> <raXP8.47926$GY.15260255@twister.nyroc.rr.com> <3D107948.B3FEE4BB@engcorp.com>
Message-ID: <aer53s$7is$1@panix1.panix.com>

Everybody have fun playing "kick the newbie"?  Maybe it's time to quit?

(Yes, yes, I've read the thread -- I still think it's time to revert
c.l.py to its usual habits.)
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From amuys at shortech.com.au  Mon Jun  3 21:00:02 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 3 Jun 2002 18:00:02 -0700
Subject: Iterating through two lists
References: <mailman.1022234790.26965.python-list@python.org>
Message-ID: <7934d084.0206031700.377c691e@posting.google.com>

"Denis S. Otkidach" <ods at fep.ru> wrote in message news:<mailman.1022234790.26965.python-list at python.org>...
> On Fri, 24 May 2002, jb wrote:
> 
> j> I have two lists, x and y with the property len(x) = len(y).
> j>
> j> I should like to achive this (x is a list of class instances)
> j>
> j>   for (a,b) in (x,y): a.f(b)
> j>
> j> Is there a fancy way of doing this or have I to introduce an
> j> auxillary
> j> counter (that is very easy but maybe not very "lispy", that
> j> is
> j> "python-like").
> 
> If lists are not very lange try this:
> 
> for a, b in zip(x, y): a.f(b)

If the lists are large then you probably need something along the lines of xzip.

def xzip(*args):
    iters = [iter(a) for a in args]
    while 1:
        yield tuple([i.next() for i in iters])

Andrae Muys


From mats at laplaza.nospam.org  Fri Jun 14 10:36:02 2002
From: mats at laplaza.nospam.org (Mats Wichmann)
Date: Fri, 14 Jun 2002 14:36:02 GMT
Subject: Shrinking Python
References: <mailman.1023999624.27508.python-list@python.org>
Message-ID: <3d09fee2.692863815@news.laplaza.org>

On Thu, 13 Jun 2002 16:15:39 -0400, "Mahrt, Dallas"
<dallasm at aiinet.com> wrote:

:I am looking into using Python in a small device with limited space for
:binaries. Due to this, I am both investigating creating a shared library
:with the interpreter and shrinking the footprint of this library. I have
:found many threads talking bout shared libraries, so I am not too concerned
:about that aspect. 
:
:I am concerned about shrinking the library's binary footprint. I did find
:several links to a "Deeply Embedded Python" based on 1.5.1 which has since
:vanished. I have stripped the library to improve size some (to ~1.13 MB) but
:would like to reduce it further. Has anyone tackled this or even attempted?
:Are there any links to resources that may help me that I haven't found? 

Have you checked what Riverbank Computing has done to trim Python down
for the Sharp Zaurus? 

riverbankcomputing.co.uk


Mats Wichmann



From markus.vonehr at ipm.fhg.de  Wed Jun 12 08:57:13 2002
From: markus.vonehr at ipm.fhg.de (Markus von Ehr)
Date: Wed, 12 Jun 2002 14:57:13 +0200
Subject: PIL _imaging
References: <Xns922AA829A8C5FHotMailcs@130.133.1.4> <3D06E694.4040103@mxm.dk> <Xns922B3150A6B32HotMailcs@130.133.1.4>
Message-ID: <3D074529.E848AA6@ipm.fhg.de>

Hi Cousin:

Lets assume that your python directory is:
k:\Python21
or maybe:
k:\program files\python21

There is a directory ..python21\DLL where you have to put the
_imaging.pyd and
_imagingtl.pyd

Put the file PIL.pth into the python21 directory,
and move the the whole PIL *.py files to k:\python21\pil\
(There are a lot of files named *ImagePlugin.py etc.)

That works!

Cousin Stanley schrieb:

> > Cousin Stanley wrote:
> >
> >
> >>   There are 2 files in K:\Python\Lib\PIL\DLLs
> >>   but apparently these are not the ones that are needed ...
> >>
> >>     _imaging.pyd
> >>     _imagingtk.pyd
> >
> >
> > Put those two files into:
> > K:\Python\DLLs
> >
> > And you should be good to go.
> >
> > regards Max M
> >
> >
>
> Cousin Max M ...
>
> Thanks for the reply ...
>
> I copied the 2 files into the  K:\Python\DLLs  folder
> and tried again, but the result is the same, leading to
> the same error I originally posted ...
>
>    python Samples/viewer.py ButtonD.jpg
>
>      File "K:\Python\Lib\PIL\PIL\Image.py", line 37, in __getattr__
>         raise ImportError, "The _imaging C module is not installed"
>      ImportError: The _imaging C module is not installed
>
> I also tried adding  K:\Python\DLLs
> to the PYTHONPATH environment variable
> in the autoexec.bat file ... no luck ...
>
> There is a file  K:\Python\Lib\PIL\PIL.pth
> that contains only the single 3 character entry PIL ...
>
> I've also tinkered with this file
> trying various path combinations
> with no luck ...
>
> Cousin Stanley
-------------- next part --------------
A non-text attachment was scrubbed...
Name: markus.vonehr.vcf
Type: text/x-vcard
Size: 245 bytes
Desc: Visitenkarte f?r Markus von Ehr
URL: <http://mail.python.org/pipermail/python-list/attachments/20020612/5690b554/attachment.vcf>

From gumuz at looze.net  Wed Jun 12 10:59:14 2002
From: gumuz at looze.net (Guyon Morée)
Date: Wed, 12 Jun 2002 16:59:14 +0200
Subject: Bits and bytes?
References: <3d074e6d$0$224$4d4ebb8e@news.nl.uu.net> <mailman.1023890477.2592.python-list@python.org>
Message-ID: <3d0760ec$0$223$4d4ebb8e@news.nl.uu.net>

ok, thanx a lot.

there's one thing i am missing though... how do i write '011010110101' as
'binary bits' to a file?




"Christopher Armstrong" <radix at twistedmatrix.com> wrote in message
news:mailman.1023890477.2592.python-list at python.org...
>>>>> "Guyon" == Guyon Mor?e <gumuz at looze.net> writes:

    Guyon> ok, this might be a strange question.... but i was wondering: how
    Guyon> can i read or write 'bits'?

    Guyon> if i have a string, how can i convert to it's basic 0's and 1's?

    Guyon> and offcourse, the other way around.... let's say i want to store
    Guyon> some bit-sequence(0011101110) to a file.


    Guyon> how do i do this?

http://twistedmatrix.com/users/moshez/binary.py

Moshe Zadka wrote this because he was annoyed at all the people in #python
asking how to do this ;-)


--
                                Chris Armstrong
                         << radix at twistedmatrix.com >>
                http://twistedmatrix.com/users/carmstro.twistd/







From steve at lurking.demon.co.uk  Tue Jun  4 15:53:12 2002
From: steve at lurking.demon.co.uk (Stephen Horne)
Date: Tue, 04 Jun 2002 20:53:12 +0100
Subject: Thought on PEP 204 and 276
References: <ckp4fuc4fskjcpilnn2mrt3665fqvld92g@4ax.com> <mailman.1022528744.19859.python-list@python.org> <gm17fu40rh6r675g3a6d1ovl49tb9pd822@4ax.com> <mailman.1022594566.27265.python-list@python.org> <slrnaf7mo4.dfe.philh@comuno.freeserve.co.uk> <6w4L8.229234$%u2.120502@atlpnn01.usenetserver.com>
Message-ID: <j36qfu4r84j8320iegrujjf8dpvdvs20bm@4ax.com>

On Tue, 4 Jun 2002 10:49:19 -0400, "Steve Holden"
<sholden at holdenweb.com> wrote:

>"phil hunt" <philh at comuno.freeserve.co.uk> wrote in message
>news:slrnaf7mo4.dfe.philh at comuno.freeserve.co.uk...

>> I think they are essentially the same. A sequence tpye is just a
>> mapping type where the keys are restricted to the integers from
>> 0 to however many elements there are minus 1.
>>
>No matter what you think, they are essentially different. Insertion into a
>mapping doesn't alter the key to which other elements are mapped, but
>insertion into a sequence (potentially) does that.

What is normally referred to as insertion in dictionaries is actually
more like overwriting a list element. This is why handling both with
the 'container [key] = value' syntax is natural.

Insertion with modification of existing keys to make room (ie list
insertion) simply does not exist in lists. The usefulness of list
insertion derives from the fact that a list is a very specialised type
of key-to-value mapping - in this respect, the type 'list' should
clearly inheret from and extend the more general key-to-value type
called a dictionary.

In essence, you are confusing the operations on the underlying data
structures with the operations on the conceptual types. These are two
separate concepts. The implementation is distinct from the
user-visible abstraction.



From vvainio at tp.spt.fi  Mon Jun  3 07:48:57 2002
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 3 Jun 2002 04:48:57 -0700
Subject: std lib suggestion: Publish docstring unindent part of inspect.getdoc() as a function
Message-ID: <ad496f8.0206030348.7ebe9f86@posting.google.com>

I often use triple-quoted strings to describe various "content", and
would like to get the results in an undindented form. inspect.getdoc()
does this for the docstrings, but I would like to see the same
algorithm as a seperate function. I would assume that it isn't too
rare to want this. Aligning those strings to the left column directly
in the source looks horrible.

Such a feature apparently doesn't exist at the moment, since
inspect.getdoc does it by hand. If if was a function (say,
string.unindent), getdoc could just be

return string.unindent(object.__doc__)

(+ handling for nonexistent __doc__, of course)

It is of course trivial to copy-paste the implementation from
inspect.getdoc(), but at least I would use such a feature all the
time. string module already has capitalize(), so my request probably
isn't way out of line ;-).

-- Ville


From claird at starbase.neosoft.com  Sat Jun  8 15:18:21 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 8 Jun 2002 14:18:21 -0500
Subject: Comparison: HTTP with Java, Perl or Python
References: <3D009911.9000904@gmx.de> <3D0127D2.2060604@gmx.de> <B9273737.4982%js_nntp@nospam_blackrocksoftware.com> <3D022634.9050304@gmx.de>
Message-ID: <6E63F7FCD191848B.20A99585505D1A67.96A3FE18E66D1AF7@lp.airnews.net>

In article <3D022634.9050304 at gmx.de>,
Ingo Linkweiler  <i.linkweiler at gmx.de> wrote:
			.
			.
			.
>The Python version is quite short, even after adding exception handling. 
>TCL seems to be same.
>The Perl version seems to be the shortest, but can you read it without 
>knowing Perl?
>
>Can you estimate the time you needed for writing it?
>Did you have any difficulties?
>My Python version needed 2 Minutes, with Java I needed 20, because I did 
>not unterstand the f******* manual.
			.
			.
			.
There's something I should mention:  the standard
distribution of Tcl includes the http package.
The standard distribution of Perl does not include
the LWP module.

Please don't generalize or conclude much from this.
Different languages manage "run-time libraries" in
different ways.  My point is that, for serious
comparisons of languages, it's important to recog-
nize they differ in *many* dimensions, not just the
"source code footprint" this thread raises.
-- 

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


From maxm at mxm.dk  Fri Jun 28 16:49:42 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 28 Jun 2002 22:49:42 +0200
Subject: incorrect upper()/lower() of UTF-8
References: <afiecp$oss$1@web1.cup.hp.com>
Message-ID: <3D1CCBE6.5070203@mxm.dk>

Jim Henry wrote:
> It appears that python does not uppercase/lowercase utf-8 strings properly:
> Am I missing something??

Try and set the locale, via the locale module

regards Max M



From msoulier at nortelnetworks.com_.nospam  Wed Jun  5 12:06:58 2002
From: msoulier at nortelnetworks.com_.nospam (Michael P. Soulier)
Date: 5 Jun 2002 16:06:58 GMT
Subject: os.popen and place in file
References: <3CFE2762.519E4344@motorola.com>
Message-ID: <slrnafsdru.12v.msoulier@pmerd071.ca.nortel.com>

On Wed, 05 Jun 2002 09:59:46 -0500, Stephen Boulet
<stephen.boulet at motorola.com> wrote:
> I'm running windows NT.

    My condolences. :)

> returns a file object 'f'. I can't however use the seek command on it:

    I can't do this on Linux either. 

>>> import os
>>> pipe = os.popen('ls /bin', "r")
>>> pipe.readlines()
['ae\n', 'arch\n', 'bash\n', 'bsd-csh\n', 'cat\n', 'chgrp\n', 'chmod\n',
'chown\n', 'cp\n', 'cpio\n', 'csh\n', 'date\n', 'dd\n', 'df\n', 'dir\n',
'dmesg\n', 'dnsdomainname\n', 'domainname\n', 'echo\n', 'ed\n', 'egrep\n',
'elvis-tiny\n', 'env\n', 'false\n', 'fdflush\n', 'fgrep\n', 'fuser\n',
'grep\n', 'gunzip\n', 'gzexe\n', 'gzip\n', 'hostname\n', 'kill\n', 'ksh\n',
'ln\n', 'loadkeys\n', 'login\n', 'ls\n', 'lspci\n', 'mkdir\n', 'mknod\n',
'mktemp\n', 'more\n', 'mount\n', 'mt\n', 'mt-gnu\n', 'mv\n', 'netstat\n',
'nisdomainname\n', 'pidof\n', 'ping\n', 'ps\n', 'pwd\n', 'rbash\n',
'readlink\n', 'rm\n', 'rmdir\n', 'run-parts\n', 'sed\n', 'setserial\n',
'sh\n', 'sleep\n', 'stty\n', 'su\n', 'sync\n', 'tar\n', 'tcsh\n',
'tempfile\n', 'touch\n', 'true\n', 'umount\n', 'uname\n', 'uncompress\n',
'vdir\n', 'vi\n', 'ypdomainname\n', 'zcat\n', 'zcmp\n', 'zdiff\n', 'zegrep\n',
'zfgrep\n', 'zforce\n', 'zgrep\n', 'zless\n', 'zmore\n', 'znew\n']
>>> pipe.seek(0)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
IOError: [Errno 29] Illegal seek

    Since the "file" in this case is a temporary buffer used by the pipe
implementation, it probably purges as you read, preventing any seeking.
Remember, it's not a real file. 

    Mike

-- 
Michael P. Soulier, QX41, SKY  Tel: 613-765-4699 (ESN: 39-54699)
Optical Networks, Nortel Networks, SDE Pegasus
"...the word HACK is used as a verb to indicate a massive amount
of nerd-like effort."  -Harley Hahn, A Student's Guide to Unix


From kragen at pobox.com  Fri Jun  7 13:53:32 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 07 Jun 2002 13:53:32 -0400
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet> <3CFDF148.AA2035FF@info.unicaen.fr> <3CFE49CA.46A5FEE1@tds.net> <838z5sgx7w.fsf@panacea.canonical.org> <mailman.1023388273.14192.python-list@python.org> <cJOL8.326812$%u2.257815@atlpnn01.usenetserver.com>
Message-ID: <83d6v3dlgj.fsf@panacea.canonical.org>

"Steve Holden" <sholden at holdenweb.com> writes:
> "Jeff Epler" <jepler at unpythonic.net> wrote ...
> > On Thu, Jun 06, 2002 at 12:58:59PM -0400, Kragen Sitaker wrote:
> > > "Edward K. Ream" <edream at tds.net> writes:
> > > > BTW, a compiled version of a Python program would be valuable
> > > > commercially, as it would allow companies to protect their source
> code,
> > > > as well as (possibly) improve performance.
> > >
> > > I'm worried about this myself; perhaps the compiler could include the
> > > source code in the resulting executable to prevent this.
> >
> > Your sarcasm is lost on me.
> 
> Lost, perhaps, but (since you recognised it) not entirely wasted ;-)

I don't know why you thought I was being sarcastic.  I was completely
serious.

The Java guys have probably a better solution: don't save the compiled
code to disk.  A traditional compiler that included the source code in
the executable (like, approximately, javac) could easily be hacked to
remove the source code; but turning a JIT compiler into a traditional
compiler is a bit more work.



From gashton at cmedltd.com  Tue Jun 25 13:50:28 2002
From: gashton at cmedltd.com (Graham Ashton)
Date: 25 Jun 2002 18:50:28 +0100
Subject: application development - separating content from function
In-Reply-To: <_fGQ8.280952$i6.18906908@amsnews02.chello.com>
References: <yroQ8.260949$i6.17986369@amsnews02.chello.com> 
	<_fGQ8.280952$i6.18906908@amsnews02.chello.com>
Message-ID: <1025027428.2449.17.camel@walter>

On Fri, 2002-06-21 at 14:56, Henry Baumgartl wrote:
> Thanks very much for the input!
> 
> I do believe it has helped me to take off my blinders, look around and
> realise what it is all about.

In that case it's probably a bad time for me to mention this link for
thoughts on how to get your GUI app to run with different toolkits (i.e.
different platforms):

  http://www.ootips.org/porting-gui.html

I totally agree with Peter -- adopting a few of the XP principles will
help you enormously, especially the idea of YAGNI; 

        1. Select a very small subset of the outstanding functionality
           on your wish list (i.e. approx 1-3 weeks worth of work).
        2. Pretend the remaining functionality on your wish list doesn't
           exist (and that it never will -- no cheating).
        3. Write code until you have implemented enough to satisify the
           functionality identified in (1).
        4. Congratulate yourself on a job well done, release to
           SourceForge, nip down the pub to celebrate...
        5. Goto 1.

For a huge improvement on the above, learn about unit tests before you
start, and how to use PyUnit. There's an excellent example here that
covers the basics of both:

  http://diveintopython.org/roman_divein.html

> Going for the ultimate design and functionality prize may well have
> (indefinitely) delayed actually getting the project under way.

I used to be terribly prone to doing that myself. Maybe I still am.

-- 
Graham Ashton





From chris.gonnerman at newcenturycomputers.net  Wed Jun 19 00:24:08 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Tue, 18 Jun 2002 23:24:08 -0500
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com>
Message-ID: <004a01c21749$2c3c5c80$0101010a@local>

----- Original Message -----
From: "George Hester" <hesterloli at hotmail.com>


> The whole issue started from here.  I did not deserve to be told to "crawl
> before I walk."  I may have responded to that inappropriately I confess.
But
> that does not excuse the implied derogatory.

Have you ever used USENET before?  You MUST be kidding!  comp.lang.python is
the
nicest group of people anywhere on USENET... we regularly help users such as
yourself.  You received EXCELLENT assistance with no derogatory comments, at
least until you lost your senses and started whining.

On most programming newsgroups you would have been told RTFM; if you were
lucky,
a web page reference or manual page would be mentioned, but in many cases
you'd
be blown off.  You got a correct, but incomplete, answer, because we
naturally
assumed you were a programmer and just needed a pointer.  Note that when you
whined that you needed a complete answer you got one; you could have gotten
the
same thing without whining.

You ARE a nerd, right?  Nerds (programmers, engineers, scientists, etc.)
tend to
be among the most blunt and direct of all people.  Expect if you ask a
simple
question, you'll get an answer which directs you to the place where you
would have
found the answer if only you had looked.  If you can't take the heat, get
out of
the kitchen.

> Since that the lambasting crecscendoed.

Since you started to call us unfair, actually.  Get a thicker skin or a more
humble attitude before you ask questions in the future.

Read the manual, already.  Learn to use Python from the interactive
prompt...
it's the best place to learn the basics.  Read that page over on ESR's site
about
asking questions on USENET, because you really need to understand it:

http://www.tuxedo.org/~esr/faqs/smart-questions.html

That page contains the RULES.  Learn them.  Expect to be treated just like
everyone
else.

And, stop whining.  The correct way for you to respond to this posting, and
all
the others, is to NOT RESPOND.  The second-best thing would be a SINGLE
apology.
Don't keep telling us how we misunderstand, or how we are unfair; that just
marks
you as a troll, and frankly that IS derogatory.  Don't live up to it (or
down to
it, whichever) and you'll never hear it.


Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net






From mdtorre at freemail.it  Fri Jun 21 08:57:28 2002
From: mdtorre at freemail.it (Matteo)
Date: 21 Jun 2002 05:57:28 -0700
Subject: What language - platform use to data acquisiton/numerical work?
References: <1c849ea9.0206130130.6cdaa36f@posting.google.com>
Message-ID: <1c849ea9.0206210457.36d494db@posting.google.com>

Thank you all for the contributions.

I have understood that Python is good for GUIs, for serial port data
acquisition and for numercal computation.

My only problem is that MatLab has some huge numerical libraries that
Phyton does not have. I have to do experiments with pattern recognitin
algoritmhs (pca, neural networks etc.) and these are already
efficently coded in MatLab. I can't afford the effort to code a radial
basis function network from scratch (I'm able to do it, but it's
tiring and very time consuming).
So I'd like to use pyhon for all except numerical computation. I've
seen Pymat to conetct MatLab to Python, but it seem a little outdated
and unsupported. Is there any other way to embed MatLab into Python?

Thank you in advance,

Matteo Della Torre


From gh_pythonlist at gmx.de  Thu Jun 13 22:03:50 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Fri, 14 Jun 2002 04:03:50 +0200
Subject: email forwarding program
In-Reply-To: <e21498ef.0206130751.2ef784a7@posting.google.com>
References: <e21498ef.0206130751.2ef784a7@posting.google.com>
Message-ID: <20020614020350.GA590@lilith.my-fqdn.de>

* Matthew Bell <mbell at cs.pitt.edu> [2002-06-13 08:51 -0700]:
> Hi,
> 
> I have a question regarding the email handling
> libraries in Python.  I'm trying to write a
> program that, based on some information it
> extracts from an email, forwards it to another
> person.  I've about figured out imaplib, but
> don't see how I can use it to forward on emails.

Use the email module to construct email messages and the smtplib module
to actually send the message.

This unfinished, but working script is lying around in my src/ dir. It's
for forwarding a message including full headers to SpamCop and shows how
to use the email module to do so.

HTH,

Gerhard

#!/usr/bin/env python2.2
#
# Written by: Gerhard H?ring (gerhard at bigfoot.de)
# License: none/public domain

import sys, os, smtplib, email
from email.Message import Message
from email.MIMEBase import MIMEBase
from email.MIMEText import MIMEText
from email.MIMEMessage import MIMEMessage

class Config:
    """Settings for Spamcop reports.
    You must at least change SUBMISSION_ADDRESS to match the one you got from
    Spamcop. You also need to provide a valid SMTP server address, localhost is
    just fine on many *nix systems."""
    SUBMISSION_ADDRESS = "Spamcop <submit.CENSORED at spam.spamcop.net>"
    SMTP_SERVER = "localhost"
    MY_ADDRESS = "CENSORED at bigfoot.de"
    SUBJECT = "Spam report"
    BODY = "Dear Spamcop! Please analyze the following spam."

def strip_spamassassin_markup(spam):
    infile, outfile = os.popen2("spamassassin -d -")
    infile.write(spam)
    infile.close()
    return outfile.read()

def main():
    forwarded_mail = strip_spamassassin_markup(sys.stdin.read())

    me = Config.MY_ADDRESS
    to = [Config.SUBMISSION_ADDRESS]
    message = MIMEBase("multipart", "mixed")
    message["Subject"] = Config.SUBJECT
    message["From"] = Config.MY_ADDRESS
    message["To"] = ", ".join(to)

    message.attach(MIMEText(Config.BODY))

    rfcmessage = MIMEBase("message", "rfc822")
    message_to_forward = email.message_from_string(forwarded_mail)
    rfcmessage.attach(message_to_forward)
    message.attach(rfcmessage)

    smtpserver = smtplib.SMTP(Config.SMTP_SERVER)
    smtpserver.sendmail(me, to, message.as_string(unixfrom=0))
    smtpserver.close() 

if __name__ == "__main__":
    main()
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 17.3 ?C      Wind: 2.0 m/s




From amckay at merlintechnologies.com  Mon Jun 10 16:44:11 2002
From: amckay at merlintechnologies.com (Andy McKay)
Date: Mon, 10 Jun 2002 13:44:11 -0700
Subject: How convert all list/tuple...
In-Reply-To: <523aguoljk3qceh4kkki7emn6283d6sqpt@4ax.com>
References: <523aguoljk3qceh4kkki7emn6283d6sqpt@4ax.com>
Message-ID: <20020610204411.GA4042@mckay.merlintechnologies.com>

On 10/06/02 20:31 +0000, Marco Stagno wrote:
> Hi
> There is a fast way to convert all item in a tuple/list into a string type,
> without do a loop?

Im not sure if you mean without looping through every element but should do:

>>> z = [1,2,'string']
>>> [str(x) for x in z]
['1', '2', 'string']

Cheers
-- 
  Andy McKay
  Merlin Technologies




From CousinStanley at HotMail.Com  Fri Jun 14 01:17:11 2002
From: CousinStanley at HotMail.Com (Cousin Stanley)
Date: 14 Jun 2002 05:17:11 GMT
Subject: PIL _imaging
References: <Xns922AA829A8C5FHotMailcs@130.133.1.4> <3D06E694.4040103@mxm.dk> <Xns922B3150A6B32HotMailcs@130.133.1.4> <3D074529.E848AA6@ipm.fhg.de> <Xns922B561EDB212HotMailcs@130.133.1.4> <3D07BCC1.2000002@mxm.dk> <Xns922BB340B788BHotMailcs@130.133.1.4> <9C_N8.11930$p56.3660854@newsb.telia.net> <Xns922C48A439D35HotMailcs@130.133.1.4> <PM8O8.43167$n4.9913522@newsc.telia.net>
Message-ID: <Xns922CE3285BF83HotMailcs@130.133.1.4>

|| like all other binary extensions (PYD files) on windows,
|| the _imaging module must be compiled for the Python
|| version you're using.
||
|| the PIL/*.py part of the library works for all versions
|| on all platforms.

Cousin Fredrik ... 

Thanks for the info ... 

I'll blow the  dust/rust  out of my C compiler
and give it a try ... 

Cousin Stanley


From marklists at mceahern.com  Thu Jun 27 10:22:39 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 27 Jun 2002 09:22:39 -0500
Subject: Var substitution
In-Reply-To: <c07d0edc.0206270557.28180893@posting.google.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBAEMPLGAA.marklists@mceahern.com>

> What i'm wanting to know, is there any way of looking at a variables
> value rather than its name.
>
> Example
>
> Unix : ${MyVar}
> Dos  : %MyVar%
>
> Both of the above will be subtituted with the values.
>
> Example python script (below causes error):
>
> >>> guy="name"
> >>> import os
> >>> os.guy
>
> This is what should happen :
>
> >>> import os
> >>> os.name
> 'nt'

eval?

<ducking, running for cover>

// m

p.s.  but seriously, do you have a more realistic example?  Surely your
motivation is more than merely to write obfuscated code.  If not, you could
just do this:

  import os
  guy = os.name
  print guy

Or are you looking for string interpolation?  search the archives of
Python-Dev because that was recently discussed.

-





From mwarden22 at hotmail.com  Tue Jun 11 14:33:15 2002
From: mwarden22 at hotmail.com (nealj)
Date: 11 Jun 2002 11:33:15 -0700
Subject: Hap debugger released
Message-ID: <d3768569.0206111033.db12ba4@posting.google.com>

Version 2.1 of the Hap python debugger (and ide) has just been
released to sourceforge.  This version improves performance, stability
and adds a number of useful features such as an interactive window,
watch tooltips and more.

We've been using it in house on our next product for six months now
and everyone is very happy with it.

Try it out at http://sourceforge.net/projects/hapdebugger


From martin at v.loewis.de  Tue Jun 11 16:41:30 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 11 Jun 2002 22:41:30 +0200
Subject: Embedding Python 2.2 in non-reentrant application
References: <ae5n1f$qtm$07$1@news.t-online.com>
Message-ID: <m3k7p5wnt1.fsf@mira.informatik.hu-berlin.de>

"Stefan Werner" <stefan1werner at hotmail.com> writes:

> I am working on embedding Python 2.2 in a large scale application. Now the
> problem is, that Python 2.2 supports multithreading, where the functions we
> want to expose to Python are not threadsafe. Is there something like a flag
> to make my whole module single-threaded for Python access or do I have to
> wrap every single function call in
> Py_BEGIN_ALLOW_THREADS/Py_END_ALLOW_THREADS macros?

No, you don't have to do anything. Inside a C function, you hold the
GIL, and any other Python thread will block while your C function is
running. It is precisely that
Py_BEGIN_ALLOW_THREADS/Py_END_ALLOW_THREADS bracket that allows
thread-safe libraries to give up the GIL, and thus allow
thread-scheduling to happen.

Regards,
Martin



From cliechti at gmx.net  Wed Jun 19 18:13:00 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 20 Jun 2002 00:13:00 +0200
Subject: old fashioned COM
References: <uh20956iiu8q52@corp.supernews.com>
Message-ID: <Xns92332A13AFAFcliechtigmxnet@62.2.16.82>

"mark" <nothere at nowherre.com> wrote in 
news:uh20956iiu8q52 at corp.supernews.com:

> I am a newbie at this, and haven't done a lot of programming in some
> years.  The book I have on Python describes a lot about programming
> internet and email apps.  I need to just select a COM port, set
> parameters like baud rate, parity, etc. and do a file capture in dumb
> terminal mode.  This may seem to simple in todays world, but I'm trying
> to log some events from a piece of equipment that uses serial I/O.  
> 
> Help?

yes :-) http://pyserial.sf.net

chris


-- 
Chris <cliechti at gmx.net>



From larooy at xtar.co.nz  Sun Jun 30 18:41:13 2002
From: larooy at xtar.co.nz (John La Rooy)
Date: Mon, 1 Jul 2002 10:41:13 +1200
Subject: [Newbie] Is the text delimiter changeable?
References: <3d1f26a4.5915886@news.texas.net>
Message-ID: <20020701104113.69a07009.larooy@xtar.co.nz>

On Sun, 30 Jun 2002 15:49:16 GMT
baf at texas.net (Ben Fairbank) wrote:

> I have to process large amounts of text and the text includes numerous
> apostrophes and quotation marks.  It would be very handy if I could
> temporarily assign a substitute for Python's use of the caracters "
> and ' to delimit text strings.  If I could use, for example, the @
> character or the # character, neither of which appears in the text I
> have to work with, and if I could disable the " and ' delimiters, then
> my task would be greatly simplified.  I realize I can backslash escape
> these characters in some circumstances, but I do not think that will
> work when a user's input is a word such as "can't" or "don't."
> Anyway, short of doing something truly drastic, such as recompiling
> the system (which I have no intention of doing), is there a workaround
> that will permit this?
> 
> Thank you,
> 
> BAF

You don't say where is the text coming from. The quotes aren't used as
delimiters except when you are working with source. 

for instance if you have a line like this in your script
s="can't"


if you are using
s=rawinput()
or
s=some_file.read()

the quote don't have any special meaning at all


just wondering as you referred to the "user's input" it doesn't sound
like the user is writing python code


John


From phd at phd.pp.ru  Thu Jun 13 10:50:23 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 13 Jun 2002 18:50:23 +0400
Subject: filter2
In-Reply-To: <aeaava$56l8p$1@ID-59885.news.dfncis.de>; from theller@python.net on Thu, Jun 13, 2002 at 04:39:51PM +0200
References: <mailman.1023975618.6474.python-list@python.org> <aeaava$56l8p$1@ID-59885.news.dfncis.de>
Message-ID: <20020613185023.G19424@phd.pp.ru>

On Thu, Jun 13, 2002 at 04:39:51PM +0200, Thomas Heller wrote:
> 
> "Oleg Broytmann" <phd at phd.pp.ru> wrote in message news:mailman.1023975618.6474.python-list at python.org...
> >    I want to have new python builtin - filter2. It is exactly like filter,
> > but it returns also a list of items that didn't pass the test.
> 
> IIRC, in Smalltalk they were named 'select' (like filter), and 'reject' (like your filter2).
> IMO these are much better names then filter and filter2.

   It is too late to rename builtin filter, so I modelled fiter2 after it.

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From aahz at pythoncraft.com  Wed Jun  5 12:31:04 2002
From: aahz at pythoncraft.com (Aahz)
Date: 5 Jun 2002 12:31:04 -0400
Subject: Urllib thread safe?
References: <mailman.1023280808.19305.python-list@python.org>
Message-ID: <adlec8$btt$1@panix1.panix.com>

In article <mailman.1023280808.19305.python-list at python.org>,
Marcel Opsteegh  <Marcel.Opsteegh at nob.nl> wrote:
>
>This message is in MIME format. Since your mail reader does not understand
>this format, some or all of this message may not be legible.

Please don't post MIME messages.  Use plain-text ASCII only.

>Does somebody know whether the urllib is thread safe? 
>I use the following call: urllib.URLopener().open(...)

It's safe as long as you don't share instances between threads.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"In the end, outside of spy agencies, people are far too trusting and
willing to help."  --Ira Winkler


From mhammond at skippinet.com.au  Mon Jun  3 18:40:55 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 03 Jun 2002 22:40:55 GMT
Subject: WIN32 Decimal point/coma
References: <68599ccc.0206030514.4364e899@posting.google.com>
Message-ID: <3CFBF0EC.2000704@skippinet.com.au>

Kaci Tizi Ouzou wrote:
> Greetings all,
> 
> I am having a problem running my application on Win32
> for if in the regional settings the user changes the
> decimal dot to say ',' then floating point operations
> do not work anymore!!!
> 
> After few experiments, I came to the conclusion that, it is
> at 'compile' time that the dot/coma is decided. This means
> if one compiles py files with dot as the decimal 'separator'
> and then changes the settings so that coma is the new separator
> then the code won't work anymore!
> 
> My question is: 
> 
> Is there a call to make so that to ignore the regional settings ?
> 
> Is there a solution/suggestion for this problem?
> 
> Thanks all

Can you please provide some code that demonstrates your problem?  I have 
seen this with external apps controlled by Python - for example, if your 
Python code tries to put a value of "1,2" into an Excel cell, this may 
be treated as 1.2 if your regional settings specify a comma as a decimal 
separator.

So please provide some pure Python code (not using external apps) that 
demonstrates your problem...

Thanks,

Mark.



From aahz at pythoncraft.com  Wed Jun 19 15:22:29 2002
From: aahz at pythoncraft.com (Aahz)
Date: 19 Jun 2002 15:22:29 -0400
Subject: Pedagogic advice needed
References: <3D104D35.E0932297@info.unicaen.fr>
Message-ID: <aeqlll$h4t$1@panix1.panix.com>

In article <3D104D35.E0932297 at info.unicaen.fr>,
Jerzy Karczmarczuk  <karczma at info.unicaen.fr> wrote:
>
>I am going to teach to some biology students (a branch called here
>"neurosciences"; you may speculate yourself what kind of computation
>proficiency they need...) - 

>From http://www.python.org/doc/Newbies.html
you will find "Bioinformatics course in Python"
at http://www.pasteur.fr/recherche/unites/sis/formation/python/

I imagine you might even be able to find some documentation in French.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From bokr at oz.net  Tue Jun 25 15:32:12 2002
From: bokr at oz.net (Bengt Richter)
Date: 25 Jun 2002 19:32:12 GMT
Subject: Suggestions for good programming practices?
References: <mailman.1024970193.27729.python-list@python.org>
Message-ID: <afagfs$3hv$0@216.39.172.122>

On Mon, 24 Jun 2002 18:55:22 -0700, "David LeBlanc" <whisper at oz.net> wrote:

>They may be returned in a tuple, but they're PUT in discrete variables and
ITYM bound to discrete variable names

>not kept in a tuple that I have to programmatically unpack.
>
Right, it's programmatically done for you ;-)

 >>> def foo(): return (1,2,3)
 ...
 >>> a,b,c = foo()
 >>> a
 1
 >>> b
 2
 >>> c
 3
 >>> import dis
 >>> dis.dis(foo)
           0 SET_LINENO               1

           3 SET_LINENO               1
           6 LOAD_CONST               1 (1)
           9 LOAD_CONST               2 (2)
          12 LOAD_CONST               3 (3)
          15 BUILD_TUPLE              3
          18 RETURN_VALUE
          19 LOAD_CONST               0 (None)
          22 RETURN_VALUE

 >>> from miscutil import disex
 >>> disex('a,b,c = foo()')
           0 SET_LINENO               0

           3 SET_LINENO               1
           6 LOAD_NAME                0 (foo)
           9 CALL_FUNCTION            0
          12 UNPACK_SEQUENCE          3
          15 STORE_NAME               1 (a)
          18 STORE_NAME               2 (b)
          21 STORE_NAME               3 (c)
          24 LOAD_CONST               0 (None)
          27 RETURN_VALUE

Note the "UNPACK_SEQUENCE" and "STORE_NAME"s.

>Language lawyer should be shot, _then_ heard.
Along with top-posters? ;-)

>
>David LeBlanc
>Seattle, WA USA
>
>> -----Original Message-----
>> From: python-list-admin at python.org
>> [mailto:python-list-admin at python.org]On Behalf Of Peter Hansen
>> Sent: Monday, June 24, 2002 18:03
>> To: python-list at python.org
>> Subject: Re: Suggestions for good programming practices?
>>
>>
>> David LeBlanc wrote:
>> >
>> > The one that always gets me is "lat,lon,timezone =
>> getLocation("seattle")".
>> > I.E. the ability to return multiple distinct values as opposed
>> to returning
>> > multiple values in the form of a struct as in C or C++.
>>
>> You _are_ returning them "in a struct".  It's spelled "tuple"
>> but other than that what's the difference?
>>
>> -Peter
>> --
>> http://mail.python.org/mailman/listinfo/python-list
>
>
>

Regards,
Bengt Richter


From cben at tx.technion.ac.il  Mon Jun 17 08:26:16 2002
From: cben at tx.technion.ac.il (Beni Cherniavksy)
Date: Mon, 17 Jun 2002 15:26:16 +0300
Subject: why not "'in' in 'in'"?
In-Reply-To: <mailman.1023989567.18847.python-list@python.org>
References: <mailman.1023989567.18847.python-list@python.org>
Message-ID: <Pine.GSO.4.33.0206171519540.8570-100000@techst02.technion.ac.il>

On 2002-06-13, Bjorn Pettersen wrote:

> Ok, how about:
>
> >>> class mystr(str):
> ...   def __contains__(self, s):
> ...     return self.find(s) >= 0
> ...
> >>> 'hello' in mystr('hello world')
> 1
> >>> 'foo' in mystr('hello world')
> 0
> >>>
>
But that's not readable to someone not familiar with your mystr()
definition.  Maybe call the class `substrings':

>>> 'hello' in substrings('hello world')
1

Or maybe `xsubstrings' to calm down the reader who would suspect that you
actually generated all substrings...

-- 
Beni Cherniavsky <cben at tx.technion.ac.il>



From rjones at ekit-inc.com  Sun Jun 23 22:04:16 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Mon, 24 Jun 2002 12:04:16 +1000
Subject: PIL, Python and CGI
In-Reply-To: <Pine.GSO.4.33.0206231958410.27090-100000@infinity.infineon.net>
References: <Pine.GSO.4.33.0206231958410.27090-100000@infinity.infineon.net>
Message-ID: <200206241204.16542.rjones@ekit-inc.com>

On Mon, 24 Jun 2002 12:00, Sascha Ferley wrote:
> import sys
> import _imaging
>
> sys.stderr = sys.stdout
> print "Content-Type: text/plain"
> print sys.path

You need to have the _imaging import after the content-type. Try:

import sys
sys.stderr = sys.stdout
print "Content-Type: text/plain"
import os
print os.environ
import _imaging


Compare the environ with your own too. Might be something there...


   Richard





From marklists at mceahern.com  Mon Jun 24 17:31:14 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 24 Jun 2002 16:31:14 -0500
Subject: Suggestions for good programming practices?
In-Reply-To: <GCEDKONBLEFPPADDJCOEKENIDLAA.whisper@oz.net>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEFFLGAA.marklists@mceahern.com>

[David LeBlanc]
> I'm curious to know how this encourages polymorphism, and how making a
> conditional test implicit rather then explicit is good practice?

Your question reveals my lack of understanding and clarity.  Let me try
again:

If you compare to None with equality, your tests may fail for objects that
overload the comparison functions and don't anticipate other being None.  An
example I encountered was using Tim Peters' FixedPoint for currency values.
It doesn't make sense to do this:

	x = FixedPoint(None)

and yet that's what will happen if you try to do this:

	x = FixedPoint(1.00)
	...
	if x == None:

In other words, this raises an error.  I think the intention behind that
comparison is really to see whether x is None, is in the specific identity
sense.  This, of course, is contrary to my statement about if x being
preferable to if x is not None.

Anyway, I gave an example of why you might not want to use this, but maybe
it seems too convoluted:

	if x == None:

How do you choose between these:

	if x is None:

	if x:

The latter is what you'd use if you wanted to work with instances that
defined __nonzero__ and or __len__.  In other words, you don't really care
what the type of x is, nor whether it's identity is zero, just whether it
considers itself nonzero.

Is one more or less polymorphic than the other?  I don't really know.  But
it sure does sound good.

Cheers,

// mark

-





From Taekyon at macchinetta.miadimora  Mon Jun  3 02:37:02 2002
From: Taekyon at macchinetta.miadimora (Tae kyon)
Date: Mon, 03 Jun 2002 06:37:02 GMT
Subject: solarwolf display ?
Message-ID: <slrnafl9gv.5h4.Taekyon@macchinetta.miadimora>

On a Mandrake 8.2 system I compiled pygame 1.5 and solarwolf 1.5.1.
I have a matrox millennium g200 8 Mb video ram graphics card with 3d support 
by XFree86 4.2.0, and a SVGA non-interlaced video with HSync 30-54,
VSync 50-90. I even get to play tuxracer at a good framerate ! But for
some reason 800x600 fullscreen displays will not work (some loki games
have to be played in a window, for instance, and nethack must be played
on a 1024x768 display).

Unfortunately it seems that solarwolf uses this type of display.
I know very little about python, nothing about pygame; anyway I tried
setting the video mode by changing the relevant options in 
pygame.display.set_mode(()) ... but nothing seems to work.

Any tips ?

-- 
Tae Kyon


From gerhard at bigfoot.de  Tue Jun 18 20:44:09 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 19 Jun 2002 02:44:09 +0200
Subject: tomorrow in yyyymmdd format
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E20192158EE1@admin56.narex.com>
References: <60FB8BB7F0EFC7409B75EEEC13E20192158EE1@admin56.narex.com>
Message-ID: <20020619004409.GA2583@lilith.my-fqdn.de>

* Bjorn Pettersen <BPettersen at NAREX.com> [2002-06-18 18:36 -0600]:
> > From: Gerhard H?ring [mailto:gerhard at bigfoot.de] 
> > Giulio Cespuglio wrote in comp.lang.python:
> > > Hello everybody,
> > > 
> > > We all know that if I want to get the present GMT date in yyyymmdd 
> > > format I should say
> > > 
> > > 	print time.strftime("%Y%m%d", time.gmtime() )
> > > 
> > > Now, what about tomorrow? [...]
> > > 
> > > 	time.strftime("%Y%m%d", time.gmtime(time.time() + 60*60*24))
> > 
> > No, there's nothing better. Unless you're using the DateTime 
> > type from the eGenix mxExtensions:
> 
> Sure there is :-) time.mktime() can take a "malformed" time tuple and
> do something sensible with it. You can therefore get a time tuple from
> gmtime, add one to the day position and call time.mktime() on the
> result:
> 
> >>> x = (2002, 2, 29, 0, 31, 42, 2, 170, 0)
> >>> time.gmtime(time.mktime(x))
> (2002, 3, 1, 7, 31, 42, 4, 60, 0)
> >>>

Pure luck that this works. Don't complain if it stops workin in Python
2.7 and your code breaks ;-)

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From djmitchell at NOSPAM.optushome.com.au  Thu Jun 20 18:14:20 2002
From: djmitchell at NOSPAM.optushome.com.au (David Mitchell)
Date: Fri, 21 Jun 2002 08:14:20 +1000
Subject: Python to MS SQL Server
References: <mailman.1024428323.19642.python-list@python.org> <nQjQ8.213357$Fp1.114650@atlpnn01.usenetserver.com> <m3vg8e9l6a.fsf@ferret.object-craft.com.au>
Message-ID: <pan.2002.06.20.22.14.19.680700.4527@NOSPAM.optushome.com.au>

I'm working on a project with a bunch of Microsoft Consulting Services
guys, and the very clear message from them is that DBLIB is either going
to be made redundant, or already is redundant (can't remember which -
dealing with MCS guys for months on end has that sort of 
"time shifting/mind altering" effect.  I've learned not to ever talk to
them about .NET, since in their minds it's proven stable technology and
millions of applications worldwide are already using it!).

In any case, DBLIB probably won't break after all these years, but it may
conceivably disappear in the next couple of releases of SQL Server.  MS
certainly won't be fixing any security problems or bugs with it; that
much was made extremely clear

You might want to consider this if you're embarking on new development
work

Regards

Dave Mitchell
> 
> I also have a module which uses the MS db library.
> 
>     http://www.object-craft.com.au/projects/mssql/
> 
> It needs more work, but might be sufficient for your needs.
> 
> - Dave


From fperez528 at yahoo.com  Thu Jun 20 14:52:36 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 20 Jun 2002 12:52:36 -0600
Subject: Problems with mayavi, pyvtk
References: <7396d2b2.0206191025.3351a888@posting.google.com> <aeqme8$avd$1@peabody.colorado.edu> <7396d2b2.0206201046.4f0530fa@posting.google.com>
Message-ID: <aet89k$jfh$1@peabody.colorado.edu>

Lemniscate wrote:

> Hey, thanks for the info.??Here's?the?only?problem...??I?still?get
> errors.??Here?is?what?happens?when?I?try?to?import?vtkpython...
> 
>>>> import vtkpython
[snip]

Well, let me suggest you post specifically to the mayavi list.  I use mayavi 
under linux, so I can't help you with dll problems in Windows. Under Linux I 
got vtk & friends running just fine as per the instructions.

Keep in mind that Prabhu doesn't use windows either (at all). Maybe someone 
on the mayavi list does use windows and can lend you a hand though.

good luck,

f.


From chris.gonnerman at newcenturycomputers.net  Wed Jun 19 08:40:11 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Wed, 19 Jun 2002 07:40:11 -0500
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com> <mailman.1024460666.24801.python-list@python.org> <B8XP8.47925$GY.15259265@twister.nyroc.rr.com>
Message-ID: <003f01c2178e$75137700$0101010a@local>

----- Original Message -----
From: "George Hester" <hesterloli at hotmail.com>


> I don't want to learn the language.

Ah, so.  Then go away, and come back when you do want to learn.  And don't
whine.


Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net






From bogus@does.not.exist.com  Wed Jun  5 09:41:23 2002
From: bogus@does.not.exist.com (Hugo van der Merwe)
Date: Wed, 05 Jun 2002 15:41:23 +0200
Subject: distutils and distributing "scripts"
Message-ID: <adl4e4$19ep$1@news.adamastor.ac.za>

I have recently started using distutils. I have a .pyw file for the windows 
users, how do I get this to be a part of my ... sdist, for example?

I have added this to the scripts list passed to "setup()", this seems to me 
the "correct" place for it (I want to eventually play with py2exe).

It seems "python setup.py sdist" does not add "scripts" by default though, 
just the py_modules. The docs seem to confirm this. How do I get my .pyw 
file distributed, do I really have to go and do all that MANIFEST.in stuff, 
just for this one file?

Thanks,
Hugo van der Merwe


From pyth at devel.trillke.net  Tue Jun 18 07:11:04 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 18 Jun 2002 13:11:04 +0200
Subject: what *is* a class?
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8BF0A396@natasha.auslabs.avaya.com>; from tdelaney@avaya.com on Tue, Jun 18, 2002 at 10:59:12AM +1000
References: <B43D149A9AB2D411971300B0D03D7E8BF0A396@natasha.auslabs.avaya.com>
Message-ID: <20020618131104.Z15079@prim.han.de>

Delaney, Timothy wrote:
> > From: holger krekel [mailto:pyth at devel.trillke.net]
> > 
> > I agree that this is a good thing. But it wouldn't hurt to have
> > 
> >     reverse, sort, extend
> > 
> > as standalone functions somewhere in the stdlib.  They would return
> > a copy...
> 
> Personally, I think this is a great idea. But I would exclude extend.

I'd even like to have something like extend (see below) ...

> I think there should be a library of "iterable" functions (perhaps in
> builtins ;) which return an iterator, and accept any iterable (where
> appropriate). Let's call it builtin module iterable (or maybe iterator ...).
> Functions defined in this module would be specified to not modify the
> original iterable (indeed, should all call iter(iterable) before doing
> anything). Of course, whee appropriate a function passed in could modify the
> original.
> 
> Examples of such functions would be:
> 
> iterable.sort(iterable [, func])
> iterable.reverse(iterable)

that would be quite expensive for common cases (like lists).  reverse and 
sort should also allow to operate on complete sequence types. 
iterable.reverse could be cheap again then and iterable.sort would be
as expensive as it is now (but not requiring any inplace-steps).

> iterable.map(func, iterable, ...)
> iterable.filter(func, iterable, ...)
> iterable.reduce(func, iterable, ...)
> iterable.zip(iterable, ...)
> iterable.range([start,] stop[, step])

yes, indeed! i'd probably do 'from iterable import *' :-)

seriously, i'd include something like concat as an iterable, too: 

    for item in iterable.concat(iter1,iter2):
        # items iterates with iter1, then iter2 

All the the iterable functions should accept sequence and 
possibly mapping types (same semantics as for 'for'-loops now).

> There have been suggestions like this before, but I don't recall one for
> sort() and reverse(), and these two have fired me :)

I think an iterable module should be done. Although there would 
inevitably be overlapping functionality with the current builtins.

Maybe something for the train ride to EuroPython :-)

> Tim Delaney

    holger krekel






From peter at engcorp.com  Sat Jun  1 11:08:50 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 01 Jun 2002 11:08:50 -0400
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org>
Message-ID: <3CF8E382.3A44F0DA@engcorp.com>

Andrei Kulakov wrote:
> 
> Hello,
> 
>     "hello world" type program from wxPython tutorial takes 6
> seconds to start on my cel 366 128mb system (running Debian). Is
> this typical? Why is it so slow? I'm looking for a graphic
> toolkit that'd be fairly quick. I don't want to use Tk because of
> the looks.

How fast do you need it to be?  (That's a serious question.
Are you planning to open and close the application so frequently
that six seconds is unacceptable?)

On my blazing 233 MHz Windows 98 machine, with Python 2.2.1
and WxPython 2.3.2.1, a hello world program I just wrote
runs in about two seconds, for what it's worth.

-Peter


From xscottgjunk at yahoo.com  Sun Jun  9 15:28:22 2002
From: xscottgjunk at yahoo.com (Scott Gilbert)
Date: 9 Jun 2002 12:28:22 -0700
Subject: array, list, performance...
References: <mailman.1023369012.10674.python-list@python.org> <3CFFB86A.4114459A@accessforall.nl> <83ofel5m8o.fsf@panacea.canonical.org>
Message-ID: <79b2b0bb.0206091128.2b54271b@posting.google.com>

Kragen Sitaker <kragen at pobox.com> wrote: 
> >>> import MetaPy.Speed
> >>> def mklist(n):
> ...     ar = []
> ...     for ii in xrange(n):
> ...             ar.append(ii)
> >>> MetaPy.Speed.seconds_per_call(lambda: mklist(0))
>  1.7825510264682151e-05
> >>> MetaPy.Speed.seconds_per_call(lambda: mklist(1))
>  2.862680041153042e-05
> >>> MetaPy.Speed.seconds_per_call(lambda: mklist(10))
>  7.8054009556034588e-05
> >>> MetaPy.Speed.seconds_per_call(lambda: mklist(100))
>  0.00057740376706547905
> >>> MetaPy.Speed.seconds_per_call(lambda: mklist(1000))
>  0.0055893541133707342
> >>> MetaPy.Speed.seconds_per_call(lambda: mklist(10000))
>  0.055918868174255905
> >>> MetaPy.Speed.seconds_per_call(lambda: mklist(100000))
> 0.59515897688580466
> 
> Looks linear to me.

In a very practical sense your test is the one that matters, since
it's usually time that we care about.  Your test could just as easily
be measuring how linear the memory allocator is or whether cache
effects are linear, since either of those could dwarf the cost of the
list operations.  Who knows, you might actually have a realloc() that
does something smart.  Also, it's pretty tough to tell the difference
between O(N) and O(N*log(N)) with such small numbers.

Here's some code that just counts the operations:

def roundupsize(n):
    "Very similar to what is actually in listobject.c"
    nbits = 0
    n2 = n>>5
    n2>>=3
    nbits+= 3
    while n2:
      n2>>=3
      nbits+=3
    return ((n>>nbits) + 1) << nbits

def count(N):
    ops = 0
    size = 0
    i = 0
    while i<N:
        i += 1
        if i>size:
            # have to pay for a copy and append
            ops += size
            ops += 1
            size = roundupsize(i)
        else:
            # only pay for append
            ops += 1
    return ops

# N, ops, N/ops for powers of 10
1 1 1.0
10 18 0.555555555556
100 724 0.138121546961
1000 12264 0.0815394651011
10000 139536 0.0716660933379
100000 1590432 0.0628759984709
1000000 18333760 0.0545441851535
10000000 165188736 0.0605368152947
100000000 1425399552 0.0701557678054


# N, ops, N/ops for powers of 2
1 1 1.0
2 2 1.0
4 4 1.0
8 8 1.0
16 24 0.666666666667
32 80 0.4
64 288 0.222222222222
128 1088 0.117647058824
256 4224 0.0606060606061
512 5888 0.0869565217391
1024 12288 0.0833333333333
2048 37376 0.0547945205479
4096 50688 0.0808080808081
8192 101888 0.0804020100503
16384 302592 0.0541455160745
32768 409088 0.0801001251564
65536 818688 0.0800500312695
131072 2424320 0.054065469905
262144 3276288 0.0800125019534
524288 6553088 0.0800062504883
1048576 19398144 0.0540554807718
2097152 26213888 0.0800015625305
4194304 52428288 0.0800007812576
8388608 155188736 0.0540542323896
16777216 209714688 0.080000195313
33554432 419429888 0.0800000976564


The powers of 10 wave up and down a bit, but the powers of 2 make a
nice .08 .08 .05 pattern.  Still looks linear, but it has bouts of
being non-linear before the next major step compensates for it.  It
doesn't look to be O(N*N), and that's what really matters for large N.


From tdelaney at avaya.com  Mon Jun 24 20:18:48 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 25 Jun 2002 10:18:48 +1000
Subject: HELP-ME pLEASE
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A3BD@natasha.auslabs.avaya.com>

> From: jubafre at zipmail.com.br [mailto:jubafre at zipmail.com.br]
> 
> i don?t know how to make a .exe,i?m atachin the file .py
> 
> SOmebody can compile for me please a need the .exe urgently!!!!

http://www.python.org/cgi-bin/faqw.py?req=show&file=faq04.028.htp

Tim Delaney




From vvainio at tp.spt.fi  Fri Jun 28 08:42:36 2002
From: vvainio at tp.spt.fi (Ville Vainio)
Date: 28 Jun 2002 05:42:36 -0700
Subject: pprinting generators
Message-ID: <ad496f8.0206280442.1fdbae42@posting.google.com>

On 2.2.1, pprint.pprint doesn't seem to be able to handle generators:


>>> from __future__ import generators
>>> def f(): yield 0
...
>>> import pprint
>>> pprint.pprint(f())
<generator object at 0x007F91D0>

Surely pprint could be more helpful with generators, iterating through
them fully or just n first elements, or whatever. Or is this a design
decision with pprint?

-- Ville


From rnd at onego.ru  Sat Jun 22 14:53:51 2002
From: rnd at onego.ru (Roman Suzi)
Date: Sat, 22 Jun 2002 22:53:51 +0400 (MSD)
Subject: Python hits the spot
In-Reply-To: <3D1482AE.8FA299D8@kfunigraz.ac.at>
Message-ID: <Pine.LNX.4.44.0206222236470.30675-100000@rnd.onego.ru>

On Sat, 22 Jun 2002, Siegfried Gonzi wrote:

>1. After starting the process from within idle the complete system
>freezes after the first wavelength (after about 20 minutes). I couldn't
>kill a process or open a different session (via F2) in order to kill
>python. The last resort: turn off the power supply.

>I cannot believe that Linux is responsible for Python's ill behavior
>under Linux. Before you start ranting about note the following: In a
>first step it has nothing to do with my programming style (admittedly
>not very good and famous) it has to do that the same calculation cannot
>be performed on Linux and Python 2.2!

Perhaps you are too nervous, Siegfried. If I understood correctly, you tried
to wrap some fortran code to use from within Python while using IDLE.
I just can't comprehend, why do you need IDLE...

I never experienced 'freezes' you are describing, but I guess they will happen
under Windows if you are using C code. (Under Linux you will get core dumps).

It's no wonder that 1000 lines of uncommented code behave like that, taking
into account that you were trying to integrate several (Tcl/Tk, Fortran, C,
Python) systems in a hurry, and, I presume, not knowing most of them.

I do not remember when I used power switch to restart my Linux box,
and you are telling it was last resort.

So, the best solution is... to flame here. It peacifies soul, probably.

In the situation you describe, I usually take a break for a day
to calm done and after that problems are solved in a matter of
an hour.


Sincerely yours, Roman Suzi
-- 
rnd at onego.ru =\= My AI powered by Linux RedHat 7.2





From christophe.delord at free.fr  Fri Jun 14 12:49:08 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Fri, 14 Jun 2002 18:49:08 +0200
Subject: I feel stoopid ... this code is to verbose
References: <3D0A1791.3060307@mxm.dk>
	<20020614183945.17cc73f7.christophe.delord@free.fr>
Message-ID: <20020614184908.5a174d7a.christophe.delord@free.fr>

Here is a 3-liner ;-)


> > I just cannot seem to find the nice 3-liner I expected from the 
> > beginning. Has anybody got a better solution ? I thought somebody might 
> > find it a fun exercise. Well I have...

def stringSlicer(st, chunkSize=3):
	def slc(st): return st and slc(st[:-chunkSize])+[st[-chunkSize:]] or []
	return ".".join(slc(st))




-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From eppstein at ics.uci.edu  Mon Jun  3 12:43:48 2002
From: eppstein at ics.uci.edu (David Eppstein)
Date: Mon, 03 Jun 2002 09:43:48 -0700
Subject: Graph distances, how can I speed it up?
References: <33803989.0206030235.4134d33e@posting.google.com> <d571ae53.0206030815.2e52aac@posting.google.com>
Message-ID: <eppstein-F307D5.09434803062002@news.service.uci.edu>

In article <d571ae53.0206030815.2e52aac at posting.google.com>,
 ellisjb at my-deja.com (Jonathan Ellis) wrote:

> The obvious idea then is to not calculate anything until all edges
> have been added.  Then use dijkstra's algorithm
> (http://ciips.ee.uwa.edu.au/~morris/Year2/PLDS210/dijkstra.html)
> to compute and store distances as needed.

There is a Python implementation of Dijkstra's algorithm at
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/119466

-- 
David Eppstein       UC Irvine Dept. of Information & Computer Science
eppstein at ics.uci.edu http://www.ics.uci.edu/~eppstein/


From jolsen at mailme.dk  Tue Jun 11 05:33:28 2002
From: jolsen at mailme.dk (Jesper Olsen)
Date: 11 Jun 2002 02:33:28 -0700
Subject: python and shared libaray
References: <a263f04a.0206101820.5461339f@posting.google.com>
Message-ID: <cf0ad9fb.0206110133.7d3ea4f3@posting.google.com>

You can not import the .so lib directly - e.g. you need to map python
types to C types when calling functions.

So there is a little more work, but if you are familiar with C it is
not
a big problem.

See the "Writing an extention module" in the standard documentation.

I recommend that you use distutils for installing/compiling your
module.
You can then specify the .so lib so that your C-wrapper code is liked
with the library (see section 3.3.4 in "Distributing Python Modules").

Cheers
Jesper



spokra at home.com (Steven Pokrandt) wrote in message news:<a263f04a.0206101820.5461339f at posting.google.com>...
> I have a vendor supplied .so file, header file and docs.  I would like
> to use this
> .so file in python but don't have the src.  Is it possible to link
> this library ie:
> 
> import filename.so
> 
> when I try the python complains about no init function..  can I add
> this to the .so  without having the src?
> 
> can someone point me to a howto?
> 
> please send message to the following address..  spokra (AT) attbi
> (dot) com
> 
> you get the meaning of the email address, and hopefully the spammers
> don't


From emile at fenx.com  Thu Jun 13 22:56:28 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 14 Jun 2002 02:56:28 GMT
Subject: Type subclassing: bug or feature
References: <aeau48$111$1@panix1.panix.com> <Xns922CEE783C10Dcliechtigmxnet@62.2.16.82> <aebijm$aq0$0@216.39.172.122>
Message-ID: <wXcO8.177829$352.9560@sccrnsc02>

Bengt Richter
> BTW, I just thought of a 'where' keyword for 'for' e.g., the above for
would become
>
>     for name,m in str.__dict__.items() where type(m) ==
type(str.find):
>
> sort of like sql select ... where ... or the list comprehension if.
>

why not use if, as in

[ name,m for name,m in str.__dict__.items() if type(m) ==
type(str.find) ]



--

Emile van Sebille
emile at fenx.com

---------



From nikolai at micon.no  Tue Jun 18 18:48:51 2002
From: nikolai at micon.no (Nikolai Kirsebom)
Date: Tue, 18 Jun 2002 22:48:51 GMT
Subject: Using calldll with mciSendString
Message-ID: <nNOP8.3825$yj7.62301@news4.ulv.nextra.no>

I'm trying to use the windll/calldll module to access the mciSendString
(winmm.dll) function.  One of the parameters to the function is a string
containing returned information.  I'm using wxPython as GUI to show a AVI /
MPEG movie (in a wxPanel).  The play command may be given a 'notify' flag,
which results in the player calling back to notify when movie is finished.
The last parameter to the mciSendString is a HANDLE (hwndCallback).

Can someone give an example how to retreive the returned information and
also how I set up the callback.


Thanks
Nikolai Kirsebom





From hesterloli at hotmail.com  Tue Jun 18 23:45:03 2002
From: hesterloli at hotmail.com (George Hester)
Date: Wed, 19 Jun 2002 03:45:03 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com>
Message-ID: <37TP8.47886$GY.15021139@twister.nyroc.rr.com>

The whole issue started from here.  I did not deserve to be told to "crawl
before I walk."  I may have responded to that inappropriately I confess.  But
that does not excuse the implied derogatory.  Since that the lambasting
crecscendoed.

--
George Hester
_________________________________
"Peter Hansen" <peter at engcorp.com> wrote in message
news:3D0E8AC8.6C5C96EE at engcorp.com...
> George Hester wrote:
> >
> > "Gerhard H?ring" <gerhard at bigfoot.de> wrote:
> > > You did not import the sys module.
>
> > Probably not.  Where do I import that from and how do it?  I am trying to
put a
> > script in ASP using Python as the script language to give me the version of
> > Python that I am running when I call the ASP.  The more the script works out
of
> > the box the better I'll be able to catch on to this language.  Thanks.
>
> It sounds to me like you should probably crawl before you walk.
> I suggest downloading the tutorial and installing Python on
> your own machine and using it interactively from the prompt
> before you advance to something like ASP scripting.  Your
> responses clearly show you aren't ready to write code in an
> environment which is doubtless going to mask from you many
> of the errors and messages which you will encounter as you learn.
>
> In the above case you would be able to experiment and learn about
> these things:
>
> C:\>python
> Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
> >>> sys.version
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> NameError: name 'sys' is not defined
>
> Here's where you would pause to learn a little more about Python....
>
> >>> import sys
> >>> sys.version
> '2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)]'
>
> -Peter




From claird at starbase.neosoft.com  Mon Jun 10 15:14:21 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 10 Jun 2002 14:14:21 -0500
Subject: activating a webpage's click command from python
References: <mailman.1023727993.12908.python-list@python.org>
Message-ID: <A4FC8A3C8DC6D7FC.72AC3E95E17B8175.32B5F0342321025E@lp.airnews.net>

In article <mailman.1023727993.12908.python-list at python.org>,
Grimes, Jessica C <grimesj at wharton.upenn.edu> wrote:
>How do I activate a click command on Yahoo's website from Python?    
>
>I want to create a  program that will ask the user what stock he/she wants to look up.  The program will then extract financial data from yahoo and import it into a spreadsheet.  
>
>My problem is that I don't know how to make the application take the user's input (the stock's ticker symbol) and place the user's input from python into the space where yahoo accepts input for a ticker symbol and activate yahoo's "Look up" click command. 
>
>I hope I have explained my question clearly.  If my proposed method is impossible, can you suggest another approach to designing the program? 
			.
			.
			.
Peter's already answered your question quite accurately
in several respects.

"Web scraping" <URL: http://cedar.intel.com/cgi-bin/ids.dll/
content/content.jsp?cntKey=GenericEditorial::ws_scraping&
cntType=IDS_EDITORIAL&catCode=CJA > is one common label
for this general idea of programmatic access to data in Web
pages.

If you don't regard yourself as a programmer, you might find
Web scraping a frustrating pursuit.  Some programmers do, too.

What you *really* want are the data available through a "Web
service" (WS), as opposed to a Web page.  I don't know how
far along Yahoo is in its WSification.

You're far from the first to write an application which re-
quires real-time market data.  Do you specifically want the
content from the Yahoo page, or would good data from anywhere
meet your requirements?
-- 

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


From irmen at NOSPAMREMOVETHISxs4all.nl  Tue Jun 11 13:28:54 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Tue, 11 Jun 2002 19:28:54 +0200
Subject: M2Crypto: select() behaves weird on SSL socket
References: <3D04ECED.2030202@NOSPAMREMOVETHISxs4all.nl> <slrnag9rie.enn.akuchlin@ute.mems-exchange.org> <3D04FBF5.8090407@NOSPAMREMOVETHISxs4all.nl> <3D04FE2F.6070302@NOSPAMREMOVETHISxs4all.nl> <1023762896.869994@yasure>
Message-ID: <3D063356.30200@NOSPAMREMOVETHISxs4all.nl>

Donn Cave wrote:

> Wrong, I'd say.  SSL's buffer, like the file object's stdio buffer,
> is user process storage.  From select's point of view (the low level
> device) you have already read that data, and it doesn't know or care
> what you did with it.

Well, that is one way of looking at it. I guess you're right.
But it breaks my code...

If there is a SSL wrapper for plain sockets, isn't there some
sort of SSL wrapper of select()?


> I see my reply earlier today followed up a separate thread, after
> you posted your initial question twice.  If you haven't seen it,
> a hint:  SSL_pending(self->ssl.)  (I don't know much more than that,
> anyway.)

I don't know if this function is exposed in M2Crypto... if it is,
I could probably get my code to work. If it isn't, I don't know
how to fix it :-(

Sorry about posting twice... must have hit 'send' once too often...

Irmen



From miracle at paradise.net.nz  Fri Jun 28 18:29:10 2002
From: miracle at paradise.net.nz (Matthew Sherborne)
Date: Sat, 29 Jun 2002 10:29:10 +1200
Subject: How to find out DNS ?
References: <3D1C79E8.22371.4709FB1@localhost>
Message-ID: <3D1CE336.4080902@paradise.net.nz>

There is a way but it's very hard. You take a list of root dns servers, 
you get your local host name, (gethostbyaddr(youripaddress)), you remove 
from it the computer name, so you've just got the domain name. Then 
taking a list of "root" dns servers, you ask one of them, what's the dns 
server for this domain?

It'l l say "I don't know, ask server X" then you ask server X, which 
will be a closer server to your host than the root server. You keep on 
going until one of the servers says "yes, This is/I am  your server" 
then that's your dns server.

You can do this with  the code in http://pydns.sf.net.

Is it an open-source project you're working on? I might take an interest :)

Blessings
Matthew Sherborne







From peter at engcorp.com  Tue Jun 11 01:31:31 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 11 Jun 2002 01:31:31 -0400
Subject: Changing the path on Win2k/XP
References: <mailman.1023768255.16926.python-list@python.org>
Message-ID: <3D058B33.8E00EA67@engcorp.com>

Bjorn Pettersen wrote:
> 
> Is there a way to change the path on Win2k/XP through Python? (I'm
> trying to write a .dll deployment script for our production department,
> and I'd like to give them a "one button" install...)

I don't think there is.  Generally speaking on most platforms 
with the PATH concept, you can't change it from within a process
(I guess since that could completely screw up the calling process'
environment).

You can play tricks.  Since a batch file (or .cmd) can
change the PATH, call the deployment script from a 
.cmd file which subsequently calls a SETPATH.CMD type
of file.  In your deployment script, modify that file
so that when it is called after your script exits, the
path gets changed.

Alternatively, install the DLL to a directory already in
the path, such as the SYSTEM32 folder or whatever...

Or just count on the fact that DLLs are found in the
same directory as the application if they are installed
there.

For that matter, on Win2K does the system really look
only to the path for loading DLLs?  I would think there
are places in the registry where pointers to these things 
are supposed to be installed.

Do you really need to install a DLL all by itself into
a new folder and add it to the path?  Why?

-Peter


From robin at execulink.com  Sat Jun 22 18:04:46 2002
From: robin at execulink.com (robin at execulink.com)
Date: Sat, 22 Jun 2002 18:04:46 -0400
Subject: Creating Dynamic Web Pages (ZOPE)
References: <46ca81a0.0206062016.789d3af6@posting.google.com> <J29M8.358696$%u2.156726@atlpnn01.usenetserver.com> <46ca81a0.0206141201.23db4a8c@posting.google.com> <kh3tgus5d6hlrt4qjsdg1odr6j9ft0nue2@4ax.com> <3D0ED3F3.2020008@mxm.dk> <u4NP8.166082$Fp1.153799@atlpnn01.usenetserver.com>
Message-ID: <vgs9hu4rrfj7ti6j0t0pttla7qih9t38p1@4ax.com>

I wrote:
> I would very much advise against DreamWeaver and similar tools that
> pretend designing web pages is a WYSIWYG activity, which it certainly
> is not.

"Max M" <maxm at mxm.dk> wrote ...
> Dreamweaver is a very fast and convenient way to edit html. If you know
> what you do, your code is just as clean as handcrafted html. There is
> nothing in html that you cannot tweak in Dreamweaver.

"Steve Holden" <sholden at holdenweb.com> wrote:
>That's been my experience too. I have certainly been no fan of WTSIWYG
>tools, FrontPage having made my life a misery for some short time before I
>abandoned it. DreamWeaver does seem to be in a rather different class,
>though not perfect. Don't knock it 'til you've tried it.

Well, I have tried it. In my role as web app designer and project
leader I have also had to deal with Dreamweavered code from
consultants. I can recognise the mess at a distance... 

To me this argument is similar to saying that Perl is a great
language, because although you can write obsfucated code you don't
have to.

But I prefer Python because for the most part it encourages good
coding practices. IMO, a text editor with templating language
encourages good markup practices.

Perl, Dreamweaver, et al. encourage write-only documents and are
especially ill-suited to teamwork.

-- robin


From marklists at mceahern.com  Wed Jun 12 08:42:54 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Wed, 12 Jun 2002 07:42:54 -0500
Subject: Zope Newbie query
In-Reply-To: <ugefre2980kj09@corp.supernews.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMEDILEAA.marklists@mceahern.com>

> I have been playing with zope and would like to use it properly now the
> problem I'm encountering is: How do I remotely install zope on a raq4 on
> which I am not the main raq4 administrator (i.e I have a subdirectory on
> which my site is hosted).? Do I have to have root priveliges to
> install and
> run zope?

This might be related:  mxCGIPython lets you install Python on a shell
account with a single binary.

Cheers,

// mark

-





From d2002xx at myrealbox.com  Sat Jun 22 08:57:30 2002
From: d2002xx at myrealbox.com (d2002xx)
Date: 22 Jun 2002 05:57:30 -0700
Subject: The best GUI library in python
Message-ID: <8d3f4438.0206220457.b820418@posting.google.com>

Hello!!

I want to learn GUI in python, who would tell me which GUI library is
the best? (don't tell me Tkinter, it's too slow......)


From usenet at thinkspot.net  Sun Jun  9 00:35:29 2002
From: usenet at thinkspot.net (Sheila King)
Date: Sat, 08 Jun 2002 21:35:29 -0700
Subject: Python MTA
References: <20020607232709.GA12947@lilith.my-fqdn.de> <mailman.1023496571.30237.python-list@python.org>
Message-ID: <adttb2.l4.1@kserver.org>

On Fri, 7 Jun 2002 17:35:33 -0700, "David LeBlanc" <whisper at oz.net> wrote
in comp.lang.python in article
<mailman.1023496571.30237.python-list at python.org>:

> <ahem> and the documentation (something more then help(smtpd) please) for
> smtpd is where? :P)
> 
> It's not in pythondocs for 2.2.1, unless it's well hidden. smtplib is
> documented, but that's a client interface.
> 
> TMDA: BTDT - Like ASK, requires an external Linux/Unix based MTA. Thanks for
> the pointer though :)

smtpd is an undocumented module thrown in as a bonus. (By Barry Warsaw)
What documentation there is would be in the source code itself.

I believe smtpd may have some Unix dependencies.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From gleki at gol.ge  Wed Jun 12 08:32:06 2002
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Wed, 12 Jun 2002 14:32:06 +0200
Subject: AttributeError? Why?
References: <3D07337A.870E82E8@gol.ge> <ae7dlv$ouk$1@wanadoo.fr>
Message-ID: <3D073F46.FAD7AFDC@gol.ge>

No, the class ARScenario is in the module ARScenario.py

Another thing is that, importing ARScenario doesn't fail.
>>> import SED
>>> import SED.ARScenario
>>>
====

However, the error maybe as follows:
I was forgetting to import SED in the modules...
Now both
ARScenario(self.xdata)
and
ARScenario.ARScenario(self.xdata)

work.

Thanx for the help.


Eric Brunel wrote:

> Giorgi Lekishvili wrote:
> > Hi all!
> >
> > I have been making a module, which, as I believed, was organized
> > correctly. It has __init__.py file, which is empty, and several modules.
> >
> > Now, what happens: import goes fine. But when I try to create a class,
> >>>> s=SimAn.SimAn([],[],None,0,0)
> > Traceback (innermost last):
> >   File "<pyshell#18>", line 1, in ?
> >     s=SimAn.SimAn([],[],None,0,0)
> >   File "D:\Python20\SED\SimAn.py", line 18, in __init__
> >     self.Scenario=ARScenario.ARScenario(self.xdata)
> > AttributeError: ARScenario
> > =====
> > Of course, ARScenario is there:
> >>>> dir(SimAn)
> > ['ARScenario', 'SimAn', '__builtins__', '__doc__', '__file__',
> > '__name__', 'copy', 'math']
> > =====
> > and, of course, the class ARScenario based on the class Scenario, has
> > __init__(self,..) method.
> >
> >
> > The question: where's my error?
>
> The answer is in the question: why do you try to create an instance of
> ARScenario with:
>
> self.Scenario=ARScenario.ARScenario(self.xdata)
>
> since your ARScenario class is in the SimAn module, as shown by:
>
> >>> dir(SimAn)
> ['ARScenario', 'SimAn', '__builtins__', '__doc__', '__file__', '__name__',
> 'copy', 'math']
>
> If your class ARScenario is in the same module than your SimAn class
> (namely SimAn), just create the ARScenario instance with:
>
> self.Scenario=ARScenario(self.xdata)
>
> It should be enough.
>
> HTH
> --
> - Eric Brunel <eric.brunel at pragmadev.com> -
> PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com



From spamtrash at softhome.net  Tue Jun 18 04:53:01 2002
From: spamtrash at softhome.net (Johannes Gijsbers)
Date: 18 Jun 2002 01:53:01 -0700
Subject: __subclasses__
References: <a9bcagho74.22t.jimrowe@optilink.com>
Message-ID: <e87ce949.0206180053.25b5fcad@posting.google.com>

jimrowe at optilink.com (James Rowe) wrote in message news:<a9bcagho74.22t.jimrowe at optilink.com>...
> So I suppose the question is: Why is __subclasses__ not documented
> or listed?  Can we safely use __subclasses__ and not worry about
> it "going away" in the future?

Last time I checked most of the work done in unifying types and
classes wasn't documented yet. I'm not sure __subclasses__ and the
unification are related, but I guess so.

The only info I've found was a thread by Bruce Eckel[1]. He also
mentions it in the draft of his book Thinking in Python.

jlg

[1] http://mail.python.org/pipermail/python-list/2001-December/075864.html
[2] http://www.mindview.net/Books/TIPython


From fredrik at pythonware.com  Sat Jun 29 17:42:32 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 29 Jun 2002 21:42:32 GMT
Subject: private
References: <60FB8BB7F0EFC7409B75EEEC13E201922151BA@admin56.narex.com> <mailman.1025383959.4322.python-list@python.org>
Message-ID: <cRpT8.45636$n4.10840411@newsc.telia.net>

Dave Reed wrote:

> > .. and of course #define private public, or just go in and edit the .h
> > file...
>
> Ok, I was thinking about doing it w/o recompiling the object or lib
> file containing the class.

the point is that you don't necessarily have to recompile anything
(except your own code, of course).

depends on the compiler, though...

</F>




From cliechti at gmx.net  Thu Jun 13 18:50:09 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 14 Jun 2002 00:50:09 +0200
Subject: returning traceback strings
References: <mailman.1024006339.11571.python-list@python.org>
Message-ID: <Xns922D8E94C2EAcliechtigmxnet@62.2.16.82>

Matthew Boedicker <mboedick at mboedick.org> wrote in
news:mailman.1024006339.11571.python-list at python.org: 
> I am looking for something exactly like traceback.print_exc(), but
> instead of writing the traceback text to a filehandle, I want it to
> return it to me as a string.  I was unable to find any way to do this
> using the traceback module, since everything there seems to be in
> terms of file handles being passed around.
> 
> The problem is I need to pass the entire multi-line traceback string
> to a file-like object's write() method in one call, not line by line.
> 
> This is the best workaround I was able to come up with:
> 
> import cStringIO
> import traceback
> 
> def print_exc_str():
>     buf = cStringIO.StringIO()
>     traceback.print_exc(file=buf)
>     return buf.getvalue()
> 
> Any suggestions for a better way?  What about the traceback module
> using strings internally, and wrapping another layer around that to
> write to files and file-like objects?

try that:
  ''.join(traceback.format_exception(*sys.exc_info()))

chris

-- 
Chris <cliechti at gmx.net>



From jbublitzNO at SPAMnwinternet.com  Sat Jun  8 13:02:54 2002
From: jbublitzNO at SPAMnwinternet.com (Jim)
Date: Sat, 08 Jun 2002 17:02:54 GMT
Subject: Detecting OS and Window Manager
References: <mailman.1023304388.9659.python-list@python.org> <ufvsco8mlhf213@corp.supernews.com> <7hjrda.ilm.ln@127.0.0.1>
Message-ID: <3D023892.60106@SPAMnwinternet.com>

Jim Richardson wrote:
> On Fri, 07 Jun 2002 00:34:30 +0100,
>  Darren Winsper <dw133 at cs.york.ac.uk.go.away> wrote:
>>In <mailman.1023304388.9659.python-list at python.org>, Sean 'Shaleh' Perry
>>wrote:

>>>the only fool proof way to detect KDE, GNOME, etc is to check for X atoms.  The
>>>bigger question is why do you care?  As long as they can display a GUI you
>>>should not be bothered about which one it is.  Especially under the X Window
>>>System it was designed so that random applications DIDN'T know what they were
>>>running under.

>>As the person writing the KDE front-end to the mentioned application, let
>>me answer your question.  The idea is that if the user is running KDE,
>>then they get a KDE front-end by default.  I also hope that should
>>somebody write a GTK/GNOME front-end, if the user was running GNOME they
>>would get a GTK-based front-end by default.  It's good UI design.

Not sure what you mean by 'running KDE' ?? Running kwm? Running kicker? 
KDE libs loaded? KDE installed? I don't think there really is a definite 
meaning for 'running KDE'.

>>>Also remember KDE is not a window manager, it is a desktop environment.  It is
>>>perfectly reasonable to run KDE and use fvwm (or even twm) as the window
>>>manager.  Same applies to GNOME.

>>OK, Andrew meant DE instead of window manager, but it's still a valid
>>question.

>>>Also, what about someone who only runs say
>>>Window Maker and never uses GNOME or KDE?

>>Well, in that case we just choose a default.  The thing is, we want to
>>detect if PyKDE is installed.  If it is, then we can do "import kdecore"
>>without risk of it throwing an exception.  If it isn't, I'd rather not
>>have to try something along the lines of attempting "import kdecore" and
>>then having to catch an exception if it's not found.

> But the only way to see if PyKDE is installed, is to check it, KDE
> doesn't require PyKDE, you can't assume that PyKDE will be present
> simply because someone is running KDE. About the only way I can see you
> toing this, is with a whole lot of try/fail checks, and even then, if
> someone is running something other than KDE/Gnome, you will be back to
> the default. Why not start the default front end first time, then ask
> the user what front end they would prefer, then save that to a dotfile? 

PyKDE normally installs in python/site-packages, so you can check for 
kdecore.py  using os.path.exists() for example, although I don't really 
understand why 'try: import kdecore ...' should be a problem. There are 
more problems with os.path.exists (like knowing for sure what the path 
is) while 'import' failure/success guarantees that Python can't/can find 
kdecore.

In fact to be 100% certain, you'd want be sure _all_ of the PyKDE 
modules you need are accessible - an aborted install could leave a lot 
orphans in site-packages, for example, or someone could have 
deleted/moved a critical file, or hacked the makefiles, or removed 
sip/PyQt and not PyKDE, or upgraded KDE and not PyKDE. 'import' also 
determines that libkdecore.so and similar are also there. It seems that 
what's important  isn't whether it's 'there', but whether it can be run 
- only way to know is to try it. It seems the easiest way is to put all 
of the PyKDE imports in a 'try' block and perhaps set a flag if they all 
succeed.


Jim



From skip at pobox.com  Fri Jun  7 15:52:26 2002
From: skip at pobox.com (Skip Montanaro)
Date: Fri, 7 Jun 2002 14:52:26 -0500
Subject: Where is Berkeley DB on your system? Can you live without DB 1.85?
Message-ID: <15617.3834.292130.446065@12-248-41-177.client.attbi.com>

I'm reworking the code in Python's setup.py and configure scripts that
locates the relevant Berkeley DB libraries and include files.  This is
important because on many systems you can install multiple incompatible
versions of this library.  Compiling the bsddb module with the include file
from one version and trying to link it with the libary from another version
is an obvious recipe for disaster.

I've figured out such information for a number of systems at my disposal but
I'd like to get as much input as possible about this for other systems so
the next Python release works out of the box for as many people as possible.
Here's what I want you to do.  If you have a system on which you use
configure to build Python *and* if your Berkeley DB installation information
doesn't match one of the following (for N in [1,2,3,4] below):

    include file location               library location
    ---------------------               ----------------
    /usr/include/dbN                    /usr/lib/lib/dbN
    /usr/local/include/dbN              /usr/local/lib/lib/dbN
    /sw/include/dbN                     /sw/lib/dbN
    /opt/sfw/include                    /opt/sfw/lib

please reply to this note off-list letting me know where it is on your
system.

Also, please let me know what version(s) of Berkeley DB you have installed
and use.  I would *really* like to stop supporting compilation and linkage
with DB 1.85 because of the long known bugs in its hash file implementation.
If you really and truly need a convenient way to enable building with
Berkeley DB 1.85, let me know and I'll figure something out.  Otherwise, I
will add the necessary lines to setup.py but comment them out.  You'll have
to uncomment them so the bsddb module will build with the now-ancient DB
1.85.

Finally, is compilation with an older version (say, version 2) important to
you in the presence of a newer installed version (version 3 or 4)?  Let me
know that as well.

Thx,

-- 
Skip Montanaro (skip at pobox.com - http://www.mojam.com/)
Boycott Netflix - they spam - http://www.musi-cal.com/~skip/netflix.html




From timr at probo.com  Wed Jun 12 01:35:33 2002
From: timr at probo.com (Tim Roberts)
Date: Tue, 11 Jun 2002 22:35:33 -0700
Subject: Removing ^M
References: <mailman.1023546912.19781.python-list@python.org> <Pine.GSO.4.30.0206111023300.6860-100000@eddie.cdf>
Message-ID: <acndgug1ls3ak9nr72o52pgbisb2t5pp5r@4ax.com>

so very tired <g0riaman at cdf.toronto.edu> wrote:

>Easiest way to do this is with sed.
>cat file.html | sed 's/^M//g' > newfile.html
>
>the trick is you have to enter ^M as the proper control character.
>You can do this at the command prompt by holding down CTRL+SHIFT and then
>hitting V and M in sequence.

Ctrl is always sufficient for this.  The Shift is merely decorative.
--
- Tim Roberts, timr at probo.com
  Providenza & Boekelheide, Inc.


From cmathur at cis.poly.edu  Thu Jun 20 14:08:42 2002
From: cmathur at cis.poly.edu (chandan mathur)
Date: Thu, 20 Jun 2002 14:08:42 -0400
Subject: HTTPSHandler ??? missing?
Message-ID: <3d121907$1@news.poly.edu>

I am trying to login to a ssl secure server using urllib2 and clientcookie
but can't find HTTPSHandler() in urllib2.py as mentioned in the manual!

Chandan Mathur
p.s = newbie




From peter at engcorp.com  Mon Jun 17 22:16:00 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 22:16:00 -0400
Subject: Problems with importing; must do it twice?
References: <fgphgus0pv2c7hp8l27uaenjlerr97tdta@4ax.com> <3D0E809F.AD9746A2@engcorp.com>
Message-ID: <3D0E97E0.E176EB19@engcorp.com>

Peter Hansen wrote:
> 
> Stew LG wrote:
> >
> >
> > My web host just switched hardware, and in
> > the process I've been moved to Solaris and
> > Python 2.1.1.
> >
> > My old scripts are breaking now, at the
> > very first line:
> >
> > import cgi;
> >
> > I think their Python install is broken,
> > but I wanted to be very sure before
> > I complained about it, since I've been
> > acting somewhat hysterically towards the
> > company's support lately - the move broke
> > a few things for me.
> 
> Presumably the hosting company is escape.com?  I just

Correction, "Cedant".  (Not sure if that's the same thing,
actually, but my friend said it was Cedant.)

-Peter


From skip at pobox.com  Mon Jun 24 18:36:16 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 24 Jun 2002 17:36:16 -0500
Subject: How to stop a SocketServer?
In-Reply-To: <3d1744d4$1_1@hpb10302.boi.hp.com>
References: <7xhejv9af2.fsf_-_@ruckus.brouhaha.com>
        <3d1744d4$1_1@hpb10302.boi.hp.com>
Message-ID: <15639.40672.304806.84777@12-248-11-90.client.attbi.com>

    Daniel> I faced this problem with a simple chat server I wrote.  I
    Daniel> expected that termination of the server with <ctrl-C> should
    Daniel> have stopped it immediately, but it instead waits until some
    Daniel> communication is received from each thread that was stopped at a
    Daniel> read().

Just make all the non-main threads daemon threads.  I believe the main
thread will be the one to catch the Ctrl-C.  It can then cleanup and exit.
Assuming it's the only non-daemon thread, the application will also exit.

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html




From K.Rdt at TU-Berlin.DE  Wed Jun 26 01:19:32 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Wed, 26 Jun 2002 07:19:32 +0200
Subject: system-output to wxpython-window
Message-ID: <OJ54SMC97OK083ZYT54JFDC41GAYXPJ.3d194ee4@FRITZweb>

---------------------------------------------------------------------
Hi

I have some batch-files, which I want to
invoke from a nice wxpython-frame. So
I did   system("0m.bat");   but the output
is coming to the DOS-shell-window. Is
there a simple way to do the job. Best case
would be one menue-window and another
for output.

		K at Rdt
---------------------------------------------------------------------







From carel.fellinger at iae.nl  Wed Jun 12 07:26:12 2002
From: carel.fellinger at iae.nl (Carel Fellinger)
Date: Wed, 12 Jun 2002 13:26:12 +0200
Subject: parse tar file with python
In-Reply-To: <ae7498$4jven$1@ID-146704.news.dfncis.de>
References: <ae7498$4jven$1@ID-146704.news.dfncis.de>
Message-ID: <20020612112612.GA13369@schaduw.felnet>

On Wed, Jun 12, 2002 at 11:29:30AM +0200, Shagshag13 wrote:
> hello,
> 
> i 'm not sure this is the right place to ask, but maybe somebody had already do some kind of python code on this :
> 
> i had hundred thousand of small files, my program is very slow. I'm thinking that i loose time waiting for opening/closing, and i

Some filesystems get very slow with lots of files in a single directory.
Maybe the simplest solution is to organize those thousands of files in
a few (hundreth) subdirectories.  Might give you all the speed-up you need.


> But in this case i don't know how to check for the end of a file and how to get its name (both in a tar archive) ...
> (i don't want to tar/untar with python, just to be able to do a sequential reading of a tar file to get its content)

Parsing the tar file is aka untarring it in memory, so what not use
the tar module?

-- 
groetjes, carel





From juanm.casillas at eresmas.com  Tue Jun 25 04:13:39 2002
From: juanm.casillas at eresmas.com (Juan M. Casillas)
Date: Tue, 25 Jun 2002 10:13:39 +0200
Subject: Creating an execution environment
Message-ID: <15640.9779.65207.940991@eresmas.com>

Hello to all !

      First of all, sorry by this long, boring mail :)

      I'm trying to do a simple execution environment with python
      embedded into a C program. I want to have some "environments"
      each one with its own data, functions, classes, and so on. But
      I don't know how to create separate environments with the same
      parser. I try to archieve that with multiple modules, but this
      is not the right focus for this problem, because the modules
      can access the code from another modules.

      It's posible to have multiple environments with a single python
      interpreter ? its posible to have multiple interpreters in a 
      C program ? Also, I want to load chunks of code for multiple
      files, and then add the symbols to the right environment...

      e.g.

      I want to create two separate execution environments, env1 and
      env2. Also, the code from preload.py should be included in the
      two environments but not as module, just as if the code was
      defined in the same file.

      -->file preload.py<--
      class PG:
	    name = ""
	    def __init__(self,name):
		self.name = name
	    def getName(self):
		return(self.name)

      -->file env1.py<--

      def env1_func(arg1):
	  return(arg1*3)

      pg = PG('hello')
		

      -->file env2.py<--	     
      
      def env2_func():
	  return('hello from env2')

      thing = PG('thing')
            
	

      So I want to call environment's functions or object methods
      from my C code, having the two environments isolated. Can be
      this done with python ? How can I merge the code from the
      files in the same execution environment ?

      I do some tests, but the unique way to have some similar is
      creating one module for each environment, and importing them
      from the C code. Any ideas ?


Thanks in advance,

--
Juan M. Casillas
assman at eresmas.com	











From fperez528 at yahoo.com  Wed Jun 19 17:00:51 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 19 Jun 2002 15:00:51 -0600
Subject: Newbie Q: File accessing
References: <aeqb18$k10$1@phxtst22.phx1.aro.allied.com>
Message-ID: <aeqre4$cu1$1@peabody.colorado.edu>

Todd Harris wrote:

> Hi, I apologize if this question has been beaten to death or if there's just
> something that I haven't been seeing, but here's my problem (with a little
> bit of background):
> 
> I have a file out on the network somewhere that stores info as text. I'm
> writing a Python script that updates that file. However, it's a multi-user
> application, and so my problem is that one user's script may be in the
> middle of writing to that file while another user's script is attempting to
> read that file or <even worse> writing to that file as well.
> 
> For those of you who use Microsoft <gulp...flame shield on> Word, when a
> document is opened by one user, and another user tries to access it, they
> can open it for read only. Obviously, Word has made some file attribute
> where it is locked to that first user, and only that first user can write to
> it while they have it open. Is there a way to do this in Python? Or what
> other possible options do I have? I've looked been using the open(
> filePath ) function as well as the os.open( ) function and os.access( ) to
> check my progress, but can't seem to figure out how to get around this.
> Thanks,
> 
> Todd

Here's a basic discussion to get you started:

http://trixie.triqs.com/pipermail/metakit/2001-December/000439.html

In summary, getting this to work portably cross-platform is highly 
non-trivial. If you need only an in-house  specific solution under 
well-specified constraints, you may be able to come up with a homebrewed 
option which works well for you. Simple file.lock locks with some process 
info in them may be enough in your environment.

But this recipe may be enough:

http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/65203


For more, just google(python file locking).

Good luck,

f.


From sholden at holdenweb.com  Wed Jun 19 16:13:47 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 19 Jun 2002 16:13:47 -0400
Subject: Python to MS SQL Server
References: <mailman.1024428323.19642.python-list@python.org>
Message-ID: <nQjQ8.213357$Fp1.114650@atlpnn01.usenetserver.com>

"Bjorn Pettersen" <BPettersen at NAREX.com> wrote ...
> > From: Mark McEahern [mailto:marklists at mceahern.com]
> >
> > > I am interested in using Python in a MS SQL Server
> > > environment, but do
> > > not see any database modules out there, especially for
> > > native versus
> > > using ODBC. Is there such a module?
> >
> > This is a total guess, because I haven't done this myself:
> >
> > Have you tried using win32com to use ADO?

> AFAIK, there is no Python database module that goes directly to MS SQL
> Server db lib. ADO does work however, so that's certainly an option.
> Personally, I would probably investigate e.g. mxODBC first (the
> interface is much easier to use from Python).


I agree that ADO is a total pain (although hesitantly, since Alex Martelli
is known to favour it) in Python. I just never got used to it in VBScript, I
guess.

Don't I remember that Dave Col's Sybase module is supposed to be usable with
SQL Server too?

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From spam at eatshit.net  Sun Jun  2 12:38:09 2002
From: spam at eatshit.net (paul colby)
Date: Sun, 02 Jun 2002 16:38:09 GMT
Subject: Help for Tkinter and python's programming
References: <mailman.1022985123.22656.python-list@python.org> <adc8kj$1ovk$1@nntp6.u.washington.edu>
Message-ID: <RRrK8.905$Ct6.719@newsread1.prod.itd.earthlink.net>

Russell E. Owen wrote:

<snip great response>

> The thing I usually do in this situation is to create a callable object.
> This is handy because an object can hold state, i.e. the "stuff I know
> now", yet can also have a __call__ method which allows you to call the
> object just like a function.
> 

I think your suggestion is great and has many uses. One that I find pleasing
is passing a member function to handle events. When a member function is
passed the "self" argument tags along for the ride thus giving your callback
access to all the members of the class as well as all the member data of the
instance. For example consider a box drawn on a canvas. Lets say we want to 
make the box red when clicked and blue when clicked again,

from Tkinter import *

canvas = Canvas(Tk(),bg='white')
canvas.pack(fill=BOTH,expand=1)

class Box:

  def __init__(self, canvas, x, y):
        self.canvas = canvas
        self.tag = canvas.create_rectangle(x-10,y-10,x+10,y+10,fill='blue')
        # here we set the callback
        canvas.tag_bind(self.tag,'<Button-1>',self.mouseClick)

  # here we define the callback. Note the "self" argument gets
  # handled by python's class mechanisms
  def mouseClick(self,event):
        color = self.canvas.itemcget(self.tag,'fill')
        if color == 'blue':
          self.canvas.itemconfigure(self.tag,fill='red')
        else:
          self.canvas.itemconfigure(self.tag,fill='blue')

b1 = Box(canvas,30,30)
b2 = Box(canvas,50,100)


Regards
Paul Colby


From oren-py-l at hishome.net  Fri Jun 28 05:42:29 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Fri, 28 Jun 2002 05:42:29 -0400
Subject: Pythoniac: Thoughts on a hardware Python processor
In-Reply-To: <affhcg$1im$1@sirius.dur.ac.uk>
References: <afc52t$2pj$1@sirius.dur.ac.uk> <mailman.1025163215.23885.python-list@python.org> <affhcg$1im$1@sirius.dur.ac.uk>
Message-ID: <20020628094229.GA96635@hishome.net>

On Thu, Jun 27, 2002 at 05:18:10PM +0000, Christopher Saunter wrote:
> Compact Objects (Oren Tirosh)
> -------------------------------
> I like this suggestion, it is something I had been thinking about in a
> slightly different context, but not thought through.  

Actually, I got the idea from an old Byte article about an interesting
processor called Rekursiv designed by the Glasgow-based audio company Linn.
This processor had hardware support for persistent objects and garbage
collection.

http://www.brouhaha.com/~eric/retrocomputing/rekursiv/rekursiv.txt
http://www.set-i.com/bruno/computing/sigarch.html

	Oren




From whisper at oz.net  Sat Jun 29 15:16:22 2002
From: whisper at oz.net (David LeBlanc)
Date: Sat, 29 Jun 2002 12:16:22 -0700
Subject: Python needs better error reporting
In-Reply-To: <afktd0$5ng$0@216.39.172.122>
Message-ID: <GCEDKONBLEFPPADDJCOEAEABDNAA.whisper@oz.net>

<snip>
> And then maybe add some AI hype and an animated GUI paper clip in 
> the shape
> of a dancing python with a policeman's cap ;-)
> 
> Regards,
> Bengt Richter

Oh yeah, and then call it "Bob" and ship it with every copy of Windows :P

- dhl




From tim.one at comcast.net  Sun Jun  2 18:00:04 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 02 Jun 2002 18:00:04 -0400
Subject: windows install put short path names to win register
In-Reply-To: <61f6bfa8.0206021304.2abedc3c@posting.google.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEONPJAA.tim.one@comcast.net>

[Toni Gaya]
> Windows python 2.2.1 installation program puts short file names to
> some path values in windows register.
>
> This makes that programs that read python path or some other
> information as python path from windows register crash
> -i'm using NTFS (maybe with FAT32, due that FAT system works with
> short file names and long file names, works, but as NTFS only works
> with long file names, some other installation programs just crash)-.
> Examples: try to install mysqldb (mysql connection) or py2exe on
> windows with a NTFS disk, and these programs crash.
>
> I'm wrong?

Beats me.  If we put long names in the registry instead, then unbounded
numbers of newbies get even more confused than Microsoft does <0.3 wink>
about how to use those names successfully.

The Python installer deliberately defaults to installing Python under a path
with a pure DOS 8.3 name, and without embedded spaces.  The "long name" form
of such a path is identical to the "short name" form, so doesn't make
trouble for anything.  If you're seeing a difference, you must have
installed into a non-default location.  If you don't like the consequences,
don't do that.

most-problems-are-self-inflicted-ly y'rs  - tim





From jarvin24 at tutor.cc.tut.fi  Thu Jun 13 06:16:37 2002
From: jarvin24 at tutor.cc.tut.fi (=?iso-8859-1?Q?J=E4rvinen?= Petri)
Date: Thu, 13 Jun 2002 10:16:37 +0000 (UTC)
Subject: Something faster then sgmllib for sucking out URLs
References: <mailman.1023913936.26808.python-list@python.org>
Message-ID: <slrnaggs86.ups.jarvin24@tutor.cc.tut.fi>

In article <mailman.1023913936.26808.python-list at python.org>, Alex Polite wrote:
> I'm working on a webspider to fit my sick needs. The profiler
> tells me that about 95% of the time is spent in sgmllib. I use sgmllib
> solely for extracting URLs. I'm looking for a faster way of doing
> this. Regular expressions, string searches? What's the way to go? I'm
> not a python purist. Calling some fast C program with the html as
> argument and getting back a list of URLs would be fine by me.
> 

*****

Thanks guys from your help! My list is now somewhat sorted :)

-Petri


From marklists at mceahern.com  Sun Jun 30 13:00:32 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sun, 30 Jun 2002 12:00:32 -0500
Subject: aspect-oriented demo using metaclasses
In-Reply-To: <pan.2002.06.29.19.23.48.841533.2265@club-internet.fr>
Message-ID: <NCBBLFCOHHDIKCAFGCFBKEEFLHAA.marklists@mceahern.com>

Here's the latest version of my attempt to implement AOP in Python with
metaclasses:

This version fixes the problem where the wrapped method's self variable was
pointing to the field descriptor instance rather than the class instance.
I've also separated the functionality into three files:

  aspect.py - the framework and the Observer interface definition
  trace.py - the simple trace Observer
  test.py - not yet a PyUnit test, just a demo for now

***
# aspect.py
#! /usr/bin/env python

__doc__ = \
"""
TODO:

  [x] Separate into appropriate modules.
  [ ] Should the framework raise errors?
  [ ] Add unit tests.
  [ ] Define scenarios for join points, point cutting, etc.:
        Register for all events, all classes, all methods.
        Register for some events, some classes, some methods.  Etc.
        What's most common?
  [ ] Explicitly specify default metaclass for framework classes.  Fixed by
      separating aspect into separate module?
  [ ] The data in the trace file is not valid xml because there is no
top-level
      document.  That may be ok, but document it.

"""

class VirtualClassError(Exception):pass

class Observer:
    """Interface for observers of Aspected classes."""

    def __init__(self, *args, **kwargs):
        if self.__class__ is Observer:
            raise VirtualClassError(self.__class__.__name__)

    def on_before(self, class_name, method_name, *args, **kwargs):
        pass

    def on_after(self, class_name, method_name, *args, **kwargs):
        pass

    def on_error(self, class_name, method_name, *args, **kwargs):
        pass

def make_event(event_name, method_name):
    def event(self, *args, **kwargs):
        class_name = self.__class__.__name__
        for o in self.observers[event_name]:
            e = getattr(o, event_name)
            e(class_name, method_name, *args, **kwargs)
    return event

def make_wrapped_method(cls, original_name, method):
    # TODO: Refactor.
    on_before_event = make_event("on_before", original_name)
    on_after_event = make_event("on_after", original_name)
    on_error_event = make_event("on_error", original_name)
    on_before_event_name = "on_before_%s" % original_name
    on_error_event_name = "on_error_%s" % original_name
    on_after_event_name = "on_after_%s" % original_name

    setattr(cls, on_before_event_name, on_before_event)
    setattr(cls, on_error_event_name, on_error_event)
    setattr(cls, on_after_event_name, on_after_event)

    on_before_event = getattr(cls, on_before_event_name)
    on_error_event = getattr(cls, on_error_event_name)
    on_after_event = getattr(cls, on_after_event_name)

    def wrapped(self, *args, **kwargs):
        on_before_event(self, *args, **kwargs)
        try:
            ret = method(self, *args, **kwargs)
        except:
            on_error_event(self, *args, **kwargs)
            raise
        on_after_event(self, *args, **kwargs)
        return ret
    return wrapped

class Aspected(type):

    def __init__(cls, name, bases, dict):
        super(Aspected, cls).__init__(name, bases, dict)
        # Skip "private" methods for now.
        skip_prefix = "__"
        observers_dict = {'on_before': [],
                          'on_error': [],
                          'on_after': []}
        setattr(cls, "observers", observers_dict)
        for k, v in dict.items():
            print type(v)
            if not k.startswith(skip_prefix):
                # TODO: This test will skip staticmethods and classmethods.
                if type(v) == type(lambda x: x):
                    new_name = "__%s" % k
                    setattr(cls, new_name, v)
                    new_method = getattr(cls, new_name)
                    setattr(cls, k, make_wrapped_method(cls, k, new_method))

***

***
# trace.py
# --------------------------------------------------------------------------
---
# Standard library imports
# --------------------------------------------------------------------------
---
import sys

# --------------------------------------------------------------------------
---
# Third party imports
# --------------------------------------------------------------------------
---
import mx.DateTime

# --------------------------------------------------------------------------
---
# Sibling imports
# --------------------------------------------------------------------------
---
import aspect

# --------------------------------------------------------------------------
---
# Classes
# --------------------------------------------------------------------------
---
class Trace(aspect.Observer):

    def __init__(self, filename=None):
        self.filename = filename

    def write(self, prefix, data, class_name, method_name, *args, **kwargs):
        # TODO: Wrap import mx in try:except ImportError and use time?
        now = mx.DateTime.now()
        if not data:
            data = ""
        s = """<event name='%(prefix)s' datetime='%(now)s'>
<call>%(class_name)s.%(method_name)s(%(args)s, %(kwargs)s)</call>
<data>%(data)s</data>
</event>
""" % locals()
        if not self.filename:
            f = sys.stdout
        else:
            f = file(self.filename, "a")
        f.write(s)
        if self.filename:
            f.close()

    def on_before(self, class_name, method_name, *args, **kwargs):
        prefix = "on_before"
        data = None
        self.write(prefix, data, class_name, method_name, *args, **kwargs)

    def on_after(self, class_name, method_name, *args, **kwargs):
        prefix = "on_after"
        data = None
        self.write(prefix, data, class_name, method_name, *args, **kwargs)

    def on_error(self, class_name, method_name, *args, **kwargs):
        prefix = "on_error"
        data = get_traceback()
        self.write(prefix, data, class_name, method_name, *args, **kwargs)

# --------------------------------------------------------------------------
---
# Helper methods
# --------------------------------------------------------------------------
---
def get_traceback():
    """get_traceback() --> string

    Return the traceback exception information as a string.
    """
    import traceback
    from StringIO import StringIO

    f = StringIO()
    traceback.print_exc(limit=None, file=f)
    s = f.getvalue()
    return s
***

***
# test.py
#! /usr/bin/env python

# --------------------------------------------------------------------------
---
# Sibling imports
# --------------------------------------------------------------------------
---
import aspect
import trace

__metaclass__ = aspect.Aspected

# --------------------------------------------------------------------------
---
# Classes
# --------------------------------------------------------------------------
---
class MyClass:

    def bar(self, *args, **kwargs):
        print "real bar(%s, %s)" % (args, kwargs)

    def raise_error(self, *args, **kwargs):
        print "real raise_error(%s, %s)" % (args, kwargs)
        1/0

    def return_something(self, *args, **kwargs):
        print "real return_something(%s, %s)" % (args, kwargs)
        return self.something

    def static_method(*args, **kwargs):
        print "this is a staticmethod"

    def class_method(cls, *args, **kwargs):
        print "this is a classmethod"
        print cls.__name__

    static_method = staticmethod(static_method)
    class_method = classmethod(class_method)

# --------------------------------------------------------------------------
---
# Demo
# --------------------------------------------------------------------------
---
def main():
    filename = "junk.txt"
    file_trace = trace.Trace(filename)
    stdout_trace = trace.Trace()
    MyClass.observers["on_before"] += [file_trace, stdout_trace]
    MyClass.observers["on_after"] += [file_trace, stdout_trace]
    MyClass.observers["on_error"] += [file_trace, stdout_trace]

    # Exercise the class to test the Aspected framework.
    f = MyClass()
    f.bar("a", "b", foo="c")
    try:
        f.raise_error("a", "b", foo="c")
    except Exception, e:
        print "An error occurred in main(): %s" % e
    f.something = "foo"
    x = f.return_something(1, 2, 3)
    print x
    MyClass.static_method()
    MyClass.class_method()

if __name__ == "__main__":
    main()
***


-





From richardt at edshk.demon.co.uk  Wed Jun 19 03:43:32 2002
From: richardt at edshk.demon.co.uk (Richard Townsend)
Date: Wed, 19 Jun 2002 08:43:32 +0100
Subject: Embedding Python - time module problem
References: <cIgHcXA4SyD9swmz@edshk.demon.co.uk>
 <3D0F99EA.7060904@SPAMnwinternet.com>
Message-ID: <R5wUQLAkYDE9swMG@edshk.demon.co.uk>

In article <3D0F99EA.7060904 at SPAMnwinternet.com>, Jim <jbublitzNO at SPAMnw
internet.com> writes
>This sounds similar to the problem I had under Linux -
>sys is builtin, time and math aren't, and the interpreter
>symbols needed aren't available to the time and math
>libs.
>
>If you're using something like 'dlopen' to import the
>interpreter, it needs to happen with the RTLD_GLOBAL
>flag set. If you're statically linking, there's probably
>an equivalent linker switch. I was using KDE which has
>a method for globally loading libs. libtool looked
>messier, but fortunately I didn't need to get into it.
>
>man dlopen (at least for Linux).
>
>Jim
>

Thanks Jim, passing -Wl,-E to the linker does the trick on HP-UX. I can
import time and math modules OK now.
-- 
Richard Townsend


From markos at elite.org  Wed Jun 19 10:48:46 2002
From: markos at elite.org (Marko Samastur)
Date: Wed, 19 Jun 2002 16:48:46 +0200
Subject: i18n/l10n and python modules
Message-ID: <200206191648.46632.markos@elite.org>

Hi all,

Does anyone have an example of how localization of modules and application 
built on top of them was supposed to look like? 

I've read Python Library Reference, but I'm obviously too dumb to make it 
work. Whatever I do, when I run the application, I get either app strings 
translated or those in modules but never both.

Cheers,

	Marko




From kragen at pobox.com  Wed Jun 26 16:26:38 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 26 Jun 2002 16:26:38 -0400
Subject: PIL, Python and CGI
References: <3D1626FD.DA934B09@infineon.net>
Message-ID: <833cv9lrb5.fsf@panacea.canonical.org>

Sascha Ferley <Sascha.Ferley at infineon.net> writes:
> Basically if i run the python program in shell mode, it works fine. But
> when i run it via cgi mode it screws up horobly.
> Looking at the error-log i see just following:
> 
> Traceback (most recent call last):
> ...
>   File "/usr/local/lib/python2.2/site-packages/PIL/Image.py", line 39,
> in __geta
> ttr__
>     raise ImportError, "The _imaging C module is not installed"
> ImportError: The _imaging C module is not installed
> 
> Anyone have any ideas?

Perhaps your CGI Python is not your shell Python, and one of them has
PIL properly installed, while the other one doesn't.


From david.bear at asu.edu  Wed Jun  5 18:42:04 2002
From: david.bear at asu.edu (David Bear)
Date: Wed, 5 Jun 2002 15:42:04 -0700
Subject: writing a filter in python
Message-ID: <adm43t$p5j$1@news.asu.edu>

I'm writing a filter to be used by lprng.  Since there is no stdout I am 
wondering two things:

1) good strategies for showing error message (was thinking of syslog)

2) an easy way to catch all exceptions and quickly return back to the 
caller (lpd)  

I'm noticing that if my script doesnt run fast enough I may end up have lpd 
try multiple times to run it.  

Any advice?
-- 
David Bear
College of Public Programs/ASU
480-965-8257
...the way is like water, going where nobody wants it to go


From ianb at colorstudy.com  Sat Jun 22 21:14:09 2002
From: ianb at colorstudy.com (Ian Bicking)
Date: 22 Jun 2002 20:14:09 -0500
Subject: Web templating/db tool with best designer/coder separation?
In-Reply-To: <af2gh6$2c2$0@216.39.172.122>
References: <af2gh6$2c2$0@216.39.172.122>
Message-ID: <1024794849.594.19.camel@lothlorien.colorstudy.net>

On Sat, 2002-06-22 at 13:43, Bengt Richter wrote:
> E.g., if you were hoping for an improvement over current practice,
> which do you consider to be the tool(s) to beat, and what would
> the killer improvement(s) be? TIA.

I think Cheetah's methods (#def) are important to real separation of
design (like MVC) -- which isn't quite the same thing as separation of
skills (like designer/coder).  Your template is an object (where you
have control of inheritance), and you can define methods for that object
inside your template.

A template should be more expressive than a primitive programming
language -- it should be able to describe how you do something more than
just how to write a page.  For instance, how to write a header, or how
to write a list.  The whole inheritance thing is pretty nice, too --
inheritance maps really well to sets of templates.


Like other templating systems that compile to Python code, it would do
well to give better error messages.  The alternative of interpretation
gives good error messages, but doesn't give good performance.  


The NameMapper in Cheetah is also spiffy (in terms of simple syntax, and
easy Python integration), but it's picky and a bit fuzzy in terms of
semantics.  Something better defined would be nice.  

NameMapper is a way of changing an expression like $a.b.c.d.e into
a.b().c['d'].e (at runtime, of course, because it depends on the types
of the objects).  It's great for adding objects and dictionaries to the
template's namespace, without burdening the non-programmer with notions
of type.

  Ian






From lennie.jarratt at bigidea.com  Mon Jun 24 17:16:39 2002
From: lennie.jarratt at bigidea.com (Lennie Jarratt)
Date: 24 Jun 2002 14:16:39 -0700
Subject: informdb-1.3 install help
Message-ID: <141d29bb.0206241316.2ae024fa@posting.google.com>

I am trying to install informixdb-1.3 with Python 2.1.  When I run make
(using gmake) I get the following error:

make[1]: Entering directory `/disk2/zope/informixdb-1.3/ext'
make[1]: Circular _informixdb.c <- _informixdb.c dependency dropped.
gcc -shared  ./_informixdb.o ./dbi.o
-L/export/home/informix/informix/lib -L/ex
port/home/informix/informix/lib/esql -lifsql -lifasf -lifgen -lifos
-lifgls -lns
l -lsocket -laio -lm -ldl -lelf
/export/home/informix/informix/lib/esql/checkapi.o -lifglx  -o
./_informixdb.so
ld: fatal: file /export/home/informix/informix/lib/esql/libifsql.so:
wrong ELF class: ELFCLASS64
ld: fatal: file /export/home/informix/informix/lib/libifasf.so: wrong
ELF class: ELFCLASS64
ld: fatal: file /export/home/informix/informix/lib/esql/libifgen.so:
wrong ELF class: ELFCLASS64
ld: fatal: file /export/home/informix/informix/lib/esql/libifos.so:
wrong ELF class: ELFCLASS64
ld: fatal: file /export/home/informix/informix/lib/esql/libifgls.so:
wrong ELF class: ELFCLASS64
ld: fatal: file /export/home/informix/informix/lib/esql/checkapi.o:
wrong ELF class: ELFCLASS64
ld: fatal: File processing errors. No output written to ./_informixdb.so

collect2: ld returned 1 exit status
make[1]: *** [_informixdb.so] Error 1
make[1]: Leaving directory `/disk2/zope/informixdb-1.3/ext'
make: *** [sharedmods] Error
2

Is this because I am using the 64 bit Informix Client or is there
another reason?
If it is the 64 bit problem, is there a version of informixdb for 64
bit?  My dilema is that I want to use Zope and python for the rewrite of
my companies inhouse application.  We use Informix 9.2 on a Solaris 64
bit machine.  Without this support I cannot do this and will have to use
Perl or JSP.

Please Help.

Thanks.

Lennie


From jblazi at hotmail.com  Fri Jun 21 04:10:01 2002
From: jblazi at hotmail.com (JB)
Date: Fri, 21 Jun 2002 10:10:01 +0200
Subject: converting an array of chars to a string
Message-ID: <3d12dda2_6@news.newsgroups.com>

How can I convert an array of chars to string?

The problem is this: I should like to perform 8 bit 
calculations in a buffer. After this has been done, the 
buffer has to ba written to a binary file (and file.write 
takes strings instead of lists).

-- 
Janos Blazi


-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From hesterloli at hotmail.com  Wed Jun 19 10:50:21 2002
From: hesterloli at hotmail.com (George Hester)
Date: Wed, 19 Jun 2002 14:50:21 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <D7NP8.46301$GY.14487785@twister.nyroc.rr.com> <slrnagvdp4.1sg.gerhard@lilith.my-fqdn.de>
Message-ID: <NS0Q8.33157$uk2.13764426@twister.nyroc.rr.com>

Because when I reply to you the email gets returned undeliverable.  That's why.

--
George Hester
_________________________________
"Gerhard H?ring" <gerhard at bigfoot.de> wrote in message
news:slrnagvdp4.1sg.gerhard at lilith.my-fqdn.de...
> George Hester wrote in comp.lang.python:
> > I just cannot believe what I ran into in this newsgroup. [...]
>
>
>                                                 .:\:/:.
>              +-------------------+             .:\:\:/:/:.
>              |   PLEASE DO NOT   |            :.:\:\:/:/:.:
>              |  FEED THE TROLLS  |           :=.' -   - '.=:
>              |                   |           '=(\ 9   9 /)='
>              |   Thank you,      |              (  (_)  )
>              |       Management  |              /`-vvv-'\
>              +-------------------+             /         \
>                      |  |        @@@          / /|,,,,,|\ \
>                      |  |        @@@         /_//  /^\  \\_\
>        @x@@x@        |  |         |/         WW(  (   )  )WW
>        \||||/        |  |        \|           __\,,\ /,,/__
>         \||/         |  |         |      jgs (______Y______)
>     /\/\/\/\/\/\/\/\//\/\\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\
> --
> mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
> web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
> public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
> reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From sarmstrong13 at mac.com  Mon Jun 17 16:17:45 2002
From: sarmstrong13 at mac.com (SA)
Date: Mon, 17 Jun 2002 15:17:45 -0500
Subject: Newbie Help Please.(OO and Tkinter)
References: <B93209DC.7281%sarmstrong13@mac.com> <r3fP8.401$uH2.132@newsread1.prod.itd.earthlink.net>
Message-ID: <B933AE19.7379%sarmstrong13@mac.com>

On 6/17/02 1:10 AM, in article
r3fP8.401$uH2.132 at newsread1.prod.itd.earthlink.net, "paul colby"
<spam at eatshit.net> wrote:
> When you set command=Saving.savin you are passing a member
> function which has no define instance pointer "self". Thus
> the "self" thingy in the definition of savin isn't there to
> be used when the button is pressed. Note that your other
> buttons likely work since you use constructs like
> command=self.clearin. Now I haven't tried it but I
> would expect if you included something like
> 
>       self.SAVE = Saving(self)
> 
> in the __init__ function definition for PyShell then you
> could use command=self.SAVE.savin for the Save Input button.
> Now when the button is pressed self = self.SAVE so there
> is an instance.
> 
Well I gave this a shot and got the following error:

Traceback (most recent call last):
  File "PyShell.py", line 54, in ?
    app = PyShell(root)
  File "PyShell.py", line 18, in __init__
    self.SAVE = Saving()
TypeError: __init__() takes at least 2 arguments (1 given)

Here is the modified script:

from Tkinter import *
import os
import commands
import tkSimpleDialog

class PyShell:
    def clearin(self):
        self.t1.delete(0.0,END)
    def clearout(self):
        self.t2.delete(0.0,END)
    def expyth(self):
        self.t2.delete(0.0, END)
        self.output = commands.getoutput(self.t1.get(0.0, END))
        self.t2.insert(END, self.output)
    def saving_callback():
        savin()    
    def __init__(self, top):
        self.SAVE = Saving(self)
        self.t1 = Text(top, height="12", width="84", font="Courier 12")
        self.t1.pack(side=TOP, pady=2)
        f = Frame(top)
        f.pack()
        self.b1 = Button(f, text="Execute", command=self.expyth)
        self.b1.pack(side=LEFT)
        self.b2 = Button(f, text="Clear Input", command=self.clearin)
        self.b2.pack(side=LEFT)
        self.b3 = Button(f, text="Clear Output", command=self.clearout)
        self.b3.pack(side=LEFT)
        self.b4 = Button(f, text="Save Input", command=self.SAVE)
        self.b4.pack(side=LEFT)
        self.t2 = Text(top, height="12", width="84", bg="lightblue",
font="Courier 12")
        self.t2.pack(side=TOP, pady=2)
        
class Saving(tkSimpleDialog.Dialog):
    def savin(self, question):
        Label(question, text="Directory:").grid(row=0)
        Label(question, text="Filename:").grid(row=1)
        self.e1 = Entry(question)
        self.e2 = Entry(question)
        self.e1.grid(row=0, column=1)
        self.e2.grid(row=1, column=1)
    def apply(self):
        self.dir = self.e1.get()
        self.fn = self.e2.get()
        self.sav = self.dir + self.fn
        self.savfile = open(self.sav, "w")
        for line in self.t1.readlines():
            self.savefile.write(line)
        self.savefile.close()
        

            
root = Tk()
app = PyShell(root)
root.mainloop()

Any ideas?

Thanks.
SA



From logiplexsoftware at earthlink.net  Fri Jun 21 13:23:47 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 21 Jun 2002 10:23:47 -0700
Subject: Newbie-Question: Oberserving an directory
In-Reply-To: <3d1352a7@news.piro.net>
References: <3d1352a7@news.piro.net>
Message-ID: <20020621102347.284166ba.logiplexsoftware@earthlink.net>

On Fri, 21 Jun 2002 18:21:59 +0200
Marcus Klein wrote:

> Hi there,
> 
> I want to observe an directory to check if a new file has been put in, if 
> that happens I want to start an event.
> 
> Does anyone have a good idea how to start, I've tried it with signal, but 
> either I am too stupid or it is the complete wrong way :-/

AFAIK, you'll have to poll the directory contents.  This solution may be
overkill for your needs, but hey.

import os, time
import threading, Queue

class DirScanner:
    def __init__(self, directory, interval = 2):
        self.dir = directory
        self.queue = Queue.Queue()
        self.running = 1
        self.interval = interval
        self.thread = threading.Thread(target = self._scan)
        self.thread.start()
        
    def _scan(self):
        lastDir = os.listdir(self.dir)
        while self.running:
            thisDir = os.listdir(self.dir)
            for f in thisDir:
                try:
                    lastDir.index(f)
                except ValueError: # new entry
                    self.queue.put(f)
            lastDir = thisDir
            time.sleep(self.interval)

    def get(self, block = 1):
        try:
            return self.queue.get(block)
        except Queue.Empty:
            return None

    def kill(self):
        self.running = 0


if __name__ == '__main__':
    ds = DirScanner("/tmp")
    while 1:
        f = ds.get(0)
        if f is not None:
            print f
            ds.kill() # stop the thread
            break


This will allow the main thread of execution to continue and you can check the
queue for new files when needed.  Alternatively, you can call ds.get() without
the 0 argument and it will wait until a new filename is put on the Queue.

Good luck,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From fredrik at pythonware.com  Thu Jun 20 04:51:30 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 20 Jun 2002 08:51:30 GMT
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <20020619154654.C619@ActiveState.com> <mailman.1024527565.30927.python-list@python.org> <lkvg8ejqrd.fsf@pc150.maths.bris.ac.uk>
Message-ID: <mIgQ8.44254$n4.10309597@newsc.telia.net>

Michael Hudson wrote:

> > You are wrong. As long as you link to the python{major}{minor}.dll you
> > are binary compatible, no matter which compiler you use.
>
> Eh, that's a bit of a stretch isn't it?  If your compilers have
> different ideas about struct layout & so on it would certainly take
> more effort than just linking the same libraries.

last time I checked, MSVC wasn't even compatible with
itself, if you mix /MD/MT/ML (i.e. runtime libraries).

different FILE* structures is a classic, for example.

</F>




From hjwidmaier at web.de  Mon Jun 24 04:43:58 2002
From: hjwidmaier at web.de (Hans-Joachim Widmaier)
Date: 24 Jun 2002 01:43:58 -0700
Subject: Advice sought on ps module
References: <6e990e29.0206200325.355f0413@posting.google.com> <aesoqg$kc9$1@scotsman.ed.ac.uk> <6e990e29.0206202356.544eb696@posting.google.com> <aev2dk$pn2$1@scotsman.ed.ac.uk>
Message-ID: <6e990e29.0206240043.414daf79@posting.google.com>

eddie at holyrood.ed.ac.uk (Eddie Corns) wrote in message news:<aev2dk$pn2$1 at scotsman.ed.ac.uk>...

> OK, that looks like a fairly complex job.  The only other suggestion I can
> make is to wonder whether you're getting too hung up on precision.  Does the
> BBox really need to be pixel perfect?

No, it doesn't have to. I'm not hung up on precision (at least I
wouldn't admit it ;-)).

>  I know myself how easy it is, when you
> know a perfect solution _should_ be possible, to get fixated on it instead of
> just saying, well take the longest string and assume len * em width and see
> whether it looks OK (eg would another person notice?).

And where do I get em in the first place?

> OK, I'm sure you
> wouldn't have gone to the effort of messing about with font metrics for
> nothing but don't forget 'good enough' solutions are just that!

My current module, crude as it is, was "good enough" until now. Alas,
this is no longer so. As I said in my first posting, I'm not going to
do the be all, end all ps module. I'm seeking a more or less elegant
solution for my current problems. (I do not care about CID fonts and
halftones and what else arcane features there are). But I'd like to
start in the right direction, so that if I want to add a few features
next year, I won't have to do it again.
 
> Misery to me is NOT having a challenging programming project :)

Naah, misery is having no clue what to do. :)

> Sorry I can't be of any help after all but good luck anyway.

Thanks. I might need it.

Hans-Joachim


From cimarron+google at taylors.org  Fri Jun 21 13:44:50 2002
From: cimarron+google at taylors.org (Cimarron Taylor)
Date: 21 Jun 2002 10:44:50 -0700
Subject: RE strings (was: Variable Interpolation - status of PEP 215)
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <4PfQ8.44245$n4.10307683@newsc.telia.net> <KhiQ8.44269$n4.10312432@newsc.telia.net> <29e28c51.0206201156.7e61a7f5@posting.google.com> <3D127B89.EE1C17F4@engcorp.com>
Message-ID: <29e28c51.0206210944.400eaedc@posting.google.com>

I should have written that to me

   def parseline1(line):
      m = re'^EMP:([^,]*),([^,]*),([^,]*),([^,]*)$'.match(line)
      if m:
          ...

is better than

   import re
   emp_re = re.compile(r'^EMP:([^,]*),([^,]*),([^,]*),([^,]*)$')

   def parseline2(line):
      m = emp_re.match(line)
      if m:
          ...
or

   import re
   def parseline3(line):
      m = re.match(r'^EMP:([^,]*),([^,]*),([^,]*),([^,]*)$', line)
      if m:
          ...


Unless I'm mistaken about how the re module works, parseline3 will
recompile the regex each time it is called.  parseline2 avoids this
but introduces an extra object.  parseline1 should compile the regex
once as well and avoid exposing the regex object.  

I should also write that while I personally like the appearance of
parseline1, I think the most significant benefit of an RE string
comes from the fact that python's parser/compiler would detect errors
in the regex before parseline1 was ever executed.  I'm not sure 
RE strings are a good idea if the implementation could not do this. 

Cim


From kragen at pobox.com  Sun Jun  2 22:42:11 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 02 Jun 2002 22:42:11 -0400
Subject: How to install Python without root priviledges ?
References: <acvmin$svjfb$1@ID-146704.news.dfncis.de> <3cf866fa@news.poly.edu>
Message-ID: <83ptz9rsl8.fsf@panacea.canonical.org>

"Chandan Mathur" <chandan_mathur at hotmail.com> writes:
> "Shagshag13" <shagshag13 at yahoo.fr> wrote in message
> news:acvmin$svjfb$1 at ID-146704.news.dfncis.de...
> > ./configure --prefix=/home/shagshag/local
> 
> you have done the right thing.., only without SU priviliges, I think you can
> run python only on the directory Python-2.2.1 as follows,  ./python
> I have the same problem... and it really sucks :-(

Add /home/shagshag/local/bin to your PATH and you will be able to run
Python without being in that directory.



From candiazoo at attbi.com  Thu Jun 27 10:15:33 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Thu, 27 Jun 2002 14:15:33 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com> <afdrlv$r83$0@216.39.172.122>
Message-ID: <3d1b1c83.294342968@netnews.attbi.com>

Here is my scenario...

1. return 3 columns/per row from a mysql db
2. extract 7 other bits of information from an oracle db
3. stuff the bits into a class which formats the data
4. write the formatted data by returning a concatenated string
   from the class object

I was guessing it was the queries to the oracle db that was causing the slowup,
but my dba was offended and besides, I have several other extraction steps that
run about 5 or 6 times faster (that write to the mysql db from items extracted
from the oracle db).

So the next thought was the file output... could be my class too...  someone
suggested trying the profiler.  I have never used it but I'll give it a shot!

On the flip side... this is my first REAL python project and I am having a LOT
of fun.  I plan on writing an application for my wife and her artist friends
once this one is done and I am learning quite a lot - thank you all for your
patience with all of my questions (here and in the mySQL email list).  Ada used
to be my favorite language, now it is Python all the way (makes working with VB
and Java at work annoying).

Mike J.

On 27 Jun 2002 02:01:35 GMT, bokr at oz.net (Bengt Richter) wrote:

>On Thu, 27 Jun 2002 01:28:22 GMT, candiazoo at attbi.com wrote:
>
>>I assume some sort of block i/o or preallocating a large block of file space
>>would be the best way to do it?  I am outputting about 700,000 records to a text
>>file and my program only spits out about 20 records per second.  This is using
>>standard file open and write...
>>
>>This is on a W2K machine... perhaps I should use the win32file interface?  Any
>>thoughts/opinions?
>>
>I doubt that that is the solution. Where are your "records" coming from and/or
>how are they being generated/mdified, and how big are they?
>
>If you are building string records by s='' followed by s+='chunk' or the like many times,
>this is a typical newbie performance killer. Try accumulating your chunks in a list,
>like sl=[] followed by sl.append('chunk') instead, and then do f.write(''.join(sl))
>instead of the f.write(s) you would have done.
>
>Just guessing...
>
>Try timing generation of output data vs writing it, and see what's happening.
>import time and use clock() to get an accurate floating point seconds time reading. E.g.,
>
>from time import clock
>tstart = clock()
># ... generate an output record here
>tendgen=clock()
># ... write the output  something like f.write(record) here
>tendwrite=clock()
>print """\
>generating took %f sec
>outputting took %f sec""" % (tendgen-tstart, tendwrite-tendgen)
>
>Regards,
>Bengt Richter



From tim.one at comcast.net  Thu Jun 13 10:09:59 2002
From: tim.one at comcast.net (Tim Peters)
Date: Thu, 13 Jun 2002 10:09:59 -0400
Subject: why not "'in' in 'in'"?
In-Reply-To: <3D089DBF.349B7A1@seebelow.org>
Message-ID: <BIEJKCLHCIOIHAGOKOLHCEGCDEAA.tim.one@comcast.net>

[Grant Griffin]
> See, that's where we bump into "practicality beats purity".
>
> note-that-every-one-of-tim's-principles-is-subjective-ly y'rs,

They only appear that way until one has developed objective good taste.

these-kids-today!-ly y'rs  - tim




From kzaragoza at attbi.com  Thu Jun 27 15:09:22 2002
From: kzaragoza at attbi.com (Kris J. Zaragoza)
Date: 27 Jun 2002 14:09:22 -0500
Subject: COM server threads/unloading
References: <10d944cb.0206261529.741da082@posting.google.com>
Message-ID: <slrnahmos0.eh5.kzaragoza@localhost.localdomain>

In article <10d944cb.0206261529.741da082 at posting.google.com>, Uncle Enzo wrote:
> Then I added the googlelogger import statement and the two lines
> initializing and calling the log method. The logger that has two
> threads. Using the same client code results in the testserver.exe
> loading after the dispatch call, but it does not unload after the
> server reference is set to None, nor does it unload after the client
> terminates.
>
> What is the proper way to get a multithreaded python COM server to
> unload when there are no external references to it?

Mark makes some good suggestions in his reply to you.  However, your
problem appears to have nothing to do with COM and everything to do
with threading.  In your googlelogger module, you create a thread that
runs in an infinite loop.  You built a mechanism to shut down the
thread by sending it a magic QUIT command.  However, your COM
component never sends it that command.  Your server doesn't shut down
because you have threads still running waiting on a queue that will
never contain more messages.

As I see it, you have two options.  First, you can build a mechanism
to explicitly send the logger the message to shut down.  Second, you
can set the logging threads to be daemon threads with setDaemon().
This will flag the thread such that it will not keep the process from
exiting when all other threads have terminated.  It's a quick and easy
way to terminate the thread without having to explicitly tell it to
shut down.

Then again, as Tim Peters writes, "Explicit is better than implicit."

-Kris

-- 
Kris J. Zaragoza       | On the face of it, Microsoft complaining about
kzaragoza at attbi.com    | the source license used by Linux is like the
                       | event horizon calling the kettle black.
                       | -- Adam Barr, article on kuro5hin.org


From Christopher.Saunter at durham.ac.uk  Thu Jun 27 13:18:10 2002
From: Christopher.Saunter at durham.ac.uk (Christopher Saunter)
Date: Thu, 27 Jun 2002 17:18:10 +0000 (UTC)
Subject: Pythoniac: Thoughts on a hardware Python processor
References: <afc52t$2pj$1@sirius.dur.ac.uk> <mailman.1025163215.23885.python-list@python.org>
Message-ID: <affhcg$1im$1@sirius.dur.ac.uk>

Dear All,
	Thanks for the replies, both email and usenet!  Well, my plans
have changed a bit, and rather than replying to each post individually, I
have grouped the whole lot together below (hope that?s okay...)

Addressing the various points / ideas:

Python->VHDL (Thomas Heller)
----------------------------
This is an interesting idea, although rather at tangents to what I would
like to do.  JHDL (www.jhdl.org) does something similar, allowing FPGA
etc. designs to be described in Java.  I haven?t played with it yet mind.
I suspect implementing something like this in Python would be a minefield,
as this is a case where strong, static typing is really quite desirable!
Of course one could always use Jython...  Guggy.....

Bytecode compatibility (Paul Rubin)
-----------------------------------
Initially I was thinking of using the same assembly as python (different
bytecode enumeration to make instruction decode more pleasant) but after
more looking into Pythons guts and various comments, I see that this is a
daft approach, as PBC is just to high level.

So Pythoniac would be some sort of processor with an inbuilt hardware
understanding of objects (stacks, registers etc. contain pointers to
objects, dedicated hardware would handle the lookup of the address of the
relevant function for the current opcode etc.), with a lower level
bytecode (P-) than Python, but one that Python compiles to in a ?natural?
way. 

?Compact Objects? (Oren Tirosh)
-------------------------------
I like this suggestion, it is something I had been thinking about in a
slightly different context, but not thought through.  

Extend the idea to registers and stack objects, so that although
register/stack/etc. would normally store pointers to objects, the core
data types would be stored directly, and flagged as
int32/float32/aaa/bbb/object pointer.  This seems a neat way of
integrating the necessary basic datatypes into an OO framework.  I think
it would also make parallel execution of bytecode less of a headache.

Performance Issues
------------------
Pipelining and Instruction/Data caches

I have given these some though, and initially I would just like to see a
working processor, not a fast one.  Pipelining sure will be interesting
with such a processor.  I suspect once some general diagrams are created
for the proc, it will be easier to get a grip on how to do this.

Perhaps the processor would benefit from a separate cache for function
lookups for objects  e.g. if ?matrix? objects are in regular use, the
function pointers for their ?add?, ?subtract? etc. operations would be
cached.  However, these are all considerations for later on... - Oren?s
?accelerated associative lookup??
 Non Pythonic processor bits: (David LeBlanc)
----------------------------
[Bus Control, Interrupts, Memory Management]

Bus Control
 - - - - - 
This should be simple.  Well, maybe not simple, but it?ll be an integral
part of the processor design.

David - thanks for the FPGA links.  Some interesting links.  I already own
a development board with a Xilinx SpartanIIE, so this is what I shall aim
to implement Pythoniac in, using the free Xilinx software.  If this is
doable, anyone should be able to run / edit the processor for free... 

Memory Management & Interrupts
 - - - - - - - - - - - - - - 
Initially I am aiming for a low speed (1MHz) processor, to allow real time
links / immersion with support software on a PC, so there will be no
caches.  Also, initially there would be no multithreading, and as far as I
can tell, no need for interrupts.  That?s not saying they wouldn?t be
added in the distant future...

The lack of threading removes some of the need for an MMU (memory
management unit)

If there was one memory space this would need to be split for different
purposes - stack, program, data etc.  However, initially I plan to use the
multiple embedded memories in Xilinx devices to give separate memories for
each purpose (this limits it to small program / stack sizes, but I?ll be
happy when I have a processor that can max them out!) 
  
On the other hand, hardware assisted reference counting / garbage
collection etc. might prove interesting. 


Lisp Processors (Paul Rubin)
----------------------------
Thanks.  Lots of useful looking Google hits to follow up there.

Bytecode Tinkering / C to P+ compiler (Steve Holden)
----------------------------------------------------
It looks like using python bytecode directly would be rather daft, so a
lower level bytecode would be used.  Hopefully it would be relatively
trivial to convert different versions of Python code to the underlying
bytecode...  As for a C,C++ to P+ compiler (I wonder if it should actually
be called P-) I?ll leave that as an exercise to someone else ;-)


Python compiler with retargetable back end (David LeBlanc)
------------------------------------------
I suspect that this is the way to go for faster Python, something I would
like to be able to use.  I stress that ?Pythoniac? is not about speed, so
much as a ?neat? factor - I have been using programmable logic for a
couple of years now, and would like to attempt a processor core.  I
thought one a bit different would be interesting...  Having said that,
Pythoniac will need a compiler...  Psyco is showing the way here, and it
looks good to me!

The current plan:
------------------

A RISC style processor that executes a more minimal bytecode than that of
cPython, but to which cPython bytecode translates in a ?natural? way.  The
processor would have hardware understanding of objects, so in general it
would store pointers to objects (compact objects not withstanding) and
would have hardware lookup to retrieve the function pointers for standard
operations on objects (+, -, compare etc etc.) based on the current
objects and opcode.  

Things to look at:
o	Caching function tables
o	Hardware assisted reference counting / garbage collection
	(nice / maybe not so nice)
o	Efficiently implemented compact objects
o	The bytecode (p-)
o	A Python -> p- compiler

My current plan of action is for a very simple, RISC style CPU that just
understands integers, and works with 16 a bit address space and 16 bit
data.  Implement a Python->p- compiler that translates a single python
function that uses just integer objects into p- and produce a processor
capable of executing the p- code.  The goals would be:

a) Something that works
b) Something that works in a way that reflects the nature of the PVM, and
that has a simple Python->p- translation stage.

I think this could be quite a large undertaking, but interesting...


Regards, and thanks again to all who replied!

Chris Saunter



From Dick.Zantow at lexisnexis.com  Tue Jun 11 10:34:36 2002
From: Dick.Zantow at lexisnexis.com (rzed)
Date: Tue, 11 Jun 2002 10:34:36 -0400
Subject: Hardware acceleration?
Message-ID: <ae51s6$1vs$1@mailgate2.lexis-nexis.com>

I read an interesting article in Javapro about Jazzelle, which appears
to be a hardware add-on to an ARM processor that allows customization
(or a sort) of the chip so that it appears to be a hardware Java
bytecode processor.

(See
<http://www.fawcette.com/javapro/2002_06/magazine/columns/javatogo/ >)

It makes me wonder whether any such thing has been tried for the
Python VM. Have there been any hardware projects taking a look at
revving up Python?

--
rzed
USAGE ends with AGE




From mhammond at skippinet.com.au  Wed Jun  5 08:47:27 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Wed, 05 Jun 2002 12:47:27 GMT
Subject: Printing under Windows
References: <mailman.1023269468.29885.python-list@python.org>
Message-ID: <3CFE08D7.1030206@skippinet.com.au>

David LeBlanc wrote:
> I don't recall - does PythonWin do printing?

Yeah, but at the MFC level, so it's not exactly obvious :)

Mark.



From marklists at mceahern.com  Fri Jun 28 15:12:17 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 28 Jun 2002 14:12:17 -0500
Subject: testing type of an object
In-Reply-To: <pan.2002.06.28.14.49.33.858327.1787@presidency.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBOEAHLHAA.marklists@mceahern.com>

> Hi,
>   I'm writing some code that depends on the type of an argument
> passed to it.

In general, this is a sign that you are using Python to program in the
paradigms of some other language.  There are exceptions, of course.  And I'm
sadly not astute enough to be able to summarize them for you.  ;-)

Anyway...

> Is there a standard trick to do this? My versions don't seem to work:
>
> if type(s) == 'str':
>   do something

Suppose s is:

  s = "foo"

You can test whether it's a string like this:

  if type(s) == type("any known string")

substituting your favorite curse word or aphorism for "any known string".

Here's why this approach is bad, by the way:

class myString(str):

  def __init__(self, s):
    self.s = s
  def __str__(self):
    return self.s

This myString can quickly be made to substitute for a string.  But your code
that does strong type checking won't allow it.

In other words, testing the type of something or whether it's
isinstance(whatever), doesn't tell you whether it walks like a duck and
quacks like a duck.  In general, Python encourages you to care how things
walk and talk not what "type" of thing they are.

Of course, your specific problem may require you to engage in the kind of
bondage you asked about.  ymmv.

Cheers,

// mark

-





From jason at tishler.net  Mon Jun  3 13:25:47 2002
From: jason at tishler.net (Jason Tishler)
Date: Mon, 03 Jun 2002 13:25:47 -0400
Subject: win32com under Cygwin
In-Reply-To: <%5JK8.180011$%u2.88943@atlpnn01.usenetserver.com>
References: <3CF9846B.579606ED@wiencko.com>
 <Xns9221CB7B32FDcliechtigmxnet@62.2.16.82>
 <%5JK8.180011$%u2.88943@atlpnn01.usenetserver.com>
Message-ID: <20020603172547.GB952@tishler.net>

On Mon, Jun 03, 2002 at 08:11:25AM -0400, Steve Holden wrote:
> > Tom Wiencko <tew at wiencko.com> wrote:
> > > I noticed that the Python distribution with Cygwin does not include
> > > the win32 extensions.  The only distribution I can find for them is
> > > a Windows executible, which does not seem to know about Cygwin.
> > >
> > > Can these libraries be loaded under Cygwin?

No (*).

> > > Do they even work under Cygwin?

No (*).  They were built against a completely different Python --
Win32 Python.

> > > Can somebody point me toward how to load them if it is possible?

One would have to build them first.  See below...

> But don't expect a cygwin port to be a simple recompile...

Agreed...

Besides converting MSVC workspaces/projects into Makefiles and dealing
with normal platform porting issues there are also ABI issues that
will make the porting of win32com even more challenging.  Unless, one
compiles using the gcc "-fnative-struct" option, it will require source
level changes to get COM to work.  IIRC, *everything* (e.g., the Cygwin
DLL, Python, win32com, etc.) must be compiled with "-fnative-struct"
or nothing.  So, until Cygwin gcc defaults to "-fnative-struct", I think
this effort may be doomed. :,(

Nevertheless, I would *love* to be able to use win32com under Cygwin
Python.  It has been on my list to try since I first got shared extensions
to work under Cygwin Python.  If anyone would like to collaborate, please
let me know.  If successful, I would definitely include win32com in the
standard Cygwin Python package.

Jason

(*) Or, at least I doubt it.




From phlip_cpp at yahoo.com  Sun Jun  2 05:20:06 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 2 Jun 2002 02:20:06 -0700
Subject: _fileNameToModuleName drek removal
References: <63604d2.0205311348.51bd4f96@posting.google.com>
Message-ID: <63604d2.0206020120.273a2176@posting.google.com>

Phlip wrote:

> Ogle this:
> 
> def _fileNameToModuleName(fileName):
>     'TODO  is there a way in Python to avoid this drek??'
> 
>     assert fileName[-3:] == '.py'
>     if fileName[0] != '.':  os.path.join('.', fileName)
>     fileName = fileName[:-3]
>     moduleName = string.split(fileName, os.sep)[1:]
>     moduleName = string.join(moduleName, '.')
>     assert moduleName[0] != '.'
>     return moduleName
> 
> That, platform neutral, converts a file name ./yak/yak/yak.py 
> into a module name yak.yak.yak 
> 
> Is there some cleaner way to do this? Preferably built-in?

Python's interpreter only goes the other way, from yak.yak.yak 
to yak/yak/yak.py. The language had no reason to do this until I
came along. So I'l have to guess there is no cleaner way than 
the above string and path manipulation.

-- 
  Phlip
                http://www.greencheese.org/YaAw
  --  Got in trouble at StarBucks. I tried to order
        "A double latte mocha and a body piercing."  --


From tack at cscs.ch  Tue Jun 25 05:22:14 2002
From: tack at cscs.ch (Davide Tacchella)
Date: Tue, 25 Jun 2002 11:22:14 +0200
Subject: Help, Compile / debug Python 2.2.1 64 bit on AIX
Message-ID: <20020625112214.72be8087.tack@cscs.ch>

I'm trying to build Python with 64 bit support on AIX, so far I've encountered 2 problems,
dynload_aix.c is not 100% 64 bit compliant (it includes some cast from pointer to (int));
this was causing python to SEGV when building extensions, after changing from int to long, 
the error is now ILL (SIGILL), there is a pointer to NULL:
The call stack from debugger is:
0x000000
initstruct (structmodule.c - line 1508)
_PyImport_LoadDynamicModule ( importdl.c - line 53)
load_module (import.c - line 1365)
import_submodule (import.c - line 1895)
load_next (import.c - line 1751)
import_module_ex (import.c - line 1602)
PyImport_ImportModuleEx (import.c - line 1643)
builtin___import__ (bltinmodule.c - line 40)
PyCFunction_Call (methodobject.c - line 80)
eval_frame (ceval.c - line 2004)
....

Any idea ?
Any help is always welcome.

Davide




From kragen at pobox.com  Fri Jun  7 13:55:20 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 07 Jun 2002 13:55:20 -0400
Subject: 64-bit Python?
References: <mailman.1023393193.10285.python-list@python.org>
Message-ID: <83adq7dldj.fsf@panacea.canonical.org>

"Bjorn Pettersen" <BPettersen at NAREX.com> writes:
> > From: Kragen Sitaker [mailto:kragen at pobox.com]=20
> > The problem is that all of the 64-bit platforms I have access=20
> > to (various Linuxes on Alpha and IA-64, Tru64 on Alpha) have=20
> > 4-byte ints and 8-byte longs, at least by default.  This is=20
> > called "LP64": longs and pointers are 64, but ints are 32, so=20
> > all the code that assumes ints are 32 will continue to work.
> 
> Are you using Win64 on IA-64?

No, of course not.  Since when is "Win64" one of "various Linuxes"?

> Last time I checked (using the MS = compiler) a long was still 32
> bits, and you had to use __int64 to get a = 64 bit integer...

Startlingly braindead.  Perhaps that's to be expected.



From sholden at holdenweb.com  Mon Jun  3 08:13:08 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 3 Jun 2002 08:13:08 -0400
Subject: HTML links in a GUI tkinter
References: <mailman.1022986323.16196.python-list@python.org> <83elfprs1x.fsf@panacea.canonical.org>
Message-ID: <C7JK8.180067$%u2.89342@atlpnn01.usenetserver.com>

"Kragen Sitaker" <kragen at pobox.com> wrote in message
news:83elfprs1x.fsf at panacea.canonical.org...
> aurel-le-gnou <aurelienviolet at noos.fr> writes:
> > How can I put in a frame or in any widget (?) an html link so that the
> > browser opens at the url indicated when the user clicks ?
> > Just simply as if it were an HTML link in an HTML document but in a
> > tkinter GUI ?
>
> Sorry about the empty reply --- in Python 2.x, the webbrowser module
> has a webbrowser.open() method that does what you want.
>
So, all Aurel needs to know now is how to run wxPython widgets inside
tkinter programs...

regards
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From phd at phd.pp.ru  Thu Jun 27 17:40:13 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 28 Jun 2002 01:40:13 +0400
Subject: Python crashed on me
In-Reply-To: <20020627160639.A23280@ibook.distro.conectiva>; from niemeyer@conectiva.com on Thu, Jun 27, 2002 at 04:06:39PM -0300
References: <20020627205042.A18986@phd.pp.ru> <20020627160639.A23280@ibook.distro.conectiva>
Message-ID: <20020628014013.A22132@phd.pp.ru>

On Thu, Jun 27, 2002 at 04:06:39PM -0300, Gustavo Niemeyer wrote:
> >    How can I debug the situation? May be my data grew too big? The pickle
> > file is about 800K.
> 
> Do you have a small testcase?

   No, smaller tests work fine.

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From jepler at unpythonic.net  Mon Jun 17 10:35:26 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Mon, 17 Jun 2002 09:35:26 -0500
Subject: what *is* a class?
In-Reply-To: <3D0DBC87.B1BDEB2D@ifib.uni-karlsruhe.de>
References: <MPG.17774343cc5a589098968a@news.rz.uni-karlsruhe.de> <Gxtt3F.upA@beaver.cs.washington.edu> <3D0DBC87.B1BDEB2D@ifib.uni-karlsruhe.de>
Message-ID: <20020617143525.GB8814@unpythonic.net>

On Mon, Jun 17, 2002 at 12:40:07PM +0200, Uwe Mayer wrote:
> I had to distinguish between creating the object model and actually
> reading the file in. Thus I decided to use the constructor for creating
> the object model and provide a parse() method. I didn't like writing:
> 
> f = file('dummy.avi','rb')
> avi = RIFF(<specify format>)
> avi.parse(f)
> 
> instead I wanted to write:
> 
> avi = RIFF(<specify format>).parse(f)
> 
> but then of course parse() must return an instance object.
> On the one side all parse() methods must explicitly "return self" (which
> is not nice). On the other side from the semantics I thought it was ok
> for parse to return an instance object of RIFF.

This is not the Pythonic style.  Even when it would sometimes be
convenient, methods do not tend to return "self".  That's why these
don't work:

    # Iterate the sorted list
    for item in l.sort(): ...

    # Create and pack a widget
    x = Label(text="Wibble").pack()

    # Dump two objects to a pickle
    p = pickle.Pickler()
    p.dump(obj1).dump(obj2)

One motivation for this is that if list.sort returns None, nobody
will write
    for item in l.sort():
thinking that a copy of l is sorted.  It'll immediately blow up.  I did
a quick Google search to try to find a clear statement of this
principle, but I didn't find it.  However, I tend to agree with the
idea that methods called for their side-effects should generally not
return the object.  This makes it clear when a method mutates the object
and when it returns a mutated object, and this distinction is often
quite important.

Jeff




From Chris.Barker at noaa.gov  Thu Jun 20 17:41:54 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Thu, 20 Jun 2002 14:41:54 -0700
Subject: logical statements (in Python and Numeric)
References: <3D11F1B2.21587F31@lanl.gov> <aet2j4$h8c$1@peabody.colorado.edu> <3D122C5D.C6985686@noaa.gov> <aetb9n$ksn$1@peabody.colorado.edu>
Message-ID: <3D124BEF.9EC71977@noaa.gov>


Fernando P?rez wrote:
> Are you sure? This is what I get:
> 
> In [11]: x = arrayrange(-4.,2.+.001,.001, Float)
> 
> In [12]: x >= -2. & x < 0.
> ---------------------------------------------------------------------------
> TypeError                                 Traceback (most recent call last)
> 
> ?
> 
> TypeError: unsupported operand type(s) for &: 'float' and 'float'

oops, that's what I get for posting untested code. You need parentheses,
the above was trying to bitwise-and -2.0 and x

(x >= -2.) & (x < 0.)
works.

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From martin at strakt.com  Mon Jun 24 03:39:41 2002
From: martin at strakt.com (Martin =?iso-8859-1?Q?Sj=F6gren?=)
Date: Mon, 24 Jun 2002 09:39:41 +0200
Subject: M2Crypto / SimpleXMLRPCServer
In-Reply-To: <slrnah4md0.r77.insyte-clp@petra.bitstream.net>
References: <slrnah2nmh.p18.insyte-clp@petra.bitstream.net> <slrnah4md0.r77.insyte-clp@petra.bitstream.net>
Message-ID: <20020624073941.GC12169@ratthing-b3cf>

On Thu, Jun 20, 2002 at 10:40:24PM +0000, Ben Beuchler wrote:
> In article <slrnah2nmh.p18.insyte-clp at petra.bitstream.net>, Ben Beuchler wrote:
> 
> > It appears that the SSL part, at least, works great.  I'm able to make a
> > connection using xmlrpclib.ServerProxy("https://blah") but the actual
> > remote method call hangs indefinitely.
> > 
> > Is my basic premise flawed?  Or just my implementation?
> 
> FWIW, strace-ing the client and the server indicates that they're both
> blocking on read()ing from the socket.  Of course, since it's all
> encrypted, I have no idea what they're waiting for...
> 
> On a related note, anyone know how to get OpenSSL (specifically via
> M2Crypto) to use a NULL cipher?  Being able to dump the traffic with
> Ethereal or something would be damn spiffy.

I don't know about M2Crypto, but in the C API, you'd do
  SSL_CTX_set_cipher_list(ctx, "eNULL");
and in pyOpenSSL you'd do
  ctx.set_cipher_list("eNULL")

my guess is that it's something similar in M2Crypto.


Regards,
Martin Sj?gren

-- 
Martin Sj?gren
  martin at strakt.com              ICQ : 41245059
  Phone: +46 (0)31 7710870       Cell: +46 (0)739 169191
  GPG key: http://www.strakt.com/~martin/gpg.html




From joe at notcharles.ca  Fri Jun  7 01:01:54 2002
From: joe at notcharles.ca (Joe Mason)
Date: Fri, 07 Jun 2002 05:01:54 -0000
Subject: Using popen interactively
Message-ID: <slrnag0fhu.6u3.joe@zork.plover.net>

I'm trying to write a CGI script which filters the output from an
executable.  The typical interaction is to print some lines to stdout
and then wait for an input on stdin.  Repeat until program finishes.
I want to take each chunk of output, wrap it in HTML and then pass it to
the web server.

So this text:

  West of House
  You are standing in an open field west of a white house, with a
  boarded front door.
  There is a small mailbox here.

  >

would get wrapped in HTML and displayed, and then I'd throw up a from
with an input box, take the text posted from the form ("open mailbox",
"go north", whatever) and send it off to the app on stdin.  ("quit"
kills the app.)

So I'm trying to use popen2.Popen3("executable-name") to get handles to
stdin and stdout.  Works fine, except that I'm not sure how to detect
when the app has stopped printing and is waiting for input.  readline()
gets the first lines ok, but then the ">" line has no newline on the
end, so it blocks.  Apparently read() is supposed to return an empty
string when it hits EOF, but it blocks too - so apparently on an
interactive file stream like this EOF isn't set when its waiting for
input.

The algorithm I'm thinking of would be something like:

  app = popen.Popen3("application")
  infile = app.fromchild
  outfile = app.tochild

  # while the app is still running...
  while app.poll() == -1:
    while infile.poll() == -1:  # except file has no poll() method
      line = infile.readline()
      # process line and output it to web form 
    # get text from web form
    outfile.write(text + "\n")

Any suggestions?

Joe 


From jepler at unpythonic.net  Mon Jun  3 19:06:14 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Mon, 3 Jun 2002 18:06:14 -0500
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
In-Reply-To: <ufnrjqb4h0ub12@news.supernews.com>
References: <06da01c208f4$d69003c0$6601a8c0@boostconsulting.com> <200206010121.g511LxX19223@pcp742651pcs.reston01.va.comcast.net> <073501c20960$23f367e0$6601a8c0@boostconsulting.com> <200206011334.g51DY1D21669@pcp742651pcs.reston01.va.comcast.net> <07d801c20970$cdebe050$6601a8c0@boostconsulting.com> <mailman.1022940122.3421.python-list@python.org> <Id5K8.171932$Po6.340972@rwcrnsc52.ops.asp.att.net> <oWIK8.179611$%u2.31494@atlpnn01.usenetserver.com> <ufnrjqb4h0ub12@news.supernews.com>
Message-ID: <20020603180609.A987@unpythonic.net>

On Mon, Jun 03, 2002 at 06:32:34PM -0400, John Roth wrote:
[about "x = ('aaa'); x[0] += 'bbb'"]
> Maybe. It might actually update the string, producing
> 'aaabbb' at a different location in memory,
> and then (properly) fail to update the tuple.

There need be no speculate.

>>> import dis
>>> def f(x): x[0] += 'bbb'
... 
>>> dis.dis(f)
          0 SET_LINENO               1

          3 SET_LINENO               1
          6 LOAD_FAST                0 (x)
          9 LOAD_CONST               1 (0)
         12 DUP_TOPX                 2
         15 BINARY_SUBSCR       
         16 LOAD_CONST               2 ('bbb')
         19 INPLACE_ADD         
         20 ROT_THREE           
         21 STORE_SUBSCR        
         22 LOAD_CONST               0 (None)
         25 RETURN_VALUE        
>>> def g(x): x[0] = x[0] + 'bbb'
... 
>>> dis.dis(g)
          0 SET_LINENO               1

          3 SET_LINENO               1
          6 LOAD_FAST                0 (x)
          9 LOAD_CONST               1 (0)
         12 BINARY_SUBSCR       
         13 LOAD_CONST               2 ('bbb')
         16 BINARY_ADD          
         17 LOAD_FAST                0 (x)
         20 LOAD_CONST               1 (0)
         23 STORE_SUBSCR        
         24 LOAD_CONST               0 (None)
         27 RETURN_VALUE        

The differences are that the two LOAD_FAST / LOAD_CONST sequences have been
optimized to LOAD_FAST / LOAD_CONST / DUP_TOPX, a ROT_THREE has been added
to get operands in the right place, and (the real difference) the operation
is INPLACE_ADD instead of BINARY_ADD in the 'meat' of the function.

So, yes, in both questions a string 'aaabbb' will be constructed somewhere
in memory, but only on the stack (cpythonically speaking, of course)

Jeff




From zopestoller at thomas-guettler.de  Thu Jun 27 07:03:09 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 27 Jun 2002 13:03:09 +0200
Subject: Interactive session via tcp
References: <mailman.1025172457.23663.python-list@python.org>
Message-ID: <3D1AF0ED.8070406@thomas-guettler.de>

Andreas Ames wrote:

> Hi all,
> 
> I have a python daemon running with access to an XML messaging system
> (jabber in this case).  I'm wondering, how I can achieve remote
> control over my daemon.
> 
> The first thing that I can think of is to start an interactive session
> via the XML stream.  Security is not my concern right now.  Following
> steps seem to be required, afaik (please correct me, if my assumptions
> are wrong).
> 
> 1) Deamon receives a command to open an interactive session.
> 
> 2) The main thread of my daemon starts a read/eval loop and redirects
>    stdout and stder.


Sorry, if this if wrong, I read your post quickly.
Doesn't "telnet jabber_port" do what you want?

thomas



From jdhunter at nitace.bsd.uchicago.edu  Wed Jun 19 10:49:21 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Wed, 19 Jun 2002 09:49:21 -0500
Subject: I have .py, I want .exe, how??
References: <mailman.1024448546.1172.python-list@python.org>
Message-ID: <m2znxrnx1q.fsf@video.paradise.lost>

>>>>> "jubafre" == jubafre  <jubafre at zipmail.com.br> writes:

    jubafre> i have a file.py how i make e executable file .exe for
    jubafre> run in windows???

In general, whenever you want to convert one file to another, go to
google and search for the two extensions joined by '2' or 'to'

Try

py2exe
jpg2gif
latex2html
ps2pdf

and so on.  You'll almost always find what you want.  In this case,
the first hit would take you to:

http://starship.python.net/crew/theller/py2exe/

Cheers,
John Hunter


From sarmstrong13 at mac.com  Fri Jun  7 14:10:01 2002
From: sarmstrong13 at mac.com (SA)
Date: Fri, 07 Jun 2002 13:10:01 -0500
Subject: Tkinter help on MacOSX please.
Message-ID: <B9266129.6930%sarmstrong13@mac.com>

Hi Everyone-

    I've installed python, tcl, tk on my MacOSX 10.1.5 system using the
--enable-frameworks configuration. I have a framework for tcl,tk,and python
under my /Library/Frameworks directory. When I run a simple button code in a
python script:

#!/usr/bin/env python

import sys
from Tkinter import *

def die(event):
    sys.exit()

root = Tk()
button = Button(root)
button["text"] = "Hello"
button.bind("<Button-1>", die)
button.pack()
root.mainloop()

I get the following output in my shell(tcsh by the way):

original argc=2
original argv[0] = "python"
original argv[1] = "./ave.py"
modified argc=2
modified argv[0] = "python"
modified argv[1] = "./ave.py"

Nothing happens.

So I then hit Ctrl-C to stop the script and I get the following Traceback
error:

Traceback (most recent call last):
  File "./ave.py", line 14, in ?
    root.mainloop()
  File 
"/Library/Frameworks/Python.framework/Versions/2.2/lib/python2.2/lib-tk/Tkin
ter.py", line 929, in mainloop
    self.tk.mainloop(n)
KeyboardInterrupt

It looks like it is calling the correct Tkinter Module, but nothing is bein
displayed. If however I enter the code line for line in the Python.app that
I compile, I get the proper button widget. Maybe I'm wrong (because I
definitely do NOT fully understand the Mac OSX Frameworks system yet), but
if the coded is properly calling the correct Tkinter module, it should not
matter whether or not it is done within the Python.app or by calling python
from within Python.app. Is this correct? Is there a way to do this on a Mac?
(other than MacPython wich works only on pre OSX systems and doe not take
advantage of the BSD under OSX)

Thanks.
SA



From fredrik at pythonware.com  Sun Jun  9 07:45:08 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 09 Jun 2002 11:45:08 GMT
Subject: another near-sighted, time-wasting syntax enhancement idea
References: <e395b886.0206051956.5f447e9b@posting.google.com> <Xns92265AD72BE0cliechtigmxnet@62.2.16.82> <e395b886.0206081735.1c1d0f4@posting.google.com>
Message-ID: <8dHM8.42399$n4.9662059@newsc.telia.net>

Carl Banks wrote:
>
> class NewsConverse:
>         def write_reply (self, message):
>                 if self.poster_misses_point (message):
>                         self.set_paragraph_reply("Code scrunched up "
>                                                  "against the right "
>                                                  "margin can be pretty "
>                                                  "unreadable itself.")

class NewsConverse:
    def write_reply(self, message):
        if not self.poster_misses_point(message):
            return
        self.set_paragraph_reply(
            "Don't use ludicrous indentation settings,"
            "and learn to refactor to keep the depth"
            "under control."
        )

</F>




From phlip_cpp at yahoo.com  Mon Jun  3 00:27:46 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 03 Jun 2002 04:27:46 GMT
Subject: prototyping good OOdesign in Python?
References: <mailman.1023074765.20704.python-list@python.org>
Message-ID: <ader82$oit@dispatch.concentric.net>

>>0. Write a project schedule for management; something like this:
>>   1. Prototype in python
>>   2. Convert to C++
>>1. Prototype in python
>>2. Give a nice demo, and convince management to skip step 2
>>3. Re-implement a few cpu intensive classes in C++

        http://www.c2.com/cgi/wiki?AlternateHardAndSoftLayers

This means that after prototyping in Python one /incrementally/ refactors 
elements into C++ (hopefully via the Boost Python Library). Of course the 
speed elements go into C++, as do the elements that bind with legacy C++ 
stuff.

It's what the Python community already does. We don't crack JPEG files in 
Python; we just call into a C layer like imlib, via a binding layer.

-- 
  Phlip
   http://www.greencheese.org/EvolutionaryPsychology
  --  Now collecting votes for a new group:
      news:alt.flame.moderated  --



From gh_pythonlist at gmx.de  Tue Jun 18 20:16:01 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 19 Jun 2002 02:16:01 +0200
Subject: procmail replacement in Python
In-Reply-To: <XFMail.20020618170437.shalehperry@attbi.com>
References: <20020618235901.GA2197@lilith.my-fqdn.de> <XFMail.20020618170437.shalehperry@attbi.com>
Message-ID: <20020619001601.GC2237@lilith.my-fqdn.de>

* Sean 'Shaleh' Perry <shalehperry at attbi.com> [2002-06-18 17:04 -0700]:
> On 18-Jun-2002 Gerhard H?ring wrote:
> > Does such a thing exist? 
>
> there is also a nice filter language from CMU called Sieve (I think).  You
> could implement a python based Sieve interpreter.

I'd just switch to Cyrus IMAP instead. I'm currently using Courier IMAP
and am quite satisfied, so I have little reason to switch. I even found
Cyrus difficult to install, even though it's also available as a FreeBSD
package (just like Courier).

All that I'm currently using from procmail are regexes on header fields,
combined with AND, and maybe OR conditions. And having two custom
filters before matching the header fields:

- spamassassin
- this one to fight stupidity in the setup of the German freemailer GMX:
  they have some braindead software that does mbox-like From_ quoting on
  their POP3 setup :-(

#!/usr/bin/env python
import re, sys

BROKEN_FROM = re.compile(r"^(>+)(from )(.*)", re.I | re.DOTALL)

for line in sys.stdin.xreadlines():
    match = BROKEN_FROM.match(line)
    if not match:
        sys.stdout.write(line)
    else:
        # Remove one quote char added by braindead GMX software
        groups = match.groups()
        escape_chars = groups[0][:-1]
        sys.stdout.write(escape_chars + groups[1] + groups[2])

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 23.1 ?C      Wind: 0.9 m/s




From david.abrahams at rcn.com  Sat Jun 15 13:29:55 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Sat, 15 Jun 2002 13:29:55 -0400
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de>
Message-ID: <aeftmj$m29$1@bob.news.rcn.net>

"Michael 'Mickey' Lauer" <mickey at tm.informatik.uni-frankfurt.de> wrote in
message news:3d0b4fd6 at nntp.server.uni-frankfurt.de...
> David Abrahams <david.abrahams at rcn.com> wrote:
> > > "Michael 'Mickey' Lauer" <mickey at tm.informatik.uni-frankfurt.de> wrote
in
> > message news:3d0a8b6e at nntp.server.uni-frankfurt.de...
> >
> >
> >> I had a quick look at the other wrappers (SILOON, GRAD, CXX, etc.)
> >> but most look either outdated or not appropriate for this task.
> >> Do you know an (automatic) wrapper generator which has this feature?
> >>
> >> I know sip can do this for c++ libraries - can sip also be used
> >> to wrap plain c-libs without "handwriting" code?
> >
> > On most platforms, Boost.Python would work for you. The determining
issue is
> > whether or not the C++ compiler treats "C" and "C++" linkage function
> > pointers the same way. If the appended program will compile with your
C++
> > compiler, you're probably OK:
>
> The attached program compiles fine here. However, regarding boost.python I
> thought this was more of a c++ library for handcrafting tightly integrated
> bindings rather than an automatic wrapper generator? Is this understanding
wrong?

Automation is a relative term. Every wrapping system requires something more
than just your source code to tell it what to wrap, and how. For example:

    struct X {
        void set(int y) { value = y; }
        int value;
    };

    struct Y {
        X& get_x() { return x; }
        X x;
    };

Now, throw this header at SWIG. What does it do with Y::get_x()? How is the
return value handled? Is the X object copied or does the resulting Python
object refer back to the x member of the containing Y object? It makes a
difference if X::set() is ever called.

With Boost.Python, what you write looks like a kind of IDL (interface
description language), which just happens to be built out of C++ source
code. Lots of jobs (function argument type checking, conversion, arity
checking) are done automatically, but the library doesn't decide which
elements to wrap. Unlike SWIG and a few others, there's no attempt to parse
your source code. Instead, the compile-time introspection capabilities of
C++ are exploited to automatically build wrappers from function and member
function pointers.

-Dave






From kragen at pobox.com  Tue Jun  4 22:31:51 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: Tue,  4 Jun 2002 22:31:51 -0400 (EDT)
Subject: Extracting data from HTML
In-Reply-To: <1023085956.8048.87.camel@localhost>
References: <82096df4.0205311152.5891a17b@posting.google.com>
        <1022883953.6150.5802.camel@localhost>
        <mailman.1022908262.30070.python-list@python.org> 
        <837klhrpv3.fsf@panacea.canonical.org>
        <1023085956.8048.87.camel@localhost>
Message-ID: <20020605023151.EDC70BDCE@panacea.canonical.org>

Ian Bicking writes:
> I think you are misinterpreting Geoff's response, and you seem to
> have a chip on your shoulder about it.

Sorry about the attitude.  I'm frustrated that, twelve years into the
World Wide Web, there are still so many people who fail to learn from
its lessons.

> He did not compare XML-RPC to HTTP, but to HTML (at least, that's
> clearly implicit because this thread was talking about HTML
> parsing).  HTML is clearly a poor way to exchange machine-readable
> information, there are too many layout-related tags that are usually
> only appreciated by humans.

I don't really agree that HTML is a poor way to exchange
machine-readable information; that is, after all, what the language is
designed for.  But if the author of the HTML doesn't have
machine-readability in mind, and many don't, the HTML usually won't be
very machine-readable.

Nevertheless, Geoff said "XML-RPC or some other protocol that's
designed to carry data".  HTTP and XML-RPC are protocols, although
they each define data formats; HTML is a language.

**

With regard to your earlier question about htmllib and HTML::Parse: I
hadn't actually tried to use htmllib, but had only read that it was
unreliable on malformed HTML.  I haven't actually been able to feed it
HTML that's malformed enough to break it, though.

Here are link-extraction scripts in Perl using standard CPAN libraries
and in Python using standard Python libraries.  The Perl script is
more featureful.  The Python script was more painful to write, partly
because htmllib.HTMLParser has a more poorly designed interface than
HTML::Parser (hard as that may be to believe), but mostly because
there's already an HTML::LinkExtor in the library.

Perl version:

#!/usr/bin/perl -w
use strict;
require HTML::LinkExtor;
my $p = HTML::LinkExtor->new(\&cb, "http://www.sn.no/");
sub cb {
    my($tag, %links) = @_;
    print "$tag @{[%links]}\n";
}
$p->parse_file("pathological.html");
__END__


Python version:

#!/usr/bin/python
import htmllib, formatter
class x(htmllib.HTMLParser):
    def dump(self, tag, attrs):
        print tag,
        for a, v in attrs:
            if a in ['action', 'src', 'href']:
                print a, v,
        print
    def do_img(self, attrs):
        self.dump('img', attrs)
    def start_a(self, attrs):
        self.dump('a', attrs)
    def start_form(self, attrs):
        self.dump('form', attrs)

y = x(formatter.NullFormatter())
y.feed(open('pathological.html').read())
y.close()


Here's pathological.html, which I guess is not very pathological,
because it didn't break either script.  I'd be very interested to see
HTML pathological enough to break one or the other, but still accepted
by Netscape >3.0 or MSIE >4.0.

<!-- look -- this is my web page, tidy finds 20 errors in it, more than 
1 per line -->
<ul>
Moroon & I are --&gt<b><B><i>GETTING MARRIED</b></i><-- next year.
<p> Look at my <a href=http://example.com/~moron>website.</a>
<body bgcolor=#ff7777>
<script>
This stuff shouldn't get displayed.
x = 1; y = 3;
if (x<y) x = y;
</script>
<IMG SRC='MYPIC.GIF'/>LOOK AT ME!
Sign my guestbook: <table><tr><td><form action=guestbook.cgi method=post>
<input name="yourname"><td><input type="submit"></form></table>

-- 
<kragen at pobox.com>       Kragen Sitaker     <http://www.pobox.com/~kragen/>
The Internet stock bubble didn't burst on 1999-11-08.  Hurrah!
<URL:http://www.pobox.com/~kragen/bubble.html>
The power didn't go out on 2000-01-01 either.  :)




From hesterloli at hotmail.com  Tue Jun 18 17:55:32 2002
From: hesterloli at hotmail.com (George Hester)
Date: Tue, 18 Jun 2002 21:55:32 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com><aegfck$6tkre$1@ID-114079.news.dfncis.de><wEPO8.16066$uk2.8447171@twister.nyroc.rr.com><3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com><gYgP8.42356$GY.12361719@twister.nyroc.rr.com><aekcug$7js8n$1@ID-114079.news.dfncis.de><9YBP8.44418$GY.13865843@twister.nyroc.rr.com><mailman.1024423525.24359.python-list@python.org><i0NP8.46300$GY.14481287@twister.nyroc.rr.com> <mailman.1024435149.22322.python-list@python.org>
Message-ID: <o%NP8.46689$GY.14560328@twister.nyroc.rr.com>

No George was asking for a script he could put in a ASP to determine his version
of Python.

--
George Hester
_________________________________
"Cliff Wells" <logiplexsoftware at earthlink.net> wrote in message
news:mailman.1024435149.22322.python-list at python.org...
> On Tue, 18 Jun 2002 20:48:14 GMT
> George Hester wrote:
>
> > Bah humbug.
>
> Ah, the Christmas spirit, all year round ;)
>
> --
> Cliff Wells, Software Engineer
> Logiplex Corporation (www.logiplex.net)
> (503) 978-6726 x308  (800) 735-0555 x308
>
>




From bernie at 3captus.com  Tue Jun 11 11:50:37 2002
From: bernie at 3captus.com (Bernard Yue)
Date: Tue, 11 Jun 2002 15:50:37 GMT
Subject: getopt
Message-ID: <3D061A9F.7CDC7132@3captus.com>

Hi,

I was expecting an exception from the following getopt.getopt() called,
but didn't.  Is it normal?  I've try with Python 2.1 and 2.2 as well. 
Result starts the same.


Python 2.3a0 (#1, Jun  7 2002, 14:47:32) 
[GCC 2.95.3 20010315 (SuSE)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import getopt
>>> args = '-a -b -c -d bar a1 a2'.split()
>>> args
['-a', '-b', '-c', '-d', 'bar', 'a1', 'a2']
>>> optlist, args = getopt.getopt(args, 'abc:d:')
>>> optlist
[('-a', ''), ('-b', ''), ('-c', '-d')]
>>> args
['bar', 'a1', 'a2']
>>> 

Thanks in advance!


Bernie


From SBrunning at trisystems.co.uk  Thu Jun 13 05:19:36 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Thu, 13 Jun 2002 10:19:36 +0100
Subject: methods to parse big foreign ascii-file?
Message-ID: <31575A892FF6D1118F5800600846864DCBD345@intrepid>

> From:	holger krekel [SMTP:pyth at devel.trillke.net]
> Igor Stroh wrote:
> > On Wed, 12 Jun 2002 18:05:51 +0200, "holger krekel"
> > <pyth at devel.trillke.net> wrote:
> > 
> > > - come up with a grammar and let a parser do
> > >   the hard work (which one?)
> > 
> > if the file is structured in a certain form, that is you can describe
> the
> > data structure with a EBNF, then have a look at 'simpleparse' (I've lost
> > the link, but google will certainly have some for you :) )...
> 
> thanks. This might be simple yet powerful enough for the task. I must
> try hard to remember EBNF-grammars from my diploma exams, though :-)
 
There is an article about SimpleParse and EBNF here -
<http://www-106.ibm.com/developerworks/linux/library/l-simple.html>.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.




From amuys at shortech.com.au  Mon Jun 24 21:50:49 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 24 Jun 2002 18:50:49 -0700
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <aer2fn$frc$1@peabody.colorado.edu> <3D122E79.A189C72E@astro.cornell.edu> <aetapt$kld$1@peabody.colorado.edu> <3D124575.2A1393FF@astro.cornell.edu>
Message-ID: <7934d084.0206241750.221bbff2@posting.google.com>

Tom Loredo <loredo at astro.cornell.edu> wrote in message news:<3D124575.2A1393FF at astro.cornell.edu>...
> Fernando P?rez wrote:
> > 
> > Don't know if there's a misunderstanding here: 
> 
> Very likely!  (On my part!)
> 
> > cygwin is a unix-like
> > environment under Windows. 
> 
> Okay, my concern was that it wouldn't work with or be able to produce
> Windows binaries---like if I had to produce binary extension modules
> to ship to others, or a McMillan-ized install that used PythonWin
> and Tk or wxWindows.  Cygwin is so huge that I got the impression
> it was a self-contained "subenvironment" so to speak, so I didn't
> realize I could use it to produce stuff that non-cygwin Windows users
> could use.
> 

It is important to differentiate between the cygwin environment, and
the cygwin platform itself.  At its core cygwin is a dll that provides
a port of most of the POSIX API to W32.  On top of this, cygwin also
offers a host of other API's common on various unix platforms ported
to the cygwin.dll 'unix platform'.  On top of these libraries the
cygwin environment is assembled out of all manner of common unix
programs, utilities, servers, etc built using these libraries and the
cygwin.dll.

The environment is huge.  Considering that a full install includes
ports of XFree86, TeX, gcc & libs, python2.2, perl5.6.1, a full
install == a LOT of bytes :).

The libraries however are a fraction of this size, consider a ls -l
*.dll on /bin (where cygwin keeps the libraries, I assume for windows
compat' sake)

$ ls -l *.dll | awk 'BEGIN { sum = 0 } { sum = sum + $5 } END { print
sum }'
8931012

$ ls -l cygwin1.dll
-rwxr-xr-x    1 Administ None       895304 Jun 24 12:33 cygwin1.dll

$ ls -l *.dll
-rwxr-xr-x    1 Administ None        46592 Jan 21 04:51 cygXpm-X4.dll
-rwxr-xr-x    1 Administ None        41984 Jan 21 04:51
cygXpm-noX4.dll
-rwxr-xr-x    1 Administ None        58880 May  7 16:36 cygbz2-1.dll
-rwxr-xr-x    1 Administ None       635904 May 17 22:25 cygcrypto.dll
-rwxr-xr-x    1 Administ None       333824 Jun 19 08:16 cygdb2.dll
-rwxr-xr-x    1 Administ None        51200 Mar 17 16:24 cygexslt-0.dll
-rwxr-xr-x    1 Administ None        46080 Apr 25  2001 cygform5.dll
-rwxr-xr-x    1 Administ None        35328 Jan  9 16:11 cygform6.dll
-rwxr-xr-x    1 Administ None        18944 Feb 20 13:06 cyggdbm.dll
-rwxr-xr-x    1 Administ None        19968 Jan 13 11:29
cyghistory5.dll
-rwxr-xr-x    1 Administ None        22016 Dec 13  2001 cygintl-1.dll
-rwxr-xr-x    1 Administ None        21504 Jun 21  2001 cygintl.dll
-rwxr-xr-x    1 Administ None        82944 Oct 20  2001 cygitcl30.dll
-rwxr-xr-x    1 Administ None        35328 Oct 20  2001 cygitk30.dll
-rwxr-xr-x    1 Administ None        45568 Feb  8 10:00 cygjbig1.dll
-rwxr-xr-x    1 Administ None       121856 Feb  9 15:21 cygjpeg6b.dll
-rwxr-xr-x    1 Administ None        26624 Apr 25  2001 cygmenu5.dll
-rwxr-xr-x    1 Administ None        19968 Jan  9 16:11 cygmenu6.dll
-rwxr-xr-x    1 Administ None       159232 Apr 25  2001
cygncurses++5.dll
-rwxr-xr-x    1 Administ None       179200 Jan  9 16:11
cygncurses++6.dll
-rwxr-xr-x    1 Administ None       230912 Apr 25  2001
cygncurses5.dll
-rwxr-xr-x    1 Administ None       206848 Jan  9 16:11
cygncurses6.dll
-rwxr-xr-x    1 Administ None        15360 Apr 25  2001 cygpanel5.dll
-rwxr-xr-x    1 Administ None        11776 Jan  9 16:11 cygpanel6.dll
-rwxr-xr-x    1 Administ None        41105 Nov 22  2001 cygpcre.dll
-rwxr-xr-x    1 Administ None        40251 Nov 22  2001
cygpcreposix.dll
-rwxr-xr-x    1 Administ None       175104 May  8 05:10 cygpng10.dll
-rwxr-xr-x    1 Administ None       179200 May 24 12:23 cygpng12.dll
-rwxr-xr-x    1 Administ None       174080 Jan 21 11:05 cygpng2.dll
-rwxr-xr-x    1 Administ None        22528 Jun  9 15:50 cygpopt-0.dll
-rwxr-xr-x    1 Administ None       123904 Jan 13 11:29
cygreadline5.dll
-rwxr-xr-x    1 Administ None        68016 Nov 21  2001 cygregex.dll
-rwxr-xr-x    1 Administ None       159744 May 17 22:25 cygssl.dll
-rwxr-xr-x    1 Administ None       399360 Oct 20  2001 cygtcl80.dll
-rwxr-xr-x    1 Administ None         4608 Oct 20  2001
cygtclpip80.dll
-rwxr-xr-x    1 Administ None        10240 Oct 20  2001
cygtclreg80.dll
-rwxr-xr-x    1 Administ None       258560 Feb 10 18:36 cygtiff3.dll
-rwxr-xr-x    1 Administ None       637952 Oct 20  2001 cygtk80.dll
-rwxr-xr-x    1 Administ None       895304 Jun 24 12:33 cygwin1.dll
-rwxr-xr-x    1 Administ None      1277440 Mar 17 13:56 cygxml2-2.dll
-rwxr-xr-x    1 Administ None       155136 Mar 17 16:23 cygxslt-1.dll
-rwxr-xr-x    1 Administ None        15360 Mar 17 16:24
cygxsltbreakpoint-1.dll
-rwxr-xr-x    1 Administ None        50688 Mar 12 14:38 cygz.dll
-rwxr-xr-x    1 Administ None        49664 Feb 25 13:33 libW11.dll
-rwxr-xr-x    1 Administ None       749056 Aug 22  2001
libperl5_6_1.dll
-rwxr-xr-x    1 Administ None       771584 Jan  1 06:26
libpython2.2.dll
-rwxr-xr-x    1 Administ None       204288 Oct 20  2001 tix4180.dll

As you can see the cygwin dll is only ~900k, while even the entire
porting environment (excluding X11R6) is < 9Mb (X11R6 is just over
4Mb).

So in most situations this is not imposing a signifigant overhead.

Andrae Muys


From kragen at pobox.com  Sat Jun  8 20:28:23 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 08 Jun 2002 20:28:23 -0400
Subject: array, list, performance...
References: <mailman.1023369012.10674.python-list@python.org> <3CFFB86A.4114459A@accessforall.nl>
Message-ID: <83ofel5m8o.fsf@panacea.canonical.org>

Ype Kingma <ykingma at accessforall.nl> writes:
> Michael Chermside wrote:
> >     li.append(x)   appending is O(1)
> >                    (actually, it SOMETIMES takes O(len(li)) as it
> >                     resizes the list. But if you grow a list by
> >                     continually appending, the amortized time is
> >                     linear (ie, constant amortized time per element)
> 
> Append is often enough linear in the length of the list
> that growing by appending is O(len(li) * len(li)).

Not when I try it:
Python 2.1.1 (#2, Dec 19 2001, 12:19:22) 
[GCC 2.95.2 20000220 (Debian GNU/Linux)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> import MetaPy.Speed
>>> def mklist(n):
...     ar = []
...     for ii in xrange(n):
...             ar.append(ii)
>>> MetaPy.Speed.seconds_per_call(lambda: mklist(0))
1.7825510264682151e-05
>>> MetaPy.Speed.seconds_per_call(lambda: mklist(1))
2.862680041153042e-05
>>> MetaPy.Speed.seconds_per_call(lambda: mklist(10))
7.8054009556034588e-05
>>> MetaPy.Speed.seconds_per_call(lambda: mklist(100))
0.00057740376706547905
>>> MetaPy.Speed.seconds_per_call(lambda: mklist(1000))
0.0055893541133707342
>>> MetaPy.Speed.seconds_per_call(lambda: mklist(10000))
0.055918868174255905
>>> MetaPy.Speed.seconds_per_call(lambda: mklist(100000))
0.59515897688580466

Looks linear to me.



From ark at research.att.com  Fri Jun  7 14:15:45 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 7 Jun 2002 14:15:45 -0400 (EDT)
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <15616.63342.559215.644086@12-248-41-177.client.attbi.com>
	(message from Skip Montanaro on Fri, 7 Jun 2002 13:11:58 -0500)
References: <yu99vg8v6orb.fsf@europa.research.att.com>
        <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
        <200206071747.g57HlbH01057@europa.research.att.com>
        <15616.62634.635544.905055@12-248-41-177.client.attbi.com>
        <200206071804.g57I4fp01212@europa.research.att.com> <15616.63342.559215.644086@12-248-41-177.client.attbi.com>
Message-ID: <200206071815.g57IFjg01303@europa.research.att.com>

Skip> Showing that you cd'd to your build directory and executed "./python" would
Skip> give me greater faith that you were executing the just-built interpreter and
Skip> importing the proper .so file.  Note that since the build process renamed
Skip> the file, you will probably have to go back and mv it back to struct.so
Skip> before trying this...

   $ ./python
   Python 2.2.1 (#4, Jun  7 2002, 13:24:38) 
   [GCC 3.1] on sunos5
   Type "help", "copyright", "credits" or "license" for more information.
   >>> import struct
   Traceback (most recent call last):
     File "<stdin>", line 1, in ?
   ImportError: No module named struct

To what would it have renamed struct.so?




From martin at v.loewis.de  Thu Jun  6 16:32:00 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 06 Jun 2002 22:32:00 +0200
Subject: 64-bit Python?
References: <87elfl3yvu.fsf@mathdogs.com> <C0FL8.41941$n4.9496108@newsc.telia.net> <j41ybk69ac.fsf@informatik.hu-berlin.de> <83660wgvrm.fsf@panacea.canonical.org>
Message-ID: <m38z5sw3lr.fsf@mira.informatik.hu-berlin.de>

Kragen Sitaker <kragen at pobox.com> writes:

> Yes, this is correct.  Even the buffer interface, which my
> arrayfrombuffer package[0] uses to provide access to mmapped files as
> Numeric arrays inexplicably uses int instead of size_t.

It can't use size_t for the object's size, since size_t is unsigned,
but some applications put negative numbers into ob_size (see
longobject.c for an example).

Regards,
Martin



From gerhard.haering at gmx.de  Thu Jun 27 11:37:26 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 27 Jun 2002 15:37:26 GMT
Subject: Var substitution
References: <c07d0edc.0206270557.28180893@posting.google.com>
Message-ID: <slrnahmc9l.2mc.gerhard.haering@lilith.my-fqdn.de>

Guy wrote in comp.lang.python:
> Hi
> 
> What i'm wanting to know, is there any way of looking at a variables
> value rather than its name.
> 
> Example
> 
> Unix : ${MyVar}
> Dos  : %MyVar%

Python: MyVar

> Both of the above will be subtituted with the values.

In Python, values are bound to names. By using a name on the left-hand
side of =, you bind a value to it. In all other cases, you access the
value with its name.

x = "foo"
Bind the value "foo" to the name x in the current scope (namespace).

print x
Access the value the name "x" in the current namespace refers to.

y.x = "foo"
Bind the value x in the namespace y (which might be a class, class
instance, module, function, ...) to the value "foo".

print y.x
Access the value of the name "x" in the namespace y.

> Example python script (below causes error):
> 
>>>> guy="name"

You bind the value "name" of type string to the name 'guy' in the
current namespace. In this case, the global namespace.

>>>> import os

You import the module os, which does have a namespace of its own.

>>>> os.guy

You try to access 'guy' in the namespace of the os module, which
doesn't exist there.

> This is what should happen :
> 
>>>> import os
>>>> os.name

Sure, the os module has "name" in its namespace, but not "guy". That's
why this works.

> 'nt'

Just for demonstration, as there's no point in it, but you _could_ do
the following:

    import os
    os.guy = "foo"
    print os.guy

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From msoulier at nortelnetworks.com_.nospam  Thu Jun 20 12:26:35 2002
From: msoulier at nortelnetworks.com_.nospam (Michael P. Soulier)
Date: 20 Jun 2002 16:26:35 GMT
Subject: do...while
Message-ID: <slrnah40l1.12v.msoulier@pmerd071.ca.nortel.com>

    Greetings.

    How do you people handle the lack of a do...while in Python? I find that
at times, I must initialize variables for a loop with identical code to the
loop, which feels like a waste to me. 

ie.

    line = filehandle.readline()
    while len(line) > 5:
        line = filehandle.readline()

    A do...while here would be nice. 

    Just curious. 

    Mike

-- 
Michael P. Soulier, QX41, SKY  Tel: 613-765-4699 (ESN: 39-54699)
Optical Networks, Nortel Networks, SDE Pegasus
"...the word HACK is used as a verb to indicate a massive amount
of nerd-like effort."  -Harley Hahn, A Student's Guide to Unix


From bdesth at nospam.free.fr  Tue Jun 25 13:59:03 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Tue, 25 Jun 2002 13:59:03 -0400
Subject: 'parent object'?
References: <3d18228c$0$221$4d4ebb8e@news.nl.uu.net> <3D182662.CBC925DB@alcyone.com>
Message-ID: <3D18AF67.2010207@nospam.free.fr>

Erik Max Francis wrote:
> "Guyon Mor?e" wrote:
> 
> 
>>Is there, like 'self, also a 'parent' object of some sort?
>>
>>The problem i'm having is that i have a class, which contains a list
>>of
>>instances of another class. How can i let this 'childclass' call a
>>function
>>which is on the 'parent class'?
> 
> 
> An instance has access to its class (.__class__ attribute), and a class
> has access to its parent classes (.__bases__ attribute).
> 

That's ok for 'parent/child' in the inheritence meaning. But the OP 
asked about something else : a container instance (the 'parent'), and 
the instances it contains (the 'childrens'). This has nothing to do with 
  subclassing.

<to the op>
You should define a 'container/child' interface. it may be very simple 
in Python, something like

class container:
     def __init__(self):
         self.childrens = []

     def add_child(self, child):
         if child not in self.childrens:
             child.parent = self     # create a 'parent' attribute
             self.childrens.append(child)

     def remove_child(self, child):
          if child in self.childrens:
	    del self.childrens[self.childrens.index(child)]
             del child.parent


Now whatever class your child objects is an instance of, it gets a new 
'parent' attribute when you add it to the container instance, and looses 
this attribute when you remove it...

Well... it could be safer to check if the 'child' object already has a 
parent before you add it, but y'll sure improve all this !-)


Hope this helps

laotseu



From merkosh at hadiko.de  Fri Jun 14 22:12:22 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Sat, 15 Jun 2002 04:12:22 +0200
Subject: file.tell()  --  obscure IOError
Message-ID: <MPG.1774c0f11ff79749989686@news.rz.uni-karlsruhe.de>

hi,

i have an object which conforms to the file protocoll. while calling the 
self.file.tell()   method, i get the following error message:

>>> s1 == s2
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "riff.py", line 153, in __cmp__
    data1 = self.read(self.value.step)
  File "riff.py", line 205, in read
    maxsize = len(self) -self.tell()
  File "riff.py", line 234, in tell
    return self.file.tell() -self.value.start
IOError: (0, 'Error')

when does tell() raise an IOError?

Thanks for any help!
Yours
Uwe


From jepler at unpythonic.net  Mon Jun 24 12:56:21 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Mon, 24 Jun 2002 11:56:21 -0500
Subject: converting an array of chars to a string
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E20192158EF5@admin56.narex.com>
References: <60FB8BB7F0EFC7409B75EEEC13E20192158EF5@admin56.narex.com>
Message-ID: <20020624165620.GD2142@unpythonic.net>

Earlier, JB [mailto:jblazi at hotmail.com] wrote:
> > Thank you very much. In principle, this was exactly, for 
> > what I was looking. But most unforunately if I declare such 
> > an array a, then
> > 
> > a[0]=210
> > a[0] += 46
> > 
> > produces an error instead of the correct result 0. Now I am 
> > a bit desperate.
> 
On Mon, Jun 24, 2002 at 09:59:05AM -0600, Bjorn Pettersen wrote:
> Python helpfully gives you an OverflowError since that's technically
> what it is. Having the value magically truncated would surprise even
> more people. The simple solution is of course:
> 
>   def byteAdd(a, b):
>       return (a + b) % 256
> 
>   a[0] = byteAdd(a[0], 46)
> 
> hth,
> -- bjorn

It seems like a "modulo arithmetic" array type would be useful.  

Hypothetically, 
>>> x = array.array("mB", [210])
>>> x[0] += 47
>>> x[0]
1

Jeff`




From cliechti at gmx.net  Sat Jun  1 15:54:18 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 1 Jun 2002 21:54:18 +0200
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org> <3CF8E382.3A44F0DA@engcorp.com> <slrnafhtdv.407.ak@ak.silmarill.org>
Message-ID: <Xns9220DF3285526cliechtigmxnet@62.2.16.82>

Andrei Kulakov <ak at silmarill.org> wrote in 
news:slrnafhtdv.407.ak at ak.silmarill.org:
> In article <3CF8E382.3A44F0DA at engcorp.com>, Peter Hansen wrote:
>> Andrei Kulakov wrote:
>>> 
>>>     "hello world" type program from wxPython tutorial takes 6
>>> seconds to start on my cel 366 128mb system (running Debian). Is
...
> . if not too much trouble, could you please try this one 
> (from wxpython tutorial)?
> 
> from wxPython.wx import *
> 
> class MyApp(wxApp):
>     def OnInit(self):
>         frame = wxFrame(NULL, -1, "Hello from wxPython")
>         frame.Show(true)
>         self.SetTopWindow(frame)
>         return true

i changed this to "return false"
 
> app = MyApp(0)
> app.MainLoop()

(2nd time, P3-600 Win32)

$ time python news_wxspeedtest.py
OnInit returned FALSE, exiting...
...
real    0m1.382s
user    0m0.020s
sys     0m0.030s

here the slowest part is loading and initilaizing the interpreter.
i think wx windows slows loading down, especialy the first time, because 
its loading the wx library. on win32 this is about 4MB.

chris

-- 
Chris <cliechti at gmx.net>



From pyth at devel.trillke.net  Fri Jun  7 11:33:55 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 7 Jun 2002 17:33:55 +0200
Subject: self
In-Reply-To: <rL3M8.85$Cf4.4420736@newnews.cc.stevens-tech.edu>; from vjovanov@stevens-tech.edu on Fri, Jun 07, 2002 at 10:51:17AM -0400
References: <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu> <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com> <mailman.1023294728.2322.python-list@python.org> <sWsL8.82$v_7.4260962@newnews.cc.stevens-tech.edu> <mailman.1023323409.9676.python-list@python.org> <jpKL8.83$HV1.4324388@newnews.cc.stevens-tech.edu> <mailman.1023409885.8528.python-list@python.org> <rL3M8.85$Cf4.4420736@newnews.cc.stevens-tech.edu>
Message-ID: <20020607173355.C6609@prim.han.de>

Vojin Jovanovic wrote:
> > Why such a
> > harsh reaction without bothering to correct the obvious 'typo'?
> 
> No, it is not the 'typo'.  I corrected that immediately.

ok. sorry for wrong accusations...

> However,  all that aside, the problems is that your code just din't work
> after I corrected those 2 'typos'.  Take a look yourself ....
> 
> Python 2.1.2 (#31, Jan 15 2002, 17:28:11) [MSC 32 bit (Intel)] on win32

hmmm. it seems to be a problem with the python version: 'type(value)!=str' 
comparison doesn't seem to work with python2.1 (works with 2.2).
So here goes my python2.1/2.2 version:

Python 2.1.3 (#1, Apr 20 2002, 10:14:34)
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "copyright", "credits" or "license" for more information.
Welcome to Lazy Python.  Type "help LazyPython" for help.
>>> class Evaluator:
      def __init__(self, eqs={}):
          self.__dict__['equations']=eqs
      def __getattr__(self, name):
          value = self.equations[name]
          if type(value)!=type(''):
              return value
          scope={}
          while 1:
              try:
                  return eval(value, scope)
              except NameError, n:
                  var=str(n).split("'")[1]
                  scope[var]=getattr(self, var)
      def __setattr__(self, name, value):
          self.equations[name]=value

>>> class Berta:
       a=5

>>> e=Evaluator({'a':3})
>>> e.berta=Berta()
>>> e.g='berta.a*a'
>>> e.g
15
>>>

now i hope we are getting close to discussing the idea, finally :-)

regards,

    holger




From larooy at xtar.co.nz  Thu Jun 13 03:46:56 2002
From: larooy at xtar.co.nz (John La Rooy)
Date: Thu, 13 Jun 2002 19:46:56 +1200
Subject: using m2crypto to encrypt credit card numbers
References: <mailman.1023826398.23398.python-list@python.org>
Message-ID: <20020613194656.04643b54.larooy@xtar.co.nz>

Whatever you use to identify the credit card number has to be guaranteed to
be unique for each card number. A hash won't give you that. Perhaps you could
use public key encryption. Encrypt the keys right away on the internet connected
machine using a key pair. The private key is kept on a separate system and
can recover the credit card numbers from the cipher. 

Every purchase made with the card will encrypt the number to the same ciphertext
which you use to identify that card everywhere in your database.

The machine with the private key need not be connected to the internet at all
if you don't mind using sneakernet to transport the transactions to it. 

John

On Tue, 11 Jun 2002 15:12:00 -0500
"Mark McEahern" <marklists at mceahern.com> wrote:

> [Graham Ashton]
> > I've not followed the thread, but I'm assuming that your problem is that
> > you don't want to store the card number in plain text, and that if you
> > encrypt and store it in a manner that will allow automatic decryption by
> > your software, then a cracker who gains access to your servers will be
> > able to decrypt the card numbers with ease.
> 
> That's it precisely.
> 
> > The only sensible solution to this that I've ever thought of involves
> > getting the payment processor (i.e. online transaction processing
> > company) to store a hash for each of your customers' credit cards.
> 
> I should have mentioned that solutions which involve getting the payment
> processor to do anything different are simply not an option.  As far as I
> can tell, I either store the credit card number or I can't do recurring
> billing.  If you know of any payment processors that support recurring
> billing, please share them.
> 
> > Consequently you wouldn't need to store the card number at all, just the
> > encrypted hash. Job's a good'un; card numbers would be nicked from far
> > fewer online web sites with shoddy security.
> 
> Also, another reason I need to store the credit card number is in the case
> of chargebacks, which don't go through the payment processor--rather, they
> go through the bank.  I don't fully understand this part, but I do believe I
> need the credit card number in order to link the chargeback to a
> transaction.
> 
> Thanks for your comments,
> 
> // mark
> 
> -
> 
> 
> 


From shagshag13 at yahoo.fr  Thu Jun  6 14:40:15 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Thu, 6 Jun 2002 20:40:15 +0200
Subject: Efficient python programming...
Message-ID: <adoa9u$vqed$1@ID-146704.news.dfncis.de>

 I'm looking for any state of art, optimizing techniques, tips, that a
python beginner couldn't know by itself, but should learn and
would be aware to write "efficient and good looking"python's code...

 (not the kind of thing which made you an obfuscated python nerd,
but that "map" should be preferred to "for in loop" if possible, that
dictionary are very well optimized, etc.)

Thanks,

s13.







From sholden at holdenweb.com  Fri Jun 28 09:39:25 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 28 Jun 2002 09:39:25 -0400
Subject: SimpleHTTPServer + Multithreading
References: <3d1b4010$1_3@nopics.sjc>
Message-ID: <_EZS8.107502$mF.30395@atlpnn01.usenetserver.com>

"Adonis" <deltapigz at telocity.com> wrote ...
> i was googling around and was able to come up with cannibalized code, and
> was wondering if there was anyway to make it more "clean" looking, or as
it
> is, is the best way:
>
> (main idea: a multithreading SimpleHTTPServer)
>
> import BaseHTTPServer, SimpleHTTPServer, SocketServer, socket, webbrowser
>
> class XFSPWebServer(SocketServer.ThreadingMixIn,
BaseHTTPServer.HTTPServer):
>     pass
>
> class ReqHandler(SimpleHTTPServer.SimpleHTTPRequestHandler):
>
>     def do_POST(self):
>         data = self.rfile.read(int(self.headers["content-length"]))
>         data = data.split('&')[0].split('=')[1]
>         self.wfile.write("HIH")
>
> PORT = 80
>
> httpd = XFSPWebServer((socket.gethostname(), PORT), ReqHandler)
>
> httpd.serve_forever()
>
> any help would greatly be appreciated.
>

Here's the threading CGI server from Chpater 7 of "Python Web Programming".
Replacing CGIHTTPServer with SimpleHTTPServer should give you what you want.
As you can see, the trick is simply to subclass the server you want, making
sure that the ThreadingMixin class is the first base class.

import SocketServer
import BaseHTTPServer
import CGIHTTPServer

class ThreadingCGIServer(SocketServer.ThreadingMixIn,
                   BaseHTTPServer.HTTPServer):
    pass

import sys

server = ThreadingCGIServer(('', 8080), CGIHTTPServer.CGIHTTPRequestHandler)
#
try:
    while 1:
        sys.stdout.flush()
        server.handle_request()
except KeyboardInterrupt:
    print "Finished"

You can get this, and lots more, code at the download page of
http://pydish.holdenweb.com/pwp/

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------





From ark at research.att.com  Fri Jun  7 14:04:41 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 7 Jun 2002 14:04:41 -0400 (EDT)
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <15616.62634.635544.905055@12-248-41-177.client.attbi.com>
	(message from Skip Montanaro on Fri, 7 Jun 2002 13:00:10 -0500)
References: <yu99vg8v6orb.fsf@europa.research.att.com>
        <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
        <200206071747.g57HlbH01057@europa.research.att.com> <15616.62634.635544.905055@12-248-41-177.client.attbi.com>
Message-ID: <200206071804.g57I4fp01212@europa.research.att.com>

Skip> Try running the just built python executable and then manually execute
Skip> "import struct".  What error message(s) do you see?

$ python
Python 2.2.1 (#1, Jun  7 2002, 12:06:38) 
[GCC 3.1] on sunos5
Type "help", "copyright", "credits" or "license" for more information.
>>> import struct
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: ld.so.1: python: fatal: relocation error: file /usr/gnu/lib/python2.2/lib-dynload/struct.so: symbol PyString_Type: referenced symbol not found
>>> 

However, please note that /usr/gnu/lib/python2.2/lib-dynload/struct.so
is left over from the previous installation, because the new one didn't
install a new struct.so at all.  Therefore, I do not think that this
diagnostic messge is relevant.





From gerhard at bigfoot.de  Tue Jun 18 13:50:05 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Tue, 18 Jun 2002 19:50:05 +0200
Subject: pychecker + XEmacs
In-Reply-To: <20020618165548.GB29257@matijek.plusseven.com>
References: <3d0de483$0$3881$e4fe514c@dreader4.news.xs4all.nl> <aekok8$7p118$1@ID-59885.news.dfncis.de> <20020618165548.GB29257@matijek.plusseven.com>
Message-ID: <20020618175005.GA613@lilith.my-fqdn.de>

* Alex Polite <m2 at plusseven.com> [2002-06-18 18:55 +0200]:
> Interesting.  Where is the CVS repository for python-mode.el?

It's in the Python CVS repository. Directory dist/src/Misc/

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From hesterloli at hotmail.com  Mon Jun 17 04:19:24 2002
From: hesterloli at hotmail.com (George Hester)
Date: Mon, 17 Jun 2002 08:19:24 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com>
Message-ID: <gYgP8.42356$GY.12361719@twister.nyroc.rr.com>

Well yours I haven't the slightest idea what you are getting at.

--
George Hester
_________________________________
"Dialtone" <dialtone#/N0SPAM/#@aruba.it> wrote in message
news:3tingu8psfmefoab93pl8gips8hg7sgpvp at 4ax.com...
> On Sat, 15 Jun 2002 22:58:36 GMT, "George Hester"
> <hesterloli at hotmail.com> wrote:
>
> >I think I need a script to do this.
>
> >>> import sys
> >>> sys.version
> '2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)]'
> >>> sys.version_info
> (2, 2, 1, 'final', 0)
> --
> try: troll = "Brain"
> except TypeError, data:
>    troll.plonk()
> Co-Responsabile Hardware Information & Technology http://hit.edengames.net




From shalehperry at attbi.com  Sat Jun 29 17:32:33 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Sat, 29 Jun 2002 14:32:33 -0700 (PDT)
Subject: can't import generators
In-Reply-To: <m2d6u9by8e.fsf@mother.paradise.lost>
Message-ID: <XFMail.20020629143233.shalehperry@attbi.com>

> 
> Any other thoughts?  The only thing slightly non-standard about by
> install is that I compiled python2.2 for large file support with 
> 
> OPT= -ggdb -O2 -Wall -Wstrict-prototypes -D_LARGEFILE_SOURCE
> -D_FILE_OFFSET_BITS=64
> 
> but I can't see that making a difference.
> 

the other thing I see different than most is you used gcc 3.1.  Not that it
should matter either.




From frederic.michoulier at prisme.fr  Tue Jun 11 12:29:13 2002
From: frederic.michoulier at prisme.fr (=?iso-8859-1?Q?Fr=E9d=E9ric_Michoulier?=)
Date: Tue, 11 Jun 2002 18:29:13 +0200
Subject: Hi, this is a test
Message-ID: <00bf01c21165$1f32fec0$4064a8c0@pcfm>

Hi, this is a test
Frederic
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: test.txt
URL: <http://mail.python.org/pipermail/python-list/attachments/20020611/e2cb1566/attachment.txt>

From loewis at informatik.hu-berlin.de  Thu Jun  6 05:33:47 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 06 Jun 2002 11:33:47 +0200
Subject: 64-bit Python?
References: <87elfl3yvu.fsf@mathdogs.com> <C0FL8.41941$n4.9496108@newsc.telia.net>
Message-ID: <j41ybk69ac.fsf@informatik.hu-berlin.de>

"Fredrik Lundh" <fredrik at pythonware.com> writes:

> Python *integer* objects are limited to that size on platforms
> where a C long is 32 bits.

I believe there is another limit: the maximum length of a string,
list, or tuple is 2**31-1 on 64-bit platforms where a C int is 4
bytes. That, in particular, includes mmapped objects as well.

Regards,
Martin


From boud at valdyas.org  Sat Jun  8 15:29:00 2002
From: boud at valdyas.org (Boudewijn Rempt)
Date: Sat, 08 Jun 2002 21:29:00 +0200
Subject: Help with Black Adder...
References: <adpc36$179jq$1@ID-139866.news.dfncis.de>
Message-ID: <3d025e96$0$3887$e4fe514c@dreader4.news.xs4all.nl>

rapskat wrote:

> Greetings All.  I'm just getting started with python and I chose Black
> Adder as a RAD IDE.  Since my background is (ugh!) VB, I'm used to the
> whole Top Down IDE thing.

Well, before I started using Python a few years ago, I was a VB programmer,
too. (And Oracle before that -- I've come a long way, I think). Currently,
I use XEmacs for everything, so the transition can be made, but you need
a strong incentive. For me that incentive was to have to write a book on
PyQt and BlackAdder in docbook sgml. And from using XEmacs for docbook to
using it for everything else (except mail and news) was an insidiuous 
process. I've now redefined KDE to accept ctrl-s for search, and ctrl-x-s 
for save :-).

> 
> My questions are:
> 
> 1) I am having problems with BA not finding the PyQT libs.  I installed
> both from the rpm's (I'm using RH 7.3).  But when I start BA it gives me
> an error that it couldn't find the PyQT modules and that they won't be
> available for the session.  I've tried removing them and reinstalling to
> no avail.  I've run ldconfig as well.  Has anyone ele seen anything like
> this and any ideas on how to resolve it?

I'm not sure that the current version of BlackAdder includes the PyQt 
library -- the commercial version does, I think, but the demo version
doesn't. Or something like that. Check whether all libs are really there;
then check whether LD_LIBRARY_PATH includes the location of the relevant
.so's, whether BLACKADDERDIR is correct, and whether BLACKADDERPYTHON is
set correctly.

(There's a layouto in my book on page 39: the second export ought to be
on the next line.)


-- 
Boudewijn Rempt | http://www.valdyas.org


From look at replyto.address.invalid  Tue Jun  4 01:31:41 2002
From: look at replyto.address.invalid (Greg Ewing)
Date: Tue, 04 Jun 2002 17:31:41 +1200
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <06da01c208f4$d69003c0$6601a8c0@boostconsulting.com> <200206010121.g511LxX19223@pcp742651pcs.reston01.va.comcast.net> <073501c20960$23f367e0$6601a8c0@boostconsulting.com> <200206011334.g51DY1D21669@pcp742651pcs.reston01.va.comcast.net> <07d801c20970$cdebe050$6601a8c0@boostconsulting.com> <mailman.1022940122.3421.python-list@python.org> <Id5K8.171932$Po6.340972@rwcrnsc52.ops.asp.att.net> <oWIK8.179611$%u2.31494@atlpnn01.usenetserver.com> <cHJK8.181202$%u2.93890@atlpnn01.usenetserver.com> <ctKK8.185481$Po6.352104@rwcrnsc52.ops.asp.att.net> <mailman.1023128707.5413.python-list@python.org>
Message-ID: <3CFC50BD.CB9CB70C@replyto.address.invalid>

Jeff Epler wrote:
> 
> Maybe the bytecode for inplace operations like
>     f[i] += v
> should implement the
> pseudocode
>     _temp = f[i]
>     _new = inplace_add(_temp, v)
>     if _new is not _temp:
>         f[i] = _new
>     del _temp, _new

That's a nice idea...

> However, this would probably double the bytecode size of the sequence,
> and hurt performance similarly....

It needn't, if:

(1) the INPLACE_xxx bytecodes leave the first operand
    on the stack

(2) there are new STORE_xxx_IF_UNCHANGED bytecodes that
    compare the value to be stored with the previous one.

The total number of bytecodes should then be the same,
and all the new work would be done in C code, which
ought to be pretty fast (a pointer comparison and
a few extra inc/decrefs).

-- 
Greg Ewing, Computer Science Dept, 
University of Canterbury,	  
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg


From op73418 at mail.telepac.pt  Fri Jun 28 08:34:21 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Fri, 28 Jun 2002 13:34:21 +0100
Subject: Pythoniac: Thoughts on a hardware Python processor
References: <F7DB8D13DB61D511B6FF00B0D0F06233015DE661@mail.hebmex.com> <mailman.1025206656.12053.python-list@python.org>
Message-ID: <1plohucs4iist5ss52957klaf9damhj6kt@4ax.com>

>
>Ahh! Haven't you heard? Forth came in Fifth and the Forth chip is dead. In
>fact I don't think anyone is still offering a stack based computer (only HP
>did AFAIK, the HP 3000).
>
>Dave LeBlanc
>Seattle, WA USA
>

Completely OT but there is that J. Joyce line in Ulysses that goes like

Lazarus, come forth! And he came fifth and lost the job.

best,
Gon?alo Rodrigues


From opus at value.net  Sun Jun 30 03:46:00 2002
From: opus at value.net (Opus)
Date: Sun, 30 Jun 2002 00:46:00 -0700
Subject: squared functions--most Pythonic way?
In-Reply-To: <3D1EB241.FF7ABD7D@cosc.canterbury.ac.nz>
Message-ID: <3D1E54C8.11630.E4859E@localhost>

Shouldn't addNumbers(5) equate to 5?  In other words, it should evaluate that 
as addNumbers(5)(0) or is that addNumbers(0)(5)?

How would you send a list of numbers (or objects that represent numbers) to 
this?

On 30 Jun 2002 at 19:24, greg wrote:

> Janto Dreijer wrote:
> > 
> > def addNumbers(k):
> >     def f(x):
> >         a = addNumbers(x + k)
> >         a.val = x+k
> >         return a
> >     return f
> > 
> > >>> addNumbers(9)(5)(2)(4)(6).val
> > 26
> > 
> > Now if only I could figure out how to use __repr__() so I don't 
> > need that ".val". It also fails when passed only one number. i.e
> > addNumbers(5). Help?
> 
> class AddNumbers:
> 
>   def __init__(self, x):
>     self.val = x
> 
>   def __call__(self, k):
>     return AddNumbers(self.val + k)
> 	
>   def __repr__(self):
>     return repr(self.val)
> 
> addNumbers = AddNumbers(0)
> 
> >>> print addNumbers(9)(5)(2)(4)(6)
> 26
> >>> print addNumbers(5)
> 0
> 
> --
> Greg
> -- 
> http://mail.python.org/mailman/listinfo/python-list


--Opus--

Dictatorship (n): a form of government under which everything 
which is not prohibited is compulsory.

--------------------------------------------------------
Get added to my Humor list:
mailto:opus at value.net?subject=ADD_HUMOR
Get added to my Neat list:
mailto:opus at value.net?subject=ADD_NEAT
Get my PGP public key:
mailto:opus at value.net?subject=PSEND&body=send%20PublicKEY.asc
Visit My Home Page:
http://value.net/~opus/







From whisper at oz.net  Mon Jun 24 18:29:49 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 24 Jun 2002 15:29:49 -0700
Subject: Suggestions for good programming practices?
In-Reply-To: <bko7fa.9t.ln@grey.aerojockey.localdomain>
Message-ID: <GCEDKONBLEFPPADDJCOEIEOBDLAA.whisper@oz.net>

<snip>
>
> * Avoid exec, execfile, eval, and input.

Might one ask why? What do you have to know to use them successfully? Does
"input" imply that "input-raw" is also to be avoided?

Avoid "from foo import *"; this avoids unnecessary namespace pollution and
other forms of confusion.

> * Use local variables whenever possible.  Important in Python because
>   local variables are much faster than global.  In practice, this
>   means that people will often put their main code in a function, and
>   call it with the "if __name__=='__main__': main()" cliche.

Does code run at global scope pay a global scope lookup cost?

And for large loops in defs that use class instance data, copying the data
to a local variable to avoid the lookup cost each time through the loop.
(Does copying globals to local also give a speed gain?)

class myclass():
	def __init__(self):
		self.min = 1
		self.max = 1000
	def loop(self):
		min = self.min
		max = self.max
		for i in range(min,max):
			print i

> * Learn how references and the object system work as soon as possible.
>   You should not be surprised when you do this:
>
>   a = [1,2,3]
>   b = a
>   b[0] = 100
>   print a
>
>   And it prints "[100,2,3]"

The one that always gets me is "lat,lon,timezone = getLocation("seattle")".
I.E. the ability to return multiple distinct values as opposed to returning
multiple values in the form of a struct as in C or C++.

I would add to the list:
* Learn to think in OO and pattern concepts.

> --
> CARL BANKS                                http://www.aerojockey.com
> "Nullum mihi placet tamquam provocatio magna.  Hoc ex eis non est."

Dave LeBlanc
Seattle, WA USA

" Curses! self.bdfl will never work! Back to the drawing board Pinky!"





From maxm at mxm.dk  Tue Jun 25 06:50:12 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 25 Jun 2002 12:50:12 +0200
Subject: Web templating/db tool with best designer/coder separation?
References: <af2gh6$2c2$0@216.39.172.122> <af8dtp$5qp$0@216.39.172.122> <23891c90.0206250232.249c8488@posting.google.com>
Message-ID: <3D184AE4.7070302@mxm.dk>

Paul Boddie wrote:

> This thread wouldn't end if we all started naming template systems,
> though. ;-) For something similar to the jonpy templating, take a look
> at http://www.clearsilver.org - a system which I was made aware of
> recently. If nothing else, that site gives some very good arguments
> against "Python in HTML" systems.


Yes but that system os only where Zope is with dtml!

Zpt has been written to not have the faults of dtml ... so Zpt is pretty 
well thought out for what it is intended to do.

regards Max m



From wurmy at earthlink.net  Sun Jun 30 00:19:10 2002
From: wurmy at earthlink.net (Hans Nowak)
Date: Sun, 30 Jun 2002 04:19:10 GMT
Subject: possibly a dumb question
References: <3d1e5eff$1_7@nopics.sjc> <3D1E6415.8128F3F8@engcorp.com> <3d1e6251$1_3@nopics.sjc>
Message-ID: <3D1E867F.8090001@earthlink.net>

Adonis wrote:
> sure:
> 
> class foo:
>     def __init__(self, value):
>         return value
> x = foo(0)
> print x ;yeilds 0

So you want to have an object, but use it like it has a certain value, like 0.

This is possible, but only to a certain extent. After x = foo(0), x will always 
be a reference to the instance of foo (unles you rebind it of course). It will 
not be a reference to the value you passed to the foo instance.

In the case of printing, you can define a __repr__ or __str__ method that 
return the value rather than the boring <__main__.Foo instance at 0x008F9EC8>:

 >>> class foo:
	def __init__(self, value):
		self.value = value
	def __repr__(self):
		return repr(self.value)
	
 >>> f = foo(0)
 >>> f
0

By defining other magic methods, you can do more, e.g. treat it like f was a 
number, etc. This may, or may not, be what you're looking for.

If not, using f.value rather than f may be your best bet...

HTH,

-- 
Hans (base64.decodestring('d3VybXlAZWFydGhsaW5rLm5ldA=='))
# decode for email address ;-)
The Pythonic Quarter:: http://www.awaretek.com/nowak/



From jhndnn at yahoo.com  Fri Jun 14 12:00:56 2002
From: jhndnn at yahoo.com (John Dunn)
Date: 14 Jun 2002 09:00:56 -0700
Subject: Standalone embedded Win32 Python
Message-ID: <2cc96dc7.0206140800.5614a0bd@posting.google.com>

I have embedded Python in my C++ Win32 app. What is the recommended
way of installation/distribuion? It seems the best would be to
statically link to Python - this would also allow me to customized the
Python environment that runs with my app. Unfortunately, Python
doesn't appear to support static linking on Win32.

I really don't want to run the whole Python installed when I install
my app. If I can't statically link, is it acceptable to just ship my
own local copy of the DLL? I assume I will have to edit some code so
it doesn't look to the registry to find the path to Python.

Is there any documentation on doing this? This seems to be important
to any app that has Python embedded - you can't very well tell your
customers to download the latest Python install just to run your
program.

Thanks for any help-

John Dunn
--
John Dunn
Peak Audio, a division of Cirrus Logic, Inc.
http://www.peakaudio.com


From whisper at oz.net  Mon Jun 24 16:44:04 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 24 Jun 2002 13:44:04 -0700
Subject: Suggestions for good programming practices?
In-Reply-To: <NCBBLFCOHHDIKCAFGCFBIEFDLGAA.marklists@mceahern.com>
Message-ID: <GCEDKONBLEFPPADDJCOEKENIDLAA.whisper@oz.net>


David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Mark McEahern
> Sent: Monday, June 24, 2002 13:22
> To: python-list at python.org
> Subject: RE: Suggestions for good programming practices?
>
>
> Dianne van Dulken wrote:
> > I'm fairly new to python, coming from a perl background, and
> was wondering
> > if anyone has a list of things that they consider as good programming
> > practice in Python (I thought this might be an interesting topic for
> > discussion, anyway)
>
> Here's one...
>
> When evaluating whether a variable is None, don't compare it like this:
>
> 	if x == None:
>
> instead, use:
>
> 	if x:
>
> or:
>
> 	if not x:
>
> The encourages a polymorphic approach.  If you must compare to
> None, use the
> identify operator rather than equality:
>
> 	if x is None:
>
> or:
>
> 	if x is not None:
>
> // mark
>

I'm curious to know how this encourages polymorphism, and how making a
conditional test implicit rather then explicit is good practice?

Regards,

Dave LeBlanc
Seattle, WA USA





From donn at u.washington.edu  Tue Jun  4 20:06:22 2002
From: donn at u.washington.edu (Donn Cave)
Date: 5 Jun 2002 00:06:22 GMT
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <3CFCF3E7.AE891EBE@noaa.gov> <adiumj$22g2$1@nntp6.u.washington.edu> <adjduf$92g$0@216.39.172.122>
Message-ID: <adjklu$18ra$1@nntp6.u.washington.edu>

Quoth bokr at oz.net (Bengt Richter):
...
| Here is a contrived case where += lets you use a
| mutable-lhs-target-expression-with-side-effect and
| get only one side effect. There are obviously other
| ways, but at least += is convenient (and presumably
| faster in a case like this).
|
|  >>> a=[0]
|  >>> def foo():
|  ...     print 'side effect'
|  ...     return a
|  ...
|  >>> foo()
|  side effect
|  [0]
|  >>> foo()[0]+=1
|  side effect
|  >>> a
|  [1]
|  >>> foo()[0] = foo()[0]+1
|  side effect
|  side effect
|  >>> a
|  [2]

I imagine you have some actual application for something like this,
but the example doesn't seem very compelling - like you say, there
are obviously other ways.  If they're less convenient, maybe they're
easier on the eyes.

But it's probably a moot point.  Whether the current behavior
of += is a wart or not, it's not likely to be removed.

	Donn Cave, donn at u.washingon.edu


From noone at here.invalid  Sun Jun 23 01:05:45 2002
From: noone at here.invalid (JohnJacob)
Date: Sun, 23 Jun 2002 00:05:45 -0500
Subject: ConfigParser & converting strings to lists
References: <3D152A28.2177B054@tds.net>
Message-ID: <af3kv701tsj@enews3.newsguy.com>

> files = config.get("recent files", "recentFiles")
>
> sets files to the _string_ "[file1, file2, ...]".  I suppose this must
> be how get should work; after all, settings could be arbitrary
> strings...

You just want to get your list back from it's string representaion, right?
Try eval:

files = eval(config.get("recent files", "recentFiles"))



I hope that's what you meant.

greg





From david.abrahams at rcn.com  Tue Jun 18 18:04:08 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Tue, 18 Jun 2002 18:04:08 -0400
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de> <aeftmj$m29$1@bob.news.rcn.net> <3d0c6d67@nntp.server.uni-frankfurt.de> <3D0CE698.1050806@SPAMnwinternet.com> <aeiu3a$4ff$1@bob.news.rcn.net> <3D0D3223.5070500@SPAMnwinternet.com> <aekrd1$4ao$1@bob.news.rcn.net> <3d0f9f70@nntp.server.uni-frankfurt.de>
Message-ID: <aeoavo$4f5$1@bob.news.rcn.net>

"Michael 'Mickey' Lauer" <mickey at tm.informatik.uni-frankfurt.de> wrote in
message news:3d0f9f70 at nntp.server.uni-frankfurt.de...
> David Abrahams <david.abrahams at rcn.com> wrote:
> > Heh, yeah. Unfortunately, documentation is often harder than coding.
Even
> > the current official release of Boost.Python is not very
well-documented.
> > We're working hard to correct that for v2, which is in pre-release.
>
> Is it completed enough to try it out for wrapping a small library?

Absolutely. People are already using it for big projects.
http://mail.python.org/pipermail/c++-sig/2002-May/001100.html to get started

> If not,
> could you estimate when it will be ?

I hope to officially retire v1 at the end of the month, replacing it with
v2.

-Dave




From gabor at realtime.sk  Thu Jun  6 13:54:47 2002
From: gabor at realtime.sk (gabor)
Date: 06 Jun 2002 13:54:47 -0400
Subject: array, list, performance...
Message-ID: <1023386087.3833.17.camel@wintermute.atriaky.sk>

hi,

how fast is list?

i mean accessing an item at an arbitrary position is O[n] or O[1]?

i need O[1].... are there any suitable classes in python?

thanks,
gabor







From chris.gonnerman at newcenturycomputers.net  Tue Jun 25 22:38:54 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Tue, 25 Jun 2002 21:38:54 -0500
Subject: Reminder of a python Project.
References: <2218.941T781T12224051threeseas@earthlink.net>
Message-ID: <007701c21cba$a4117ee0$0101010a@local>

----- Original Message -----
From: "Timothy Rue" <threeseas at earthlink.net>


> First off, it's ok that I got such a high level of resistance to my
> initial posting here about this project. Thanks to some research that /.
> had an article on, regarding the size of GPL project teams.
>
> But anyway.... Remember the autocoding project I mentioned?

Gah.  I thought I had this guy in my message rules under "Delete from
Server."

I will remedy this immediately...

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net





From peter at engcorp.com  Fri Jun 21 09:38:45 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 21 Jun 2002 09:38:45 -0400
Subject: [newbie] import fails first time, then works :-[
References: <3D1379D1.5020705@nospam.free.fr>
Message-ID: <3D132C65.34B38F8C@engcorp.com>

laotseu wrote:
> 
> Hi Everybody
> 
> I've got this strange problem :
> when importing modules that wrap external modules (like socket and
> _socket etc), the first time, I get an ImportError
> (no module named _[module]).
> If i try a second time, it works ok ?-o
> 
> Exemple :
>  >>> import socket
> Traceback (most recent call last):
>    File "<stdin>", line 1, in ?
>    File "/usr/local/lib/python2.2/socket.py", line 41, in ?
>      from _socket import *
> ImportError: No module named _socket
>  >>> import socket
>  >>> socket
> <module 'socket' from '/usr/local/lib/python2.2/socket.pyc'>

See recent postings in the archives for explanations about why this
is actually _not_ working the second time, but only appears to.

In this case, despite that you say the external modules are in the
right place, there is probably something wrong with the _socket
external module or it cannot be found and that's the root cause
of your trouble.

-Peter


From gerhard at bigfoot.de  Fri Jun 21 10:05:53 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 21 Jun 2002 14:05:53 GMT
Subject: a Tree data-structure class?
References: <b16e4ef7.0206210510.77a956b6@posting.google.com>
Message-ID: <slrnah6chh.154.gerhard@lilith.my-fqdn.de>

Stephen Ferg wrote:
> I do a lot of processing of tree-structured data, so I'm thinking it
> would be useful to develop a generic Tree class that I could adapt to
> various applications.  Or perhaps a Tree class and a generic Node
> class.

I've once quickly hacked this as a demonstration (just skip the German):
http://starship.python.net/pipermail/python-de/2001q4/003178.html

Btw. Zope CVS contains highly optimized BTree classes.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From nessus at mit.edu  Tue Jun  4 16:05:19 2002
From: nessus at mit.edu (Douglas Alan)
Date: 04 Jun 2002 16:05:19 -0400
Subject: manipulating child processes under Windows
References: <lcptz7zw0m.fsf@gaffa.mit.edu> <uadqb3q9v.fsf@online.no>
Message-ID: <lcy9duhksg.fsf@gaffa.mit.edu>

Syver Enstad <syver-en+usenet at online.no> writes:

> Hmm, my first reaction is to think in terms of window handles in
> some way, there's a FindWindow function in win32 (and win32all) that
> let's you search for a window by name, I'll have to check the msdn
> library to find out how to get the owning process handle from a
> window handle.

What if there were more than one process with the same name.  How
would I tell which one is the right one?

|>oug


From flynt at gmx.ch  Fri Jun  7 11:50:18 2002
From: flynt at gmx.ch (flynt)
Date: Fri, 07 Jun 2002 17:50:18 +0200
Subject: Session Cookies
References: <000d01c20e38$da94fdb0$4100a8c0@KARIMXP1>
Message-ID: <3D00D63A.51940D20@gmx.ch>

Karim Shehadeh wrote:
> 
>    Part 1.1    Type: Plain Text (text/plain)
>            Encoding: 7bit


Hello Karim

For an example look into the *New Riders* book:
"Python Web Programming" by Steve Holden and David Beazley
ISBN: 0-7357-1090-2
you find an example of a web-site framework based on python. They call
it *AWeFUL*; chapter 17 in that book.
Session handling with cookies is part of it.

Greetings
Flynt




From garrett at bgb.cc  Thu Jun  6 16:48:35 2002
From: garrett at bgb.cc (Don Garrett)
Date: Thu, 06 Jun 2002 20:48:35 GMT
Subject: constructive critisism on pythonwin IDE
References: <6bd9f01b.0206050736.f469082@posting.google.com> <admuke$gu8$1@newsreaderg1.core.theplanet.net>
Message-ID: <3CFFCA79.4050905@bgb.cc>

   Thomas Heller wrote a wonderful script named 'autoreload' that helps with 
this problem quite a bit. I don't think this is an official distribution 
location (and thus not current?), but I found a link to it here:

   http://starship.python.net/crew/zack/

F. GEIGER wrote:

> This a Python issue, not a PythonWin IDE issue.
> 
> Cheers
> Franz
> 
> "sameer" <sameer_ at email.com> schrieb im Newsbeitrag
> news:6bd9f01b.0206050736.f469082 at posting.google.com...
> 
>>First of all, let me start of by saying that Pythonwin is a great IDE.
>> Now for the constructive critisism.  The thing I hate about PythonWin
>>is, that despite the fact that it's a development environment, there
>>is a lot of caching done of modules, and the cached copy is not
>>checked against a potentialy modified copy.  Whenever I change the
>>contents of a file and try to run a module that depends on that file,
>>I always end up with the old copy.  I have to then import and then
>>reload the changed module in the interactive window for it to refresh.
>> It would greatly increase my productivity, if I can specify which
>>modules can be cached.  I understand that there is a speed and
>>development time issue, but I would like a little more fine grained
>>control on how caching works on PythonWin, as the way it's currently
>>implemented now is not acceptable.
>>
> 
> 



From never at mind.info  Sun Jun  2 17:32:43 2002
From: never at mind.info (G. Willoughby)
Date: Sun, 2 Jun 2002 22:32:43 +0100
Subject: HTML links in a GUI tkinter
References: <mailman.1022986323.16196.python-list@python.org>
Message-ID: <ade2tq$3q4$1@newsg1.svr.pol.co.uk>

> How can I put in a frame or in any widget (?) an html link so that the
> browser opens at the url indicated when the user clicks ?

I think you can use a Tkinter Text widget, and then specify tags around the
web address:
http://www.pythonware.com/library/tkinter/introduction/x7883-concepts.htm

then:

import os
os.startfile(page.html)

to execute the html file.

--G. Willoughby






From marklists at mceahern.com  Mon Jun 24 16:22:01 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Mon, 24 Jun 2002 15:22:01 -0500
Subject: Suggestions for good programming practices?
In-Reply-To: <bko7fa.9t.ln@grey.aerojockey.localdomain>
Message-ID: <NCBBLFCOHHDIKCAFGCFBIEFDLGAA.marklists@mceahern.com>

Dianne van Dulken wrote:
> I'm fairly new to python, coming from a perl background, and was wondering
> if anyone has a list of things that they consider as good programming
> practice in Python (I thought this might be an interesting topic for
> discussion, anyway)

Here's one...

When evaluating whether a variable is None, don't compare it like this:

	if x == None:

instead, use:

	if x:

or:

	if not x:

The encourages a polymorphic approach.  If you must compare to None, use the
identify operator rather than equality:

	if x is None:

or:

	if x is not None:

// mark

-





From gh_pythonlist at gmx.de  Tue Jun 11 19:37:40 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 12 Jun 2002 01:37:40 +0200
Subject: RfD: Automagically making a class instance 'synchronized' (draft impl.)
In-Reply-To: <20020612000425.U6609@prim.han.de>
References: <20020611205423.GA28113@lilith.my-fqdn.de> <20020612000425.U6609@prim.han.de>
Message-ID: <20020611233740.GA28544@lilith.my-fqdn.de>

* holger krekel <pyth at devel.trillke.net> [2002-06-12 00:04 +0200]:
> Gerhard H?ring wrote:
> > For one of my projects, I'll need something not quite unlike the
> > following code. It's not tested, yet, but it looks ok and complete to
> > me. I thought that it could be useful to other developers who need to
> > use multithreading, so I'm posting it here. Maybe something like this
> > would even be a useful addition to the Standard Library?
> 
> note that Ian Kjos some weeks ago send a nice module 'morethreading.py'
> to the dev-list. IIRC, it still waits for beeing reviewed and extended
> with other ideas (like synchronized methods). 

Sounds interesting, I'll have a look at it. I also just found this
recipe: http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/65202
which has a few more interesting additions. It's possible to do without
my methodlist, for example.

> [...] this can be safely substituted and then even returns the correct value<wink> with
> try: ... finally:

Whoops. I missed that originally, thanks.

Btw. I've only recently started to seriously experiment with
multithreading. Have you read my recent post

Message-ID: <slrnafpeei.p8.gerhard at developer2.opus-gmbh.net>
with the title "Need advice on multithreading problem"

? I find it really nifty. It's about wrapping a C library function that
uses callbacks to Python generators. I'd welcome any comments on the
approach I've taken there (Queue.Queue, a deferred producer thread with
the generator in the main thread being the consumer).

Unfortunately, this opened a whole can of worms, which I think I solved
now. Now I'm trying to solve them more elegantly.

If anybody wants to try this out in real life, it's in the PySQLite CVS
as the alternative IterCursor implementation.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 14.7 ?C      Wind: 2.6 m/s




From oren-py-l at hishome.net  Sun Jun 30 14:01:39 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Sun, 30 Jun 2002 14:01:39 -0400
Subject: yield equivalent in C/JavaScript?
In-Reply-To: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk>
References: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk>
Message-ID: <20020630180139.GA38821@hishome.net>

On Sat, Jun 29, 2002 at 07:12:46PM +0100, Robin Becker wrote:
> Is there any way to do yielding easily in C? I have a bunch of programs
> which are normally connected together using unix pipes. I would like to
> merge them and the problem is then how to get them to communicate nicely
> using a single process.

It's possible to implement this in C using Duff's Device:

#define BEGIN_GENERATOR static int _GenState=0; \
  switch(_GenState) { \
    case 0:

#define END_GENERATOR } \
  _GenState = 0;

#define YIELD(x) _GenState=__line__; return (x); case __line__:


int generator(arguments) {

  static int foo, bar, etc;	/* Make all variables static */

  BEGIN_GENERATOR 
  for(...) {
      if(...) {
        ...
        YIELD(x)
        ...
      } else {
        ...
      }
      ...
      YIELD(y)
      ...
   }
   END_GENERATOR
}

If you want this to be reentrant use a struct passed as an argument
to the function instead of static variables.

	Oren





From jolsen at mailme.dk  Sun Jun  2 08:23:30 2002
From: jolsen at mailme.dk (Jesper Olsen)
Date: 2 Jun 2002 05:23:30 -0700
Subject: How to play a .wav sound?
References: <ad8jo7$ohl$1@wanadoo.fr>
Message-ID: <cf0ad9fb.0206020423.7dd2de36@posting.google.com>

I see that your question has already been answered.

An alternative solution is to install the snack toolkit:

http://www.speech.kth.se/snack/

and from python do this:

import tkSnack
tkSnack.initializeSnack(root)
sound=tkSnack.Sound(load="file.wav")
sound.play()

Cheers
Jesper




Bruno Bellamy <bellamy at neverland.net> wrote in message news:<ad8jo7$ohl$1 at wanadoo.fr>...
> I guess the question has probably been asked many times, sorry for 
> bothering...
> But I searched here and there, and I couldn't find how to do that, simply 
> play a .wav sound from a python program using Tkinter.
> 
> Moreover, I need to do that in a program that would run similarly under 
> Linux and Windows.
> 
> I saw there's a python module called wave, but it seems it can only read 
> .wav files. I coulnd't find a way to actually play the sound. Maybe it's 
> hidden somewhere?
> 
> If anybody can help me, that would be cool. :)


From irmen at NOSPAMREMOVETHISxs4all.nl  Mon Jun 10 14:16:55 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Mon, 10 Jun 2002 20:16:55 +0200
Subject: M2Crypto: select() behaves weird on SSL sockets
Message-ID: <3D04ED17.5070606@NOSPAMREMOVETHISxs4all.nl>

Hi
I'm using m2crypto 0.07 snapshot 3 with Python 2.2.1
and I'm experiencing weird behavior when using select().

Basically, it seems that a SSL socket is no longer in ready
state when entering select() again, after reading *a part*
of the available data on the socket.

Using plain sockets, the select() will keep reporting that
the socket is ready for reading, until all data has been read.
With SSL sockets, the second select() blocks eternally...

Any clue? I really need the regular select() behavior,
also on SSL sockets...


Thanks,
Irmen de Jong

(test code available on request.)



From fcordonier at myrealbox.com  Fri Jun 21 04:22:13 2002
From: fcordonier at myrealbox.com (=?ISO-8859-15?Q?Fr=E9d=E9ric_Cordonier?=)
Date: Fri, 21 Jun 2002 10:22:13 +0200
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <20020619154654.C619@ActiveState.com> <20020619225800.GB12907@lilith.my-fqdn.de>
Message-ID: <3D12E235.6070304@myrealbox.com>

Gerhard H?ring wrote:

>* Trent Mick <trentm at ActiveState.com> [2002-06-19 15:46 -0700]:
>  
>
>>My understanding is that for C extensions to work they must be built
>>with the same compiler as that used to build the Python with which they
>>will run. (I may be wrong. Someone please correct me if I am.)
>>    
>>
>
>You are wrong. As long as you link to the python{major}{minor}.dll you
>are binary compatible, no matter which compiler you use. It's just a
>matter of providing an additional compiler option to distutils, like in:
>
>    python setup.py build --compiler=mingw32
>
>Gerhard
>  
>
Hello,

I am no windows or C++ specialist, but I happened to write extensions 
for Python in the scope of unitary testing code for embedded systems.

I did extensions with both Borland C++ 5.01 and mingw32. Under windows, 
it is always easier to link the project with a .lib file rather than 
directly connecting to the .dll. The problem with Borland is that the 
.lib for Pythonxx.dll is tricky to generate. In fact, Borland adds some 
'_' to the exported symbols.  You have to:
 > impdef Python22.def Python22.dll

modify the Python22.def to add leading '_' to the symbols (I use gvim 
that is really handy for this kind of tasks).

then
 > implib Python22.lib Python22.def

For Mingw32, the .lib file can be generated with 'pexports' and 'dlltool'.

Moreover, I wrote extensions with SWIG using Borland. One problem I 
encountered is that the main module entry is not exported correctly 
(when importing the module, an exception is raised), because SWIGEXPORT 
is not adapted (maybe specific to BC++5.01?).
You have to redefine it to
#define SWIGEXPORT(a)   a _export FAR PASCAL

Last, using distutil with Borland is certainly possible, but not with 
version 5.01 (my customer only has this version). Some linker options 
used with '--compler=bcpp' are not known by 5.01. I had to patch 
distutils to make it work. If ever someone needs it, I will go back to 
my notebook to check what was done...

Fred.






From juanm.casillas at eresmas.com  Thu Jun 27 10:06:59 2002
From: juanm.casillas at eresmas.com (Juan M. Casillas)
Date: Thu, 27 Jun 2002 16:06:59 +0200
Subject: Module Help
In-Reply-To: <3D1B0FA6.9080101@thomas-guettler.de>
References: <mailman.1025164967.15063.python-list@python.org>
	<3D1B0FA6.9080101@thomas-guettler.de>
Message-ID: <15643.7171.903855.772185@eresmas.com>

>>>>> "Thomas" == Thomas Guettler <zopestoller at thomas-guettler.de> writes:

    Thomas> Juan M. Casillas wrote:
    >> Hello!
    >> 
    >> I want to create an empty module from C, and then add functions
    >> and definitions stored in another file, and some code that I
    >> will create on-the-fly, from C.


    Thomas> I don't understand you. Want to you mean with "from C"?
    Thomas> Why not use python for code generating?

    Hello!

	I have a C program that gets a file that gets some code
	from various places; then I want to create different
	execution environments, and the way I think I can do it
	is creating various modules on the fly, and executing
	code from them. 


    Kind Regards,
    Juan M. Casillas



    >> The basic problem I have is that I don't know how to create the
    >> empty module (this module doesn't have any representation in
    >> the filesystem) and then how can I add the code and save the
    >> "environment" for this newly created module, how can I manage
    >> the globals and locals dictionaries ? for each module ?

    Thomas> Do you want to create a text file containing python code,
    Thomas> or create it at runtime dynamically?

    Thomas> thomas

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





From pyth at devel.trillke.net  Tue Jun 11 16:28:44 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 11 Jun 2002 22:28:44 +0200
Subject: Why does Python mix OO concepts and non OO concepts for operation s  on basic types?
In-Reply-To: <Pine.LNX.4.33.0206120008180.9668-100000@rnd.onego.ru>; from rnd@onego.ru on Wed, Jun 12, 2002 at 12:20:00AM +0400
References: <3D054A98.3090708@bgb.cc> <Pine.LNX.4.33.0206120008180.9668-100000@rnd.onego.ru>
Message-ID: <20020611222844.T6609@prim.han.de>

Roman Suzi wrote:
> I only forgot how will I do this:
> 
> map(string.split, s)
> 
> if GvR will deprecate string module...

what about

    map(str.split, s) 

:-) holger




From op73418 at mail.telepac.pt  Wed Jun 26 14:03:08 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 26 Jun 2002 19:03:08 +0100
Subject: question od default args
References: <o3vjhuguen6gfm8vnq7rbk5volmi80shl5@4ax.com> <3D19FEA3.6020804@earthlink.net>
Message-ID: <ta0khusfa3q2g9lj0aglrutrahf9o8a931@4ax.com>

On Wed, 26 Jun 2002 17:50:42 GMT, Hans Nowak <wurmy at earthlink.net>
wrote:

>Gon?alo Rodrigues wrote:
>> Hi,
>> 
>> When I want default args I usually do
>> 
>> def (arg = None):
>>     ...
>> 
>> But what if I want to make None a reasonable argument too? That is, I
>> want to know if the user has in fact passed an argument, and if not to
>> do something special, with None (or whatever object) being a reasonable
>> arg to be passed.
>
>Hmm, so use a default argument that is not reasonable to pass... some weird 
>string, or a complex number, or even an instance of a (singleton?) class 
>written especially for this... although that may be a bit overkill. Of course, 
>the question is what is "reasonable" here and what not...

In principle everything is reasonable. But (some variation of) "an
instance of a (singleton?) class written especially for this" is worth a
try.

Thanx,
Gon?alo Rodrigues



From fredrik at pythonware.com  Sat Jun 29 15:05:18 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 29 Jun 2002 19:05:18 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com> <mailman.1025336262.12245.python-list@python.org> <3d1dc547.468619140@netnews.attbi.com> <afkv73$f2274$1@ID-139865.news.dfncis.de>
Message-ID: <OxnT8.45616$n4.10835602@newsc.telia.net>

Dave Kuhlman wrote:

> > 1) "import sys" became "import exit from sys"
> > 2) "import os" became "import makedires from os"
>
> Does importing a function from a module (as opposed to importing
> the whole module) save a look-up each time the function is called?

no.

but evaluating "exit" instead of "sys.exit" saves one lookup.

(the latter first looks up "sys" in the global namespace, and
then "exit" in the sys namespace).

(not that it matters for these functions; it didn't sound like
his performance problem was that Python couldn't exit fast
enough ;-)

> > 3) "import time" bacame "import asctime from time" (this one is
> > used frequently)

"from time import asctime", I hope.

> Is finding things in globals slower than finding them in locals?

yes.  Python assigns integer indices to local names, and
stores the corresponding objects in an array.

globals (and ordinary object attributes) are stored in a
dictionary.

if you read enough code, you'll find that "rebinding" is a pretty
common pydiom in performance-critical code.  some examples:

    import time

    def myfunction(...):

        # bind asctime to a local name
        asctime = time.asctime

        for ... in ...:
            x = asctime(...)

this also works for methods:

    def myotherfunction(...):
        result = []
        append = result.append
        for .. in ...:
            append(...) # appends to result
        return result

since arguments are also treated as local variables, and
default values are evaluated once (when the function
object is created), you'll also find things like:

    def myfunction(..., asctime=time.asctime):
        for ... in ...:
            x = asctime(...)

> Did this change in Python 2.2?

no.

</F>




From hst at empolis.co.uk  Tue Jun 11 12:05:03 2002
From: hst at empolis.co.uk (Harvey Thomas)
Date: Tue, 11 Jun 2002 17:05:03 +0100
Subject: Cross platform spawning
Message-ID: <8FC4E7C302A6A64AAD5DB1FA0E825DEB220BF9@hendrix.empolisuk.com>

Peter Scott wrote:
> 
> I am trying to make a script spawn a seperate process, like you would
> do like this on unix:
> 
> os.system('python myscript.py myargument &')
> 
> But the & at the end (to make a separate process) doesn't work on
> Windows. I want the script to run on both Linux (*BSD, whatever) and
> Windows. I didn't want to have to find out the path to the python
> interpreter, so os.spawnlp (with no blocking) sounded perfect until I
> saw that it wasn't available on Windows.
> 
> Does anyone out there know what I can do in this situation?

Not sure if it's exactly what you want but under Win NT you could use

os.system('start python myscript.py myargument')

(see the NT help for details of the start command)

_____________________________________________________________________
This message has been checked for all known viruses by the MessageLabs Virus Scanning Service.




From edream at tds.net  Fri Jun 21 09:43:26 2002
From: edream at tds.net (Edward K. Ream)
Date: Fri, 21 Jun 2002 13:43:26 GMT
Subject: Inhibiting duplicate apps in Windows?
References: <3D11F9E8.DFFFC8E4@tds.net>
Message-ID: <3D132D7E.35CBF952@tds.net>

> What I would like is for the second copy of Leo to detect that another
> copy of leo.py is already running, send a message to the first copy and
> then exit. I wonder if anyone knows how this might be done?

This morning I realize that a workaround would be to implement a "Recent
Files" menu. This would be especially useful in Leo, as .leo files
typically represent many other files.  This doesn't solve the stated
problem, and will be good enough in practice.

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


From BPettersen at NAREX.com  Mon Jun 10 19:00:30 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 10 Jun 2002 17:00:30 -0600
Subject: Why does Python mix OO concepts and non OO concepts for operation  s on basic types?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158EBD@admin56.narex.com>

> From: James T. Dennis [mailto:jadestar at idiom.com] 
> 

[snip]

> 
> 	Personally I think it would be generally nice if I could
> 	call most of the list and string methods in both ways:
> 
> 	list.append(somelist,someitem) would be the same as:
> 	somelist.append(someitem).
> 
> 	And the obvious:
> 
> 	string.join('',something) would be as: ''.join(something).

Your wish is granted:

>>> x = []
>>> list.append(x, 5)
>>> x
[5]
>>> str.join(' ', ['a', 'b', 'c'])
'a b c'
>>>

not-that-I-would-ever-do-something-like-that'ly y'rs
-- bjorn




From mjackson at wrc.xerox.com  Tue Jun 18 07:49:38 2002
From: mjackson at wrc.xerox.com (Mark Jackson)
Date: 18 Jun 2002 11:49:38 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com> <gYgP8.42356$GY.12361719@twister.nyroc.rr.com> <aekcug$7js8n$1@ID-114079.news.dfncis.de> <9YBP8.44418$GY.13865843@twister.nyroc.rr.com> <Xns92317F85D48EDduncanrcpcouk@127.0.0.1>
Message-ID: <aen6oi$fpj$1@news.wrc.xerox.com>

Duncan Booth <duncan at NOSPAMrcp.co.uk> writes:
> "George Hester" <hesterloli at hotmail.com> wrote in 
> news:9YBP8.44418$GY.13865843 at twister.nyroc.rr.com:
> 
> > I think you guys don't how to do it.
> 
> And I think you are one of the most impolite posters we have seen around 
> here.
> 
> Try this:

[snip]

Slipping George the old "reformats your hard drive" script - that was
*mean*.

Fair, but mean.

-- 
Mark Jackson - http://www.alumni.caltech.edu/~mjackson
	Until we test our beliefs, we can't say for sure if we have
	leeches or we have aspirin.	- David Faigman




From logiplexsoftware at earthlink.net  Thu Jun 13 04:00:51 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: 13 Jun 2002 01:00:51 -0700
Subject: How to call a function using apply with keyword args?
In-Reply-To: <j4ofef7h2v.fsf@informatik.hu-berlin.de>
References: <ae9fud$u23$00$1@news.t-online.com> 
	<j4ofef7h2v.fsf@informatik.hu-berlin.de>
Message-ID: <1023955252.7453.15.camel@localhost.localdomain>

On Thu, 2002-06-13 at 00:50, Martin v. L?wis wrote:
> "Achim Domma" <achim.domma at syynx.de> writes:
> 
> > I have a function like this:
> > 
> > def MyFkt(A,B,C): pass
> > 
> > and some parameters from a config file:
> > 
> > params = {'A':5,'B':8,'C':9}
> > 
> > how can I call MyFkt with this parameters? 
> 
> I recommend
> 
> MyFkt(**params)
> 
> > I tried apply(MyFkt,[],params) 
> 
> If you want to use apply, it is
> 
> apply(MyFkt, (), params)
> 
> The arguments need to be a tuple.

Actually, a list will work as well:

Python 2.1.1 (#1, Aug 13 2001, 19:37:40) 
[GCC 2.96 20000731 (Red Hat Linux 7.1 2.96-96)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> def MyFkt(A,B,C): 
...     print A, B, C
... 
>>> params = {'A':5,'B':8,'C':9}
>>> apply(MyFkt,(),params)
5 8 9
>>> apply(MyFkt,[],params)
5 8 9
>>> 







From ryanmorillo at hotmail.com  Tue Jun 18 06:44:59 2002
From: ryanmorillo at hotmail.com (Ryan)
Date: 18 Jun 2002 03:44:59 -0700
Subject: tray icon using tkinter
References: <mailman.1023950718.20009.python-list@python.org> <e08b1d4.0206171530.563e67fd@posting.google.com>
Message-ID: <4a7fde99.0206180244.11eea205@posting.google.com>

john at johnnypops.demon.co.uk (John Popplewell) wrote in message news:
> Anoop P B <anoop79 at myrealbox.com> wrote in message news:
> > 	is there a way that i can provide a tray icon ie. next to the clock (on 
> > win32) for my program (using tkinter) ? i know that there is a way of 
> > doing this in wxpython but i need to know whether its possible in tkinter.
> > 
> > thanks,
> > anoop
> 
> 
> Have you had a look at "Tray It!" ?
> 
> http://www.teamcti.com/trayit/index.html
> 
> cheers,
> John.

also look at the win32 extensions, I remember going through all the
demo code that there is a way to do it (MFC I think but don't know for
sure).  It is simple to do a try except block around the import and
logic to use it since it will only be for windows.  It is used for the
windows IDE also.
  Also check out TKinter3000 at 
http://www.pythonware.com/products/tkinter/tkinter3000.htm, they may
have something (or you can ship the one from win32 extensions with
your app)


From rpm1deletethis at nospamfrontiernet.net  Sun Jun  2 08:19:38 2002
From: rpm1deletethis at nospamfrontiernet.net (RPM1)
Date: Sun, 2 Jun 2002 08:19:38 -0400
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org>
Message-ID: <ufk3bdnh675sc1@corp.supernews.com>

"Andrei Kulakov" <ak at silmarill.org> wrote
> Hello,
>
>     "hello world" type program from wxPython tutorial takes 6
> seconds to start on my cel 366 128mb system (running Debian). Is
> this typical? Why is it so slow? I'm looking for a graphic
> toolkit that'd be fairly quick. I don't want to use Tk because of
> the looks.

I found wxPython to be slow also.  I wrote a program to syntax
highlight a specific type of log file with Tkinter.  I thought wxPython
would look better and it has more controls available so I ported it
to wxPython.  The performance was so bad I went back to Tkinter.

For example, loading a 1MB file into the text box and syntax highlighting
it in Tkinter takes about 2 seconds.  In wxPython it took more than
30 seconds.

Patrick M.





From marklists at mceahern.com  Tue Jun 11 09:05:17 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Tue, 11 Jun 2002 08:05:17 -0500
Subject: using m2crypto to encrypt credit card numbers
In-Reply-To: <4620daca.0206110429.1417b389@posting.google.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBOEPLLDAA.marklists@mceahern.com>

[gyromagnetic]
> Managing keys is a fundamental problem with encryption. I would highly
> recommend that you not store the credit card numbers at all. Instead,
> store a (SHA, MD5) hash of the number, and then validate against the
> hash.

I'm not using the credit card number to validate--that's what the password
is for.  (I don't store the password, but a hash of it.)  I'm using the
credit card number to collect payment.

If I don't store the credit card number, how do I send it to the payment
processor to collect payment?

// mark

-





From aahz at pythoncraft.com  Thu Jun 27 12:59:53 2002
From: aahz at pythoncraft.com (Aahz)
Date: 27 Jun 2002 12:59:53 -0400
Subject: doc for threading?
References: <mailman.1025188254.19336.python-list@python.org>
Message-ID: <affga9$lfe$1@panix1.panix.com>

In article <mailman.1025188254.19336.python-list at python.org>,
Klaus Reinhardt  <K.Rdt at TU-Berlin.DE> wrote:
>
>Is there a more elaborated doc (with examples)
>for threading than the distribution is providing?

See my web page.  Still not great, but the examples should be useful.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From pyth at devel.trillke.net  Wed Jun  5 20:41:12 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 6 Jun 2002 02:41:12 +0200
Subject: self
In-Reply-To: <926F937512224245A616323693D3F16B1C01EF@ctmsg01.corp.fitlinxx.com>; from db3l@fitlinxx.com on Wed, Jun 05, 2002 at 08:19:05PM -0400
References: <926F937512224245A616323693D3F16B1C01EF@ctmsg01.corp.fitlinxx.com>
Message-ID: <20020606024112.G30696@prim.han.de>

David Bolen wrote:
> [me]
> > everything is a google-search away if you know the right codewords :-)
> > I don't in this case, though ...
> 
> Sorry, I must admit to being temporarily overcome with a fit of laziness.  I
> had actually fumbled around a bit before finding the posts I mentioned, so I
> should really have tried to include a reference.
> 
> Lest the codewords remain secret, I managed to wind my way there with a
> search on "eval dictionary instance" in comp.lang.python (which referenced a
> post of Tim's on the fourth page of results).
> 
> The messages I had found occurred deep into a "Dictionary from list?" thread
> from last October which crossed over into the subject of subclassing
> dictionaries:
> 
> http://groups.google.com/groups?th=2338421ec8397bec&seekm=mailman.1004338951
> .27778.python-list%40python.org

very nice, thanks!

[GvR in the above link]
  IMO, fixing these internal accesses to always use the overrides would
  remove much of the attractiveness of subclassing built-in types; I
  believe it would slow things down too much given the current state of
  the implementation.

  If you want a vaguely dictionary-like object, write a class that
  doesn't derive from dictionary but implements the mapping protocol; if
  you want a base class that implements most of the operations already,
  start with UserDict.

This seems reasonable except that 'exec' and 'eval' require a real dict-type.
Due to the generic nature of these execution statements i guess that
fixing it is roughly equivalent to fixing it all over the place unless
they could be somehow special-cased. Oh well...

  holger




From tim.one at comcast.net  Tue Jun 11 21:32:14 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 11 Jun 2002 21:32:14 -0400
Subject: Sorting list (maybe stupid)
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8BF0A37A@natasha.auslabs.avaya.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAEEEPMAA.tim.one@comcast.net>

[Delaney, Timothy]
> ...
> Personally, I think the implementation should be documented to
> use a stable sort - presumably mergesort (stable, minimal number of
> comparisons which is very important in Python).

Write a mergesort as fast and memory-efficient as the sort we've got and
we'll consider it.  I wasn't able to.  Great progress in memory-efficient
mergesort algorithms has been made over the last decade (the ancient one
<wink> in a Knuth Vol 3 exercise achieves memory efficiency at the cost of
vastly more comparisons), but I don't know whether they're truly practical
yet.  Count the # of comparisons the samplesort/binary_insertion hybrid
actually does and you'll find that mergesort doesn't have enough of an
advantage there to make up for the speed it loses due to greater data
movement.





From juanm.casillas at eresmas.com  Thu Jun 27 04:01:13 2002
From: juanm.casillas at eresmas.com (Juan M. Casillas)
Date: Thu, 27 Jun 2002 10:01:13 +0200
Subject: Module Help
Message-ID: <15642.50761.524492.167169@eresmas.com>

Hello!

	I want to create an empty module from C, and then add
	functions and definitions stored in another file, and
	some code that I will create on-the-fly, from C.
	
	The basic problem I have is that I don't know how to create the
	empty module (this module doesn't have any representation
	in the filesystem) and then how can I add the code and
	save the "environment" for this newly created module, how
	can I manage the globals and locals dictionaries ? for
	each module ?

Thanks in advance
Juan M. Casillas





From altis at semi-retired.com  Sun Jun  2 17:20:25 2002
From: altis at semi-retired.com (Kevin Altis)
Date: Sun, 2 Jun 2002 14:20:25 -0700
Subject: convert string to a method of a class at runtime?
Message-ID: <5YvK8.285$Vm3.130161@news.uswest.net>

I would like to be able to take a string containing the definition of a
method, compile the method and then add it to an instance variable at
runtime, so that the method defined in the string is part of the class just
as if it had actually been in the class definition to begin with when the
module was first used. So, instead of a class definition like:

class H:
    def hello(self, s):
        print s

   def world(self, s):
        print s

I would like to have just the 'hello' method defined beforehand and add
'world' at runtime.

Here's a shell transcript of my attempt to add the 'world' method by
compiling it and using the 'new' module to turn the compiled string into a
method of the class. As the transcript shows, my attempt failed, so I'm
hoping some Python language experts will be able to tell me what steps I got
wrong. Perhaps, I need to use a different method to define the argument list
or change where the global namespace comes from; I thought the namespace
should be the same as that already defined for the instance or class?

>>> class H:
...     def hello(self, s):
...         print s
...
>>> h = H()
>>> h.hello('world')
world
>>> s = """def world(self, s):
...     print s
...
... """
>>> s
'def world(self, s):\n    print s\n\n'
>>> c = compile(s, '', 'exec')
>>> c
<code object ? at 0164B620, file "", line 1>
>>> import new
>>> f = new.function(c, globals())
>>> h.__class__.__dict__['world'] = new.instancemethod(f, None, h.__class__)
>>> h.world('new world')
Traceback (most recent call last):
  File "<input>", line 1, in ?
TypeError: ?() takes no arguments (2 given)
>>>

Fingers crossed, that this kind of dynamic method addition is doable. I know
I can do it by just typing a function into the shell like this:

>>> def world2(self, s):
...     print s
...
>>> f2 = new.instancemethod(world2, None, h.__class__)
>>> f2
<unbound method H.world2>
>>> h.__class__.__dict__['world2'] = f2
>>> h.world2('yea')
yea

But I need to be able to do the addition outside the shell, so compiling the
string is necessary. In more complex examples, the method needs the
namespace to be right so it can make use of whatever imports and other
global variables have been setup.

Thanks,

ka
---
Kevin Altis
altis at semi-retired.com




From msoulier at nortelnetworks.com_.nospam  Wed Jun  5 07:29:32 2002
From: msoulier at nortelnetworks.com_.nospam (Michael P. Soulier)
Date: 5 Jun 2002 11:29:32 GMT
Subject: building on hpux 10.20
Message-ID: <slrnafrtjn.12v.msoulier@pmerd071.ca.nortel.com>

    Hello. I have Python 1.5.2 and Python 2.0 both built on HP-UX 10.20 with
no problems. However, I just tried building 2.2.1, and I got this error
immediately in the compile:

gcc -c -DNDEBUG -g -O3 -Wall -Wstrict-prototypes -I. -I./Include
-DHAVE_CONFIG_H  -o Modules/python.o Modules/python.c
In file included from /usr/include/limits.h:6,
                 from
/opt/corp/unsupported/lib/gcc-lib/hppa1.1-hp-hpux10.20/2.8.1/include/
limits.h:116,
                 from
/opt/corp/unsupported/lib/gcc-lib/hppa1.1-hp-hpux10.20/2.8.1/include/
syslimits.h:7,
                 from
/opt/corp/unsupported/lib/gcc-lib/hppa1.1-hp-hpux10.20/2.8.1/include/
limits.h:11,
                 from Include/Python.h:27,
                 from Modules/python.c:3:
/usr/include/sys/stdsyms.h:244: #error "Large Files (ILP32) not supported in
strict ANSI mode."
In file included from Include/Python.h:62,
                 from Modules/python.c:3:
Include/pyport.h:480: #error "LONG_BIT definition appears wrong for platform
(bad gcc/glibc config?)."
gmake: *** [Modules/python.o] Error 1

    Has anyone seen this before? We won't be moving away from HP-UX 10.20 for
a while yet, so it would be nice to get this working.

    Mike

-- 
Michael P. Soulier, QX41, SKY  Tel: 613-765-4699 (ESN: 39-54699)
Optical Networks, Nortel Networks, SDE Pegasus
"...the word HACK is used as a verb to indicate a massive amount
of nerd-like effort."  -Harley Hahn, A Student's Guide to Unix


From fperez528 at yahoo.com  Wed Jun 19 13:44:48 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Wed, 19 Jun 2002 11:44:48 -0600
Subject: Pedagogic advice needed
References: <3D104D35.E0932297@info.unicaen.fr>
Message-ID: <aeqfuh$88n$1@peabody.colorado.edu>

Jerzy Karczmarczuk wrote:

> Then I began to digest all that, and the obvious question was: why not
> Python?

Indeed :)

> 1. Does anybody here have some teaching experience in a similar context?

I've been invited to give a talk to a similar group of biology students, at 
the end of the course, and have also given informal talks on 'python for 
science' to mathematicians. My topic for the biologists was C/python 
integration, but the whole course was given to a mix of graduate/undergrads 
with no computing background. I've spoken a fair bit to the professor and his 
experience (even though this was his first time teaching it) seems to have 
been very positive.

You can find more about his course at: http://mcdb.colorado.edu/courses/6440/


> 2. Assuming that the visualisation issues, all kind of plots, graphs
>    *and animations* are very important, how would you organize with
>    Python such a work?

For this kind of work, my current solution is using python/Numeric for the 
matlab-like numerical work, with Gnuplot for normal 2d/3d plotting (I have 
available a customized interface for Gnuplot access which runs on top of M. 
Haggerty's Gnuplot.py -- http://www-heller.harvard.edu/~mhagger/download/). 
This works very well for all kinds of 'typical' scientific plotting needs.

For fancier 3d visualization needs (isosurfaces of functions of 3 variables, 
volume rendering, etc) I use the amazing MayaVi 
(http://mayavi.sourceforge.net/). It's a full featured GUI visualizer, but 
all written in python and scriptable 'from inside'. With it you'll also get 
ivtk, a very nice system for interactive use of VTK.

Here are a few quick examples of what mayavi can do (from my own stuff, 
there's plenty more out there):

http://windom.colorado.edu/~fperez/pub/topo_viz_sample.jpg -> isosurfaces and 
colormap cut planes (with and without transparency) of a 3-d slice of a 4-d 
dataset.

http://windom.colorado.edu/~fperez/pub/topo_vol_ray.jpg -> a fancy ray-traced 
volume rendering of one 3-d slice. This kind of rendering is SLOW, but gives 
pretty pictures :) You have to write your own translucency tables obviously, 
and that takes some experimenting. You can grab from there the code to make 
that image if you want to play with it 
(http://windom.colorado.edu/~fperez/pub/vtk_volume.tgz)

http://windom.colorado.edu/~fperez/pub/topo_2412.b585m008.36.1000.slices/ -> 
the full 4-d dataset sliced into a webpage. The html generation code is about 
15-20 lines of very simple python. It's a very convenient way of visualizing 
complex datasets and automatically sharing the results with ohters: all they 
need is a web browser.


> Of course I know Numeric Python, Scientific Python modules, and other
> standard stuff permitting to do all kind of graphic exercices and
> demonstrations (eg., all the wx bazar).

Do you know about SciPy ()? It is a complete 'framework' that wraps around 
Numeric to provide very impressive overall functionality.

> But I *must* avoid the low-level programming, we won't have time for that.
> We will need a reasonable complete scientific 3D plotting package usable
> by people without too much experience.

If someone is there to do the install for them, I'd say scipy could satisfy 
your needs. Installing it is a major chore right now, but once up and running 
it works very well. You need to run off of CVS though, all releaseed code is 
outdated now.

> (I checked the Obvious Suspects, the Vault of Parnassus, etc., I am veryfing
> all that stuff, but perhaps some of you know something really succulent and
> full of vitamines. I need *your experience*, NOT just standard Web links.)

Well, besides the above, I would plug here my own project IPython 
(http://windom.colorado.edu/~fperez/ipython/). It's a command-line shell much 
more powerful than the default one, with many features specifically designed 
for scientific computing work (inspired/stolen from environments like 
Mathematica and IDL). If you install it and start 'ipython -p numeric' 
(assuming you have both Numeric and Gnuplot.py installed), you'll have an 
environnment very much like Matlab. Its gnuplot access functions are enhanced 
versions of the originals in Gnuplot.py, to make day to day plotting easier 
(I use it constantly, so I do eat my own dog food :)

I use IPython everyday for exactly the kind of thing you have in mind, so 
I've tuned it to be as convenient as possible in that kind of environment. If 
you are interested in using it, I'll be glad to help you out with any snags 
you encounter.

Cheers,

f.



From mhammond at skippinet.com.au  Fri Jun 14 19:40:44 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 14 Jun 2002 23:40:44 GMT
Subject: PythonWin Problem!
References: <1a52c7f1.0206131455.6df50cf8@posting.google.com> <3D0973C5.60407@skippinet.com.au> <1a52c7f1.0206141158.1c03b3f5@posting.google.com>
Message-ID: <3D0A7F8E.3090606@skippinet.com.au>

Ethan wrote:
> Thanks Mark. It worked like a charm. One thing though, the key I had
> to delete was in HKCU rather than in HKLM. The second method would not
> work with ME since like I said in my earlier post, I don't have a
> right click menu for items in the taskbar.
> 
Yes, I meant HKCU - sorry.

I haven't used ME, but I know that Win9x, 2k and XP all have a 
right-click menu on the taskbar.

Mark.



From andy47 at halfcooked.com  Tue Jun 11 01:57:47 2002
From: andy47 at halfcooked.com (Andy Todd)
Date: Tue, 11 Jun 2002 05:57:47 +0000 (UTC)
Subject: MySQLdb warnings problem
References: <mailman.1023595512.23403.python-list@python.org>
Message-ID: <Xns922AA23E3DE4Dandy47halfcookedcom@203.109.252.31>

Terry Hancock <hancock at anansispaceworks.com> wrote in 
news:mailman.1023595512.23403.python-list at python.org:

> Hi all,
> 
> I'm in the middle of writing a Zope/MySQL project, and
> I use a Python/MySQL crawler to do some database
> maintenance tasks.
> 
> I'm writing this crawler now, and it's my first
> experience with "Python Database API 2.0", which seems
> really overblown and complicated compared to the simple
> interface described in O'Reilly's MySQL/mSQL book, or
> with the ZSQL interface in Zope. I suppose this is
> for the benefit of more complicated RDBMs than MySQL.
> 
> Anyway, I hope I'm not making a mistake using it --
> I'm just trying to use the most "standard" Python
> interface to MySQL, and I'm under the impression that
> this is it. Please advise if I'm wrong.
> 
> I couldn't find a mysqldb specific mailing list, so I
> thought I'd try asking my question here:
> 
> The program is fairly long, but what happens is that
> it scans the database for entries in a database table
> called "user" and calculates some statistics for them
> based on other tables.  All the retrievals work fine,
> but I get a funny traceback when I write the data
> back into the user's row in the table:
> 
> Traceback (most recent call last):
>   File "Styx.py", line 258, in ?
>     styx.crawl( styx.params['P_Styx'],  styx_tasks)
>   File "Crawler.py", line 193, in crawl
>     task[1](argval)                     # Operate on it
>   File "Styx.py", line 165, in compute_user_stats
>     dbcursor.execute(sqlcmd)
>   File
> "/usr/local/narya/lib/python2.1/site-packages/MySQLdb/cursors.py", line
> 61, in execute
>     r = self._query(query)
>   File
> "/usr/local/narya/lib/python2.1/site-packages/MySQLdb/cursors.py", line
> 168, in _query
>     rowcount = self._BaseCursor__do_query(q)
>   File
> "/usr/local/narya/lib/python2.1/site-packages/MySQLdb/cursors.py", line
> 118, in __do_query
>     self._check_for_warnings()
>   File
> "/usr/local/narya/lib/python2.1/site-packages/MySQLdb/cursors.py", line
> 150, in _check_for_warnings
>     raise Warning, self._info
> _mysql_exceptions.Warning: Rows matched: 1  Changed: 0  Warnings: 1
> 
> For reference, here's a snippet of the code where this
> broke:
>      sqlcmd= """
>           UPDATE user SET nposts=%d, nrecent=%d, keytopic='%s'
>           WHERE username='%s'
>           """ % ( n_posts, n_recent, keytopic, user[0] )
> 
>      #print sqlcmd
>      #try:
>      dbcursor.execute(sqlcmd)
>      #except:
>      #     pass
> 
> As the commented-out lines indicate, I tried ignoring this problem.
> Doing so seems to make the program work fine, but I can't figure
> out why it should be an issue in the first place (and I generally
> prefer not to just ignore warnings, especially without understanding
> them).
> 
> The thing is, what's the warning message?  Nothing was changed
> this time, because a previous run already updated the stats (the
> first run said "Rows matched: 1  Changed: 1  Warnings: 1").
> 
> Uncommenting the print statement shows what the actual SQL was:
> 
>   UPDATE user SET nposts=7, nrecent=0, keytopic='/'
>   WHERE username='anonymous'
> 
> which looks legit to me, though I'm not really an SQL expert.
> 
> So maybe it's raising a warning every time I write to the
> database? Why?
> 
> Should I just stick the try-except code back in and forget
> about it, or am I missing a step somewhere? Any ideas would
> be much appreciated.
> 
> Thanks!
> Terry
> 

Terry,

There isn't (as far as I know) a MySQLdb mailing list, but questions of 
this type are very welcome at the db-sig mailing list (db-sig at python.org or  
http://mail.python.org/mailman/listinfo/db-sig).

Regards,
Andy
-- 
Contents free posts a speciality


From hesterloli at hotmail.com  Wed Jun 19 04:23:30 2002
From: hesterloli at hotmail.com (George Hester)
Date: Wed, 19 Jun 2002 08:23:30 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <D7NP8.46301$GY.14487785@twister.nyroc.rr.com> <ouu1o0djnx.fsf@andlx-anamika.cisco.com>
Message-ID: <6cXP8.47927$GY.15261354@twister.nyroc.rr.com>

You ought to be

--
George Hester
_________________________________
"Ganesan R" <rganesan at myrealbox.com> wrote in message
news:ouu1o0djnx.fsf at andlx-anamika.cisco.com...
> >>>>> "George" == George Hester <hesterloli at hotmail.com> writes:
>
> > Don't worry I won't return here.  This is not a group to get help but a
> > group of individuals that think the world owes them something when they
> > participate here.
>
> Oh, my! I am simply overcome with grief *sob*
>
> Ganesan
>
> --
> Ganesan R (rganesan at debian.org>)     | http://www.debian.org/~rganesan/
> 1024D/5D8C12EA, fingerprint F361 84F1 8D82 32E7 1832  6798 15E0 02BA 5D8C 12EA




From hancock at anansispaceworks.com  Sat Jun  8 12:03:26 2002
From: hancock at anansispaceworks.com (Terry Hancock)
Date: Sat, 08 Jun 2002 09:03:26 -0700
Subject: MySQLdb warnings problem
Message-ID: <3D022ACE.A545BF98@anansispaceworks.com>

Hi all,

I'm in the middle of writing a Zope/MySQL project, and
I use a Python/MySQL crawler to do some database
maintenance tasks.

I'm writing this crawler now, and it's my first
experience with "Python Database API 2.0", which seems
really overblown and complicated compared to the simple
interface described in O'Reilly's MySQL/mSQL book, or
with the ZSQL interface in Zope. I suppose this is
for the benefit of more complicated RDBMs than MySQL.

Anyway, I hope I'm not making a mistake using it --
I'm just trying to use the most "standard" Python
interface to MySQL, and I'm under the impression that
this is it. Please advise if I'm wrong.

I couldn't find a mysqldb specific mailing list, so I
thought I'd try asking my question here:

The program is fairly long, but what happens is that
it scans the database for entries in a database table
called "user" and calculates some statistics for them
based on other tables.  All the retrievals work fine,
but I get a funny traceback when I write the data
back into the user's row in the table:

Traceback (most recent call last):
  File "Styx.py", line 258, in ?
    styx.crawl( styx.params['P_Styx'],  styx_tasks)
  File "Crawler.py", line 193, in crawl
    task[1](argval)                     # Operate on it
  File "Styx.py", line 165, in compute_user_stats
    dbcursor.execute(sqlcmd)
  File
"/usr/local/narya/lib/python2.1/site-packages/MySQLdb/cursors.py", line
61, in execute
    r = self._query(query)
  File
"/usr/local/narya/lib/python2.1/site-packages/MySQLdb/cursors.py", line
168, in _query
    rowcount = self._BaseCursor__do_query(q)
  File
"/usr/local/narya/lib/python2.1/site-packages/MySQLdb/cursors.py", line
118, in __do_query
    self._check_for_warnings()
  File
"/usr/local/narya/lib/python2.1/site-packages/MySQLdb/cursors.py", line
150, in _check_for_warnings
    raise Warning, self._info
_mysql_exceptions.Warning: Rows matched: 1  Changed: 0  Warnings: 1

For reference, here's a snippet of the code where this
broke:
	sqlcmd= """
		UPDATE user SET nposts=%d, nrecent=%d, keytopic='%s'
		WHERE username='%s'
		""" % ( n_posts, n_recent, keytopic, user[0] )

	#print sqlcmd
	#try:
	dbcursor.execute(sqlcmd)
	#except:
	#	pass

As the commented-out lines indicate, I tried ignoring this problem.
Doing so seems to make the program work fine, but I can't figure
out why it should be an issue in the first place (and I generally
prefer not to just ignore warnings, especially without understanding
them).

The thing is, what's the warning message?  Nothing was changed
this time, because a previous run already updated the stats (the
first run said "Rows matched: 1  Changed: 1  Warnings: 1").

Uncommenting the print statement shows what the actual SQL was:

  UPDATE user SET nposts=7, nrecent=0, keytopic='/'
  WHERE username='anonymous'

which looks legit to me, though I'm not really an SQL expert.

So maybe it's raising a warning every time I write to the
database? Why?

Should I just stick the try-except code back in and forget
about it, or am I missing a step somewhere? Any ideas would
be much appreciated.

Thanks!
Terry

-- 
------------------------------------------------------
Terry Hancock
hancock at anansispaceworks.com       
Anansi Spaceworks                 
http://www.anansispaceworks.com 
------------------------------------------------------




From egbert at bork.demon.nl  Fri Jun  7 04:48:22 2002
From: egbert at bork.demon.nl (Egbert Bouwman)
Date: Fri, 7 Jun 2002 10:48:22 +0200
Subject: Program to an interface, not an implementation
In-Reply-To: <adnjee$p88$1@wanadoo.fr>; from eric.brunel@pragmadev.com on Thu, Jun 06, 2002 at 02:15:41PM +0000
References: <mailman.1023358448.9091.python-list@python.org> <adnjee$p88$1@wanadoo.fr>
Message-ID: <20020607104821.A227@bork.demon.nl>

All your comments are crystal clear. Thanks.

My next problem is on nearly the same page (17) of the GoF book: 
class versus interface (== type) inheritance, but in the python context.

"In C++ and Eiffel inheritance means both interface and implementation
inheritance" and 
"in Smalltalk inheritance means just implementation inheritance".
I don't know any of these languages.

Now in my simplistic worldview, if you inherit from a class,
you inherit everything, implementation and interface, 
because the implementation defines the interface as well.
Even if all methods of the superclass only contain 'pass'.
But I may be wrong.
egbert
-- 
Egbert Bouwman - Keizersgracht 197 II - 1016 DS  Amsterdam - 020 6257991
========================================================================




From occeanlinux at linuxmail.org  Mon Jun  3 02:09:43 2002
From: occeanlinux at linuxmail.org (Gold Fish)
Date: Mon, 03 Jun 2002 06:09:43 GMT
Subject: List
References: <20020602205610.31575.qmail@linuxmail.org> <009b01c20a7e$12a5ad80$0101010a@local> <mailman.1023062343.28782.python-list@python.org>
Message-ID: <3cfb07cf_1@news.iprimus.com.au>

Mike C. Fletcher wrote:

> I don't know if you'll consider this a more elegant way to divide up the
> lists, but I think it's a little more beautiful than the index-mucking
> approaches IMO:
> 
> def into( source, count=3 ):
>      source = source[:]
>      return intoDestructive( source, count )
> 
> def intoDestructive( source, count=3 ):
>      if source and not count:
>          raise ValueError( """Cannot divide into 0-length sub-elements"""
>          )
>      elif not source:
>          return []
>      results = []
>      if count > 0:
>          while source:
>              results.append( source[:count])
>              del source[:count]
>      else:
>          while source:
>              results.append( source[count:])
>              del source[count:]
>      return results
> 
> It will be very slow for very large lists with positive slice values
> (something the index-mucking approach doesn't suffer from), but for most
> reasonably sized lists, should be fairly good.  It handles negative
> steps as well, so you can slice up the list from the right _or_ the left
> side :) .
> 
> Enjoy all,
> Mike
> 
> Chris Gonnerman wrote:
> ...
>> 
>>     Lists = []
>>     N = 2
>> 
>>     adj = 0
>> 
>>     if len(BigList)%N:
>>         adj = 1
>> 
>>     for i in range(len(BigList)/N+adj):
>>         Lists.append(BigList[i*N:(i+1)*N])
>> 
>> (If anyone knows a more elegant solution please post it,
>> as I am getting headaches trying to think of one.)
> ...
> _______________________________________
>    Mike C. Fletcher
>    http://members.rogers.com/mcfletch/

Sorry for the very stupid question but i still don't get my goal, that's is 
using Mike C. Fletcher method i will generate the list of elements again 
and don't know what size it is. For example :
[['p1','p2','p3'],['p4','p5','p6'],['p7','p8','p9'],['p10']]
I want to seperate this list again in order to using them later .It look 
like you go to the realestate agent website and find the list of property 
then you just want to display 3 properties per pages. And you can go to 
other pages which also contain 3 properties as well until the last pages 
which has only 1 items. For example:
Page 1: ['p1','p2','p3']
Page 2: ['p4','p5','p6']
Page 3: ['p7','p8','p9']
Page 4: ['p10']
I didn't sleep well due to this problem Can any help me.



From cmkim at shinbiro.com  Thu Jun 27 00:17:25 2002
From: cmkim at shinbiro.com (Kim Chul Min)
Date: Thu, 27 Jun 2002 13:17:25 +0900
Subject: [Q] How Can I call python function from C source?
Message-ID: <afe3ji$gvk$1@news.kreonet.re.kr>

Hello there,

I want to call python function from c source.

for example,

in file "main.c"
--------------------------------------------------------------
int main(void)
{

    /** want to call python function add(a,b) that return a+b **/

}
--------------------------------------------------------------

in the file "add.py"
--------------------------------------------------------------
def add(a,b):
    return(a+b)
--------------------------------------------------------------

Can I call any function that written by python from C source ?

If possible, can anyone show me simple example like above??

Plz,,, Help me....Thankx in advance...

Ciao,
Farangan Xavio
cmkim at shinbiro.com






From tdelaney at avaya.com  Wed Jun 19 00:01:48 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 19 Jun 2002 14:01:48 +1000
Subject: python version?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A3A0@natasha.auslabs.avaya.com>

> From: George Hester [mailto:hesterloli at hotmail.com]
> 
> The whole issue started from here.  I did not deserve to be 
> told to "crawl
> before I walk."  I may have responded to that inappropriately 

Actually, that is a very common phrase which is used to suggest that you
probably don't know enough yet to do what you're trying to do, and so should
work on something simpler first.

So it was indeed very appropriate here.

Every time I learn a new programming language I need to crawl before I walk
- I can't do something complicated until I understand the syntax and
semantics.

In any case, this all started when you did not try to understand the useful
suggestions you were given right from the very first reply to this thread.

I refer you once again to:

http://www.tuxedo.org/~esr/faqs/smart-questions.html

which contains *very* useful information for *anyone* when asking questions
(especially in a technical forum). It is not an attempt to ridicule you - it
is an attempt to help you learn what is appropriate and what is expected.

I've already spent far too much time on this myself, but I'm still hoping
you will help yourself.

Tim Delaney




From johnroth at ameritech.net  Tue Jun  4 08:03:30 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 4 Jun 2002 08:03:30 -0400
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com>
Message-ID: <ufpb3sn6332ub4@news.supernews.com>

"Huaiyu Zhu" <huaiyu at gauss.almadan.ibm.com> wrote in message
news:slrnafo7qv.37c.huaiyu at gauss.almadan.ibm.com...
> Gustavo Cordova <gcordova at hebmex.com> wrote:

>
> Great explanation!  The problem stems from the fact that the current
+=
> operator implements two semantically distinct operations, chosen
according
> to some properties of the operands.  This is not unlike the situation
of the
> division operator with distinct behaviors on floats and ints.

This is not an issue with '+='. It's an issue with assignment,
regardless
of whether it is standalone or combined with some other operation.

And I disagree that assignment implements two distinct operations.
Assignment always rebinds. Always. In the examples cited, it
attempted to rebind into a tuple, because that's what was asked
for when the left hand side had a subscript.

The reason that the examples given at the top of the thread fail
in such strange ways is simply that '+=' does not check whether
the assignment is possible before doing the addition (or concatination,
in the specific cases cited.)

To see this, you could try the first example, using explicit '+' and
'='. I suspect it would fail in exactly the same way, with exactly
the same side effect.

John Roth





From peter at engcorp.com  Wed Jun  5 10:44:43 2002
From: peter at engcorp.com (Peter Hansen)
Date: Wed, 05 Jun 2002 10:44:43 -0400
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet>
Message-ID: <3CFE23DB.994F8A6C@engcorp.com>

TuxTrax wrote:
> 
> Hello all
> 
> I am new to this forum, and to Python, please excuse the newbie question.
> 
> I have started reading the o'reilly book, "learning python" and find python
> to be an amazing language so far. My first impression of it is a language that
> is easier than basic to learn and use, and more powerful than many other
> high level languages such as pascal.
> 
> My question is this; I realize that Python compiles to bytecode automatically
> as part of the interpreters tokenization run, but is there a way to permanently
> compile python code outputting to an executable binary?

Why do you want to do that?  For presumably increased speed, or because
you don't want the hassle of multiple .py files all over the place (in
other words, better packaging)?

-Peter


From whisper at oz.net  Mon Jun 24 21:55:22 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 24 Jun 2002 18:55:22 -0700
Subject: Suggestions for good programming practices?
In-Reply-To: <3D17C148.6CC8B3D2@engcorp.com>
Message-ID: <GCEDKONBLEFPPADDJCOEAEOKDLAA.whisper@oz.net>

They may be returned in a tuple, but they're PUT in discrete variables and
not kept in a tuple that I have to programmatically unpack.

Language lawyer should be shot, _then_ heard.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Peter Hansen
> Sent: Monday, June 24, 2002 18:03
> To: python-list at python.org
> Subject: Re: Suggestions for good programming practices?
>
>
> David LeBlanc wrote:
> >
> > The one that always gets me is "lat,lon,timezone =
> getLocation("seattle")".
> > I.E. the ability to return multiple distinct values as opposed
> to returning
> > multiple values in the form of a struct as in C or C++.
>
> You _are_ returning them "in a struct".  It's spelled "tuple"
> but other than that what's the difference?
>
> -Peter
> --
> http://mail.python.org/mailman/listinfo/python-list





From sjmachin at lexicon.net  Tue Jun 11 16:20:31 2002
From: sjmachin at lexicon.net (John Machin)
Date: 11 Jun 2002 13:20:31 -0700
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
References: <Xns9229EE722467Fgustafl@195.100.94.182> <mailman.1023746539.13373.python-list@python.org> <Xns922AD3C74280gustafl@195.100.94.182>
Message-ID: <c76ff6fc.0206111220.57a381a7@posting.google.com>

Gustaf Liljegren <gustafl at algonet.se> wrote in message news:<Xns922AD3C74280gustafl at 195.100.94.182>...
> Cliff Wells <logiplexsoftware at earthlink.net> wrote: 
> 
> Thanks both of you. I tried both solutions. Since some intervals are very 
> high (1000000+), it takes too long to add them to a list. Cliff's solution 
> was really fast, but it took a while to understand. :)
> 

Sorry, Mr Customer, I thought I heard you say "thousands & elegant";
now I'm hearing "millions & fast". Let's get serious:

How many intervals do you have (a) max (b) average? How many of the
intervals are single integers? Between the lowest possible number and
the highest possible number, what percentage of integers are in the
set? What is the frequency distribution of membership tests? [I.e. is
it biassed towards small/large integers or is it "random"?] What is
your real speed requirement? With Cliff's solution, what proportion of
your app's run-time is spent testing set membership? How many tests
per set?


From m.faassen at vet.uu.nl  Sun Jun 30 16:58:39 2002
From: m.faassen at vet.uu.nl (Martijn Faassen)
Date: 30 Jun 2002 20:58:39 GMT
Subject: Reminder of a python Project.
References: <2218.941T781T12224051threeseas@earthlink.net>
Message-ID: <afnrdv$els$1@newshost.accu.uu.nl>

Timothy Rue <threeseas at earthlink.net> wrote:
> First off, it's ok that I got such a high level of resistance to my
> initial posting here about this project. Thanks to some research that /.
> had an article on, regarding the size of GPL project teams.

> But anyway.... Remember the autocoding project I mentioned?

FOR IMMIADIATE DISTRIBUTION TO ALL PSU MEMBERS -- ULTRA SECRET

The Python Secret Underground does not allow any functional autocoder 
to exist in this particular timeline at this particular time, outside of
the various bots.

The ruebot is (obviously) a rogue mutant version of the timbot. The mutation
has however wiped out its autocoding ability, deleting any knowledge of
the 10th autocoding command, retaining only 9.

Detecting this lack in itself but not exactly understanding it,
the ruebot is now attempting regain this lost ability. Luckily its
post-singularity knowledge is virtually incomprehensible
to most humans before the year 2053.

All PSU members are instructed to either AVOID or MISLEAD the ruebot in a 
subtle way. At all costs, do not leak the information of the 10th command
necessary for true autocoding. Let it focus on the 9 commands and let it
think those are all that are needed for autocoding.



From hesterloli at hotmail.com  Tue Jun 18 16:41:09 2002
From: hesterloli at hotmail.com (George Hester)
Date: Tue, 18 Jun 2002 20:41:09 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <KiEP8.44424$GY.13974606@twister.nyroc.rr.com> <ou8z5cg7ft.fsf@andlx-anamika.cisco.com>
Message-ID: <FVMP8.46296$GY.14475470@twister.nyroc.rr.com>

Thought so.  Since I have been treated so nicely I will make it a point to
return with the solution.  Thanks for all the help.

--
George Hester
_________________________________
"Ganesan R" <rganesan at myrealbox.com> wrote in message
news:ou8z5cg7ft.fsf at andlx-anamika.cisco.com...
> >>>>> "George" == George Hester <hesterloli at hotmail.com> writes:
>
> > "If you can't get an answer, please don't take it personally that we don't
> > feel we can help you. Sometimes the members of the asked group may simply
> > not know the answer."
>
> Yeah, right. Nobody in the group knows the answer to your question. We would
> really appreciate if you do some research and figure it out for yourself. We
> would be even more happy when you come back and enlighten us with your
> knowledge. Thank you.
>
> Ganesan
>
> --
> Ganesan R




From egbert at bork.demon.nl  Mon Jun 10 06:19:52 2002
From: egbert at bork.demon.nl (Egbert Bouwman)
Date: Mon, 10 Jun 2002 12:19:52 +0200
Subject: Program to an interface, not an implementation
In-Reply-To: <D54BF26DB5A7D511B36B00508B9596A402F8EC63@sdgcaexch2.peregrine.com>; from gscott@peregrine.com on Fri, Jun 07, 2002 at 08:43:23AM -0700
References: <D54BF26DB5A7D511B36B00508B9596A402F8EC63@sdgcaexch2.peregrine.com>
Message-ID: <20020610121951.B473@bork.demon.nl>

Again I learned a lot from both of you, and reading a couple of hours 
in a C++ book helped as well. But it is not all crystal clear any more. 

Eric says:
> So, as you see, you probably can live without knowing anything about all 
> this and still be able to do great stuff with whatever language you like 

You should see my programs already. However I have the impression that 
assimilating the GoF book will improve things enormously. And for that
I have to understand their and other terminology. 
Compared to the terminologies the basic mechanisms look simple.

In Gordon's reply the venom is in the end:
> For C++ interface inheritance REQUIRES class inheritance.  For
> Python/Smalltalk, it does not.

This I read as:
For Python interface_inheritance does not require class_inheritance.

So you can inherit an interface without inheriting the class,
but interface_inheritance by way of class_inheritance is possible as well.
That means at least two things:
- python has interface inheritance (which some people deny),
- interface_inheritance without class_inheritance is a kind of cultural 
  (not genetical) inheritance: give it the same name and all is well.
  But for me that is only polymorphism !

I do all this hair splitting because the GoF say (still page 17) 
that many patterns depend on the distinction between interface- and 
implementation inheritance. How do you define in python the
interface inheritance the GoF talk about ?
egbert
-- 
Egbert Bouwman - Keizersgracht 197 II - 1016 DS  Amsterdam - 020 6257991
========================================================================




From ykingma at accessforall.nl  Thu Jun  6 15:30:55 2002
From: ykingma at accessforall.nl (Ype Kingma)
Date: Thu, 06 Jun 2002 20:30:55 +0100
Subject: array, list, performance...
References: <mailman.1023369012.10674.python-list@python.org>
Message-ID: <3CFFB86A.4114459A@accessforall.nl>


Michael Chermside wrote:
> 
> Here is my understanding of it:
> 
>     li = [x0, x1, x2, ...]
> 
>     li[n]          element access is O(1)
>     li.append(x)   appending is O(1)
>                    (actually, it SOMETIMES takes O(len(li)) as it
>                     resizes the list. But if you grow a list by
>                     continually appending, the amortized time is
>                     linear (ie, constant amortized time per element)

Append is often enough linear in the length of the list
that growing by appending is O(len(li) * len(li)).
You can prevent that by creating a list a large enough list
when starting, eg:

li = [None] * neededSize

>     li.insert(0)   prepending is O(len(li))
>                    (so grow the list from the end!)
>     del li[0]      deleting from the front (or middle) is O(len(li))
>     li.pop()       deleting from the end is O(1)
>     li[n] = x      assignment to the middle is O(len(li))

Assigment to the middle of a list is O(1), ie. same as element access.

>     len(li)        determining length is O(1)
>     li.index(x)    search is O(len(li)) (NOT efficient)
>     li.sort()      sort is O(len(li)*ln(len(li)))
>                    (I believe it uses quicksort, but if a python
>                     function has to be executed for the compare
>                     that may take some time.)
> 
> It's really just your basic auto-growing array, but with a bunch of
> clever things done to speed up common cases, and I don't know those well
> enough to comment on them.
> 
> -- Michael Chermside

Have fun,
Ype

-- 
email at xs4all.nl


From occeanlinux at linuxmail.org  Sun Jun  2 16:35:20 2002
From: occeanlinux at linuxmail.org (Gold Fish)
Date: Sun, 02 Jun 2002 20:35:20 GMT
Subject: string generator
Message-ID: <3cfa8039_1@news.iprimus.com.au>

I like to create a list of file automatical based on series. For example
user input the number of file say 5. So i want to generate the file 
according to this range , this mean the generated files could be 
filename1
filename2
filename3
filename4
filename5
I using the for loop like that

for file in range(5):
        lis = [file]
        string.join('filename',lis[1:])
how can it didn't work. I am confusing how to do it.



From feldspar at ix.netcom.com  Sat Jun 22 13:45:21 2002
From: feldspar at ix.netcom.com (Antaeus Feldspar)
Date: Sat, 22 Jun 2002 14:45:21 -0300
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at>
Message-ID: <3D14B7B1.7040602@ix.netcom.com>

Siegfried Gonzi wrote:

	[snip]

> Thank you Billy Boy for a forgiving user system Windows XP which in turn
> relieves Python's "wreak havoc" behavior.


	[snip]


You know, a month or so ago, I was listening to complaints from a webpage builder 

that of all the Web browsers out there, only Internet Explorer rendered his 

friend's HTML code "correctly".  When he was pressed for what the other browsers

did "incorrectly", it turned out that the would-be web-maven wanted to 
be able to use the digit "0" and the letter "O" interchangeably in 
hexadecimal values -- not because there was any particular benefit to 
mixing-and-matching, merely because they were used to be careless about 
such things.  Internet Explorer was the only one that accepted such 
behavior without even an alert that it was wrong.

"Forgiving"?  Or enabling?

One hard-won piece of wisdom I've learned about computer interfaces over 
the years is "people learn by doing; if you allow them to do it the 
wrong way, they learn to do it wrong."  Even if you say "In this case, 
it doesn't make a difference, so why not?" that only builds in bad 
habits that compound on each other until the time comes that it *does* 
make a difference.  That's the hard lesson that has been learned about 
safety systems, is that building in six layers of safety to protect 
workers from the consequences of their actions will in fact train 
workers to get all six wrong at once.

	-jc




From niemeyer at conectiva.com  Thu Jun 27 15:06:39 2002
From: niemeyer at conectiva.com (Gustavo Niemeyer)
Date: Thu, 27 Jun 2002 16:06:39 -0300
Subject: Python crashed on me
In-Reply-To: <20020627205042.A18986@phd.pp.ru>
References: <20020627205042.A18986@phd.pp.ru>
Message-ID: <20020627160639.A23280@ibook.distro.conectiva>

Hello Oleg!

>    Recently Python (one program that I am dbugging) started to crash.
> FreeBSD kills it with "Bus error", Linux with "Segmentation fault".
> 
>    I think the program crashed in the cPickle.dump(file, 1), because after
> the crash the file is corrupted and cannot be load back (upon loading
> cPickle complain EOFile). The error is reproduceable.
> 
>    How can I debug the situation? May be my data grew too big? The pickle
> file is about 800K.
[...]

Do you have a small testcase?

-- 
Gustavo Niemeyer

[ 2AAC 7928 0FBF 0299 5EB5  60E2 2253 B29A 6664 3A0C ]




From paoloinvernizzi at dmsware.com  Tue Jun 25 13:24:31 2002
From: paoloinvernizzi at dmsware.com (Paolo Invernizzi)
Date: Tue, 25 Jun 2002 17:24:31 GMT
Subject: Python and Eclipse
References: <3D17FAF4.35CC9F1B@earthlink.net> <81CC01E58BF2A99B.DFCD5EC6E4654907.32D731380951B9CA@lp.airnews.net> <3D189A8A.577C8186@earthlink.net>
Message-ID: <3D18A6D2.5080404@dmsware.com>

> Eclipse is open source, so no matter what
> IBM decides, interested parties could pick
> it up and run if necessary. 

I've read the licence, I've understood that I can release product based 
on Eclipse, either commercially, either compiled, preserving the 
copyright note and a link to original eclipse source... but I'm a 
terrible reader of license-style  document! Someone can confirm this?

> I have not tried JBuilder. 

I worked with jbuilder 5 for about 6 mnts, and I feel Eclipse is far 
superior. I love the refactoring tools, but I still whisper for a form 
builder plugin...(and Borland is a master in this field!)

Concerning elipse and python, I think there are 2 roads.
The first one is using jpe (http://sourceforge.net/projects/jpe) the 
python-java framework. I've tried it some time ago... and I managed to 
  execute python code from a java VM (a jython prompt in reality! ;) but 
I stopped there. Perhaps Frederic Giacometti and jython folks can give 
some help!

The second one is to use socket (is possible?) to connect to a python 
server application... (with the plus that switching various python 
versions is very easy)... and move some logic in python (fast 
development times ;)

The problem, as usual It's that I have no time ;(

Cya
Paolo Invernizzi



From never at mind.info  Fri Jun 28 08:47:26 2002
From: never at mind.info (G. Willoughby)
Date: Fri, 28 Jun 2002 13:47:26 +0100
Subject: Is Python growing?
References: <3D1B92FB.59B590A0@pop.ntlworld.com> <slrnahnif7.1c6.matt@happy-hour.mondoinfo.com> <7x66042j50.fsf@ruckus.brouhaha.com>
Message-ID: <afhlsj$adv$1@news6.svr.pol.co.uk>

> Well, you could have said...

Theres always one isn't there?  ;-)

--G. Willoughby




From karczma at info.unicaen.fr  Tue Jun 11 08:55:02 2002
From: karczma at info.unicaen.fr (Jerzy Karczmarczuk)
Date: Tue, 11 Jun 2002 14:55:02 +0200
Subject: Defining a method final
References: <6glN8.35376$86.884025@twister1.libero.it> <mailman.1023796095.1427.python-list@python.org> <phmN8.36164$K4.893500@twister2.libero.it>
Message-ID: <3D05F326.478E20B9@info.unicaen.fr>

Fabio Corneti wrote:
> 
> I would like to define a method for a class which has
> not to be overriden. Of course, since the developer of
> this project are two, this could be established by
> convention, but if the project will grow consistently
> another developer could override the method by mistake,
> causing unpleasant side effects.

Hm. A classical problem.

Will you ask tomorrow for private methods and private inheritance
as well?

Python, like Smalltalk was never designed for paranoid [*NO* disrespect
meant or implied] projects, whose members need to protect themselves
from themselves. If such a danger is real, use qualified names.


Jerzy Karczmarczuk
Caen, France


From donn at u.washington.edu  Tue Jun 11 20:10:22 2002
From: donn at u.washington.edu (Donn Cave)
Date: 12 Jun 2002 00:10:22 GMT
Subject: M2Crypto: select() behaves weird on SSL socket
References: <3D04ECED.2030202@NOSPAMREMOVETHISxs4all.nl> <slrnag9rie.enn.akuchlin@ute.mems-exchange.org> <3D04FBF5.8090407@NOSPAMREMOVETHISxs4all.nl> <3D04FE2F.6070302@NOSPAMREMOVETHISxs4all.nl> <1023762896.869994@yasure> <3D063356.30200@NOSPAMREMOVETHISxs4all.nl>
Message-ID: <ae63he$20h8$1@nntp6.u.washington.edu>

Quoth Irmen de Jong <irmen at NOSPAMREMOVETHISxs4all.nl>:

| If there is a SSL wrapper for plain sockets, isn't there some
| sort of SSL wrapper of select()?

I don't know.  Probably not, but I don't know what's in M2Crypto.

| > I see my reply earlier today followed up a separate thread, after
| > you posted your initial question twice.  If you haven't seen it,
| > a hint:  SSL_pending(self->ssl.)  (I don't know much more than that,
| > anyway.)
|
| I don't know if this function is exposed in M2Crypto... if it is,
| I could probably get my code to work. If it isn't, I don't know
| how to fix it :-(

Assuming I guessed right about what the function does, you either
need it - or you have to redesign your application to use a separate
thread here, where the thread will just read and not care whether
it blocks or not.  (This may be the first time I have ever proposed
threads as a solution in c.l.p., which goes to show what a stupid
problem this is.)

	Donn Cave, donn at u.washington.edu


From vze3myxh at verizon.net  Sun Jun  2 13:37:20 2002
From: vze3myxh at verizon.net (vze)
Date: Sun, 02 Jun 2002 17:37:20 GMT
Subject: GUI Toolkit!
References: <1a52c7f1.0206011720.7c3ac7ea@posting.google.com>
Message-ID: <kJsK8.4908$Oj4.3590@nwrddc03.gnilink.net>

If you need to make something very quickly, try pythoncard at
pythoncard.sourceforge.net. It is a layer above wxWindows and comes with
tons of samples - including an interface builder written in itself.

<ethanw at rediffmail.com> wrote:

> Hi,
>   I need to do some GUI programming with Python. I found that there
> are quite a few options available for the toolkits . Apart from TkInter,
> I have seen bindings for Qt, wxWindows, FOX and FLTK. I want to pick one
> with the best open visual tools available. From what I have looked I am
> really impressed with the screen shots of Boa Constructor. Apparently it
> even generates the event procs. But I doubt that I have seen everything.
> Can anyone please tell me about any other options that I might have. I
> am developing on Windows.
> 
> Thanks


From gerhard.haering at gmx.de  Sun Jun 30 23:33:16 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Mon, 1 Jul 2002 05:33:16 +0200
Subject: Q: Status of MySQL & Python?
In-Reply-To: <69hvhu4otkrcncmu0d4vudfhv74o7maiul@4ax.com>
References: <69hvhu4otkrcncmu0d4vudfhv74o7maiul@4ax.com>
Message-ID: <20020701033316.GA6829@lilith.my-fqdn.de>

* John Hall <wweexxsseessssaa at telusplanet.net> [2002-07-01 03:12 +0000]:
> When trying to run the dbBrowser from Samples, MySQLdb/__init__.py
> does a traceback in line #27 because _mysql is not found. Sure 'nuff,
> I don't have that. I do have souce C source: _mysql.c

Obviously you downloaded the source distribution. It's dead simple to
build this, if you have Visual C++ installed:
- Edit setup.py to point to your MySQL installation
- Enter python setup.py install

But you can just download the binary installers for Windows, which are
available for Python versions from 1.5.2 to 2.2 at the same place as the
source tgz: http://sourceforge.net/project/showfiles.php?group_id=22307
If you find any Windows-specific problems with these, complain to me - I
built them.

The source distribution is still useful, because it contains the fine
documentation, which is missing in the binaries.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From rayed at awal.net.sa  Sun Jun 30 03:50:04 2002
From: rayed at awal.net.sa (rayed)
Date: Sun, 30 Jun 2002 10:50:04 +0300
Subject: Preforked server in Python
Message-ID: <3D1EB82C.1000501@awal.net.sa>

Hi,

I am looking for module similar to SocketServer that prefork chlidren 
instead of forking on each request.

Thanks in advance

- rayed



From bjarne_christiansen at hotmail.com  Wed Jun 26 04:08:26 2002
From: bjarne_christiansen at hotmail.com (Bjarne Christiansen)
Date: Wed, 26 Jun 2002 08:08:26 +0000
Subject: Python file upload
Message-ID: <F201A2DreBBsQwY6YHk00000017@hotmail.com>

Hi,
I have some problems uploading binary files though the web browser. It seens 
to work fine when uploading ACII file but but binary files seems to be 
currupted. The begining of the binary file seems fine, but some data is 
missing....

Here is my file upload script:

#!c:\Python\python -d

import cgi

print "content-type: text/html\n\n"

form = cgi.FieldStorage()
if not form:
  print """
<form action="http://localhost/cgi-bin/upload.py" method="POST" 
enctype="multipart/form-data">
<input type="file" name="filename">
<input type="submit">
</form>
  """
elif form.has_key("filename"):
  item = form["filename"]
  if item.file:
    data = item.file.read()
    print cgi.escape(data)
    data1 = cgi.escape(data)
    f = open("file1.jpg","wb")
    f.write(data1)

Any help will be greatly appriciated!

Best Regards,
Bjarne Christiansen
bjarne_christiansen at hotmail.com


_________________________________________________________________
MSN Photos is the easiest way to share and print your photos: 
http://photos.msn.com/support/worldwide.aspx





From plugin at supercable.es  Mon Jun 24 18:25:51 2002
From: plugin at supercable.es (Ugo García)
Date: Tue, 25 Jun 2002 00:25:51 +0200
Subject: Threads on embedded Python
References: <1024952585.111238@seux119> <Xns9237EECAA9CC5cliechtigmxnet@62.2.16.82>
Message-ID: <1024957605.771843@seux119>

Let see if I get something clear... I don't have any main loop in Python. I
only have some functions that 'control' certains object in my operation. To
be a little more clear, the task of the functions are for example, move an
object in the world generated by a game with a little intelligence. For
example: (a simple example)

FUNCTION 1:
    - Wait until main character is in correct position
    - Play a sound
    - Move an enemy from pos X1,Y1 to X2,Y2
    - Play an animation
    - ....
    - ....

So, I don't have a main loop, only function like that (and other kind of
functions) that a run as a thread to have all them running simultaneous. Is
there other way to do that without having threads? Do I have to have a main
loop in order to have the threads running? Perhaps running Python in a
separate thread could be a good idea.

Thanks,
--ugo
"Chris Liechti" <cliechti at gmx.net> escribi? en el mensaje
news:Xns9237EECAA9CC5cliechtigmxnet at 62.2.16.82...
> "Ugo Garc?a" <plugin at supercable.es> wrote in
> news:1024952585.111238 at seux119:
>
> > Hi all. I have an applicattion who uses Python to control some
> > processes. I've made a little script library to work with threads and
> > each proccess is a Python thread. The main application make calls to
> > Python in order to create these proccesses. The problem is that the
> > proccess doesn't run unless I call Python to do something. Now in the
> > main loop I have a line like this:
> >     PyRun_SimpleString("dummy=1");
> >
> > This make Python the threads to run. The problem is that this is a
> > little slow 'cause if I call Python from tha main application too many
> > times the proccesses run well but not the main application, and if a
> > make a big pause between calls the proccesses doesn't work at a good
> > speed. Is posible to have some threads running in an embedded Python
> > without making callings to it continuosly? It would be a better idea
> > to have another thread who's only task is to call Python (a line that
> > the one written before) in order to have the Python proccesses
> > running?
>
> why not run python in a separate thread (created by your app) and doing
the
> loop in python, like PyRun_SimpleString("while 1: mainloop()") or so?
> you do have a main loop in python, don't you? otherwise you wouldn't need
> python threads, right? :-)
>
> you could possibly also release the GIL, but don't forget to grab it again
> when accessing the python interpreter. i'm not sure about this one, so i
> would try the above.
>
> chris
>
> --
> Chris <cliechti at gmx.net>
>




From d2002xx at softhome.net  Fri Jun  7 09:50:33 2002
From: d2002xx at softhome.net (nobody)
Date: 7 Jun 2002 06:50:33 -0700
Subject: Native code compiler for python?
Message-ID: <baf24bd4.0206070550.84f1376@posting.google.com>

Is there any existing native code compiler for python? Or Python to C
translator? Or someone is doing this?


From dfackrell at DELETETHIS.linuxmail.org  Thu Jun 13 09:59:00 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Thu, 13 Jun 2002 07:59:00 -0600
Subject: parse tar file with python
References: <ae7498$4jven$1@ID-146704.news.dfncis.de> <ae7o6s$4oh4v$1@ID-50400.news.dfncis.de> <aea0hs$54qq4$1@ID-146704.news.dfncis.de>
Message-ID: <3d08a527$1_3@hpb10302.boi.hp.com>

Just tossing out some ideas here that may or may not pay off (not in any
particular order).

1. Try splitting the files between different directories like another post
mentioned.
2. Try using the tar module on one large tar file.
3. Try using the tar module on several medium-sized tar files.
4. (Was going to suggest open()ing a bunch of files at once in case there is
some delay there and separating the open() and read() calls might show some
speed improvement, but I think I won't.)
5. Consider threads?  They can help in some cases where I/O is the
speed-limiting factor.
6. Post a sample of the code you have right now so that we can see what
you're doing, perhaps including data extracted with the profile module.

--
Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.

"Shagshag13" <shagshag13 at yahoo.fr> wrote in message
news:aea0hs$54qq4$1 at ID-146704.news.dfncis.de...
> In fact, i don't want to tar/untar files, and especially not in main
memory !
>
> I wish i could read the tar-ed file line by line (f.readline) and be able
to check when i find the beginning of an "inside file" and
> get some info about it like name, how and so on... (that's because my
original files are plain text file, and i think that tar will
> let them unchanged)
>
> In my tar file there is, for example, a kind of separator like this (but
with everything in one long line) with :
>
>
shag.py_0100744_0002033_0001750_00000004414_07500237361_0015314_0_ustar_00_s
hagshag_user_0000040_0000417_beginofmyfilehere
>
> where _ stand for a variable amount of another ascii code that i can't
cut/paste...
>
> Do you kwow what each means (for example the first one is undoubtly the
file name, but then...) ?
> And what are the fixed position of each of theses ?
> Have a clue ?
>
> Many thanks,
>
> s13.
>
>
>
>
>




From thomas.mangin at free.fr  Mon Jun  3 15:13:50 2002
From: thomas.mangin at free.fr (Thomas Mangin)
Date: 3 Jun 2002 12:13:50 -0700
Subject: Created a "file" like command in python (see man 4 magic) and face a showstoper bug
Message-ID: <359e75aa.0206031113.6a8bfed1@posting.google.com>

Hi,

I wrote some code perform "file" like operation and I am facing a
nasty bug.

the files can be found at : ftp://www.slhan.org/source/circle/

It produce the same output that "file":

# ./magic.py core
core: ELF 32-bit LSB core file of 'vi' (signal 11), Intel 80386,
version 1 (SYSV)

# ./magic.py jpg_good.jpg
jpg_good.jpg: JPEG image data, JFIF standard 1.01, resolution (DPI),
50 x 50

# ./magic.py jpg_bad.jpg
jpg_bad.jpg: JPEG image data, EXIF standard %d.73, 10752 x 2048

Note the %d !

The code causing problem is :

if replace:
  try:
    mime = mime % replace
  except:
    pass

In both case mime is containing the string "%d." and replace is in
both case of type int and valued 1 and 0.

Any idea ?

Thomas Mangin


From jeff at sasmor.com  Mon Jun 17 17:53:37 2002
From: jeff at sasmor.com (Jeff Sasmor)
Date: Mon, 17 Jun 2002 21:53:37 GMT
Subject: If you use Python -OO is it a permanent condition??
References: <rXpP8.32471$ks6.25445@nwrddc02.gnilink.net> <aelamd$fo1$1@panix1.panix.com>
Message-ID: <BTsP8.57000$uX3.41630@nwrddc01.gnilink.net>

'Fascinating'
I'm sure there's a logical reason... somehow. Anyway, I found
that if I renamed all the .pyo files to .pyc everything works fine.
I guess that's because there's little difference between them. It's
too bad that there's no way to switch optimized mode on and
off programmatically - would be useful for apps that use compile().

--
#--------------------------------
Jeff Sasmor
jeff at sasmor.com
"Aahz" <aahz at pythoncraft.com> wrote in message
news:aelamd$fo1$1 at panix1.panix.com...
> In article <rXpP8.32471$ks6.25445 at nwrddc02.gnilink.net>,
> Jeff Sasmor <jeff at sasmor.com> wrote:
> >
> >So  I tried:
> >
> >python -OO P1.pyo
> >
> >and it works just fine.  Does this mean that the interpreter must
> >always be started with -OO to execute files compiled with -OO?  I can't
> >imagine why this would be the case, but it seems to be -- at least as
> >far as I can see from my little bit of experimentation.
>
> Python doesn't search for .pyo files unless it is running in optimized
> mode.  Try setting the PYTHONOPTIMIZE env var.
> --
> Aahz (aahz at pythoncraft.com)           <*>
http://www.pythoncraft.com/
>
> Project Vote Smart: http://www.vote-smart.org/




From s.keim at laposte.net  Wed Jun 26 06:32:01 2002
From: s.keim at laposte.net (sebastien)
Date: 26 Jun 2002 03:32:01 -0700
Subject: 'parent object'?
References: <3d18228c$0$221$4d4ebb8e@news.nl.uu.net>
Message-ID: <7da8d8f8.0206260232.44587981@posting.google.com>

> Is there, like 'self, also a 'parent' object of some sort? 
> The problem i'm having is that i have a class, which contains a list of
> instances of another class. How can i let this 'childclass' call a function
> which is on the 'parent class'?

As have said other posters, you should probably avoid this kind of design.
But if you absolutely need it, here is a short hack for Python 2.2 that 
use attribute descriptors:

class Parent(object):
    class Child(object):
        def __get__(self, obj, cls):
            self.__parent = obj
            return self
        def hello(self):
            print "Hello, I'm",self,"and my father is",self.__parent
    child = Child()

x = Parent()
x.child.hello()


From baf at texas.net  Thu Jun 27 08:43:01 2002
From: baf at texas.net (Ben Fairbank)
Date: Thu, 27 Jun 2002 12:43:01 GMT
Subject: [Newbie]  How to manipulate variable _names_ ?
Message-ID: <3d1b0723.6179705@news.texas.net>

I have a program that requires the user to enter a file name, such as
"othello," which the program will then open, process, and create a
dictionary based on the contents of the file.  I would like to name
the dictionary "OthelloCount," and then go on and create more
dictionaries, each with a name made up of the inputfile name as
provided by the user, concatenated with "Count," but I do not know how
to make a variable name based on input.

Any suggestions?

BAF.  


From thomas.mangin at free.fr  Tue Jun  4 04:14:23 2002
From: thomas.mangin at free.fr (Thomas Mangin)
Date: 4 Jun 2002 01:14:23 -0700
Subject: "file" like command in python (see man 4 magic) download
References: <359e75aa.0206031113.6a8bfed1@posting.google.com> <Xns9222E32F2F52Fcliechtigmxnet@62.2.16.82>
Message-ID: <359e75aa.0206040014.3e3fa8de@posting.google.com>

> 1. initialize replace with None
> 
> > if replace:
> 
> use this instead:
>   if replace is not None:
> 
> and have fun progamming python!
> chris

It was the source of the problem. Feel free to download the code (GPL)
if you need it at ftp://www.slhan.org/source/magic.tar.gz

Thomas


From look at sig.invalid  Wed Jun 26 06:27:16 2002
From: look at sig.invalid (Simo Salminen)
Date: Wed, 26 Jun 2002 10:27:16 +0000 (UTC)
Subject: Python file upload
References: <mailman.1025079034.19051.python-list@python.org>
Message-ID: <slrnahj5o4.rli.look@pikkukorppi.cs.tut.fi>

* Bjarne Christiansen [Wed, 26 Jun 2002 08:08:26 +0000]
> Hi,
> I have some problems uploading binary files though the web browser. It seens 
> to work fine when uploading ACII file but but binary files seems to be 
> currupted. The begining of the binary file seems fine, but some data is 
> missing....
> 
> Here is my file upload script:
> 
> #!c:\Python\python -d

see http://www.python.org/cgi-bin/faqw.py?req=show&file=faq07.012.htp

>     data1 = cgi.escape(data)
>     f = open("file1.jpg","wb")
>     f.write(data1)

>>> cgi.escape.__doc__
"Replace special characters '&', '<' and '>' by SGML entities."

why would you cgi.escape the data?

--
simo <dot> salminen <at> iki <dot> fi


From spammers.do.not.bother at void.bogus  Sun Jun  2 07:33:24 2002
From: spammers.do.not.bother at void.bogus (Magnus)
Date: Sun, 02 Jun 2002 11:33:24 GMT
Subject: How to call functions inside a cgi script from the html interface?
References: <adc4sv$vjrdc$2@ID-49758.news.dfncis.de> <2bkcda.c25.ln@10.0.0.1> <adcq91$vmqig$1@ID-49758.news.dfncis.de>
Message-ID: <8onK8.10334$p56.3020410@newsb.telia.net>

Ken wrote:

<cut>

> Where can I find good references with both python and HTML explained in
> detail on how they work together?


Maybe if you use our friend Google?
http://www.google.com/search?hl=en&lr=&q=%2Bpython+%2Bcgi+%2Bhtml+%2Btutorial+%2Bhowto

Regards,
Magnus



From occeanlinux at linuxmail.org  Sun Jun  2 17:43:20 2002
From: occeanlinux at linuxmail.org (Gold Fish)
Date: Sun, 02 Jun 2002 21:43:20 GMT
Subject: List
References: <3cfa6ed8_1@news.iprimus.com.au> <addrjh$imq$1@news.tudelft.nl> <3cfa7619_1@news.iprimus.com.au> <addths$jb9$1@news.tudelft.nl>
Message-ID: <3cfa8ef5_1@news.iprimus.com.au>

Charl P. Botha wrote:

> In article <3cfa7619_1 at news.iprimus.com.au>, Gold Fish wrote:
>> Charl P. Botha wrote:
>> 
>>> In article <3cfa6ed8_1 at news.iprimus.com.au>, Gold Fish wrote:
>>>> Assume we got 2 list
>>>> list 1 = ['ABC','CDG','DAV','FE','FECA','FEAFA','FEA']
>>>> list 2 = []
>>>> how can we copy element from list 1 to list 2 say i just want to copy
>>>> the 3 element 'CDG','DAV','FE' to list2 only.
>>> 
>>> list1 = ['ABC','CDG','DAV','FE','FECA','FEAFA','FEA']
>>> list2 = list1[1:4]
>>> 
>>> Is that what you meant?
>>> 
>> No i mean if i got from list1 i wanna take 3 elements in any position to
>> list2 but i don't know how many item in that list.
> 
> Could you rephrase that?  The example I posted does exactly what your
> example showed.  Maybe if you gave another example and tried to explain
> more clearly, I'd know what you were trying to do.
> 
What i am trying to do is that i got one biglist don't know how big is it 
say list1 then i like to divided into some small sublists. These sublists 
will have some elements in the biglist for example 3 elements maximum in 
each sublist. So if the biglist got 10 elements then it'll be divided into 
4 sublists the first 3 will have 3 elements and the last one will have one 
elements.


From jepler at unpythonic.net  Tue Jun  4 09:57:52 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 4 Jun 2002 08:57:52 -0500
Subject: Looking for a list subclassing example...
In-Reply-To: <adi6lb$107klr$1@ID-146704.news.dfncis.de>
References: <adi4dv$118tko$1@ID-146704.news.dfncis.de> <3cfc994e$0$3851$e4fe514c@dreader4.news.xs4all.nl> <adi5hc$10ulbe$1@ID-146704.news.dfncis.de> <adi6lb$107klr$1@ID-146704.news.dfncis.de>
Message-ID: <20020604085747.A22766@unpythonic.net>

On Tue, Jun 04, 2002 at 01:01:14PM +0200, Shagshag13 wrote:
> Is is possible to write a method in an object subclassing from list and which "frees" it ?

You could 'del self[:]' (slice deletion) to remove all the entries from self.

Jeff




From derek at wedgetail.com  Fri Jun 21 07:41:22 2002
From: derek at wedgetail.com (Derek Thomson)
Date: Fri, 21 Jun 2002 11:41:22 GMT
Subject: I'd give up Perl tomorrow if only...
References: <mailman.1024655248.4124.python-list@python.org>
Message-ID: <3D1310E0.9070803@wedgetail.com>

Gerhard H?ring wrote:

> 
>>[...] What we *really* need to do is dynamically create CORBA
>>interfaces for Perl modules, but that might be impossible given that
>>Perl is like Python in that module and class interfaces aren't
>>pre-declared :(
> 
> 
> Why is that a problem? Why not just override __getattr__ or its
> Perl-equivalent and switch to SOAP/XML-RPC instead of CORBA?

I'm not sure what you mean. The calls are coming from the client, 
remember. On the server (Perl side), you have to unmarshall the request, 
decide which object the request is for, unmarshall the method name and 
the arguments, then make the right method call on the correct object. I 
can't see AUTOLOAD (Perl's _getattr_ equivalent) entering into this 
process anywhere.

Besides, how do I deal with distributed object references, or map an 
object reference on the client to an object on the server in SOAP or 
XML-RPC? Lots of Perl modules create objects as you use them, as you 
would expect. Then there's the question of making invocations back to 
the client to support the push model common in serious distributed 
computing.

You need the infrastructure CORBA provides to do real distributed 
objects, as far as I can see. SOAP and XML-RPC alone are good only for 
fairly simple RPC, as they only specify a marshalling scheme.

To demonstrate this, I've used XML-RPC as a pluggable marshalling scheme 
for Fnorb. So, you're using the CORBA infrastructure (object adaptors, 
object references, connection management and so on), but you're 
communicating with XML on-the-wire. Best of both worlds?

Check it out on the Fnorb CVS repository. It's on the branch tagged 
xmlrpc. I may merge it into the trunk if enough people are keen. It's 
pretty nice, you can have some objects talking XMLRPC, and some 
CORBA-IIOP, and some talking both, without changing any implementation 
code at all.

> You're
> restricted to simple data types, then, but that would probably be the
> same for CORBA, right?

I can't recall how complex SOAP structures can be, but in CORBA-IIOP and 
XML-RPC you can represent arbitrarily complex structures, just as long 
as they're not self-referential i.e. only trees not graphs.

Regards,
Derek.



From whisper at oz.net  Thu Jun 27 17:07:39 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 27 Jun 2002 14:07:39 -0700
Subject: Pythoniac: Thoughts on a hardware Python processor
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F06233015DE6BA@mail.hebmex.com>
Message-ID: <GCEDKONBLEFPPADDJCOEOEJADMAA.whisper@oz.net>

> -----Original Message-----
> From: Gustavo Cordova [mailto:gcordova at hebmex.com]
> Sent: Thursday, June 27, 2002 12:54
> To: David LeBlanc
> Cc: Python List
> Subject: RE: Pythoniac: Thoughts on a hardware Python processor
>
>
> >
> > Ahh! Haven't you heard? Forth came in Fifth and the Forth
> > chip is dead. In fact I don't think anyone is still offering
> > a stack based computer (only HP did AFAIK, the HP 3000).
> >
> > Dave LeBlanc
> > Seattle, WA USA
> >
>
> Which is truly a shame; it's more because of commercial
> and market pressures, and nothing to do with true
> throughput or technical merit.

Actually, it's hard to make zero address machines effeciently in hardware
(on chip). At least a (substantial) part of the problem is having to keep
pushing and popping parts of the hardware stack into memory. That might be
less of a problem these days given VVLSI, but back then 4k of 16 bit stack
would have been a big deal - and yet not enough for multi-tasking. Would
have been great for embedding though. I agree that theoretically, a stack
machine is a thing of grace and beauty. To be honest though, the IBM System
34 tagged architecture is one I would have preferred to have survived. IMO,
_that_ could have made Python really really go fast! Personally, i'd settle
for a good 3-address CISC machine (which Intel X86 is not), but IIRC, those
never went far back when there was still a good deal of experimentation in
the MPU world.

> Let's see if Python can be ported to Forth, like it was
> to Java. If it can, *then* python'll run at hardware speed,
> don't you think?

I don't know about your Python, but my Python definately runs just as fast
as the hardware can make it go ;-). It's going to run a whole lot faster
when the AMD "Hammer" 64 bit hummer hits the bricks at a price that mere
mortals can afford. Seriously though, there are some aspects of Python that
don't lend themselves to efficient implementation in hardware.

> Anyhow, as I said, it's a shame that Forth has been lambasted
> so much that it's not really possible to find much info
> and code on the net. It's actually a quite nice tool.

I don't think Forth got lambasted so as to cause it to be hard to find info
on - I think it, like many many other computer and natural languages, had
it's day and has left the stage. Too bad Lisp hasn't followed it ;-)

> :-/
>
> -gus

Whoever asked if the Intel/AMD FP stack counted: nah - it's not the primary
mode of the CPU like it was on the HP 3000 and the Forth chip. BTW, speaking
of the Forth chip, I think Forth, Inc. is still offering the IP for that if
anyone wants to burn a few hunderd k$ on new silicon.

While I think that creating a Python chip is a great and wonderful idea, I
doubt it will or should happen:

* The CPU vendors are ahead of the power curve and I don't see an Open
Source group doing as well, let alone better. I figure a volunteer effort
would take on the order of $1,000,000,000 in time/services to get a 1-2 GHZ
chip. Intel spent about 4x that for the Itanic...err Itanium ;) and they had
the capital, IP and bodies already on hand. (Coming soon: "Itanic 2 - The
Resurrection!" (really, Itanium 2 is coming).)

* Important IP for modern CPU performance is under patent. Plus, the guys
who have the knowledge are probably paid extreme sums and kept in obscure
out of the way places, not to mention being contractually obligated up the
ying-yang to not disclose, compete or otherwise ply their trade for other
then their designated keepers for a long term of years.

* Efficient explotation of current and proposed CPU architectures is
predicated upon software tools that do a good bit of (proprietary)
optimization and organization during code generation before the CPU ever
sees the bits.

* Burning Python into silicon will tend to fossilize the language. Is Guido
done enough with it for that yet? Guido and the gang can whip out a patch
for the PVM in a matter of hours. How long would that take for a silicon
change and what would it cost?

For all these reasons and more, I think a good compiler with a retargetable
backend is a better expenditure of time/effort and would yield a result far
faster then a silicon development effort.

Dave LeBlanc
Seattle, WA USA





From maxm at mxm.dk  Fri Jun 14 06:48:57 2002
From: maxm at mxm.dk (Max M)
Date: Fri, 14 Jun 2002 12:48:57 +0200
Subject: Using the locale module on Windows under hte IIS
Message-ID: <3D09CA19.5050709@mxm.dk>

I have an app that I am putting the final touches on. Including 
localisation, so that the users can use danish nummeric formats etc.

But I am very much in doubt as to how I should use the locale module. I 
have found out easy enough that I need to set::

     from locale import setlocale

     setlocale(locale.LC_ALL, 'danish')

It kind of puzzles me as it seems that it is a global switch that is set 
for the entire programme.

But I wonder exactly where to do this. It's a web app running under the 
IIS, and so there is no natural starting point for the programme. Should 
I encapsulate the functionality an a few wrapper functions?

Is there some kind of global variable that I can set?

Perhaps I should set it in the gloabl.asa ?

Anybody's got a clue?


regards Max M



From PoulsenL at capecon.com  Thu Jun 27 13:21:34 2002
From: PoulsenL at capecon.com (PoulsenL at capecon.com)
Date: Thu, 27 Jun 2002 13:21:34 -0400
Subject: Most elegant python program to date.
Message-ID: <72F73B808A78D511A73600034771D9FF717DCF@dc_exchange2.howrey.com>

Just curious.  I find it much easier to learn a language by reading lucid
examples of its prose.  Are there any nominations for the most elegant
example of a python program to date.  Not sure under what criteria, I leave
that up to you.  I will say that I use python for matrix manipulation more
than anything, but that shouldn't prevent you from nominating your favorite
regardless of purpose.

Loren




From mwh at python.net  Thu Jun 13 08:36:22 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 13 Jun 2002 12:36:22 GMT
Subject: Extension objects in Python 1.5 and 2.2
References: <20020613071813.GA16764@strakt.com> <15624.19045.700125.215173@12-248-41-177.client.attbi.com> <mailman.1023953777.18976.python-list@python.org> <lk3cvry1pi.fsf@pc150.maths.bris.ac.uk> <mailman.1023961320.665.python-list@python.org>
Message-ID: <lky9djmk44.fsf@pc150.maths.bris.ac.uk>

Martin Sj?gren <martin at strakt.com> writes:

> On Thu, Jun 13, 2002 at 09:20:53AM +0000, Michael Hudson wrote:

[Misc/pymemcompat.h]

> > Pleasure!  Be sure to complain if it doesn't work, won't you?
> 
> So far it's looking good. I still had to ifdef on the Python version
> since I'd used PyModule_AddObject, but that was easily fixed.

Right.  This header is only about the memory interface.  I guess it
wouldn't be *that* hard to add many of the new API functions to a
compat library, but that project would have much wider scope...

Actually, the real benefit of pymemcompat.h as I see it are the
comments it contains laying down the law about which interfaces to
use.

Cheers,
M.

-- 
  > Touche! But this confirms you are composed of logic gates.
  Crud, I thought those were neurons in there.
                    -- Thomas F. Burdick, Kenny Tilton, comp.lang.lisp


From uwe at rocksport.de  Mon Jun  3 09:59:52 2002
From: uwe at rocksport.de (Uwe Schmitt)
Date: 3 Jun 2002 13:59:52 GMT
Subject: different gui-toolkits pros/cons...
Message-ID: <adfsoo$dmik4$1@hades.rz.uni-sb.de>

Hi,

I'm looking for an appropriate toolkit for GUI programming,
which do you use ? what are pros and cons ???
I'm specially interested in wxPython, pyQt and pyGTK...
and I'm intersted in a windows-like look and feel on win platforms...
Finally I'm intersted in execution speed.

So, I await your answers, greetings, Uwe

-- 

Dr. rer. nat. Uwe Schmitt      ICQ# 159647634   Uwe.Schmitt at num.uni-sb.de


From gaya at infomail.lacaixa.es  Sun Jun  2 17:04:44 2002
From: gaya at infomail.lacaixa.es (Toni Gaya)
Date: 2 Jun 2002 14:04:44 -0700
Subject: windows install put short path names to win register
Message-ID: <61f6bfa8.0206021304.2abedc3c@posting.google.com>

Windows python 2.2.1 installation program puts short file names to
some path values in windows register.

This makes that programs that read python path or some other
information as python path from windows register crash
-i'm using NTFS (maybe with FAT32, due that FAT system works with
short file names and long file names, works, but as NTFS only works
with long file names, some other installation programs just crash)-.
Examples: try to install mysqldb (mysql connection) or py2exe on
windows with a NTFS disk, and these programs crash.

I'm wrong?


From emile at fenx.com  Tue Jun 18 10:03:52 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 18 Jun 2002 14:03:52 GMT
Subject: why len(list) instead of list.len()  ?
References: <EhBP8.1$Qn1.1124@news.ecrc.de>
Message-ID: <c5HP8.243362$cQ3.10419@sccrnsc01>

"Markus Jais" <mjais at web.de> wrote in message
news:EhBP8.1$Qn1.1124 at news.ecrc.de...
> hello
>
> I way wondering why one has to write
>
> list = [2, 3, 5]
> print len(list)
>
> instead of list.len()
>
> in Ruby I can write
>
> array = [2, 3, 4]
> puts array.length()
>
>

Because a two-line wrapper goes a long way?

Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> class myList(list):
...     len = list.__len__
...
>>> l = myList([1,2,3,4])
>>> l
[1, 2, 3, 4]
>>> l.len()
4
>>> l.extend([5,6,7])
>>> l.len()
7



--

Emile van Sebille
emile at fenx.com

---------



From pyth at devel.trillke.net  Mon Jun 17 19:10:32 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Tue, 18 Jun 2002 01:10:32 +0200
Subject: inherrited private class varriables
In-Reply-To: <MPG.17788662b0a1b67a98968c@news.rz.uni-karlsruhe.de>; from merkosh@hadiko.de on Tue, Jun 18, 2002 at 12:51:36AM +0200
References: <MPG.17788662b0a1b67a98968c@news.rz.uni-karlsruhe.de>
Message-ID: <20020618011032.W15079@prim.han.de>

Uwe Mayer wrote:
> hi,
> 
> why can't i access a class private varriable from within a predecessor?
> f.e.
> 
> class parent:
>   __length = 5
>   def __init__(self): pass
>   def action(self):
>     print self.__class__.__length
> 
> class child(parent):
>   __length = 10
>   def __init__(self): pass
> 
> test = child()
> test.action()
> 
> will print 5, instead of 10. why?

short answer: 

private names are obtained by textual replacement.  
child.__length and parent.__length may look similar 
but the interpreter really sees [*]

    child._child__length  and   parent._parent__length

and refuses to override.  Think 'macro-preprocessor'.
              
long answer from http://www.python.org/doc/1.5/tut/node67.html :

    Any identifier of the form __spam (at least two leading 
    underscores, at most one trailing underscore) is now textually 
    replaced with _classname__spam, where classname is the current 
    class name with leading underscore(s) stripped.  This mangling 
    is done without regard of the syntactic position of the
    identifier, so it can be used to define class-private instance 
    and class variables, methods, as well as globals, and even to 
    store instance variables private to this class on instances 
    of other classes.

regards,

    holger



[*] if you want a proof 'import dis ; dis.dis(parent.action)'




From loewis at informatik.hu-berlin.de  Fri Jun 14 06:05:27 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 14 Jun 2002 12:05:27 +0200
Subject: Exheritance and Python
References: <aecaov$3ji$1@mordred.cc.jyu.fi>
Message-ID: <j4ptyu2n14.fsf@informatik.hu-berlin.de>

Jonne Itkonen <ji at tarzan.it.jyu.fi> writes:

> For new type classes, those that inherit object, this doesn't work,
> as __bases__ is read-only. Why? 

There is a technical problem: If changing the bases would change the
layout of instances (due to additional inherited slots, say), then
adding new bases is not implementable.

> So, is there a way to change the __bases__-attribute of a new type
> class? Am I to write a PEP?

In this specific case, I'd say a PEP won't help that much - write a
patch instead.

Regards,
Martin



From gmcm at hypernet.com  Fri Jun  7 09:10:21 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 07 Jun 2002 13:10:21 GMT
Subject: Gordon McMillan's Installer - Data files in the archive
References: <y9e0xy2x.fsf@morpheus.demon.co.uk> <uofewjgv4.fsf@ctwd0143.fitlinxx.com> <lm9sl8pw.fsf@morpheus.demon.co.uk>
Message-ID: <Xns92265D7BF8D4Dgmcmhypernetcom@199.171.54.216>

[posted and mailed]

Paul Moore wrote:

[snip]

> That worked remarkably easily. A trivial example:
> 
> from carchive import CArchive
> me = CArchive(sys.executable)
> readme_file = me.extract('README')[1]
> 
> One thought occurs - how "supported" is the CArchive class interface?

I expect advanced users to do things like this.
I also expect them to manipulate TOCs and put
extra processing in their spec files. Installer's
only a black box if you don't open it :-).

-- Gordon
http://www.mcmillan-inc.com/
 

 



From stephen.boulet at motorola.com  Fri Jun  7 19:02:27 2002
From: stephen.boulet at motorola.com (Stephen Boulet)
Date: Fri, 07 Jun 2002 18:02:27 -0500
Subject: Confidence intervals and other stats
Message-ID: <3D013B83.2E6F0E91@motorola.com>

Are there any python modules out there for computing confidence
intervals on data?

Thanks.

-- Stephen


From rjones at ekit-inc.com  Sun Jun 23 20:00:40 2002
From: rjones at ekit-inc.com (Richard Jones)
Date: Mon, 24 Jun 2002 10:00:40 +1000
Subject: PIL, Python and CGI
In-Reply-To: <3D165BB7.B901F064@infineon.net>
References: <3D1626FD.DA934B09@infineon.net> <1HrR8.44835$n4.10519234@newsc.telia.net> <3D165BB7.B901F064@infineon.net>
Message-ID: <200206241000.40596.rjones@ekit-inc.com>

On Mon, 24 Jun 2002 09:37, Sascha Ferley wrote:
> strange reason it can't find the _imaging extention..
> It is sort of strange that it wouldn't pick it up then.. any more ideas?

I'd say there's a chance that the _imaging module has linked to a shared 
library and the web server doesn't have the appropriate lib dir in its 
LD_LIBRARY_PATH. That'd cause an ImportError.

Unfortunately, PIL is swallowing the exact error - try removing the try/except 
around the "import _imaging" and see what the real error is.


   Richard





From emile at fenx.com  Sat Jun 29 16:03:32 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sat, 29 Jun 2002 20:03:32 GMT
Subject: ImportError: cannot import name mm_cfg
References: <mailman.1025375559.4917.python-list@python.org>
Message-ID: <oooT8.349712$352.41005@sccrnsc02>

"Pedro Bados" <pedrobados at eresmas.net> wrote in message
news:mailman.1025375559.4917.python-list at python.org...
>
> Hi :
>
>  I am not a python user so I have no idea at all of how I can resolve
this
> problem with mailman. Currently, I get this error each minute ...
>
>    File "/home/mailman/cron/qrunner", line 90, in ?
>     from Mailman.Bouncers import BouncerAPI
>   File "/home/mailman/Mailman/Bouncers/BouncerAPI.py", line 32, in ?
>     from Mailman import mm_cfg
> ImportError: cannot import name mm_cfg
>

Are you a Mailman user?  Or is it there as the result of an RH7x
install?
You can turn off qrunner if you don't use Mailman.

If you use mailman, did it ever work? or are you trying to get it going?
Did it just start breaking?  Did you change something? Which version of
python are you using?  Is there more than one installed?

The reason you got no responses the first time through is that there are
more questions open in your post than any of us can assume to know the
answers to, rendering any response non-responsive.  (as this is ;-)

Provide details.  Your error means the python path can't find
mm_cfg.(py|pyc).  Is there one one your system?  Where is it?  Is it
reachable from python?  Are the permissions OK? How about the
owner:group settings? Etc, etc, etc...

--

Emile van Sebille
emile at fenx.com

---------



From jimrowe at optilink.com  Thu Jun 13 13:57:47 2002
From: jimrowe at optilink.com (James Rowe)
Date: 13 Jun 2002 12:57:47 -0500
Subject: __subclasses__ 
Message-ID: <a9bcagho74.22t.jimrowe@optilink.com>

A search for __subclasses__ on the www.python.org
documentation search pages turns up empty. I can
find not mention of it in my Python books or even
in the "What's new in Python 2.2" document.  But:

class Foo(object):
    def __init__(self):
        object.__init__(self)

class Bar(Foo):
	pass

dir(Foo) # returns:

 ['__class__', '__delattr__', '__dict__', '__doc__',
 '__getattribute__', '__hash__', '__init__', '__module__',
 '__new__', '__reduce__', '__repr__', '__setattr__',
 '__str__', '__weakref__']

But there is a __subclasses__ method:

Foo.__subclasses__() # returns:

 [<class '__main__.Bar'>]


So I suppose the question is: Why is __subclasses__ not documented
or listed?  Can we safely use __subclasses__ and not worry about
it "going away" in the future?




From gleki at gol.ge  Fri Jun 28 04:38:18 2002
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Fri, 28 Jun 2002 10:38:18 +0200
Subject: squared functions--most Pythonic way?
References: <mailman.1025207218.19740.python-list@python.org>
Message-ID: <3D1C207A.74F24164@gol.ge>

One of the biggest advantages of Python is its nice and elegand mixture of
procedural, OO and functional programming... I would prefer lambda stile as the
most concise, but it is a matter of personal "phylosophy";)

Grtz,
Giorgi

Sean 'Shaleh' Perry wrote:

> >
> > functional programming:
> >
> >     # with an internal named function
> >     def fsquare(f):
> >         def f2(x):
> >             return f(x)**2
> >         return f2
> >
>
> this seems to be the most pythonic (lambda is not the preferred approach
> usually).
>
> Of course someone will mention taking this to the next step with a compose()
> function which take f(x) and g(x).



From usenet at thinkspot.net  Sat Jun  1 16:42:43 2002
From: usenet at thinkspot.net (Sheila King)
Date: Sat, 01 Jun 2002 13:42:43 -0700
Subject: How to open a HTML file when the python cgi program is executing?
References: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de> <ad95l1.i0.1@kserver.org> <ad9vsm$vhjnu$1@ID-49758.news.dfncis.de> <9m6ada.31b.ln@10.0.0.1> <adaeh8$uum7t$1@ID-49758.news.dfncis.de> <ada3k7.lo.1@kserver.org> <83n0uevjvx.fsf@panacea.canonical.org>
Message-ID: <adaj0j.13g.1@kserver.org>

On 01 Jun 2002 16:12:02 -0400, Kragen Sitaker <kragen at pobox.com> wrote in
comp.lang.python in article <83n0uevjvx.fsf at panacea.canonical.org>:

> Sheila King <usenet at thinkspot.net> writes:
> > print "Location: http://replacement-url\n"
> > 
> > You should have nothing before this, and there's no point in having
> > anything after it, since the page is being redirected.
> 
> Some browsers (in particular old versions of Opera) have problems with
> redirects every once in a while; that's why redirects traditionally
> contain HTML, in case the browser is broken.

Printing the Location header is not a "redirect" and should contain no
HTML.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From arosado at softhome.net  Tue Jun  4 19:59:57 2002
From: arosado at softhome.net (Andres Rosado)
Date: Tue, 04 Jun 2002 19:59:57 -0400
Subject: HTML links in a GUI tkinter
Message-ID: <5.1.0.14.0.20020604195947.00be66b0@mail.softhome.net>

At 03:21 PM 6/3/2002 +0000, you wrote:
>Message: 10
>    Date: Mon, 3 Jun 2002 15:23:47 +0000
>    From: Eric Brunel <eric.brunel at pragmadev.com>
>Subject: Re: HTML links in a GUI tkinter
>
>Steve Holden wrote:
> > So, all Aurel needs to know now is how to run wxPython widgets inside
> > tkinter programs...
>
>Why? Tkinter has everything to do it:
>
>----------------------------------
>from Tkinter import *
>
>root = Tk()
>t = Text(root)
>t.pack()
>
>def openHLink(event):
>   start, end = t.tag_prevrange("hlink",
>                                t.index("@%s,%s" % (event.x, event.y)))
>   print "Going to %s..." % t.get(start, end)
>
>t.tag_configure("hlink", foreground='blue', underline=1)
>t.tag_bind("hlink", "<Control-Button-1>", openHLink)
>t.insert(END, "This is a link\n")
>t.insert(END, "http://www.python.org", "hlink")
>t.insert(END, "\nAnd text goes on...\n")
>root.mainloop()
>----------------------------------
>
>Control-click on the link and you should see the message "Going to <link
>text>", which may be replaced by a webbrowser.open to actually open the
>URL. To insert a new link, just do:
>t.insert(<pos>, <url>, "hlink")

By far a more simpler solution. Thanks!

>HTH

-----------------------------------
Andres Rosado
Email: andresr at despammed.com
ICQ: 66750646
Homepage: http://andres980.tripod.com/

You should emulate your heros, but don't carry it too far.
Especially if they are dead.





From fredrik at pythonware.com  Thu Jun  6 04:26:10 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 06 Jun 2002 08:26:10 GMT
Subject: 64-bit Python?
References: <87elfl3yvu.fsf@mathdogs.com>
Message-ID: <C0FL8.41941$n4.9496108@newsc.telia.net>

Mike Coleman wrote:

> Has anyone looked at making Python 64-bit clean?

It is 64-bit clean, and has been so for many years.

> I was looking at trying to use the new mmap module to map in a large (>2G)
> file, and the comments in the source note that the mmap regions are limited to
> int-size (31 bits), primarily because Python objects are also limited to this
> size.

Python *integer* objects are limited to that size on platforms
where a C long is 32 bits.

Python runs perfectly fine on tru64, itanium, and other 64-bit
platforms.  On most of them, Python ints hold 64-bit values.

</F>




From nessus at mit.edu  Mon Jun  3 21:13:13 2002
From: nessus at mit.edu (Douglas Alan)
Date: 03 Jun 2002 21:13:13 -0400
Subject: manipulating child processes under Windows
Message-ID: <lcptz7zw0m.fsf@gaffa.mit.edu>

Hi.  I know lots about Python and lots about Unix, but unfortunately
I've been given the task of developing some software for Windows 2000,
and I'm having a problem.

The problem I'm having is this:  I want the Python script to run an
executable and then wait for the executable to terminate before
continuing.

Now, actually, I have no problem doing precisely this under Windows
2000 -- I know how to use wp.CreateProcess() to start up a program
under Python and wait for the program to terminate.  The problem I am
having is that the program itself is apparently just a stub that
spawns off the real program and then terminates, while the real
program then continues to run.  So my Python script ends up continuing
right away (because the stub process terminates right way), rather
than waiting for the real program to terminate.

So, I have two questions:

(1) Is this thing where a stub process fires off the real process a
    typical thing to do under Windows?  If so, it seems to make
    synchronizing processes difficult.

(2) What's the best way for me to solve the problem.  I.e., what's the
    best way to get my Python script to wait for the real program to
    terminate?

|>oug


From jimrowe at optilink.com  Thu Jun 13 18:40:52 2002
From: jimrowe at optilink.com (James Rowe)
Date: 13 Jun 2002 17:40:52 -0500
Subject: newbie: from .. import *
References: <56e1eff0.0206131437.4ca8ed9e@posting.google.com>
Message-ID: <a9bcagi8q0.22t.jimrowe@optilink.com>

On 13 Jun 2002 15:37:38 -0700, Jean-S?bastien Bolduc <jseb at cs.mcgill.ca> wrote:
>Hi All
>
>I've been working with Python for a while, but never really felt
>confortable with the way namespaces are imported.
>
>I'm working on a simple application that is to be used in the
>interactive mode. The idea is that the behavior of a function is
>determined, in part, by some global(?) flags. So I can change my
>flags, and explicitely call my function. Really simple. Except that
>for some reason, it does not work.
>
>In "test.py", I have the following:
>
>  flag = 1
>  def foo():
>    print flag
>  foo()
>  flag = 0
>  foo()
>  
>In the interactive mode, if I import the code above as:
>>>> from test import *
>1
>0
>
>The output is as expected. However, if I continue:
>>>> flag = 1
>>>> foo()
>0
>

# bar.py

flag = 1
def foo():
   print flag
foo()
flag = 2
foo()


>>> from bar import *
1
2
>>> import bar
>>> foo()
2
>>> flag = 3
>>> foo()
2
>>> bar.flag = 4
>>> foo()
4




From aahz at pythoncraft.com  Wed Jun 12 19:12:59 2002
From: aahz at pythoncraft.com (Aahz)
Date: 12 Jun 2002 19:12:59 -0400
Subject: Wanted:  Potential Python Authors
References: <ae7gke$mhb$1@ccis12.baylor.edu> <db3fgu40sbnt3pno5v1i2h9lnmpcoti2bk@4ax.com> <3d07d4a8@news.eol.ca>
Message-ID: <ae8khr$nnn$1@panix1.panix.com>

In article <3d07d4a8 at news.eol.ca>,
William Park  <opengeometry at NOSPAM.yahoo.ca> wrote:
>
>There is no such thing as "trained as writer", whether it's for computer
>books or fiction/nonfiction novels.  You either can write or you can't.

That's not true.  Writing is a skill -- just like programming -- that
can be learned and/or taught.  Of course, just like programming, innate
talent counts for something, but there are few people on this newsgroup
who are incapable of writing a book.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From twister at raplounge.de  Tue Jun 11 15:19:07 2002
From: twister at raplounge.de (Ben Thursfield)
Date: Tue, 11 Jun 2002 21:19:07 +0200
Subject: Newbie question communicating between 2 progs
References: <3D046BF0.DC36542C@raplounge.de> <3D04A938.627EE705@engcorp.com> <d9deee30.0206101041.4e8de6a2@posting.google.com>
Message-ID: <3D064D2B.43D73510@raplounge.de>

JS schrieb:

> Don't fully understand the question, but could you write the list to a
> temporary file. It could then be read as needed??

I was thinking about this, but i thought it would be bad to because i would have to set the
attributes of a folder to write, which I thought would be probably be a security hole...
Thanks for your answer!

Ben



From bh at intevation.de  Fri Jun 21 08:53:27 2002
From: bh at intevation.de (Bernhard Herzog)
Date: 21 Jun 2002 14:53:27 +0200
Subject: Advice sought on ps module
References: <6e990e29.0206200325.355f0413@posting.google.com> <aesoqg$kc9$1@scotsman.ed.ac.uk> <6e990e29.0206202356.544eb696@posting.google.com> <aev2dk$pn2$1@scotsman.ed.ac.uk>
Message-ID: <6qn0tohjy0.fsf@abnoba.intevation.de>

eddie at holyrood.ed.ac.uk (Eddie Corns) writes:

> P.P.S. Actually it occurs to me there is another solution, namely
> creating the diagram with something else (eg a canvas) and generating
> (presumably bitmapped) PS from that. Though I'm sure you've already
> rejected that idea as being unworthy!

I'd also say, that for good results with more complex layouts you'd need
a more object oriented approach than going straight to postscript.

Personally, I'd do this sort of thing with Sketch, of course, where you
could hook your code into in the GUI or simply use Sketch as a highlevel
vector graphics library for a command line program.

IIRC, reportlab nowadays also has an object oriented graphics/layout
library on top of the pdf generator. ISTR it even can generate EPS
directly.

   Bernhard

-- 
Intevation GmbH                                 http://intevation.de/
Sketch                                 http://sketch.sourceforge.net/
MapIt!                                           http://www.mapit.de/


From brian at sweetapp.com  Sat Jun 29 04:42:29 2002
From: brian at sweetapp.com (Brian Quinlan)
Date: Sat, 29 Jun 2002 01:42:29 -0700
Subject: Python needs better error reporting
In-Reply-To: <GCEDKONBLEFPPADDJCOEOEPADMAA.whisper@oz.net>
Message-ID: <01c501c21f48$e7ce0460$bd5d4540@Dell2>

David LeBlanc wrote:
> Cute anthropomorphism is lots of fun, but doesn't _quite_ obscure the
fact
> that computers don't really "tell" anybody anything, programmers do.
IIRC,
> Python throws this "Syntax Error: invalid syntax" in a number of
places.
> In the place that I used as an example, looking at the language
reference,
> an appropriate error would be more like:
> 	Syntax Error: expected "and", "or", "xor" or ":"

There are a lot more tokens than that allowed. What about "+", "*", "\",
etc?
 
> This would at least narrow down what should be looked at to resolve
the
> problem.

> Vague error reporting doesn't help new user's self confidance or
> acceptance of the language.
> 
> I really DO have a point!

Probably. Figure out how to add more meaningful syntax errors and submit
a patch. Good luck!

Cheers,
Brian





From cliechti at gmx.net  Sun Jun 30 19:47:35 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 1 Jul 2002 01:47:35 +0200
Subject: I'd give up Perl tomorrow if only...
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au> <mailman.1024681648.27687.python-list@python.org> <Pine.LNX.4.44.0206301550120.7807-100000@palanthas.neverending.org> <Xns923DF22CCC54Ccliechtigmxnet@62.2.16.82> <Pine.LNX.4.44.0206301756130.22298-100000@palanthas.neverending.org>
Message-ID: <Xns923E12B2373FDcliechtigmxnet@62.2.16.82>

Frank Tobin <ftobin at neverending.org> wrote in
news:Pine.LNX.4.44.0206301756130.22298-100000 at palanthas.neverending.org: 

> Chris Liechti, on 2002-06-30, wrote:
> 
>> well it has. the module unittest provides all thats needed. the
>> module just needs a class derrived from unittest.TestSuite and the
>> unittest module can find the tests.
> 
> No it doesn't, for two reasons.
> 
> 1) unittest is merely a testing framework, and there is no standard
> for using it from an extracted archive.
> 
> 2) unittests are only one way to test; it is generally best suited for
> small-unit testing.  I personally use it in conjunction with
> whole-program blackbox tests.
> 
> Perl also has some fairly standard testing suites, but there is a wide
> variety you can choose from.  However, the *interface* from an
> extracted Perl module is *always* "make test".

yes, so all we need is a convention (for the interface) and a small 
addition to distutils to run the tests, right?
point 2) is irrelevant when there is a standarized interface to invoke the 
tests and return the results.

>> its also a convention to have a test() function, but the former is
>> more powerful. yes, it should be used more widely.
>> 
>> > 3) a standard installation mechanism (Perl and Python do have)
>> > 4) a standard README system (Perl has, Python doesn't)
>> 
>> i don't know the perl way but with python every object has a __doc__ 
>> attribute, every module, class, function and method etc. and with
>> pydoc there is a standard extraction tool right in the standard
>> distro. ever tried it?
> 
> I'm very familiar with docstrings.  Perl also has its POD.  But each
> of these is API information.  README's are a different sort. 
> dostrings do not substitute for what a README is for:
> 
> 1) description of the module
> 2) describes dependencies
> 3) other information that is not API-oriented

absolutely nothing speaks againts a readme (altough i would call it 
README.txt, windows frendly, ya know ;-)...  you can add it to an archive 
made by distutils etc. and most people have one when they distribute a 
module.  i think the only diffrence is that you want to declare it as a 
_must_.
a description is already in the pgk-info file if you build with distutils. 
i think the dependencies belog in there too as this is a file for the 
machine and the readme should be a file for the user (freely formated, no 
need to automaticaly extract vital pakage information from there).

>> as metioned above, python has all it needs, now if you talk of an
>> organization that runs test on non-standard extension modules, no we
>> don't have that right now. the python distro has good tests for the
>> standard lib, but as there is no organization that distributes
>> extensions, everyone is responsible for testing its modules. (there
>> are collections of modules, like The Vaults of Parnassus:
>> http://www.vex.net/parnassus/ or ActiveStates Programmer's Package
>> Manager (PPM) but thats not as strong or established as CPAN)
> 
> Yes, I'm talking about non-bundled extension modules.  That's all
> we're talking about here.  The standard lib is irrevelant to the
> discussion. 
> 
> I suggest that you be a Perl programmer for a while, and see what
> having CPAN accessible and authoring for it really is like.  You
> explicitly state you aren't familiar with it.  I've been author of a
> couple CPAN modules for about 4 years now, and an extensive user for
> just as long, and have been a Python module writer for just over a
> year now.  Python is a great language, but in terms of a reusable code
> repository, *nothing* comes close to the jewel that is CPAN.  There is
> a reason you'll hear this over and over from people coming to Python
> from Perl. 

belive me, i would also like something like CPAN. i have a module out 
in the wild too. and while i understand that it is nothing for the standard 
distro i see that a couple of people use it (at least a bunch of people 
download it each day :-) and a standard way to get it would be cool.

i think that an packaging system with dependecy- and downloadmanager, 
testing and more would be of great value. there are some beginnings of such 
a system by different people, but nothing lifted off until now.
i think, if a group of programers donated enough of their spare time and 
found a machine to run it on, a prototype could be set up fairly quick (as 
python has all (99%) it needs to build such a system. i guess required are 
some conventions for packages and a bit of code). but that needs two things 
1) someone who does the work on it 2) the community must use it (test it, 
make it usable, fill it with modules).

maybe, later when we have such a system, everybody will ask how we could 
have lived without it for such a long time... ;-)

> The Vaults and PPM just don't hold a candle to CPAN.  For all
> practical purposes, they are just big FTP repositories.

yes i know and agree, i didn't wanted to express something different. my 
point was that there are organizations that build collections of modules, 
even if they don't provide testing, dependecy management etc. (not sure 
about PPM's capabilities, though). i mean that i must have mentioned them 
in the context and the way i wrote above.

chris

-- 
Chris <cliechti at gmx.net>



From tim at remove_if_not_spam.digitig.cix.co.uk  Fri Jun 28 08:13:56 2002
From: tim at remove_if_not_spam.digitig.cix.co.uk (Tim Rowe)
Date: Fri, 28 Jun 2002 12:13:56 GMT
Subject: Problems configuring 4suite
Message-ID: <3d1c4f89.1205192@usenet.plus.net>

(I thought I'd already posted something on this, but it's not
appeared. Sorry if this is a duplicate).

I'm having trouble getting 4Suite installed on an MS Windows / Python
2.2 setup. So much so that I've even gone to the extent of removing
all traces of Python from my machine, then reinstalling just Python
and 4Suite.

Having done that, the 4Suite howto says to check it by typing
	from xml.xslt import Processor
from a Python prompt. Unfortunately I just get a "no module named
xslt" error. Python seems to be looking at the standard Lib\xml
package, and isn't finding 4Suite's Lib\site-packages\Ft\Xml directory
(from Xml.Xslt doesn't work either, BTW). 

Could anybody help me with the correct way to tell Python to look for
the 4Suite stuff? Pretty please?


From ken at hotmail.com  Sat Jun  1 04:16:28 2002
From: ken at hotmail.com (Ken)
Date: Sat, 1 Jun 2002 18:16:28 +1000
Subject: How to open a HTML file when the python cgi program is executing?
References: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de> <ad9uv4$g2n$2@agate.berkeley.edu>
Message-ID: <ad9vo8$utftp$1@ID-49758.news.dfncis.de>

"Daniel Yoo" <dyoo at hkn.eecs.berkeley.edu> wrote in message
news:ad9uv4$g2n$2 at agate.berkeley.edu...
> Ken <ken at hotmail.com> wrote:
> : How do I open a HTML file when the cgi program is executing?
>
> : The only method I can think of, which doesn't work is this:
> : <meta HTTP-EQUIV="REFRESH" CONTENT="0; URL=main.html">
>
> Not quite sure I understand yet.
>
>
> Do you have a file called 'main.html' that you're trying to display
> from your CGI program?  I just want to make sure we're work on the
> same problem here before babbling.  *grin*
>
>
> Talk to you later!

Yes, I want to display the content of another html file (main.html in this
example) onto the browser. How do I do that? And also, how do I set target
to which frame to display?

Thanks




From chris.gonnerman at newcenturycomputers.net  Sun Jun  2 01:23:51 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sun, 2 Jun 2002 00:23:51 -0500
Subject: writing to a file (atomic operation)
References: <20020602050909.GA7192@mercury.senux.com>
Message-ID: <002801c209f5$b2da8ca0$0101010a@local>

----- Original Message ----- 
From: "Brian Lee" <senux at senux.com>


> Is this a atomic operation?
> 
> open('/tmp/test.log', 'a').write('test\n')
> 
> from py newbie.

No.  Why exactly do you need it to be?  "Atomic" means that it can't be
interrupted by another process before completion; not many operations
are truly atomic in modern OS design. 

Very little of what you do in Python will be truly atomic.

Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net





From kragen at pobox.com  Sun Jun  2 21:29:43 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 02 Jun 2002 21:29:43 -0400
Subject: variable X procuct - [(x,y) for x in list1 for y in list2]
References: <1022619068.34271@newsmaster-04.atnet.at> <1022686332.297659@newsmaster-04.atnet.at> <20020531215914.15202822.larooy@xtar.co.nz>
Message-ID: <834rgltaig.fsf@panacea.canonical.org>

John La Rooy <larooy at xtar.co.nz> writes:
> Perhaps the translation is correct, but the example you give is a
> cartesian product, not a cross product.  From my algebra book...

Cartesian products are often called cross products.



From sholden at holdenweb.com  Tue Jun 18 16:50:13 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 18 Jun 2002 16:50:13 -0400
Subject: python cgi + IIS
References: <3D0F1813.3010304@gmx.de>
Message-ID: <x0NP8.165989$Fp1.91529@atlpnn01.usenetserver.com>

"Axel Grune" <NineOfSix at gmx.de> wrote in message
news:3D0F1813.3010304 at gmx.de...
> Hi, I'm trying to get my IIS (Win2k Pro) execute my python scripts.
> I added the 'App Mapping' info for '.pycgi' files in the properties of
> IISAdmin (Computermanagement) as I read in the internet. The string
> for the executable file is 'D:\Python22\python.exe -u %s %s' (the path
> is valid).
> But when i try to call it within my browser, it just displays it like a
> normal text file instead of executing it.
>
>
> the script (its called test.pycgi):
> #!/usr/bin/python
>
> print "Content-Type: text/plain\n\n"
> print "Hallo Welt"
>
>
> has anyone an idea?
>
You have almost certainly made a typo, or you didn't apply the changes to
the directory you are trying to run the CGI from. I just followed the steps
at

    http://www.e-coli.net/pyiis.html

to install .cgipy as a Python-executing extension, and everything worked --
the following program printed out a complete list of environment variables.
#
# Python CGI script
#
import os
print """Content-Type: text/html

<html>
<head>
  <title>Python Web Page</title>
</head>
<body>
"""

for k, v in os.environ.items():
    print "%s=%s<br>\n" % (k, v)

print """
</body>
</html>
"""

Personally I'd rather run CGI than use ASP under IIS, so I hope you can get
it to work. Clearly at the moment IIS isn't recognising the .cgi extension
as needing to run the Python interpreter.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From zopestoller at thomas-guettler.de  Wed Jun 26 04:03:08 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Wed, 26 Jun 2002 10:03:08 +0200
Subject: Parsing strings (\n and \\)
References: <mailman.1025016731.901.python-list@python.org>
Message-ID: <3D19753C.9070104@thomas-guettler.de>

Mark McEahern wrote:

>>Is there a function for parsing c-like
>>strings? I need to parse strings containing
>>'\n' and '\\' '\"'.
>>
> 
> I suppose it's obvious to anyone from a C background, but I find myself
> wondering, "What do you mean by parse?"
> 
> I might be able to help if you provide a little more detail.


Look at the two functoins quote and unquote. I wrote them
without regular expression because I think it faster.

def quote(string):
     '''
     foo\bar --> foo\\bar
     foo"bar --> foo\"bar
     '''

     new_string=[]
     for char in string:
         if char=='\\':
             new_string.append('\\\\')
         elif char=='"':
             new_string.append('\\"')
         else:
             new_string.append(char)
     return ''.join(new_string)

def unquote(string):
     '''
     foo\\bar --> foo\bar
     foo\"bar --> foo"bar
     '''
     new_string=[]
     i=0
     len_str=len(string)
     while 1:
         if i==len_str:
             break
         char=string[i]
         if char=='\\':
             i=i+1
             if i==len_str:
                 raise 'Parse Error: Backslash at end of string'
             char=string[i]
             if char=='\\':
                 new_string.append('\\')
             elif char=='"':
                 new_string.append('"')
             else:
                 raise 'Parse Error: Unsupported character after backslash'
         else:
             new_string.append(char)
         i=i+1
     return ''.join(new_string)

def test_quote():
     strings=['foo', '', '\\', ' ', '"', '\\"', '\\\\']
     for s in strings:
         if unquote(quote(s))!=s:
             print "Error: s: %s quoted: %s unquoted: %s" % (s, quote(s),
                                                             unquote(quote(s)))




From HH at attbi.com  Thu Jun 27 17:48:58 2002
From: HH at attbi.com (HH at attbi.com)
Date: Thu, 27 Jun 2002 21:48:58 GMT
Subject: Distutils RPM Requires tag
Message-ID: <slrnahn22d.1no.id@c1299689-a.saltlk1.ut.home.com>

Is there any way to get distutils to insert a proper Requires: tag in
a bdist_rpm generated RPM for packages which Require more than one
package?  For example, I'd like a Requires tag that looks like:
Requries: foo >= .96, bar >= 1.0.2

but when I try
python setup.py bdist_rpm --requires 'foo >= .96, bar >= 1.0.2'

the generated Requires tag turns the comma into a space, and RPM
subsequently makes a package which Requires only foo.  There does not
seem to be any way around this.  Is this a bug, or is there a solution
I have overlooked?

Thanks for your help.
-- 
HH


From a at a.com  Mon Jun 17 08:13:59 2002
From: a at a.com (Adam)
Date: Mon, 17 Jun 2002 13:13:59 +0100
Subject: urllib2 getting a webapage authenticating proxy...
Message-ID: <3D0DD287.7B221A42@a.com>

this code doesnt work for me... i have replaced the proxy details for
security... can anyone help... as i havent been able find any helpfull
websites/examples for urllib2... is urllib2 what i realy want to use for
this...

import urllib2

def get_page(url):
 proxy =
urllib2.ProxyHandler({'http':'http://userid:pass at proxy.com:port/'})
 opener = urllib2.build_opener(proxy)
 site = urllib2.urlopen(url).read()
 print(site)

get_page('http://www.python.org/')



From aldo at nullcube.com  Fri Jun 28 06:04:22 2002
From: aldo at nullcube.com (Aldo Cortesi)
Date: Fri, 28 Jun 2002 20:04:22 +1000
Subject: metaclasses vs. inheritance
In-Reply-To: <ghTS8.58749$mh.1826422@news1.telusplanet.net>
References: <ghTS8.58749$mh.1826422@news1.telusplanet.net>
Message-ID: <20020628100422.GA18469@nullcube.com>

Thus spake Ian McMeans (imcmeans at home.com):

> Can someone give an example of a problem that is solved
> with metaclasses, which would be difficult or impossible
> to solve with inheritance?


There are some things that can only be done with metaclasses
- changing the text representation of a class, for instance.


class Extended(type):
    def __repr__(self):
        attrs = []
        for i in dir(self):
            attrs.append("%-20s\t\t%s" % (i, type(self.__getattribute__(self, i))))
        return "%s:\n\t%s" % (type.__repr__(self), "\n\t".join(attrs))
                                            
class Foo(object): 
    __metaclass__ = Extended
    pass
print Foo



Baroque, I know, but it illustrates the point. 



> PS - what's the antonym of meta?

Hmmm... the meta- in "metaclass" or "metaphysics" really has
no relation to the original Greek, which means "with", or
"among" (as in "metatarsus" - the bones in the foot which
lie next to the tarsus). So whatever antonym you come up
with can't have anything to do with the original etymology
of the prefix...  Erm... How about mundane? Or actual?
Intrinsic? Essential? Or even "per se"....?




Cheers,




Aldo





-- 
Aldo Cortesi
aldo at nullcube.com




From shredwheat at attbi.com  Sun Jun  2 12:44:20 2002
From: shredwheat at attbi.com (Pete Shinners)
Date: Sun, 02 Jun 2002 16:44:20 GMT
Subject: SolarWolf 1.1 Released
Message-ID: <3CFA4B52.9020901@attbi.com>

SolarWolf is a frantic arcade game written in Python, using the Pygame library.

Version 1.1 adds easier gameplay and packs many new features to make the 
game funner. New is stereo sound, support for 8bit display, in-game 
"comments", and better levels. We've also moved to a website that will be 
online more than 2 days per month.

SolarWolf is the recreation of the classic game, SolarFox (as played on my 
Atari 2600). You pilot a ship through 48 levels. You must collect the boxes 
on screen to advance, and each level provides more challenging layouts. To 
keep things difficult, enemies patrol the edges of the screens. They seem 
to really enjoy shooting, and in the later levels they become more determined.

SolarWolf requires Pygame-1.5 and Python-2.0. Pygame requires the use of 
SDL, and the SDL support libraries SDL_ttf, SDL_image, and SDL_mixer. It 
runs on just about every platform available. (with osx support coming as 
soon as the latest pygame is compiled for osx)

SolarWolf Website:
http://pygame.org/shredwheat/solarwolf



From thenault at nerim.net  Mon Jun 17 04:27:22 2002
From: thenault at nerim.net (Sylvain =?iso-8859-1?Q?Th=E9nault?=)
Date: Mon, 17 Jun 2002 10:27:22 +0200
Subject: [ANN] PyReverse 0.3
In-Reply-To: <GCEDKONBLEFPPADDJCOECEECDJAA.whisper@oz.net>
References: <mailman.1024086967.9145.clpa-moderators@python.org> <GCEDKONBLEFPPADDJCOECEECDJAA.whisper@oz.net>
Message-ID: <20020617082722.GA2359@ntw23.fr>

On Friday 14 June ? 14:38, David LeBlanc wrote:
> On windows, "setup install" terminates with:
> error: file 'logilab\pyreverse\bin\py2tests.bat' does not exist
> 
oops ! a little a packaging problem... It will be fixed as soon as
someone at logilab can commit the required fix and put a 0.3.1 on the
ftp, since I have no access to do this right now. Until this moment 
you can put the files joined to this message in the 
logilab/pyreverse/bin directory and the installation process should 
work correctly.

sorry for this annoyance

-- 
Sylvain Th?nault 
-------------- next part --------------
#!/usr/bin/env python

from logilab.pyreverse import py2tests
import sys 
py2tests.Run(sys.argv[1:])
-------------- next part --------------
A non-text attachment was scrubbed...
Name: py2tests.bat
Type: application/x-msdos-program
Size: 118 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020617/ed326cd8/attachment.bat>
-------------- next part --------------
#!/usr/bin/env python

from logilab.pyreverse import py2xmi_uml
import sys 
py2xmi_uml.Run(sys.argv[1:])
-------------- next part --------------
A non-text attachment was scrubbed...
Name: py2xmi.bat
Type: application/x-msdos-program
Size: 122 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020617/ed326cd8/attachment-0001.bat>

From marklists at mceahern.com  Fri Jun 28 11:20:43 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 28 Jun 2002 10:20:43 -0500
Subject: metaclasses vs. inheritance
In-Reply-To: <ghTS8.58749$mh.1826422@news1.telusplanet.net>
Message-ID: <NCBBLFCOHHDIKCAFGCFBKEPLLGAA.marklists@mceahern.com>

> Can someone give an example of a problem that is solved with metaclasses,
> which would be difficult or impossible to solve with inheritance?

You can intervene in class creation to do aspect-oriented type stuff.  E.g.,
before each method call, do this, after each method call, do that.  Etc.  I
will post a more detailed example as the start of a separate thread shortly.

// m

-





From bokr at oz.net  Thu Jun  6 16:37:02 2002
From: bokr at oz.net (Bengt Richter)
Date: 6 Jun 2002 20:37:02 GMT
Subject: Using AT on XP to run python script - MARK HAMMOND PLEASE READ
References: <mailman.1023320948.13063.python-list@python.org>
Message-ID: <adoh5e$3lt$0@216.39.172.122>

On Wed, 5 Jun 2002 16:48:03 -0700, "David LeBlanc" <whisper at oz.net> wrote:

>This is most curious, but I believe the commands _are_ getting executed!
>
>I tried the following (assume time is 16:30) "at 16:32 pythonw
>j:\python22\tools\idle\idle.py" (without the qoutes of course). At the
>appointed time, there was disk activity, but no idle window popped up (idle
>was chosen just to get something to pop up!). However, there _was_ a pythonw
>process running in the task manager! I think there might be a way to attach
>a console or get access to the GUI, but i'm not sure - maybe Mark Hammond
>might have a clue (he being the Python Win guru :) ).
>
>I tried the following:
>#at_test.py
>import sys
>import time
>
>tst = open("testfile.txt", 'a+')
>
>tst.write(time.asctime() + "\n")
>
>tst.close()
>
>put into an 'at' job as: at xx:xx python j:\python22\at_test.py
>
>I expected the output from this to be in j:\python22, but to my surprise, it
>was in c:\winnt\system32 where cmd.exe resides! testfile.txt contained the
>exact time the 'at' job was set to run at too!
>
>I suspect your last example that produced no visible output did in fact work
>:)
>
>In any case, it works - now if one could just get visible output!
>
I's suggest wrapping your python invocation in a .cmd file something like:
--< todoat.cmd >--
c:
cd \pywk
cmd/k D:\python22\python hw.py
------------------

Here hw.py is in c:\pywk
--< hw.py >-------
print 'Hello World from hw.py\n'
------------------

Then you can do this:
==========
[13:19] C:\pywk>at 13:20 /interactive c:\pywk\propt\todoat.cmd
Added a new job with job ID = 1

[13:19] C:\pywk>at
Status ID   Day                     Time          Command Line
-------------------------------------------------------------------------------
        1   Next 6                  13:20         c:\pywk\propt\todoat.cmd

[13:19] C:\pywk>
==========
and have a separate console window pop up (and stay), with this in it:
(Your paths may vary, of course)
==========

C:\WINNT\system32>c:

C:\WINNT\system32>cd \pywk

C:\pywk>cmd/k D:\python22\python hw.py
Hello World from hw.py


C:\pywk>
==========

Some things to note:

The console window popped up, but the path and default directory
have nothing to do with where I did the AT command. The title
of the new console window is C:\WINNT\system32\MSTask.exe.
and the user is the system, with all that may or may not imply ;-)


>David LeBlanc
>Seattle, WA USA
>
>> -----Original Message-----
>> From: python-list-admin at python.org
>> [mailto:python-list-admin at python.org]On Behalf Of Chris Stromberger
>> Sent: Tuesday, June 04, 2002 17:35
>> To: python-list at python.org
>> Subject: Using AT on XP to run python script
>>
>>
>> I'm having trouble getting AT on XP to run a python script for me.
>> Have tried:
>>
>> at xx:xx "c:\python21\python.exe c:\scripts\script.py"
>> at xx:xx "start c:\scripts\script.py"
>> at xx:xx "c:\scripts\script.py"
>>
>> and get "the system cannot find the file specified" for the first two.
>> The last version does nothing apparently--no error indication or
>> anything.
>>
>> I imagine it is related to user accounts and such.  Anyone have any
>> pointers?
Your environment and path are not used by the AT command, so use a .cmd
file that sets things up as needed and then invokes python or pythonw
as needed, depending on whether you want a silent background job or
interaction with whoever happens to be around ;-)
(I'm not sure what happens if no one is logged in. Presumably /interactive
shouldn't work?)

Regards,
Bengt Richter


From mickey at tm.informatik.uni-frankfurt.de  Sun Jun 16 06:50:15 2002
From: mickey at tm.informatik.uni-frankfurt.de (Michael 'Mickey' Lauer)
Date: 16 Jun 2002 12:50:15 +0200
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de> <aeftmj$m29$1@bob.news.rcn.net>
Message-ID: <3d0c6d67@nntp.server.uni-frankfurt.de>

David Abrahams <david.abrahams at rcn.com> wrote:
> With Boost.Python, what you write looks like a kind of IDL (interface
> description language), which just happens to be built out of C++ source
> code. Lots of jobs (function argument type checking, conversion, arity
> checking) are done automatically, but the library doesn't decide which
> elements to wrap. Unlike SWIG and a few others, there's no attempt to parse
> your source code. Instead, the compile-time introspection capabilities of
> C++ are exploited to automatically build wrappers from function and member
> function pointers.

I see. So it seems in order to get a nice object oriented Python interface,
I can 1.) use SWIG to parse the header file and automatically produce a wrapper
      1.1.) and then use this low-level wrapper to write some high-level python classes around it or
      2.) use Boost.Python and do the two steps in one by writing a "handcrafted" interface?

How would sip come into this picture? Is it more like SWIG for C++ or more like Boost.Python?
(Apart from not beeing concerned about the total lack of documentation ;)

Yours,

:M:



From llothar at web.de  Wed Jun 12 18:27:19 2002
From: llothar at web.de (Lothar Scholz)
Date: Thu, 13 Jun 2002 00:27:19 +0200
Subject: Wanted:  Potential Python Authors
References: <ae7gke$mhb$1@ccis12.baylor.edu>
Message-ID: <db3fgu40sbnt3pno5v1i2h9lnmpcoti2bk@4ax.com>

On Wed, 12 Jun 2002 08:03:35 -0500, "Jeff Donahoo"
<Jeff_Donahoo at Baylor.edu> wrote:

>I'm looking for somebody to author a short book on Python in my programming
>technology series.  Who do you think would make a great author that hasn't
>already written a book on Python?  Candidates include:

>- Competent and frequent posters to this newsgroup
This can be an indicator. 

>- Author of articles in the various trade magazines
You really mean "trade" magazines ?
Normally they are written by marketing peoples who don't
have the competence to write more then a 4 page overview.
Hands off !!!!

>- Expert programmer
No, definitely not. Most programmers are not trained on writing
tutorials or books - and normally they are really bad.
Look more for someone trained as technical writer.




From cpbotha at i_triple_e.org  Tue Jun 11 06:19:18 2002
From: cpbotha at i_triple_e.org (Charl P. Botha)
Date: Tue, 11 Jun 2002 10:19:18 +0000 (UTC)
Subject: Sorting list (maybe stupid)
References: <slrnagbh7m.h03.jarvin24@lehtori.cc.tut.fi> <ae4gnp$l18$1@news.tudelft.nl> <ae4hn9$jm4$1@wanadoo.fr>
Message-ID: <ae4ir6$lq0$1@news.tudelft.nl>

In article <ae4hn9$jm4$1 at wanadoo.fr>, Eric Brunel wrote:
> Charl P. Botha wrote:
> 
>> In article <slrnagbh7m.h03.jarvin24 at lehtori.cc.tut.fi>, Jarvinen Petri
>> wrote:
>>> I have a problem and can't solve it:
>>> 
>>> There is lines of text like:
>>> 
>>> 2. Header2
>>> 1. Header1
>>> 102. Header102
>>> 
>>> and so on...
>>> 
>>> And I need to sort them according to the number (1,2,3,4...)
>> 
>> Put each line in a tuple, first element the number, second element the
>> header.  Make a list of these tuples, and then use list.sort() with a comp
>> function that compares using the first element in the tuple.
> 
> I think it's useless. Just try:

I would NOT use the word "useless" here.  You were thinking of
"unnecessary", probably.  Please don't go into politics, you are bound to
cause wars with your choice of words. ;)

> l = [(102, 'foo'), (2, 'bar'), (1, 'baz')]
> l.sort()
> print l
> 
> Apparently, the default sort function already sorts tuples according to 
> their first element...

Making an assumption based on this single example is very dangerous.
However, it seems you were lucky in this case.  See:
http://manatee.mojam.com/~skip/python/fastpython.html (the section on
Sorting).

-- 
charl p. botha http://cpbotha.net/ http://visualisation.tudelft.nl/


From rdsteph at earthlink.net  Thu Jun 27 20:00:42 2002
From: rdsteph at earthlink.net (Ron Stephens)
Date: Fri, 28 Jun 2002 00:00:42 GMT
Subject: Be the first on your block to register your computer.
References: <mailman.1025211444.21440.python-list@python.org> <afg1ks$e6lj5$1@ID-71831.news.dfncis.de>
Message-ID: <3D1B6E69.1090204@earthlink.net>

Well, Python just might not be one of the Microsoft approved programs 
allowed to run. In that case, you would not be able to use Python  at all.

Secondly, the Microsoft plan is so diabolically evil, and yet at the 
same time it is conceivable that it could be pushed into reality by a 
powerful corporate oligarchy in conjunction with a corrupt US government 
and an apathetic populace, that one  might suppose all decent humans 
would find the whole thing to be a gross affront to human dignity, and 
thus appreciate being made aware of the whole mess by David Leblanc's 
timely heads-up on comp.lang.python...



From opengeometry at NOSPAM.yahoo.ca  Wed Jun 26 19:04:10 2002
From: opengeometry at NOSPAM.yahoo.ca (William Park)
Date: 26 Jun 2002 23:04:10 GMT
Subject: Integrals with Python
References: <3D188551.6BA00DED@lanl.gov> <3D189307.1417ECD5@info.unicaen.fr> <3D18A010.CAF3FA9E@lanl.gov>
Message-ID: <afdh99$drql8$1@ID-99293.news.dfncis.de>

Nathan Given <ngiven at lanl.gov> wrote:
> Matlab has quad, quadl, and quad8...
> 
> 
> does Numeric have anything like those?
> 
> 
> 
> Here is my current (lame) project...
> 
> I am comparing the Monte Carlo method for single integrals against the
> answers given by quad, quadl, and quad8 in Matlab.  I coded everything
> in matlab, and was able to do it just fine... the problem is that it was
> SLOW... so, someone suggested that I run python, so I converted all my
> code to python (except the quad, quadl, and quad8 parts), and was able
> to run my code and get the monte carlo answer much faster....
> 
> now, i am going to be doing some double integrals, and I was just
> wondering if python had some built in functions that will calculate some
> double integrals (and single integrals) for me... if there are different
> methods, great... I will be able to compare the monte carlo answer to
> different methods of calculating the integral...

Since the key routine in Monte-Carlo is evaluation of f(), your program
would remain the same for 1-D and 2-D.

As for different integration methods, there are all sorts of C/Fortran
routines for 1-D which you can translate to Python.  Long ago, I wrote 2
Simpson's rule in Python.  For 2-D, however, your choice is a bit limited.
I only know of few approximations based on nicely divisible domains.

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin


From d2002xx at myrealbox.com  Tue Jun 11 23:05:09 2002
From: d2002xx at myrealbox.com (d2002xx)
Date: 11 Jun 2002 20:05:09 -0700
Subject: JIT Complier for python?
Message-ID: <8d3f4438.0206111905.7934664b@posting.google.com>

Is there any JIT Compiler for Python?


From duncan-news at grisby.org  Wed Jun 12 10:10:23 2002
From: duncan-news at grisby.org (Duncan Grisby)
Date: Wed, 12 Jun 2002 14:10:23 GMT
Subject: ANN: omniORB 3.0.5 / omniORBpy 1.5 released
Message-ID: <jDIN8.12297$sC4.1584555@news6-win.server.ntlworld.com>

I am pleased to announce the release of omniORB 3.0.5 and omniORBpy 1.5.

The releases are minor updates to the previous versions, containing
bug fixes and ports to new platforms, but no significant new features.

These releases are the first new versions of omniORB since it became
an independent project, following the closure of AT&T Laboratories
Cambridge.

omniORB and omniORBpy are high performance, robust CORBA ORBs for C++
and Python. They are available free under the terms of the GNU LGPL
and GPL. You can find more information at

  http://omniorb.sourceforge.net/

You can download the new versions by following links from that web
page, or directly from

  http://sourceforge.net/project/showfiles.php?group_id=51138

Many thanks to all the people who have contributed to these releases,
either by reporting bugs, providing ports to new platforms, or by
building binaries.


Enjoy!

Duncan.

-- 
 -- Duncan Grisby         --
  -- duncan at grisby.org     --
   -- http://www.grisby.org --


From tchur at optushome.com.au  Sat Jun  8 05:17:05 2002
From: tchur at optushome.com.au (Tim Churches)
Date: Sat, 08 Jun 2002 19:17:05 +1000
Subject: Creating Dynamic Web Pages (ZOPE)
References: <Pine.LNX.4.21.0206081919430.1782-100000@localhost.localdomain>
Message-ID: <3D01CB91.8B98F2D7@optushome.com.au>

Michael Hall wrote:
> 
> I notice no-one has suggested good ol' CGI. For a relatively simple job
> such as this, a few basic Python CGI scripts, some templates and MySQLdb
> could also be used ... oh, just noticed the MS-SQL database. There is
> probably a module around to cope with MS-SQL, no?

Object Craft have a Python wrapper for MS SQL Server. See
http://www.object-craft.com.au
And while you are there, have a look at their Albatross CGI-based
stateful Web application 
framework, which could be just what is required for the rest of the Web
site envisaged.
In my experience, nearly all Web applications and database-driven Web
sites end up needing 
to retain state information sooner or later. Might as well start
with a framework which can go the distance. Albatross is especially nice
in this respect because
you can choose alternative application models, starting with "SimpleApp"
which is just what the
name suggests, and later progress to more complex application models
through mixins without
having to start from scratch or throw away your existing code. 

Tim C

> 
> Mick
> 
> On 7 Jun 2002, Bill Tate wrote:
> 
> > hwcowan at hotmail.com (Hugh Cowan) wrote in message news:<46ca81a0.0206062016.789d3af6 at posting.google.com>...
> >
> > [snip]
> > > My scenario is typical where I need to put together a company Intranet
> > > with a few front end Web-Pages where users can select information,
> > > various links, and enter search criteria.  The information is then
> > > retrieved from the MS-SQL Server and the results are sent back to the
> > > user.  It's nothing fancy, no shopping carts, or remembering sessions
> > > states, etc.. just really a simple front-end interface for
> > > non-technical people to easily retrieve information.
> > >
> > > I don't know if ZOPE would fit my situation or not.  I know that it is
> > > / can be used to create complex E-commerce sites, but I am wondering
> > > if it is overkill for what I need done.  Would I be better off trying
> > > to find some sort of RAD Web Tool instead (anyone have any good
> > > suggestions) ?
> >
> > I think its safe to say that there are any number of tools that could
> > fit the bill here including Zope.  There are aspects of zope that can
> > make your life much easier (e.g., zsql methods) but it does take some
> > time to get used to it and everything it offers.  Several recent books
> > on the subject I think would be very helpful to you in this regard.
> > If you go to zope.org, there is some information on using WYSIWYG
> > tools such as Dreamweaver with zope that may also be of interest to
> > you as well.
> >
> > I think zope works bests for you over the long haul where it is
> > important to keep things well organized and well structured.  The
> > ability to collaborate amongst individuals is also a great strength.
> > If these issues are important to you, then I would say that Zope would
> > be an excellent choice.
> >
> > Good luck!!!
> >
> 
> --
> --------------------------------
> n   i   n   t   i  .   c   o   m
> php-python-perl-mysql-postgresql
> --------------------------------
> Michael Hall     ninti at ninti.com
> --------------------------------
> 
> --
> http://mail.python.org/mailman/listinfo/python-list




From Chris.Barker at noaa.gov  Mon Jun  3 13:30:29 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Mon, 03 Jun 2002 10:30:29 -0700
Subject: Speeding up the Canvas Widget
References: <ad8jif$dki$1@netnews.upenn.edu>
Message-ID: <3CFBA74E.C5EB9213@noaa.gov>

Salim Zayat wrote:
> I need to draw, and almost constantly redraw, upwards of like
> 10,000 or so things to the canvas (lines, text, and rectangles, etc.).
> The larger the number of things to be drawn, the slower it is.
There are apparently some tricks that can be used to speed some of thee
kinds of operations up. I'm not a TKuser, but there wasa a discusson
quite a while back (2 yrs?, I cna't find it in the archives, it may
predate them) on the wxPython newsgroup about drawing 10,000 rectangles,
and someone had found some tricks to make it go much faster with TK.

Since then, wxPython has added a couple of "Draw***List()" methods, that
draw a whole bunch of points or lines as a single Python call. It speeds
up the process of drawing losts of similar items considerably, by
putting the loop in C rather than Python. At the moment, only Points and
Lines are covered, but there is talk of adding all the draw objects,
someone just needs to get around to writing the code..(hint, hint)

I have no idea if pyGTK or pyQT support a call to draw lots of objects
with a single Python call, but that may be neccesary to get the speed
you want.

Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From tjreedy at udel.edu  Wed Jun 12 16:45:29 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 12 Jun 2002 20:45:29 GMT
Subject: methods to parse big foreign ascii-file?
References: <mailman.1023899956.13534.python-list@python.org>
Message-ID: <JpON8.146413$4i.13338500@bin2.nnrp.aus1.giganews.com>

"holger krekel" <pyth at devel.trillke.net> wrote in message
news:mailman.1023899956.13534.python-list at python.org...
> Hello,
>
> tommorrow i want to convince an old skool C-programmer to
> use python. We are going to make a one-day project :-)
>
> It involves reading 50MB+ files of a strange ascii-format
> containing names, floats and flags all around. These items
> span multiple lines and should be grouped in lists or objects.
>
> How would you go about it? Options that come to my mind:

IF the fields are comma (delimiter) separated, but you need to ignore
commas (delimiters) within "quoted text fields", there is at least one
Python module that should help (including dealing with multiline
fields).  Search google clp archive or ask.  IF not, consider the
other suggestions.

TJR





From cliechti at gmx.net  Sat Jun 22 20:42:41 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 23 Jun 2002 02:42:41 +0200
Subject: TimeoutQueue.py
References: <mailman.1024698810.22387.python-list@python.org>
Message-ID: <Xns92361C045B3C6cliechtigmxnet@62.2.16.82>

Tim Peters <tim.one at comcast.net> wrote in
news:mailman.1024698810.22387.python-list at python.org: 

> [Chris Liechti]
>> Anyone ever needed a Queue with timeout?
> 
> Yes.  I encourage you to submit a patch to SourceForge, ...

patch 572628 :-) that one uses additional 'timeout' args to get() and new 
to put() too. the old behaviour is still maintained when not using the new 
argument (unlike the original post which was not backwards compatible).

chris

-- 
Chris <cliechti at gmx.net>



From tim.one at comcast.net  Tue Jun  4 19:50:17 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 04 Jun 2002 19:50:17 -0400
Subject: 'for every' and 'for any'
In-Reply-To: <20020604072437.GA21777@strakt.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEIAPKAA.tim.one@comcast.net>

[Andrae Muys]
> Now I just don't understand how this position can possibly make sense?
> C dosn't have any 32-bit int types standardised, so what do _you_ use
> when you need one?  Please don't tell me you don't use int, or I'll
> have to hunt you down and put you out of your misery ;).

[Martin Sj?gren]
> Doesn't C99 specify <stdint.h>?

Yes.

> With int8_t, int16_t, ..., uint8_t, uint16_t, ...

Not necessarily.  All of the exact-size stdint gimmicks are optional -- an
implementation need not supply any of them.  This simply reflects reality.
There's a slew of other names, like int_least8_t and int_least32_t, that are
required, and those guarantee to resolve to an integer type *at least* big
enough.

In the Python source, we spell int_least32_t "long", since even C89 required
that the platform long hold at least 32 bits' worth of stuff.  People who
think they need a type exactly N bits wide are people who've never thought
about it, or are missing some neurons <wink -- but >= N is easy to live with
provided you're not trying to optimize for space, or relying on woefully
unportable casting tricks).

Well, OK, the Python source does assume in several places that "a byte" is
an 8-bit thingie.  The saving grace there is that any platform on which
that's false will never get popular enough to attract a significant user
base.

19-bit-microcomputers-are-a-thing-of-the-past-ly y'rs  - tim






From jonathan at onegoodidea.com  Wed Jun 26 05:20:38 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Wed, 26 Jun 2002 10:20:38 +0100
Subject: Suggestions for good programming practices?
References: <mailman.1024951532.5616.python-list@python.org> <af8n9i$3at$1@panix1.panix.com> <afafee$fc$0@216.39.172.122> <afaohi$682$1@panix1.panix.com> <afb3a2$skm$0@216.39.172.122>
Message-ID: <B93F45F6.CEA6%jonathan@onegoodidea.com>

On 26/6/2002 1:53, in article afb3a2$skm$0 at 216.39.172.122, "Bengt Richter"
<bokr at oz.net> wrote:

> Well, I meant using *args to allow detecting actual "no default-overriding
> argument passed", in place of x=None. The arg list items (only x here)
> don't have to remain unnamed long. E.g.,
> 
>>>> def foo(*args):
> ...     if len(args)==1:
> ...         print "We know we have one non-default arg: %s" % `args[0]`
> ...         x = args[0]
> ...     elif len(args)==0:
> ...         print "We can supply a default arg, even None"
> ...         x = None
> ...     else: raise TypeError,'foo() takes 0 or 1 arguments (%d given)' %
> len(args)
> ...     print 'Effective arg was: %s' % `x`
> ...

I find this stuff much easier to write with a helper function:

>>> def shift():
...     global _0
...     __, _0 = _0[0], _0[1:]
...     return __
... 

Then I can write the function 'foo' much more naturally as:

>>> def foo( *args ):
...     global _0; _0 = args
...     try:
...         x = shift()
...     except IndexError:
...         print 'We can supply a default arg, even None'
...         x = None
...     else:
...         try:
...             shift()
...         except IndexError:
...             print 'We know we have one non-default arg: %s' % `x`
...         else:
...             raise TypeError( 'foo() takes 0 or 1 arguments' )
...     print 'Effective arg was: %s' % `x`
...

This can be further simplified by removing the '*args' nonsense altogether
and defining another helper function:

>>> def p( f ):
...     def _p( *args ): global _0; _0 = args; f()
...     return _p
...
>>> def foo():
...     try:
...         x = shift()
[etc]
...
>>> p(foo)( 'hello world' )



How did we get onto this from 'Suggestions for good programming practices'?

Jonathan



From mwh at python.net  Fri Jun  7 10:20:27 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 7 Jun 2002 14:20:27 GMT
Subject: Memory leaks when  embedding python (debug-version)  ?
References: <1eed5b9.0206070157.6d281f65@posting.google.com>
Message-ID: <lkptz3kw5v.fsf@pc150.maths.bris.ac.uk>

Sarah_L at gmx.net (Sarah Mayer) writes:

> Hi,
> maybe someone could help me,
> 
> I just embedded python 2.2.1 into a dll , 
> 
> Program looks like:
> 
> MyClass::MyClass()
> {
>     Py_Initialize();
>     Py_Finalize();
> }
> 
> 
> and after the program finishes I get a lot of memory-leaks.
> What is missing ???

The interned string dict?  There's some call for getting rid of that,
but I can't remember it off hand.

Cheers,
M.

-- 
  If you're talking "useful", I'm not your bot.
                                            -- Tim Peters, 08 Nov 2001


From bdesth at nospam.free.fr  Sun Jun 23 16:12:17 2002
From: bdesth at nospam.free.fr (laotseu)
Date: Sun, 23 Jun 2002 16:12:17 -0400
Subject: [newbie] import fails first time, then works :-[
References: <3D1379D1.5020705@nospam.free.fr> <3D132C65.34B38F8C@engcorp.com> <3D138E31.4070409@nospam.free.fr> <3D13D104.A49AF73C@engcorp.com>
Message-ID: <3D162BA1.9040103@nospam.free.fr>

Peter Hansen wrote:
> laotseu wrote:
> 
>>Well... I added '/usr/local/lib/lib-dynload' to my PYTHONPATH and
>>'import' does not complain anymore. Everything seems just fine now. But
>>that is puzzling me ! [being a python newbie,] I thought the
>>'/lib-dynload/' directorie, that is in the '/lib/python<version>/'
>>directories, was automagically searched for external modules. Was I wrong ?
> 
> 
> Searching the executable with "strings" and "grep", I find references
> to /usr and /lib/python2.0 (in my case) and both /lib/lib-dynload 
> and /lib-dynload.  When I run python and check sys.path I get
> 
> ['', '/usr/bin', '/usr/lib/python2.0', '/usr/lib/python2.0/plat-linux2', 
> '/usr/lib/python2.0/lib-tk', '/usr/lib/python2.0/lib-dynload', 
> '/usr/lib/python2.0/site-packages'] 
> 
> All with default values, under RedHat 7.1.  Not sure why you've
> got troubles... Linux is still Black Magic much of the time to me. :)
> 
> -Peter

I guess I'd better make a new clean install of python... Anyway version 
2.2.1 is out so !-)

Ok, thanks for your help.

laotseu
--
Ye Old Master Said :
"Computers are still black magic much of the time"
--



From whisper at oz.net  Thu Jun 27 03:12:43 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 27 Jun 2002 00:12:43 -0700
Subject: How to find out operating system
In-Reply-To: <3D1ACBC5.22482.D4313A4@localhost>
Message-ID: <GCEDKONBLEFPPADDJCOECEHEDMAA.whisper@oz.net>

You might try os.environ:

os.environ['WINOS'] -> WIN2000

I don't know if WINOS is defined on '98 or ME, but it should be good for NT,
2K and XP

os.environ.has_key('WINOS') will tell you that. However, MS os's are so
stupid that there must be some env variable to tell themselves what they
are.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of A
> Sent: Wednesday, June 26, 2002 23:25
> To: python-list at python.org; tutor at python.org;
> activepython at listserv.activestate.com; python-help at python.org
> Subject: How to find out operating system
>
>
> Hi,
> What is the best way of  finding out the kind operating system?
> I can use os.name but if the system is Windows I would like also
> know if the system is Windows98 or Windows ME or W2K or
> Windows XP.
> Thanks for help.
> Ladislav
>
>
> _______________________________________________
> ActivePython mailing list
> ActivePython at listserv.ActiveState.com
> To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
> Other options:
> http://listserv.ActiveState.com/mailman/listinfo/ActivePython
>
>
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From gerhard.haering at gmx.de  Wed Jun 19 18:59:21 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Thu, 20 Jun 2002 00:59:21 +0200
Subject: GOTO w/ Python?
In-Reply-To: <aer1gr$fj1$1@peabody.colorado.edu>
References: <3d10ed95@xpl-sdf1-sec1.> <um7Q8.44210$n4.10292249@newsc.telia.net> <aer1gr$fj1$1@peabody.colorado.edu>
Message-ID: <20020619225921.GC12907@lilith.my-fqdn.de>

* Fernando P?rez <fperez528 at yahoo.com> [2002-06-19 16:44 -0600]:
> Fredrik Lundh wrote:
> 
> > 
> > you can find more Python tutorials here:
> > 
> >     http://www.python.org/doc/Newbies.html
> > 
> 
> I've always been curious. Why isn't the standard tutorial linked to here? It 
> seems like the obvious first link to have, no?

Because it's not well suited for _programming_ newbies? It's, however,
good material if you already know to program.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From mcherm at destiny.com  Fri Jun 14 09:22:52 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Fri, 14 Jun 2002 09:22:52 -0400
Subject: Medium to Large Scale Python Deployments
Message-ID: <3D09EE2C.2070908@destiny.com>

Michael Chermside writes:
> In fact, it might be really nice to collect a list of large (successful)
> Python projects and post it somewhere so the Python community could use
> it for advocacy.

Domenic Merenda writes:
>    Michael, I agree with you.  I was formerly the Vice President of Business
> Development for BeOpen.com, while we had PythonLabs there, and I can tell
> you that we were compiling such a list for that very reason.  Using the
> WayBack Machine, I was able to grab the list.  Here ya go:
       [... excellent list of companies using python ...]

Wow... that's GREAT.

I think this should probably be set up as a list that people can easily 
contribute to to enhance and update. With that in mind, the entries 
should probably have dates, indicating how recently they were updated. 
Do you have an estimate of when these were entered?

Michael Chermside writes:
> Domenic's
>    ERP system   >100      --

Domenic Merrenda writes:
>    Let's not call it that. :-)  And I will check upstairs when we get back
> from our vacation to see if it's okay to release the number of lines in the
> code.  Don't be surprised if it's treated like a National Security issue.
> (Shh!  If they know how many lines are in the code, they can CRACK US!)

This may be a more widespread problem... projects where the project 
owner (the corporation) does not want to release the information. We 
could simply ommit them, or we could list under "<name withheld by 
request>". I would welcome any other suggestions. Of course, the best 
solution is convincing them that publicizing thier efforts is really to 
their benefit, although this may not work in cases like the CIA.

-- Michael Chermside






From martin at v.loewis.de  Fri Jun  7 02:05:37 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 07 Jun 2002 08:05:37 +0200
Subject: minidom and encoding problem
References: <17aafe08.0206051514.4af054d2@posting.google.com> <m37klcubdu.fsf@mira.informatik.hu-berlin.de> <17aafe08.0206061419.6366b12b@posting.google.com>
Message-ID: <m3zny7aaj2.fsf@mira.informatik.hu-berlin.de>

ehab_teima at hotmail.com (Ehab Teima) writes:

> > This is a bug in your code. You must not insert (byte) string in a DOM
> > tree; always use Unicode objects.
> 
> I do not have control over the sent text. 

[I assume that the "sent text" is also the one that you pass to
createTextNode].

Even if you don't have that control, you still need to know what
encoding it uses. If you don't know the encoding, you cannot put it
into XML documents.

> The issue started when some bullets were copied from a word document
> and pasted into a file and the whole file was passed to my
> classes. I cound not find a way to convert this text to UTF-8 or
> anything else.

You don't need to convert it to UTF-8, you need to convert it to
Unicode objects. You can use the unicode() builtin to do that.

> Is there a way to prevent this from happening?

What is "this", and why do you want to prevent it from happening?

Regards,
Martin


From bh at intevation.de  Mon Jun 17 05:35:25 2002
From: bh at intevation.de (Bernhard Herzog)
Date: 17 Jun 2002 11:35:25 +0200
Subject: Using Python to script a game?
References: <cdc2589b.0206141718.7dc2ae13@posting.google.com> <mailman.1024106376.21850.python-list@python.org> <aeedci$29e$1@panix1.panix.com> <3D0D9317.6371F155@del_cygnus-software_del.com>
Message-ID: <6qbsaab63m.fsf@abnoba.intevation.de>

Bruce Dawson <del_brucedawson_del at del_cygnus-software_del.com> writes:

> I did a Game Developers Conference talk on using Python in games.
[...]
> Ultima Online 2 was also rumoured to be using Python.

For this, see
http://python9.org/p9-cdrom/13/index.htm

   Bernhard

-- 
Intevation GmbH                                 http://intevation.de/
Sketch                                 http://sketch.sourceforge.net/
MapIt!                                           http://www.mapit.de/


From vjovanov at stevens-tech.edu  Tue Jun  4 12:15:22 2002
From: vjovanov at stevens-tech.edu (Vojin Jovanovic)
Date: Tue, 4 Jun 2002 12:15:22 -0400
Subject: self
Message-ID: <BI5L8.78$Vv1.4155449@newnews.cc.stevens-tech.edu>

I have been working on a project trying to implement some advanced
functionality.  Now, one thing that has been a constraint for me is the
concept of self.  Why is it that within a class one has to use self. (or
"any word". ) to refer to instance variables? Would it not be more logical
to just use the name of the variable like it is for example in C++?

Vin




From sholden at holdenweb.com  Fri Jun 28 09:51:29 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 28 Jun 2002 09:51:29 -0400
Subject: self destruction :)
References: <f4KS8.1326$qq1.14832@castor.casema.net> <7x6604ctb1.fsf@ruckus.brouhaha.com> <3d1c2d6a$0$231$4d4ebb8e@news.nl.uu.net>
Message-ID: <kQZS8.107949$mF.62458@atlpnn01.usenetserver.com>

"Guyon Mor?e" <gumuz at looze.net> wrote ...
> >   x = [a, b, c]
> >
> > and you want to remove b from the list?
> >
> > Just say "del x[1]".
>
>
> yep this is actually what i mean, but the problem is is that this action
> should come from the object within the list. this is object 'knows' when
> it's done, the list doesn't
>
> see my problem?
>

Indeed. If your objects don't "know" which list they are being held on, or
what element they are in that list then you have a problem: objects in
Python can't tell which variables or other items are bound (refer) to them

It's even trickier with a list, because you can't just keep the index as an
object attribute: ince you delete one list element, all indexes above it
would need to be reduced by one.

Perhaps, rather than assuming a particular type of answer is required to
your problem, you could just step back a little and describe in application
terms what it is you need to do. c.l.py contains enough inventive and gifted
programmers you are almost certain to get a workable solution to the real
problem.

regards
----------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From loewis at informatik.hu-berlin.de  Sun Jun  9 07:03:39 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 09 Jun 2002 13:03:39 +0200
Subject: Socket error problem
References: <mailman.1023614413.20191.python-list@python.org>
Message-ID: <j4adq44stw.fsf@informatik.hu-berlin.de>

"David LeBlanc" <whisper at oz.net> writes:

> I have this problem on windows where socket.socket(...) keeps returning
> 10022, 'Invalid argument'. It's actually dieing in the
> poplib.POP3.__init__(...) method. I have verified that host is reasonable
> (mymail.myisp.net or something like that ;)), and port is 110.
> 
> If I use the test script in poplib.__main__(...) with the correct host, user
> and password, all is well, but if I call it from this other program i'm
> trying to get working I get the above error.
> 
> I can't figure out what's wrong - any pointers?

You make it sound like something impossible happens, so it is hard to
recommend a solution (since clearly you are overlooking some piece of
information, which you could not report for the very reason that you
are overlooking it).

I recommend to put a print statement immediately before the call to
socket.socket inside poplib, to see what arguments it actually does
pass to that call.

Please report any information you gain by doing so.

Regards,
Martin



From d2002xx at myrealbox.com  Sun Jun 23 11:13:09 2002
From: d2002xx at myrealbox.com (d2002xx)
Date: 23 Jun 2002 08:13:09 -0700
Subject: The best GUI library in python
References: <8d3f4438.0206220457.b820418@posting.google.com> <3D1481DE.C0331204@engcorp.com> <8d3f4438.0206221422.6f71287a@posting.google.com> <3D14FFB5.90808865@engcorp.com>
Message-ID: <8d3f4438.0206230713.535169aa@posting.google.com>

I hope the GUI could work well when I compiling a program in
background (in my K6-2 400, running program other than compiler in
this kind of situation is allocated only about 10% to 15% CPU
resource)

Most programs such as opera and emacs works well, except for programs
using tk/tkinter.

Currently, I know only wxPython and PyQT, they are fast enough, but I
want to know are there any GUIs better than the two.


From aahz at pythoncraft.com  Sat Jun 29 16:59:27 2002
From: aahz at pythoncraft.com (Aahz)
Date: 29 Jun 2002 16:59:27 -0400
Subject: can't import generators
References: <m2hejlc060.fsf@mother.paradise.lost> <afl5je$4iu$1@panix1.panix.com> <Xns923CE7C804E6Fcliechtigmxnet@62.2.16.82>
Message-ID: <afl73f$6j6$1@panix1.panix.com>

In article <Xns923CE7C804E6Fcliechtigmxnet at 62.2.16.82>,
Chris Liechti  <cliechti at gmx.net> wrote:
>aahz at pythoncraft.com (Aahz) wrote in news:afl5je$4iu$1 at panix1.panix.com:
>> In article <m2hejlc060.fsf at mother.paradise.lost>,
>> John Hunter  <jdhunter at nitace.bsd.uchicago.edu> wrote:
>>>
>>>I am having trouble working with generators.  Anyone know what might
>>>cause this behavior:
>>>
>>>Python 2.2.1 (#1, Apr 22 2002, 21:20:54) [GCC 3.0.4] on linux2
>>>Type "help", "copyright", "credits" or "license" for more information.
>>>>>> from __future__ import generators
>>>Traceback (most recent call last):
>>>  File "<stdin>", line 1, in ?
>>>ImportError: cannot import name generators
>> 
>> Can't do this in interactive mode.  Use a script.
>
>?!?!? since when?
>
>$ python
>Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
>Type "help", "copyright", "credits" or "license" for more information.
>>>> from __future__ import generators
>>>> def g(x):
>...     for p in x: yield p
>...
>>>> [x for x in g(range(3))]
>[0, 1, 2]

Huh.  That'll teach me to trust the combination of screen output and my
memory.  Here's some more weird stuff.  In a script, this works:

    x=1
    from __future__ import generators

this fails with SyntaxError:

    x=1
    print x
    from __future__ import generators

but it works in interactive mode:

    >>> x=1
    >>> print x
    1
    >>> from __future__ import generators

Of course, none of this explains why John gets an ImportError, unless
he's got a completely screwed up installation.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From whisper at oz.net  Tue Jun 18 17:41:02 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 18 Jun 2002 14:41:02 -0700
Subject: python version?
In-Reply-To: <FVMP8.46296$GY.14475470@twister.nyroc.rr.com>
Message-ID: <GCEDKONBLEFPPADDJCOEIEAADKAA.whisper@oz.net>

George;

Watch out for that tree!

- dhl




From karczma at info.unicaen.fr  Tue Jun 11 09:42:30 2002
From: karczma at info.unicaen.fr (Jerzy Karczmarczuk)
Date: Tue, 11 Jun 2002 15:42:30 +0200
Subject: Defining a method final
References: <6glN8.35376$86.884025@twister1.libero.it> <mailman.1023796095.1427.python-list@python.org> <phmN8.36164$K4.893500@twister2.libero.it> <KwmN8.153269$352.7754@sccrnsc02> <ae4tvv$rvd$1@wanadoo.fr>
Message-ID: <3D05FE46.3E9B8430@info.unicaen.fr>

Eric Brunel wrote:

> Python now have an "official" means to make attributes private by prefixing
> them with a double-underscore. And it *does* prevent their use or
> overloading:
> 
> class A:
>   def __private(self): pass
> class B(A):
>   pass
> o = B()
> o.__private()
> 
> results in:
> 
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> AttributeError: B instance has no attribute '__private'

Some newbies may understand this wrongly. The attribute is still
accessible, it has just a different name: _A__private etc.

(You don't even need the class B to find out that 
o=A()
o.__private()

won't work. But launch dir(A) and look.)

Jerzy Karczmarczuk


From news at nospam.majid.fm  Tue Jun 25 23:28:17 2002
From: news at nospam.majid.fm (Fazal Majid)
Date: Tue, 25 Jun 2002 20:28:17 -0700
Subject: Installation problems DCOracle2 & W2000
References: <talon34-C09E81.19332125062002@netnews.attbi.com>
Message-ID: <3D1934D1.3000707@nospam.majid.fm>

Talon wrote:
> I am having a heck of a time getting DCOracle2 to funcion properly.  I 
> am running on a W2K machine, python2.1.3 and Zope 2.5.1, both of which 
> run just fine.  I installed DCOracle2 and the ZOracleDA under Zope and 
> it works fine.  But standalone python scripts don't see the dco2.py 
> module.  So I tried to install DCOracle2 in the python21/lib folder.  I 
> got that latest version from Zope.org and it said it installed properly.  
> But it doesn't run.  If you try to import it into a script, you get 
> complaints that it can't find this or that module, all of which are part 
> of the Zope path.

I have gotten DCOracle2 to work correctly on my Windows 2000 setup with 
2.1.1 (but I haven't tried it with Zope). You do need to set your 
Pythonpath environment variable correctly, as well as copy the correct 
dco2.pyd (DLL) version for your Python and Oracle versions.

-- Fazal Majid

If you want to respond by email, take the nospam out of my email address.



From jblazi at hotmail.com  Mon Jun 24 13:40:17 2002
From: jblazi at hotmail.com (JB)
Date: Mon, 24 Jun 2002 19:40:17 +0200
Subject: converting an array of chars to a string
References: <3d12dda2_6@news.newsgroups.com> <aevj41$fg2$0@216.39.172.122> <3d17038e_6@news.newsgroups.com> <af7jei$3kr$0@216.39.172.122>
Message-ID: <3d1757c7_1@news.newsgroups.com>

Bengt Richter wrote:
> No cause for desperation ;-)
> 
> Are you saying that you want modulo 256 arithmetic on
> unsigned 8-bit data? No problem. But what does your data
> represent? And what does adding a number to such a data
> item mean in the real world?

Of course you are right. I really needed only some 8-bit 
shifting. The point is, that there is no need for the 
overhead that is involved. If anybody uses the array module 
which, you can argue, is a low level module, he should know 
what he is doing.

-- 
Janos Blazi


-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From whisper at oz.net  Sun Jun  9 14:19:56 2002
From: whisper at oz.net (David LeBlanc)
Date: Sun, 9 Jun 2002 11:19:56 -0700
Subject: Socket error problem
In-Reply-To: <j4adq44stw.fsf@informatik.hu-berlin.de>
Message-ID: <GCEDKONBLEFPPADDJCOEMELBDHAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Martin v. Lowis
> Sent: Sunday, June 09, 2002 4:04
> To: python-list at python.org
> Subject: Re: Socket error problem
>
>
> "David LeBlanc" <whisper at oz.net> writes:
>
> > I have this problem on windows where socket.socket(...) keeps returning
> > 10022, 'Invalid argument'. It's actually dieing in the
> > poplib.POP3.__init__(...) method. I have verified that host is
> reasonable
> > (mymail.myisp.net or something like that ;)), and port is 110.
> >
> > If I use the test script in poplib.__main__(...) with the
> correct host, user
> > and password, all is well, but if I call it from this other program i'm
> > trying to get working I get the above error.
> >
> > I can't figure out what's wrong - any pointers?
>
> You make it sound like something impossible happens, so it is hard to
> recommend a solution (since clearly you are overlooking some piece of
> information, which you could not report for the very reason that you
> are overlooking it).
>
> I recommend to put a print statement immediately before the call to
> socket.socket inside poplib, to see what arguments it actually does
> pass to that call.
>
> Please report any information you gain by doing so.
>
> Regards,
> Martin

Martin and Steve;

In one part of the program it does this:

	serv = Pop3Server(adressaux, aux_username, aux_passwd, serverport)

and the args are: adressaux: mail.oz.net aux_username: whisper aux_passwd:
password serverport: 110 (with aux_passwd changed to protect my innocence).

Then, it invokes POP3Server's connect method as:
def process (self, serv):
	"""Code actually executed by every thread except the main thread"""
	config = self.config # Very used
	log_index = 0
	if config.optnotlog:
		log_index= 0
		logAgent = None
	else:
		# logAgent collects stadistics and show them at the end (see logger.py)
		# It has an entry for each server, which get indexed by log_index
		logAgent = self.logAgent
		log_index = logAgent.new(serv.alias, thread.get_ident())

	# If configured, send logging info to the system log
	# aprint() (see general.py) takes care of this automatically
#     if  config.optsyslog:
#     	import syslog
#     	syslog.openlog("Animail", 0, syslog.LOG_MAIL)

try:
	messages = 0
	aprint(_("Trying to connect (%s)") % serv.alias, general.YELLOW)
	v = serv.connect()
	aprint (_('connected!\n'), serv.alias, general.YELLOW)
	aprint(_("Authenticating (%s)") % serv.alias, general.YELLOW)
	serv.auth()
	.
	.
	.
except smtplib.SMTPException, x:
	self.error_msg("SMTP Error (%s): " % serv.adress, x, log_index)
	self.__exit_thread()

except socket.error, x:
	self.error_msg(_("Connection error (%s): ") % serv.name, x, log_index)
	self.__exit_thread()

except poplib.error_proto, x:
	self.error_msg(_("Protocol Error (%s): ") % serv.adress, x, log_index)
	self.__exit_thread()
----------------------
The general.* constant references are to colorize the output which is turned
off but still results in the "31" in the connection error line in the
program output below. NOTE: I didn't write this code, so I have no idea why
it's returning the username in the socket.error instead of the host. FWIW,
this code is running in a thread, but there is only one thread trying to get
mail.

POP3Server's connect method:
    def connect(self):
        self.lmsg = []
        if self.adress == '':
            raise ConfigError, _("Server address not given!!")

        if self.ssl:
            import popslib
            self.objpop = popslib.POP3S(self.adress, self.port)
        else:
            self.objpop = poplib.POP3(self.adress, self.port)
        welcome = self.objpop.getwelcome()
        if self.protocol == 'APOP':
            lwelcome = welcome.split()
            lwelcome.reverse()
            self.timestamp = lwelcome[0]
        return welcome
        # end if
    # end def connect
----------------------
"self.ssl" is false, so objpop becomes a poplib.POP3
"self.protocol is POP3, so the APOP stuff isn't executed.

At this point, self.adress and self.port are mail.oz.net and 110

In poplib.POP3.__init__(...) they are also mail.oz.net and 110
    def __init__(self, host, port = POP3_PORT):
        self.host = host
        self.port = port
        msg = "getaddrinfo returns an empty list"
        self.sock = None
        for res in socket.getaddrinfo(self.host, self.port, 0,
socket.SOCK_STREAM):
            af, socktype, proto, canonname, sa = res
            try:
                self.sock = socket.socket(af, socktype, proto)
                self.sock.connect(sa)
            except socket.error, msg:
                if self.sock:
                    self.sock.close()
                self.sock = None
                continue
            break
        if not self.sock:
            raise socket.error, msg
        self.file = self.sock.makefile('rb')
        self._debugging = 0
        self.welcome = self._getresp()
----------------------

The output of the program is:
K:\tmp\animail>animail -v
Opening Config file: c:/user/dave/.animail/animailrc

Trying to connect (mail.oz.net)
("Connection error (whisper):  (10022, 'Invalid argument')", 31)

####### Operation summary #######

Sun Jun 09 11:01:04 2002

--- Server: mail.oz.net

* Total messages: 0
* Total volume on server: 0.00 Bytes
* Filtered messages: 0
* Downloaded messages: 0
* Postergated messages: 0
* Incidents:
        -Connection error (whisper):  (10022, 'Invalid argument')

Dave LeBlanc
Seattle, WA USA





From mcalla at insightbb.com  Thu Jun 13 08:27:55 2002
From: mcalla at insightbb.com (Mike Callahan)
Date: Thu, 13 Jun 2002 12:27:55 GMT
Subject: Question about asksaveasfilename
Message-ID: <ed0O8.23891$nZ3.2215@rwcrnsc53>

I am using asksaveasfilename from tkFileDialog. I want to restrict my users
to save filenames with these extensions .unl or .sql. This is my call:

from tkFileDialog import asksaveasfilename
fn = asksaveasfilename(filetypes=[('Unloaded', '*.unl'),
('Queries','*.sql')])

This gives me the correct dialog box, but how do I force the correct
extension on fn?  Thanks.

Mike C.





From dallasm at aiinet.com  Thu Jun 13 16:15:39 2002
From: dallasm at aiinet.com (Mahrt, Dallas)
Date: Thu, 13 Jun 2002 16:15:39 -0400
Subject: Shrinking Python
Message-ID: <638AA0336D7ED411928700D0B7B0D75B682046@aimail.aiinet.com>

I am looking into using Python in a small device with limited space for
binaries. Due to this, I am both investigating creating a shared library
with the interpreter and shrinking the footprint of this library. I have
found many threads talking bout shared libraries, so I am not too concerned
about that aspect. 

I am concerned about shrinking the library's binary footprint. I did find
several links to a "Deeply Embedded Python" based on 1.5.1 which has since
vanished. I have stripped the library to improve size some (to ~1.13 MB) but
would like to reduce it further. Has anyone tackled this or even attempted?
Are there any links to resources that may help me that I haven't found? 


---
Dallas S. Mahrt               614-923-1194
Software Engineer             Network Management Systems
Applied Innovation Inc.       Dublin, Ohio





From guido at python.org  Wed Jun 19 11:07:36 2002
From: guido at python.org (Guido van Rossum)
Date: Wed, 19 Jun 2002 11:07:36 -0400
Subject: [Python-Dev] Tcl adept wanted for Stackless problem
In-Reply-To: Your message of "Wed, 19 Jun 2002 15:38:46 +0200."
             <3D108966.9050402@tismer.com> 
References: <3D108966.9050402@tismer.com> 
Message-ID: <200206191507.g5JF7ag02088@pcp02138704pcs.reston01.va.comcast.net>

> My big question is:
> When does Tcl use C stack entries as globals, which
> are passed as function arguments to interpreter calls?

It's a performance hack, just as stackless :-).

Tcl's interpreter data structure has a return value field which can
receive a string of arbitrary length.  In order to make this
efficient, this is initialized with a pointer to a limited-size array
on the stack of the caller; when the return value is longer, a
malloc()'ed buffer is used.  There is a little dance you have to do to
free the malloc()'ed buffer.  The big win is that most calls return
short strings and hence you save a call to malloc() and one to free()
per invocation.  This is used *all over* the Tcl source, so good luck
getting rid of it.

--Guido van Rossum (home page: http://www.python.org/~guido/)





From h.baumgartl at chello.NOSPAM.nl  Thu Jun 20 13:39:42 2002
From: h.baumgartl at chello.NOSPAM.nl (Henry Baumgartl)
Date: Thu, 20 Jun 2002 17:39:42 GMT
Subject: application development - separating content from function
Message-ID: <yroQ8.260949$i6.17986369@amsnews02.chello.com>

For eight months now, I am preparing for the development of an open source
Python application to support individuals in Life Long Learning. It will be
a mixture of PIM, extended diary, groupware, project management, and modules
to interact with centralised knowledge / skill exchange servers.

Most of the user activity is done on individual boxes. However, when and if
needed, the application must provide intuitive and standardised interaction
with fellow network members (e.g. project workgroups, mentor, and such) and
centralised server modules.

What needs to be done is:
1. Develop GUIs and modules that are interchangeable on, at least, Linux,
Windows, and Mac.
2. Use the modularity of Python to allow personalised application
adaptation, thus requiring an effective way of integrating personal
variables into almost all scripts.
3. And, last but not least ;-) the application needs to be multi-lingual.

The more I have tried to solve these requirements in an elegant and
efficient manner, the further I seem to stray from anything remotely
resembling a viable solution. In other words, I can't see the forest for the
trees anymore.

Anyone working on something similar? Or, just plain brilliant in all things
Python, and ready and willing to point me in the right direction and make it
all worth while again :-)

Help!

Henry Baumgartl
Dynamic Knowledge Network

*Please remove .NOSPAM from my e-mail address when abswering me directly.*




From shalehperry at attbi.com  Fri Jun  7 20:04:56 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 07 Jun 2002 17:04:56 -0700 (PDT)
Subject: Detecting OS and Window Manager
In-Reply-To: <ug2hcuoq7egm6d@corp.supernews.com>
Message-ID: <XFMail.20020607170456.shalehperry@attbi.com>

> Alas, I don't think there is an easy answer to this.  I think we'll just
> have to either:
> 1) Settle for a default
> 2) Provide different commands such as gfunky and kfunky and just provide
> menu entries to point to the appropriate one depending upon which DE you
> are running.
> 

yep, that about covers it.




From emile at fenx.com  Thu Jun  6 09:04:17 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 06 Jun 2002 13:04:17 GMT
Subject: check on numerical value?
References: <3cff56b4.511834453@news.skynet.be>
Message-ID: <l5JL8.123663$cQ3.3519@sccrnsc01>

<pixie888 at hotmail.com> wrote in message
> Is there any way to discover when a value is numerical or not in
> Python? I want a check in my code on a given value. If somebody enters
> or gives me something like "akjlvq?irq" this is surely not a numerical
> value.
>

int(value) or float(value) depending on what values you'd except.

--

Emile van Sebille
emile at fenx.com

---------



From aahz at pythoncraft.com  Mon Jun  3 14:05:15 2002
From: aahz at pythoncraft.com (Aahz)
Date: 3 Jun 2002 14:05:15 -0400
Subject: thread module under Windows
References: <59d8a5f6.0206030730.1be14ffb@posting.google.com>
Message-ID: <adgb4r$6ls$1@panix1.panix.com>

In article <59d8a5f6.0206030730.1be14ffb at posting.google.com>,
Dmitri <dm_top at hotmail.com> wrote:
>
>Reading the conference I see that many people use the python in
>multi-threading. How they get the python version for windows with
>treading support on ( I am not talking about other optional services
>now... )???

Almost everyone downloads the pre-built binary for Windows instead of
compiling it themselves; the binary on www.python.org comes with
threading enabled.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"In the end, outside of spy agencies, people are far too trusting and
willing to help."  --Ira Winkler


From jonathan at onegoodidea.com  Fri Jun 28 05:07:33 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Fri, 28 Jun 2002 10:07:33 +0100
Subject: metaclasses vs. inheritance
References: <ghTS8.58749$mh.1826422@news1.telusplanet.net>
Message-ID: <B941E5E5.D1AD%jonathan@onegoodidea.com>

On 28/6/2002 7:23, in article ghTS8.58749$mh.1826422 at news1.telusplanet.net,
"Ian McMeans" <imcmeans at home.com> wrote:

> Can someone give an example of a problem that is solved with metaclasses,
> which would be difficult or impossible to solve with inheritance?
> 
> I (very barely) understand metaclasses, but to me it just seems like
> inheritance, where the classes who get the benefit of the meta-class get the
> metaclass' features.

Interesting you should ask this as I've spent the last day reading up on
metaclasses and playing with them to better understand what can be done with
them. I find the best way to understand them is with a little graph:

           metaclass                 metaclass
              ^                         ^
              |                         |
           instance                  instance
              |                         |
            class B ---inheritance--> class A
              ^                         ^
              |                         |
           instance                  instance
              |                         |
              b                         a

This shows that metaclasses are "orthogonal" to normal inheritance. By which
I mean when you define a (new-style) class like so:

>>> __metaclass__ = type
>>> 
>>> class A: 
...     pass
... 

an instance of class 'A' has the type 'A':

>>> a = A()
>>> 
>>> a
<__main__.A object at 0x3f8cf0>
>>> type(a)
<class '__main__.A'>
>>>

But 'A' itself is an instance of the metaclass, in this instance 'type':

>>> A
<class '__main__.A'>
>>> type(A)
<type 'type'>
>>> 

Thus the metaclass is used in the construction and definition of classes. A
metaclass is a class itself (with the mind-bending result that the type of
'type' is 'type') and new ones can be defined using the same syntax. The
easiest way to define a new metaclass is to inherit from 'type' as that
defines all of the magic necessary to construct and manage classes.

The main reason one would want to define a new metaclass is to override some
of this magic in order to bend the way classes work. A useful method that
one might want to override in a metaclass is '__init__', which is called to
initialise a new class. An example will show what happens:

>>> class mytype( type ):
...    def __init__( cls, name, bases, dict ):
...        print '__init__'
...        print '    new class:', cls
...        print '    name:', name
...        print '    base classes:', bases
...        print '    dictionary:', dict
...        type.__init__( cls, name, bases, dict )
... 
>>> __metaclass__ = mytype
>>> 
>>> class Foo:
...     def foo( self ):
...         print 'foo'
... 
__init__
    new class: <class '__main__.Foo'>
    name: Foo
    base classes: ()
    dictionary: {'__module__': '__main__', 'foo': <function foo at
0x409768>}
>>> 

You can see from the above example that the 'mytype.__init__' is called
immediately after we finish defining the class 'Foo'. It is passed as
arguments the information used to initialise a new class, which is: the
class name, a tuple of the base classes (none in this example), and the
class dictionary which contains the method functions and class instance
variables.

Most people won't actually have any particular purpose for metaclasses, but
for altering the way that the language works they are invaluable. The
example I worked on yesterday is a mechanism for declaring methods and
classes final (in the Java sense) - prompted by an example someone else
wrote a while ago on this group. My metaclass intervenes in the creation of
new classes to check that final constraints aren't being broken:

>>> from final import *
>>> 
>>> class Foo:
...     def foo( self ):
...         print 'foo'
...     foo = finalmethod( foo )
... 
>>> class Bar( Foo ):
...     def foo( self ):
...         print 'bar'
... 
Traceback (most recent call last):
[..]
final.FinalError: finalmethod 'foo' of class 'Foo' may not be overriden
>>> 

and also:

>>> class Foo( finalclass ):
...     pass
... 
>>> class Bar( Foo ):
...     pass
... 
Traceback (most recent call last):
[...]
final.FinalError: finalclass 'Foo' may not be subclassed
>>> 

There's no easy way to do this sort of magic without the use of metaclasses.

Hope this helps a little. I'd better go do some real work ;-)

Jonathan



From emile at fcfw.fenx.com  Mon Jun 24 17:20:47 2002
From: emile at fcfw.fenx.com (Emile van Sebille)
Date: 24 Jun 2002 16:20:47 -0500
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Jun 24)
Message-ID: <F5158407F83E4D2B.81555A54E3C8D462.5EDAA20E7844B1F0@lp.airnews.net>

QOTW:  "I read with pleasure all the thread saying that `roundup' has an
email interface, is actively being improved, and could replace the SF
tracker. Let us hope it will be more usable than its precessors!  You
know, the real goal of all this is allowing for simple and humble
communication between humans, about the knowledge of a problem.  I
surely used to be a very active reporter for all problems I saw
everywhere, at the time maintainers were still reachable.  When the
effort gets too frustrating, sadly, one might feel less inclined to
offer contributions, and rather choose to enjoy more of the sun, music,
and life! :-)" -- Fran?ois Pinard on python-dev, Jun 1 2002


    Travis N. Vaught announces a Python for Scientific Computing Workshop
    at CalTech, Pasadena, CA September 5-6, 2002.  This workshop provides
    a unique opportunity to learn and affect what is happening in the
    realm of scientific computing with Python.  Space is limited.
       http://groups.google.com/groups?selm=mailman.1024366882.23776.python-list at python.org
    
    On the question of text to html conversion, Gilles point to
    StructuredText in Zope as a package that can be used "standalone".
    He also provides a dozen lines of code as a start:
       http://groups.google.com/groups?selm=aen4n1$35b$1 at norfair.nerim.net
    
    The European Python and Zope Conference 2002 happens June 26-28 in
    Charleroi, Belgium.  It promises to be a great show.
    ...for general information, see
        http://europython.zope.nl
    ...and for schedule and information on the talks being given see
       http://europython.zope.nl/sessions/talks
    
    Fernando Perez on tools for python in science, in response to a question
    of experiences teaching Biology students with python support.
       http://groups.google.com/groups?selm=aeqfuh$88n$1 at peabody.colorado.edu
    
    Lamy Jean-Baptiste releases TreeWidget 0.2 a pure-Python tree widget
    for Tkinter...
       http://groups.google.com/groups?selm=mailman.1024601110.13018.clpa-moderators at python.org
    ...and EventObj 0.2, a pure Python module that allows to add modification
    callback (=event) to any Python object (including lists and dicts).
       http://groups.google.com/groups?selm=mailman.1024601109.12981.clpa-moderators at python.org
    
    Gerhard Haring asks if a python based procmail replacement exists, and
    Radovan Garabik reminds us that he's written one...
       http://groups.google.com/groups?threadm=mailman.1024444823.19953.python-list at python.org
    ...Francois Pinard also posts a summary of additional options...
        http://groups.google.com/groups?selm=mailman.1024451876.25700.python-list at python.org
    
    Tom Loredo needs to move into Windows and asks for advise on proceeding:
       http://groups.google.com/groups?threadm=3D110B55.4D8AD7FA at astro.cornell.edu
    
    Steven M. Castellotti gets help connecting to the Microsoft Speech API
    to synthesize Text-To-Speech output, then posts the working result:
       http://groups.google.com/groups?threadm=U_6Q8.39803$ks6.8500 at nwrddc02.gnilink.net
    
    Here's a discussion on implementing and representing infinity (eg 1e400)
    in python but if you use it, be sure to test on the target platforms.
       http://groups.google.com/groups?threadm=3d11ee9b$1 at news.swissonline.ch
    
    Anyone ever needed a Queue with timeout?  Chris Liechti posts a working
    implementation.
       http://groups.google.com/groups?selm=Xns9234F249D7D85cliechtigmxnet at 62.2.16.82
    
    Angela Williams asks for clarification on Python's FAQ 6.26. Why no goto?
    One gem I found in this thread was a counter proposal for COME FROM...
      * Example:
           10 J=1
           11 COME FROM 20
           12 WRITE (6,40) J STOP
           13 COME FROM 10
           20 J=J+2
           40 FORMAT (14)
      * Explanation:
    In this example, J is set to 1 by statement 10. Statement 13 then
    causes control to be passed to statement 20, which sets J to 3.
    Statement 11 then causes control to be passed to statement 12, which
    writes the current value of J. The STOP statement then terminates the
    program.
    ...For more like this, see (Thank Carl Banks for the pointer)
       http://www.fortran.com/fortran/come_from.html
    ...For the current discussion on GOTO and python, start here:
       http://groups.google.com/groups?threadm=uh1s1qdcl5b1b0 at news.supernews.com
    
    One lump or two?  If you're easily distracted by the rules of
    punctuation, you may want to follow this thread on python-dev:
       http://mail.python.org/pipermail/python-dev/2002-June/025074.html
    

========================================================================

Everything you want is probably one or two clicks away in these pages:

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
	 http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Michael Hudson continued Andrew Kuchling's marvelous tradition
    of summarizing action on the python-dev mailing list once every
    other week, into July 2001.  Any volunteers to re-start this
    valuable series?
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

    Much of Python's real work takes place on Special-Interest Group
    mailing lists
        http://www.python.org/sigs/

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
	http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
	http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
	http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

    Links2Go is a new semi-automated link collection; it's impressive
    what AI can generate
        http://www.links2go.com/search?search=python

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning
(approximately), ask <claird at neosoft.com> to subscribe.  Mention
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

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


From fredrik at pythonware.com  Sat Jun 29 08:06:24 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 29 Jun 2002 12:06:24 GMT
Subject: How to get rid the new line
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org> <3D1AD534.A74EC9F4@ipm.fhg.de> <3D1B1A76.563E5394@engcorp.com> <pan.2002.06.28.00.14.48.789371.16627@linuxmail.org> <afg513$2426$1@news.idiom.com> <3D1BA535.1EE16646@engcorp.com>
Message-ID: <4phT8.45578$n4.10822996@newsc.telia.net>

Peter Hansen wrote:

> >  Could try this:
> >
> >         def chomp(line):
> >                 if line[-1]=='\n':
> >                         line=line[:-1]
> >                 return line
>
> Just make sure you pass chomp() only lines that are not equal
> to the empty string ''.

or add the single character needed to make chomp a
bit more robust:

    def chomp(line):
        if line[-1:] == '\n':
            line = line[:-1]
        return line

</F>




From pereira at cis.upenn.edu  Mon Jun  3 12:43:34 2002
From: pereira at cis.upenn.edu (Fernando Pereira)
Date: Mon, 03 Jun 2002 12:43:34 -0400
Subject: Graph distances, how can I speed it up?
References: <33803989.0206030235.4134d33e@posting.google.com> <d571ae53.0206030815.2e52aac@posting.google.com>
Message-ID: <B92114F6.B79E%pereira@cis.upenn.edu>

On 6/3/02 12:15 PM, in article
d571ae53.0206030815.2e52aac at posting.google.com, "Jonathan Ellis"
<ellisjb at my-deja.com> wrote:
> The obvious idea then is to not calculate anything until all edges
> have been added.  Then use dijkstra's algorithm
> (http://ciips.ee.uwa.edu.au/~morris/Year2/PLDS210/dijkstra.html)
> to compute and store distances as needed.  (You can probably get some
> savings over the classic algorithm if you know you are going to NEED
> distances for ALL vertices -- otherwise you are better off doing lazy
> (as-needed) computation.)
Dijkstra's algorithm is not a good choice here, because distances from all
sources appear to be needed, not from a single source. There is no good,
general way of saving parts of the work done for one source node to speed up
computing the distances from another source node (but see Johnson's
algorithm). So, he may need an all-sources algorithm like Floyd-Warshall,
which unfortunately is cubic on the number of nodes. Since the WordNet graph
is sparse, and in fact "almost" a tree, Johnson's algorithm might be better.
If the graph were a tree, the following much simpler algorithm would do. As
always, Cormen, Leiserson and Rivest (+ Stein in the new edition)
http://theory.lcs.mit.edu/~clr/ are your friends.

-- F



From asgarbutt at ucdavis.edu  Thu Jun 20 15:30:30 2002
From: asgarbutt at ucdavis.edu (ANdrew)
Date: 20 Jun 2002 12:30:30 -0700
Subject: PyQT Complie problems
Message-ID: <a5384db6.0206201130.1e08b42a@posting.google.com>

First off,

using:

Windows 2000

MSVC 6.0 service pack 5
QT eval 3.0.4
Active State Python 2.2
PyQt-3.2.4 (source)


I am trying to compile for Qt 3.0.4 eval.  Following the win32
instructions, i am able to compile the first makefile
(sip_helper.exe).  However, when i change dir to the "qt" dir and
modify the makefile for the correct version of the .lib file and fix
the one type-o in the sipqtDeclqt.h file ( changed #include <sipQt.h>
to #include <sipQtQT.h> ), i get a synatx error followed by an
unexpected end of file error. This is stated as being in the file
"sipqtQPaintDevice.h".  The error is "sipqtQPaintDevice.h(44) : error
C2146: syntax error : missing ';' before identifier 'sipLazyAttrDef'"

I have downloaded the latest snapshot (PyQt-snapshot-20020619.zip)and
there is nothing in there about the qt section.  I am at a loss and
any help would be greatly appreciated.

-Andrew


From ark at research.att.com  Fri Jun  7 15:02:47 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 7 Jun 2002 15:02:47 -0400 (EDT)
Subject: Python 2.2.1 vs. gcc 3.1?
In-Reply-To: <15616.64407.863059.729351@12-248-41-177.client.attbi.com>
	(message from Skip Montanaro on Fri, 7 Jun 2002 13:29:43 -0500)
References: <yu99vg8v6orb.fsf@europa.research.att.com>
        <15616.57963.519636.730290@12-248-41-177.client.attbi.com>
        <200206071747.g57HlbH01057@europa.research.att.com>
        <15616.62634.635544.905055@12-248-41-177.client.attbi.com>
        <200206071804.g57I4fp01212@europa.research.att.com>
        <15616.63342.559215.644086@12-248-41-177.client.attbi.com>
        <200206071815.g57IFjg01303@europa.research.att.com> <15616.64407.863059.729351@12-248-41-177.client.attbi.com>
Message-ID: <200206071902.g57J2lJ01644@europa.research.att.com>

Skip> Dunno.  Try "find . -name 'struct*'.  It should be in your build/lib*
Skip> directory (where "*" matches your configuration name).

Not as far as I can see...

$ find . -name '*struct*' -print
./Doc/lib/libstruct.tex
./Include/structmember.h
./Include/structseq.h
./Lib/test/test_struct.py
./Lib/test/test_structseq.py
./Modules/structmodule.c
./Objects/structseq.c
./Objects/structseq.o
./Python/structmember.c
./Python/structmember.o
./Tools/modulator/Templates/object_structure




From aahz at pythoncraft.com  Sun Jun 30 21:04:49 2002
From: aahz at pythoncraft.com (Aahz)
Date: 30 Jun 2002 21:04:49 -0400
Subject: I'd give up Perl tomorrow if only...
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au> <Pine.LNX.4.44.0206301550120.7807-100000@palanthas.neverending.org> <Xns923DF22CCC54Ccliechtigmxnet@62.2.16.82> <Pine.LNX.4.44.0206301756130.22298-100000@palanthas.neverending.org>
Message-ID: <afo9rh$jlb$1@panix1.panix.com>

In article <Pine.LNX.4.44.0206301756130.22298-100000 at palanthas.neverending.org>,
Frank Tobin  <ftobin at neverending.org> wrote:
>
>Yes, I'm talking about non-bundled extension modules.  That's all we're 
>talking about here.  The standard lib is irrevelant to the discussion.

Yes and no.  Thing is, historically Python has done better than Perl at
pulling publicly recognized libraries into the standard library.  This
tends to reduce the interest in a CPAN-like construct.  Perl has more
recently recognized the importance of this, and Perl's standard library
has been growing faster than Python's, I think, due to the backlog of
stuff in CPAN that is being moved.

>I suggest that you be a Perl programmer for a while, and see what having
>CPAN accessible and authoring for it really is like.  You explicitly state
>you aren't familiar with it.  I've been author of a couple CPAN modules
>for about 4 years now, and an extensive user for just as long, and have
>been a Python module writer for just over a year now.  Python is a great
>language, but in terms of a reusable code repository, *nothing* comes
>close to the jewel that is CPAN.  There is a reason you'll hear this over
>and over from people coming to Python from Perl.

But somehow one doesn't hear it very often from people who've been using
Python for a long time -- and certainly nobody has truly stepped up to
the plate to put all the finishing details into place.  Sean
Reifschneider came pretty close, then dropped the ball.  Where's the
itch-scratching that induces people to do the work?

You might be able to get some traction if you concentrate on just the
testing issue (and possibly the README issue).  That's the kind of thing
that gets at least public verbal support from Guido.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From paul at boddie.net  Tue Jun 25 06:32:11 2002
From: paul at boddie.net (Paul Boddie)
Date: 25 Jun 2002 03:32:11 -0700
Subject: Web templating/db tool with best designer/coder separation?
References: <af2gh6$2c2$0@216.39.172.122> <af8dtp$5qp$0@216.39.172.122>
Message-ID: <23891c90.0206250232.249c8488@posting.google.com>

bokr at oz.net (Bengt Richter) wrote in message news:<af8dtp$5qp$0 at 216.39.172.122>...
> 
> No one mentioned ZPT (Zope Page Templates),
>     http://www.zope.org//Wikis/DevSite/Projects/ZPT/FAQ

This thread wouldn't end if we all started naming template systems,
though. ;-) For something similar to the jonpy templating, take a look
at http://www.clearsilver.org - a system which I was made aware of
recently. If nothing else, that site gives some very good arguments
against "Python in HTML" systems.

> and from the description, this would seem to have a real advantage over all of
> the above, in that it uses attributes of html tags as opposed to special tags
> that browsers or design tools won't generally understand and render usefully.

It should work well with XML-capable tools. However...

> This is touted as enabling the "round trip" from graphic designer to template
> logic programmer and back using the same html source. I assume this is
> the "round trip" Paul was referring to. To me, it seems a winning point for
> ZPT, though I don't know how ZPT works out in practice.

Yes, that was what I was referring to. However...

> They have a list of design tools that apparently pass through the ZPT markup
> they don't understand. According to the FAQ:
> 
>    "Q8: What HTML editors work well with ZPT?
>         We are still compiling a list, but on the commercial (industrial
>         strength) front, GoLive and Dreamweaver 4 work well with ZPT. Recent
>         versions of Amaya work well in HTML mode."
> 
> Opinions on ZPT?

In my experience, "recent versions of Amaya" either don't work with
ZPT, or the secret incantation to make it work hasn't been widely
publicised - "HTML mode" doesn't really tell me anything. What happens
when I try to load ZPT into Amaya is a large number of warnings about
invalid attributes, followed by the loss of those attributes in the
saving process.

On the other hand, if I use some 'id' attributes in HTML to give
"structure" to my template, Amaya will only intervene when two such
attributes are identical. Consequently, it's possible to do like
Enhydra's XMLC and make a template system operating around those
principles.

Paul


From derek at wedgetail.com  Sat Jun 15 09:18:47 2002
From: derek at wedgetail.com (Derek Thomson)
Date: Sat, 15 Jun 2002 23:18:47 +1000
Subject: Integer conversion
References: <3D0B3257.E8572AE5@sympatico.ca>
Message-ID: <3D0B3EB7.2020304@wedgetail.com>

Colin J. Williams wrote:
> int(string, base) => integer
> 
> For int('127',  8), I had expected this function to deliver 87.

It does deliver 87. I enter int('127', 8) at the Python prompt, and I 
get 87, which is the integer (decimal) value of '127' if interpreted as 
a base 8 number.

> 
> The example below correctly gives errors at the top and bottom of the
> range and
> for int('127', 10) delivers 127. 

Right.

 > The other cases are not what I expected.

The output looks correct to me. For example, '127' as a base 16 integer 
is 295 (decimal). What output are you seeing, exactly?

Regards,
Derek.



From roy at panix.com  Tue Jun 25 07:48:06 2002
From: roy at panix.com (Roy Smith)
Date: Tue, 25 Jun 2002 07:48:06 -0400
Subject: Suggestions for good programming practices?
References: <mailman.1024971992.26429.python-list@python.org>
Message-ID: <roy-2CF124.07480625062002@reader2.panix.com>

"David LeBlanc" <whisper at oz.net> wrote:
> Structs and tuples are different. They're both of the same general category
> of aggregate data, but they are different as Brian notes.

I've come to really dislike using tuples for data structures, and 
instead create small data classes.  Even if the class is no more 
complicated than:

class myData:
   def __init__ (self, foo, bar):
      self.foo = foo
      self.bar = bar

you've added quite a bit of functionality.  First, the code becomes a 
little more self-documenting.  "power = ckt.volts * ckt.amps" is a lot 
easier to read and understand than "power = ckt[0] * ckt[1]".

Secondly, it makes it easier to catch programming errors.  I was working 
with somebody the other day who was using tuples to store data.  His 
program was doing something strange.  When I made him rewrite one of his 
data structures as a simple data class instead of a tuple, the reason 
instantly popped out.  He had gotten confused and was using the wrong 
piece of data.  When he had everything as tuples, it just happily took 
the i'th element of the wrong tuple and gave him garbage.  But, when he 
tried to access an attribute that didn't exist, it immediately raised an 
exception.


From fredrik at pythonware.com  Tue Jun 11 17:01:38 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 11 Jun 2002 21:01:38 GMT
Subject: tk gui for python
References: <e0b1d6f6.0206110802.7e38179c@posting.google.com>
Message-ID: <SytN8.42840$n4.9774222@newsc.telia.net>

Martin Krallinger wrote:

> I would appreciate any sample code to generate tk gui for python.
> i just need to load a file, of past instead a text (protein sequence)
> and run on it a c programm and return the content of it's output
>
> any sample code is welcomed,

start here:

    http://www.python.org/topics/tkinter/doc.html

</F>




From marklists at mceahern.com  Thu Jun 20 12:06:10 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 20 Jun 2002 11:06:10 -0500
Subject: Function to add commas in numbers?
In-Reply-To: <Xns923372EFBBC5Dmdkmdkmdkmdk@130.133.1.4>
Message-ID: <NCBBLFCOHHDIKCAFGCFBAEKMLFAA.marklists@mceahern.com>

> Is there a function that will take, for example, 1234567 and return 
> 1,234,567?

Short answer:  Look into the locale module.

Demo:

  mynum = 123456
  import locale
  locale.setlocale(locale.LC_ALL, "")
  locale.format('%.0f', mynum, 3)
  >>> '123,456'

// m
-





From see at reply.address.invalid  Tue Jun 25 02:06:31 2002
From: see at reply.address.invalid (Greg Ewing)
Date: Tue, 25 Jun 2002 18:06:31 +1200
Subject: __getattr__ feature
References: <3D180042.6000902@earthlink.net>
Message-ID: <3D180867.4020703@reply.address.invalid>

Hans Nowak wrote:
> I didn't know this was possible:
> 
> class Foo:
>     def __init__(self, obj):
>         self.obj = obj
>     def __getattr__(self, name):
>         return getattr(self.obj, name)
> 
> foo = Foo([1,2,3])
> print foo[2]    # prints 3

I think what's happening here is that invoking
the [] is causing a lookup for a __getitem__
method, and your __getattr__ is catching this
and redirecting it to self.obj, which does
have a __getitem__ method.

 > A quick test learns that this is apparently a
> feature new to 2.2;

The new feature is that built-in types now
have all the relevant special methods accessible
by __xxx__ names. Previously they only worked
on user-defined classes.

> Does anybody 
> know how/why this works and why it was added?

I don't think it was explicitly added -- it's
just a side effect of the type/class unification
stuff.

Greg Ewing, Computer Science Dept,
University of Canterbury,	
Christchurch, New Zealand
http://www.cosc.canterbury.ac.nz/~greg



From never at mind.info  Sun Jun  2 16:50:32 2002
From: never at mind.info (G. Willoughby)
Date: Sun, 2 Jun 2002 21:50:32 +0100
Subject: Reading coords from a scrolled Tkinter Canvas widget?
References: <addeiv$df1$1@news6.svr.pol.co.uk> <addjad$3cp@dispatch.concentric.net>
Message-ID: <ade0em$1u9$1@newsg1.svr.pol.co.uk>

>Your code appears not to grok canvasx and canvasy. They translate event.x
>style coordinates to canvas model coordinates.

Hmmm i know, have you any code to grab the canvasx and canvasy?

>BTW you could be using a Python Megawidgets Pmw.ScrolledCanvas here. More
>features less pain.

I could do but im trying to keep imports to a minimum and keep its
portability.

--G. Willoughby




From phd at phd.pp.ru  Mon Jun  3 10:57:31 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Mon, 3 Jun 2002 18:57:31 +0400
Subject: mySQL on Solaris .8
In-Reply-To: <79d1869.0206030646.512af52c@posting.google.com>; from dryose@juno.com on Mon, Jun 03, 2002 at 07:46:25AM -0700
References: <79d1869.0206030646.512af52c@posting.google.com>
Message-ID: <20020603185731.H29392@phd.pp.ru>

On Mon, Jun 03, 2002 at 07:46:25AM -0700, yose wrote:
> I am trying to install a mySQL database in solaris 2.8. I have a copy
> of the distro for mySQL-python-0.9.1. When I try to run setup.py, I am
> told that I am missing distutils.core
> 
> Where do I get this? What package is it part of?

   Do you use Python 1.5.2? Then download Distutils from
http://www.python.org/sigs/distutils-sig/download.html.

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From andreas.ames at Tenovis.com  Thu Jun 27 04:43:46 2002
From: andreas.ames at Tenovis.com (Andreas Ames)
Date: 27 Jun 2002 10:43:46 +0200
Subject: Interactive session via tcp
Message-ID: <u8z513yd9.fsf@tenovis.com>

Hi all,

I have a python daemon running with access to an XML messaging system
(jabber in this case).  I'm wondering, how I can achieve remote
control over my daemon.

The first thing that I can think of is to start an interactive session
via the XML stream.  Security is not my concern right now.  Following
steps seem to be required, afaik (please correct me, if my assumptions
are wrong).

1) Deamon receives a command to open an interactive session.

2) The main thread of my daemon starts a read/eval loop and redirects
   stdout and stder.

Is step 2 easily implemented in python?  Is it reasonably easy to have
more than one concurrent interactive sessions (say one local and one
remote or several remote)?

If there can be more than one concurrent interactive sessions how is
this implemented (different threads, different stdout/stderr etc.)?

How do several concurrent interactive sessions possibly interfere
(sideeffects etc.), when they use the same interpreter instance (or
don't they share the same one)?


TIA,

Andreas




From huaiyu at gauss.almadan.ibm.com  Thu Jun 20 23:14:53 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 21 Jun 2002 03:14:53 +0000 (UTC)
Subject: How to represent the infinite ?
References: <3d11ee9b$1@news.swissonline.ch> <20020620193230.53d58f8f.christophe.delord@free.fr> <aet4p6$33kk$1@nntp6.u.washington.edu> <20020620202732.68e67873.christophe.delord@free.fr> <slrnah4dnd.8m7.huaiyu@gauss.almadan.ibm.com> <20020620225355.7b8dd9db.christophe.delord@free.fr>
Message-ID: <slrnah56hd.aa1.huaiyu@gauss.almadan.ibm.com>

Christophe Delord <christophe.delord at free.fr> wrote:
>On Thu, 20 Jun 2002 20:11:25 +0000 (UTC)
>huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu) wrote:
>
>> Christophe Delord <christophe.delord at free.fr> wrote:
>> [regarding 1e1000 giving inf]
>> >With Python 2.2 and Linux it works. The example I gave is a copy-paste from IDLE. It also works with my Python 1.5.2 and 2.3 (I'm under Linux on an Intel processor)
>> 
>> Does 1e200**2 also work for you out of the box?  I had to get rid of the
>> artificial check for overflow in the source to really use inf in
>> computations.  I don't remember if 1e1000 works on my box without patch.
>
>
>>>> 1e200**2
>Traceback (most recent call last):
>  File "<pyshell#1>", line 1, in ?
>    1e200**2
>OverflowError: (34, 'Numerical result out of range')
>>>> 1e400
>inf
>
>When the overflow occurs, the OverflowError exception is thrown. inf can't be the result of a computation (without patch and without catching exception).

OK, that's what I thought.  I still don't understand why it should default
to disabling inf and nan in computation.  I've lived without the overflow
checking for a year, without noticeable bad effect (with much benefit, of
course).

Huaiyu


From siegfried.gonzi at kfunigraz.ac.at  Sat Jun 22 10:43:58 2002
From: siegfried.gonzi at kfunigraz.ac.at (Siegfried Gonzi)
Date: Sat, 22 Jun 2002 16:43:58 +0200
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <3D148CD0.4050008@NOSPA_M.online.no>
Message-ID: <3D148D2E.3020806@kfunigraz.ac.at>

Erlend J. Leiknes wrote:
> Siegfried Gonzi wrote:

> What do you mean with
> "free_software_garbage_2_weeks_installation_nightmare",
> Ive never had any problem installing python/idle...

Consult the deja news archive and see the many hundred of posts with 
Tkinter/Tcl/idle problems. And then look for the accompanying Readmes 
with the Python distribution...


>> SWIG helped me to make my external Python C function (with gcc); and
>> F2PY helped me to create my external Python Fortran 90 function (with
>> the Intel Fortran 95 compiler).
> 
> 
> Perhaps it wasnt python's fault?
> Perhaps it was swig or f2py?

No. I have tested the two external functions before and they work.


On Windows after the first wavelength the system goes down remarkably 
but it keeps over water; on Linux Python goes down.


S. Gonzi





From talon34 at hotmail.com  Wed Jun 26 13:30:31 2002
From: talon34 at hotmail.com (Talon)
Date: Wed, 26 Jun 2002 17:30:31 GMT
Subject: Installation problems DCOracle2 & W2000
References: <mailman.1025101359.15602.python-list@python.org>
Message-ID: <talon34-E85D1F.10285926062002@netnews.attbi.com>

In article <mailman.1025101359.15602.python-list at python.org>,
 brueckd at tbye.com wrote:

> On Wed, 26 Jun 2002, Talon wrote:
> 
> > I am having a heck of a time getting DCOracle2 to funcion properly.  I 
> > am running on a W2K machine, python2.1.3 and Zope 2.5.1, both of which 
> > run just fine.  I installed DCOracle2 and the ZOracleDA under Zope and 
> > it works fine.  But standalone python scripts don't see the dco2.py 
> > module.  So I tried to install DCOracle2 in the python21/lib folder.  I 
> > got that latest version from Zope.org and it said it installed properly.  
> > But it doesn't run.  If you try to import it into a script, you get 
> > complaints that it can't find this or that module, all of which are part 
> > of the Zope path.
> 
> Hi Talon,
> 
> Some of our development machines use the same configuration you list 
> above. You're probably _very_ close, so I'd stick with DCOracle2 if you 
> can. Would you mind posting the specific error messages you're getting? 
> Also, you mentioned that standalone scripts don't work, but just to be 
> clear, using DCOracle2 in Zope _does_ work without problems? (your 
> connection object is open, you can browse the database, you can run test 
> queries, ZSQL methods, etc.)
> 
> If it's working for Zope, then standalone scripts should be able to use
> DCO2 if you include the right directory in your PYTHONPATH environment
> variable, so please tell us what your PYTHONPATH is too.
> 
> Good luck,
> Dave
> 
> 
> 

Hi Dave,

OK, here you go:

Path = C:\Python21\;  (I don't have a specific PYTHONPATH variable set. 
This is in the general PATH variable)

The error I get is

File "C:\Python21\myfiles\htsa_data_loader_v2.py", line 9, in ?
  import os,sys,re,string,DCOracle2
File "C:\Python21\DCOracle2\__init__.py", line 91, in ?
  import DA
File "C:\Python21\DCOracle2\DA.py", line 90, in ?
  from db import DB
File "C:\Python21\DCOracle2\db.py", line 89, in ?
  import DCOracle2, DateTime
ImportError: No module named DateTime

DateTime is located in C:\Program Files\ZopeDev\lib\python\DateTime

are you suggesting that I put the Zope path into the PATH variable as 
well?  DCOracle2 DOES work in Zope, I can do all the things you listed, 
I have DTML that queries a db and returns the data.

By the way, I got things to work well with cx_Oracle this morning, so I 
do have another option, but I'm very interested in solving this problem.

Thanks for your assistance.

Mark


From garrett at bgb.cc  Wed Jun 19 01:23:49 2002
From: garrett at bgb.cc (Don Garrett)
Date: Wed, 19 Jun 2002 05:23:49 GMT
Subject: What If.....    Strong Types
References: <mailman.1024271960.27442.python-list@python.org>
Message-ID: <3D1014DC.7090600@bgb.cc>

brueckd at tbye.com wrote:
> On Sun, 16 Jun 2002, Don Garrett wrote:
> 
> I never understood all the hooplah over ints not being classes. I've only 
> ever seen a handful of cases where it might be useful, and lots of cases 
> where it'd get abused, but some people really seem to want it.

   My desire for it is simple desire for orthagonality. The curse of a CS 
background, I guess. But I'm curious, how could you really abuse it?

>>   My belief is that almost all the convenience of Python would be
>>maintained, but that compilation, optimization, and automated code
>>analysis would be simpler.
> 
> 
> I'm _sort of_ in favor of some optional type annotation for optimization 
> purposes, and sort of against it, the idea being that after your code is 
> done you could go back and annotate (or even have a tool suggest 
> annotation) variables with type information such that you are saying that 
> a particular variable name will always refer to integers, for example. My 
> fear, though, is that it would be used too often, i.e. people would 
> provide type information in places where it wouldn't help at all in terms 
> of execution speed. One of the beauties of Python is the clean syntax, the 
> readable code. A good portion of that is the code is not encumbered with 
> all sorts of variable and type declarations. I like the approach of Pyrex 
> because it's close to Python but still separate enough to prevent people 
> from using it too much - if you make it too easy to optimize the code then 
> people will optimize it too often. Bleh.

   I have very similar feelings on this. My hope is that high quality 
compilation could be done without type tagging. I really don't feel 
comfortable about that, but have some ideas about how to approach it.

>>   I'm just wondering if the change would be good or bad. Would it really 
>>break anything important in Python today?
> 
> 
> You might not consider it important, but it'd break a lot of _my_ code, so 
> it'd be important to me! :) Python doesn't force you to use all its 
> dynamic abilities, but I regard that as one of the tools of the Python 
> toolbox, so yes, I'd be hurt by its removal.

   What kinds of code would it break? That's what I'm trying to understand. I 
find that my code doesn't take advantage of these features at all.

   I use function pointers a lot, but I never change/add/delete methods at run 
time, and I always define all members in __init__, even if it's just to a 
default value (or None). Whenever I start to do anything else, it feels hacky 
and hard to follow. Obviously other people feel differently.

>>Would it help with automated code analysis (compilation and optimization 
>>included) as much as I think it would?
> 
> 
> Dunno - how much do you think it would? ;-) A better question, is how
> often do we really need compilation and optmization at all anyway? If it's
> less than 10% (which I believe it is), then the payoff generally wouldn't
> be worth language-wide changes. A plain extension module (especially with
> the help of Boost, SWIG, PyInline, etc.) isn't too painful, and projects
> like Pyrex might have a permanent niche too, but in all cases they are one 
> step removed from standard Python - which is exactly where they belong.

   Since what I'm trying (badly) to suggest is discussion of a new language 
based on Python, but which is compilable..... I'd say it's really needed.

   As a note, I'm pretty much against adding interfaces to Python. I think 
it's a matter of adding a feature when it's not needed just because everyone 
else has them.


> -Dave
> 
> 
> 



-- 
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc



From sholden at holdenweb.com  Mon Jun  3 08:40:51 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 3 Jun 2002 08:40:51 -0400
Subject: Why no open(f, "w").write()?
References: <mailman.1023067625.14591.python-list@python.org> <20020603181332.2975c585.larooy@xtar.co.nz>
Message-ID: <CxJK8.180902$%u2.92785@atlpnn01.usenetserver.com>

"John La Rooy" <larooy at xtar.co.nz> wrote ...
>
> >
> > It is explicitly not a documented part of the interface - the documented
> > part is that it is not safe to rely on this behaviour.
> >
> > When it is explicitly documented that this behaviour is guaranteed now
and
> > forever, for all implementations of Python after a specified version, I
will
> > *consider* using this behaviour. Until then it is not worthy of
> > consideration.
> >
> > Tim Delaney
> >
> >
>
> It's unlikely to be explicitly documented. from the docs...
>
> The file() constructor is new in Python 2.2. The previous spelling,
open(),
> is retained for compatibility, and is an alias for file().
>
> So you shouldn't be using open() in new programs anyway
>

You might want to consider using it for exactly the reason it's been
retained: backward compatibility to older versions of Python in use by your
potential users.

regards
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From claird at starbase.neosoft.com  Tue Jun 25 17:39:50 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 25 Jun 2002 16:39:50 -0500
Subject: Memory Monitor
References: <pan.2002.06.25.12.14.06.994352.17223@pobox.com> <3D1894B5.720EB7F7@bioreason.com>
Message-ID: <DD62FD80F1B154BC.D2610489CF54187F.736A27538AC8EB0B@lp.airnews.net>

In article <3D1894B5.720EB7F7 at bioreason.com>,
Mitch Chapman  <Mitch.Chapman at bioreason.com> wrote:
			.
			.
			.
>	inf = open("/proc/self/status", "r")
>	print inf.read()
>	inf.close()
>
>
>'course, you probably want more concise output, e.g.:
>
>        os.system("cat /proc/self/status | grep VmSize")
			.
			.
			.
Rather than
  os.system("grep VmSize /proc/self/status")
?
-- 

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


From dyoo at hkn.eecs.berkeley.edu  Fri Jun 21 15:46:23 2002
From: dyoo at hkn.eecs.berkeley.edu (Danny Yoo)
Date: Fri, 21 Jun 2002 12:46:23 -0700 (PDT)
Subject: [Tutor] How to find all emails (on the same line)
In-Reply-To: <3D12F30D.2071.204353@localhost>
Message-ID: <Pine.LNX.4.44.0206210949310.20729-100000@hkn.eecs.berkeley.edu>


On Fri, 21 Jun 2002, A wrote:

> I need to find ALL emails in a file. I can use re module but I do not
> know how to find all emails if there are more emails on the same
> line because re module ( as far as I know )finds only one/first
> occurrence on that line.

Ah!  Try the findall() method from the 're' module:

###
>>> def findAllNumbers(line):
...     return re.findall(r'\d+', line)
...
>>> findAllNumbers("this is a test 42")
['42']
>>> findAllNumbers("this is a test 42 but 43 and 44 are also nice")
['42', '43', '44']
###


Hope this helps!






From gerhard at bigfoot.de  Mon Jun 17 20:51:06 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Tue, 18 Jun 2002 02:51:06 +0200
Subject: Python Executable Code
In-Reply-To: <20020618022953.X15079@prim.han.de>
References: <1024357771.569c0ff8rchopra@myrealbox.com> <20020618000621.GA3119@lilith.my-fqdn.de> <20020618022953.X15079@prim.han.de>
Message-ID: <20020618005105.GA3468@lilith.my-fqdn.de>

* holger krekel <pyth at devel.trillke.net> [2002-06-18 02:29 +0200]:
> Gerhard H?ring wrote:
> > Ger'URL-bot'hard
> 
> some day there will be a quick search engine where
> people can just type in some keywords and get a good URL. 
> 
> oh wait ...

Some day there will be method to collect the frequently asked questions
from a newsgroup into a searchable interface.

http://www.python.org/cgi-bin/faqw.py
http://www.python.org/cgi-bin/faqw.py?query=py2exe&querytype=simple&casefold=yes&req=search

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From R.Barrett at ftel.co.uk  Wed Jun 26 04:56:33 2002
From: R.Barrett at ftel.co.uk (Richard Barrett)
Date: Wed, 26 Jun 2002 09:56:33 +0100
Subject: Python file upload
In-Reply-To: <F201A2DreBBsQwY6YHk00000017@hotmail.com>
Message-ID: <5.1.0.14.2.20020626095102.02a11c78@pop.ftel.co.uk>

I think you will find that your calls to cgi.escape the incoming data are 
the cause of your problem.

You do not need to escape incoming multipart/form data. The documentation 
for cgi.escape says:

<quote>
Use this if you need to display text that might contain such characters in HTML
</quote>

which is clearly not your purpose in handling the incoming data, 
particularly when it is binary data.

At 08:08 26/06/2002 +0000, Bjarne Christiansen wrote:
>Hi,
>I have some problems uploading binary files though the web browser. It 
>seens to work fine when uploading ACII file but but binary files seems to 
>be currupted. The begining of the binary file seems fine, but some data is 
>missing....
>
>Here is my file upload script:
>
>#!c:\Python\python -d
>
>import cgi
>
>print "content-type: text/html\n\n"
>
>form = cgi.FieldStorage()
>if not form:
>  print """
><form action="http://localhost/cgi-bin/upload.py" method="POST" 
>enctype="multipart/form-data">
><input type="file" name="filename">
><input type="submit">
></form>
>  """
>elif form.has_key("filename"):
>  item = form["filename"]
>  if item.file:
>    data = item.file.read()
>    print cgi.escape(data)
>    data1 = cgi.escape(data)
>    f = open("file1.jpg","wb")
>    f.write(data1)
>
>Any help will be greatly appriciated!
>
>Best Regards,
>Bjarne Christiansen
>bjarne_christiansen at hotmail.com
>
>
>_________________________________________________________________
>MSN Photos is the easiest way to share and print your photos: 
>http://photos.msn.com/support/worldwide.aspx
>
>
>
>--
>http://mail.python.org/mailman/listinfo/python-list





From whisper at oz.net  Mon Jun 17 19:04:53 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 17 Jun 2002 16:04:53 -0700
Subject: coding python by web?
In-Reply-To: <3eeda89d.0206170317.100549db@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEGEMADJAA.whisper@oz.net>

Medusa (http://oedipus.sourceforge.net/medusa/) has an example
"scriptserver" that might be what you're looking for. I don't know how it
works for sure though.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Thinkit
> Sent: Monday, June 17, 2002 4:17
> To: python-list at python.org
> Subject: coding python by web?
>
>
> Is there any way to test out simple python code with only a web
> browser (no input or files, just printing text)?  I'm guessing a
> simple cgi setup could do this, but has it been implemented anywhere?
> --
> http://mail.python.org/mailman/listinfo/python-list





From phd at phd.pp.ru  Sat Jun  8 07:22:26 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Sat, 8 Jun 2002 15:22:26 +0400
Subject: Where is Berkeley DB on your system? Can you live without DB 1.85?
In-Reply-To: <GCEDKONBLEFPPADDJCOEAEGJDHAA.whisper@oz.net>; from whisper@oz.net on Fri, Jun 07, 2002 at 01:14:59PM -0700
References: <15617.3834.292130.446065@12-248-41-177.client.attbi.com> <GCEDKONBLEFPPADDJCOEAEGJDHAA.whisper@oz.net>
Message-ID: <20020608152226.E31320@phd.pp.ru>

On Fri, Jun 07, 2002 at 01:14:59PM -0700, David LeBlanc wrote:
> If Sleepycat will allow it, i'm all in favor of updating to the latest
> version! However, they want money for commercial use of versions > 1.85
> AFAIK.

   They don't. You can freely incorporate BSDDB into an opensource project.
   http://www.sleepycat.com/faq.html#A22

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From zopestoller at thomas-guettler.de  Fri Jun 28 10:25:17 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Fri, 28 Jun 2002 16:25:17 +0200
Subject: GUI-Control for HTML
References: <3D1C4B40.6090803@thomas-guettler.de> <slrnahok93.js.gerhard@lilith.my-fqdn.de>
Message-ID: <3D1C71CD.6040409@thomas-guettler.de>

Gerhard Haering wrote:

> In article <3D1C4B40.6090803 at thomas-guettler.de>, Thomas Guettler wrote:
> 
>>Hi!
>>
>>Is there are GUI-control which can display HTML?
>>
>>The HTML that will be used will be simple, no javascript or
>>form support is needed.
>>
>>Is there are cross-plattform solution? (Unix + Win32)
>>
> 
> Yes, wxPython is available on these platforms, and has a HTML component. There
> might be a HTML viewer add-on for Tkinter, too.


Has someone tryed gtkHTML?

  thomas




From radix at twistedmatrix.com  Thu Jun  6 10:13:37 2002
From: radix at twistedmatrix.com (Christopher Armstrong)
Date: 06 Jun 2002 10:13:37 -0400
Subject: array, list, performance...
In-Reply-To: <3CFF5F98.2060406@destiny.com>
References: <3CFF5F98.2060406@destiny.com>
Message-ID: <87sn40xzou.fsf@twistedmatrix.com>

> Michael Chermside <mcherm at destiny.com> writes::
>   li[n] = x      assignment to the middle is O(len(li))

This doesn't seem right. Shouldn't that be O(1), just like
access-time?

-- 
                                Chris Armstrong
                         << radix at twistedmatrix.com >>
                http://twistedmatrix.com/users/carmstro.twistd/





From mongiaamit at yahoo.com  Tue Jun  4 00:05:00 2002
From: mongiaamit at yahoo.com (Amit Mongia)
Date: 3 Jun 2002 21:05:00 -0700
Subject: Fatal Python error: GC object already in linked list
Message-ID: <cf43c01d.0206032005.647e3355@posting.google.com>

Hi,
  I am running a program using pycurl, to upload files on a server. I
am using Python 2.2.1. I am using threads and lists in my program. the
application just keeps waiting for new files to arrive and uploads
them. There are times my program errors out and crashes with this
error message.

Fatal Python error: GC object already in linked list

I am not using the Garbage Collector anywhere in my program. Any idea
what can cause this error? Any solutions. I desperately need one.
There is nothing given anywhere in any of the books. I am not even
sure if its an error with the Garbage Collector. Any comments?

Thanks,
Amit Mongia


From whisper at oz.net  Wed Jun 26 18:59:57 2002
From: whisper at oz.net (David LeBlanc)
Date: Wed, 26 Jun 2002 15:59:57 -0700
Subject: Pythoniac: Thoughts on a hardware Python processor
In-Reply-To: <afc52t$2pj$1@sirius.dur.ac.uk>
Message-ID: <GCEDKONBLEFPPADDJCOEGEGLDMAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Christopher Saunter
> Sent: Wednesday, June 26, 2002 3:30
> To: python-list at python.org
> Subject: Pythoniac: Thoughts on a hardware Python processor
>
>
> Dear All,
>
> I have seen the occasional post about creating a specific hardware 'Python
> Processor', that could then be built and used in programmable logic or the
> like (Valves, relays, core memory - Pythoniac ;-) - in other words, create
> a processor that executes Python bytecode directly.  There would be
> various different reasons for doing this (the 'neat' factor etc...)

<snip>

> I would like to try and implement this, initially in a simulation
> environment, then in a real FPGA, with hooks back to a PC allowing the
> loading, execution, single step execution and monitoring etc of PBC.
> Ideally I would like to reach a state where an interactive interpreter
> runs over a serial port...
>
> So, anyone else interested in this?  Comments, suggestions, gaping holes
> in my understanding of things as outlined above etc are welcome!
>
> Regards,
> 	Chris Saunter

Have you considered the need for:

* Bus control
* Interrupts
* Memory Management

(without even considering pipelining or other performance enhancing
techniques)

Xilinix makes available a basic FPGA/FPLA design tool for free - an entry
level of their full-on product. They are also known to support projects with
full copies of their tool and their (not-inexpensive) chips for free.
Presumably they do this if they think there might be commercial potential.

Here's a starting place: http://www.etek.chalmers.se/~e8cal1/jamcpu.html

Introductory survey piece: http://www.eetimes.com/story/OEG20010201S0050

Motherlode of information: http://www.opencores.org/ Includes info on open
design tools, most for Linux.

64 bit super-pipelined CPU dev project: http://www.f-cpu.org/

HTH,

Dave LeBlanc
Seattle, WA USA

P.S. Personally, I think a kick-ass true compiler with a retargetable
backend would be a better thing to persue.





From insyte-clp at emt-p.org  Thu Jun 20 18:40:24 2002
From: insyte-clp at emt-p.org (Ben Beuchler)
Date: 20 Jun 2002 22:40:24 GMT
Subject: M2Crypto / SimpleXMLRPCServer
References: <slrnah2nmh.p18.insyte-clp@petra.bitstream.net>
Message-ID: <slrnah4md0.r77.insyte-clp@petra.bitstream.net>

In article <slrnah2nmh.p18.insyte-clp at petra.bitstream.net>, Ben Beuchler wrote:

> It appears that the SSL part, at least, works great.  I'm able to make a
> connection using xmlrpclib.ServerProxy("https://blah") but the actual
> remote method call hangs indefinitely.
> 
> Is my basic premise flawed?  Or just my implementation?

FWIW, strace-ing the client and the server indicates that they're both
blocking on read()ing from the socket.  Of course, since it's all
encrypted, I have no idea what they're waiting for...

On a related note, anyone know how to get OpenSSL (specifically via
M2Crypto) to use a NULL cipher?  Being able to dump the traffic with
Ethereal or something would be damn spiffy.

-Ben

-- 
Ben Beuchler                                           There is no spoon.
insyte at emt-p.org                                            -- The Matrix


From aahz at pythoncraft.com  Thu Jun  6 10:36:56 2002
From: aahz at pythoncraft.com (Aahz)
Date: 6 Jun 2002 10:36:56 -0400
Subject: array, list, performance...
References: <mailman.1023369012.10674.python-list@python.org>
Message-ID: <adns28$7jb$1@panix1.panix.com>

In article <mailman.1023369012.10674.python-list at python.org>,
Michael Chermside  <mcherm at destiny.com> wrote:
>
>    li.sort()      sort is O(len(li)*ln(len(li)))
>                   (I believe it uses quicksort, but if a python
>                    function has to be executed for the compare
>                    that may take some time.)

Nope.  I forget what Uncle Timmy used as his base, but while the
algorithm is O(NlogN) like quicksort, worst-case time is significantly
improved over quicksort's worst-case O(N^2), and the algorithm is
specifically designed to minimize compares precisely because even if a
Python function isn't listed for L.sort(), if L contains Python class
instances, it's quite likely to call instance methods.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From mwh at python.net  Thu Jun 20 04:32:04 2002
From: mwh at python.net (Michael Hudson)
Date: Thu, 20 Jun 2002 08:32:04 GMT
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <20020619154654.C619@ActiveState.com> <mailman.1024527565.30927.python-list@python.org>
Message-ID: <lkvg8ejqrd.fsf@pc150.maths.bris.ac.uk>

Gerhard H?ring <gerhard.haering at gmx.de> writes:

> * Trent Mick <trentm at ActiveState.com> [2002-06-19 15:46 -0700]:
> > My understanding is that for C extensions to work they must be built
> > with the same compiler as that used to build the Python with which they
> > will run. (I may be wrong. Someone please correct me if I am.)
> 
> You are wrong. As long as you link to the python{major}{minor}.dll you
> are binary compatible, no matter which compiler you use.

Eh, that's a bit of a stretch isn't it?  If your compilers have
different ideas about struct layout & so on it would certainly take
more effort than just linking the same libraries.

> It's just a
> matter of providing an additional compiler option to distutils, like in:
> 
>     python setup.py build --compiler=mingw32

I do think this works, though.  I guess all compiler implementors on
Windows try to be compatible with MSVC.

Cheers,
M.

-- 
  Well, you pretty much need Microsoft stuff to get misbehaviours
  bad enough to actually tear the time-space continuum.  Luckily 
  for you, MS Internet Explorer is available for Solaris.
                              -- Calle Dybedahl, alt.sysadmin.recovery


From arosado at softhome.net  Sat Jun  8 21:03:32 2002
From: arosado at softhome.net (Andres Rosado)
Date: Sat, 08 Jun 2002 21:03:32 -0400
Subject: Distributing Packages / os independent installation folder
Message-ID: <5.1.0.14.0.20020608210324.00bcff38@mail.softhome.net>

At 06:01 AM 6/7/2002 -0400, you wrote:
>I'm building a python interface to ImageMagick. I already succeded in
>building an installer with distutils, but have some questions regarding
>portability:
>
>- where should my package and shared librarie go to? On windows I put the
>dll into python22/DLLs and the rest into a folder
>python22/Lib/site-packages/ImageMagick. What to do on other plattforms? I
>don't expect a DLLs folder on Linux. ;-)
>
>- In my __init__.py I have to change one environment variable. Currently I
>do it like this:
>
>import os
>os.environ['MAGICK_HOME'] = r'C:\Python22\Lib\site-packages\ImageMagick'
>from _ImageMagick import *
>
>Is there an plattform independent way to get the installation folder of my
>package? On windows I can get the python folder from registry, but this will
>also break on linux.

I'm new at Python, but you could use wrap any platform-dependant code in if 
statements using the os.name like this:

if os.name == 'nt':
         #Whatever you are going to do in WinNT family
         pass
elif os.name == 'posix':
         #Whatever you are going to do in Linux
         pass
else:
         print '''Sorry, we don't support your platform.
Maybe continue message here.
'''

For the other questions, I don't know. :(


-----------------------------------
Andres Rosado
Email: andresr at despammed.com
ICQ: 66750646
Homepage: http://andres980.tripod.com/

All the world's a stage and most of us are desperately unrehearsed.
                 -- Sean O'Casey





From gustav at morpheus.demon.co.uk  Wed Jun 12 05:47:21 2002
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: Wed, 12 Jun 2002 11:47:21 +0200
Subject: Python Virtual File System: A spike test
References: <mailman.1023872902.2540.python-list@python.org>
Message-ID: <fzzsg76f.fsf@morpheus.demon.co.uk>

"Mike C. Fletcher" <mcfletch at rogers.com> writes:

> Purpose:
>
> A class for representing file system paths which supports both 
> method-based operations, and abstraction through object encapsulation to
> eventually allow for Virtual File System operations (for instance, 
> transparently accessing the contents of zip files).

[...]

> 	open( ) method for path specifiers

[...]

> If you're interested in working on such a project (or even using it), 
> let me know, so I can decide whether to devote any more resources to it.

I'm interested, in the sense that I would have a use for a way of
doing open() on an arbitrary "path-like" thingy, and getting back a
relatively full-featured file-like object, as efficiently as
possible. (ie, preferably without grabbing the whole contents of the
file into memory and using cStringIO :-))

Most of the rest of the path-manipulation functions are of little
interest to me, personally.

I would be willing to help in coding the parts I have a use for, but I
have limited time.

Paul.



From mlh at vier.idi.ntnu.no  Sun Jun  9 18:57:20 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sun, 9 Jun 2002 22:57:20 +0000 (UTC)
Subject: email: msg.add_payload not synonymous with msg.attach?
References: <slrnafvtu6.7q0.mlh@vier.idi.ntnu.no>  <slrnag5j96.17b.quinn@lira.ugcs.caltech.edu>
Message-ID: <slrnag7nag.5n0.mlh@vier.idi.ntnu.no>

In article <slrnag5j96.17b.quinn at lira.ugcs.caltech.edu>, Quinn Dunkan wrote:
[snip]
>In 1.0 (comes with python 2.2.1), attach is defined as 'attach = add_payload'
>so it's synonymous.  It also doesn't have any mention of 'add_payload" being
>deprecated.  I hope it's not really, since I use it all the time.
>
>CVS is probably not at 1.0 anymore.

Probably not. There is a deprecation warning, and they behave
differently. Maybe it's a bug in the package -- maybe I'm using it
wrong.

[snip]
>What the error msg probably means is that your message somehow has multiple
>payloads, but doesn't have a multipart/* type.  Maybe they changed attach
>to always make _payload a list or something.

Arf... But I don't want a multipart message :|

>>(The code is very simple -- basically just instantiating Message and
>>adding a payload.)
>>
>>I'm using the email package from CVS (downloaded everything a day or
>>two ago).
>
>I wouldn't be surprised if the docs are not up to date with the
>latest CVS.

I'm sure they're not... But this sort of breakage is unfortunate
anyway...

> You should probably either use 1.0 or check the
>Message.py source.

I guess so.

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From dreed at capital.edu  Sat Jun 29 16:49:17 2002
From: dreed at capital.edu (Dave Reed)
Date: Sat, 29 Jun 2002 16:49:17 -0400
Subject: can't import generators
In-Reply-To: <afl5je$4iu$1@panix1.panix.com> (aahz@pythoncraft.com)
References: <m2hejlc060.fsf@mother.paradise.lost> <afl5je$4iu$1@panix1.panix.com>
Message-ID: <200206292049.g5TKnHx28907@localhost.localdomain>

> From: aahz at pythoncraft.com (Aahz)
> Newsgroups: comp.lang.python
> Organization: The Cat & Dragon
> Xref: news.baymountain.net comp.lang.python:170335
> Sender: python-list-admin at python.org
> Date: 29 Jun 2002 16:33:50 -0400
> 
> In article <m2hejlc060.fsf at mother.paradise.lost>,
> John Hunter  <jdhunter at nitace.bsd.uchicago.edu> wrote:
> >
> >I am having trouble working with generators.  Anyone know what might
> >cause this behavior:
> >
> >Python 2.2.1 (#1, Apr 22 2002, 21:20:54) 
> >[GCC 3.0.4] on linux2
> >Type "help", "copyright", "credits" or "license" for more information.
> >>>> from __future__ import generators
> >Traceback (most recent call last):
> >  File "<stdin>", line 1, in ?
> >ImportError: cannot import name generators
> 
> Can't do this in interactive mode.  Use a script.


Works for me:

Python 2.2.1 (#1, May  7 2002, 17:07:14) 
[GCC 2.96 20000731 (Red Hat Linux 7.3 2.96-110)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> from __future__ import generators
>>> 

This is a Red Hat 7.3 system with 2.2.1 compiled by me from the .tgz
file.

Dave




From candiazoo at attbi.com  Thu Jun 27 13:38:15 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Thu, 27 Jun 2002 17:38:15 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com>
Message-ID: <3d1b4d29.306797031@netnews.attbi.com>

Based on my profile, it looks like several of my functions could be revised.
The fix_date_range function converts a string (a year) to an integer, performs a
series of tests on it, and ends up returning a string date range (YYYY-YYYY)
based on the values...

to_string is where I built my output record.  I'll try a few things with it...

Mike

On Thu, 27 Jun 2002 01:28:22 GMT, candiazoo at attbi.com wrote:

>I assume some sort of block i/o or preallocating a large block of file space
>would be the best way to do it?  I am outputting about 700,000 records to a text
>file and my program only spits out about 20 records per second.  This is using
>standard file open and write...
>
>This is on a W2K machine... perhaps I should use the win32file interface?  Any
>thoughts/opinions?
>
>Thank you!
>
>Mike J.



From tdelaney at avaya.com  Sun Jun  2 20:31:25 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 3 Jun 2002 10:31:25 +1000 
Subject: Thanks!
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A33C@natasha.auslabs.avaya.com>

> From: Banar-Tul [mailto:tonyclick at mindspring.com]
>
> Ah! That sound you hear is me repeatedly slapping my forehead. Thanks.

What type of fish?

Tim Delaney




From cliechti at gmx.net  Sun Jun 30 12:02:26 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 30 Jun 2002 18:02:26 +0200
Subject: [Newbie] Is the text delimiter changeable?
References: <3d1f26a4.5915886@news.texas.net>
Message-ID: <Xns923DB7F943FF7cliechtigmxnet@62.2.16.82>

baf at texas.net (Ben Fairbank) wrote in news:3d1f26a4.5915886 at news.texas.net:

> I have to process large amounts of text and the text includes numerous
> apostrophes and quotation marks.  It would be very handy if I could
> temporarily assign a substitute for Python's use of the caracters "
> and ' to delimit text strings.  If I could use, for example, the @
> character or the # character, neither of which appears in the text I
> have to work with, and if I could disable the " and ' delimiters, then
> my task would be greatly simplified.  I realize I can backslash escape
> these characters in some circumstances, but I do not think that will
> work when a user's input is a word such as "can't" or "don't."
> Anyway, short of doing something truly drastic, such as recompiling
> the system (which I have no intention of doing), is there a workaround
> that will permit this?

are these text you wan't to process python sources in any way or just 
normal text? in the later you can load these texts in python strings and 
there you can have any character without having a special effect.

e.g. to replace all "is" in a text with "isn't":
out = file("output.txt","w")
for line in file("input.txt"):
    	out.write(line.replace("is", "isn't")) #process lines

HTH
chris

-- 
Chris <cliechti at gmx.net>



From bokr at oz.net  Thu Jun  6 15:17:31 2002
From: bokr at oz.net (Bengt Richter)
Date: 6 Jun 2002 19:17:31 GMT
Subject: for , ordering
References: <mailman.1023377649.29116.python-list@python.org> <yu99u1og1g9b.fsf@europa.research.att.com> <yu99ptz41eiu.fsf@europa.research.att.com>
Message-ID: <adocgb$j2m$0@216.39.172.122>

On Thu, 6 Jun 2002 17:51:53 GMT, Andrew Koenig <ark at research.att.com> wrote:

>gabor> for obj1 in objects1:
>gabor> objects2.append(obj2(obj1.x))
>
>
>gabor> 1. isn't there a more elegant way to do this?
>
>ark> Isn't this the same as saying
>
>ark>         objects2 = objects1[:]
>
>ark> ?  Am I missing something obvious?
>
Misplaced coffee mug? I know that one ;-)

>The answer is yes:  I *am* missing something obvious:  I didn't
>notice the obj2(obj1.x).  But why not
>
>             objects2 = map(lambda x: obj2(obj1.x), objects1)
Um,
              objects2 = map(lambda y: obj2(y.x), objects1)
?
E.g., (for others ;-)

 >>> class X:
 ...     def __init__(self,v): self.x=v
 ...     def __str__(self): return '<from X(%s)>' % self.x
 ...     __repr__ = __str__
 ...
 >>> objects1 = [X(i) for i in xrange(5)]
 >>> objects1
 [<from X(0)>, <from X(1)>, <from X(2)>, <from X(3)>, <from X(4)>]
 >>> def obj2(x): return '<obj2(%s)>' % x
 ...
 >>> objects2  = map(lambda y: obj2(y.x), objects1)
 >>> objects2
 ['<obj2(0)>', '<obj2(1)>', '<obj2(2)>', '<obj2(3)>', '<obj2(4)>']

>
>?
 >>> map(lambda x: obj2(obj1.x), objects1)
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "<stdin>", line 1, in <lambda>
 NameError: global name 'obj1' is not defined

I'd second Eric's suggestion though:

 >>> objects2_eric = [obj2(obj1.x) for obj1 in objects1]
 >>> objects2_eric
 ['<obj2(0)>', '<obj2(1)>', '<obj2(2)>', '<obj2(3)>', '<obj2(4)>']

Regards,
Bengt Richter


From xscottgjunk at yahoo.com  Fri Jun 28 16:42:42 2002
From: xscottgjunk at yahoo.com (Scott Gilbert)
Date: 28 Jun 2002 13:42:42 -0700
Subject: feeding a prompt with python
References: <3D1C5341.9070001@huno.net>
Message-ID: <79b2b0bb.0206281242.131c385e@posting.google.com>

thomas <tom at huno.net> wrote in message news:<3D1C5341.9070001 at huno.net>...
> hello,
> 
> is there a way with python to feed a prompt with data? i call sth. with 
> system($some_cmd) and the cmd gives me a prompt which of course waits 
> for user input. however i want to automate this, is there a way?
> i browsed through the essential reference 2nd edt. but didn't find anything.
> 
> thx,
> thomas

Look at os.popen(some_cmd).  On unixy like places, it will give you a
file handle that you can write to.


From peter at engcorp.com  Sun Jun  2 17:42:38 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 02 Jun 2002 17:42:38 -0400
Subject: string generator
References: <3cfa8039_1@news.iprimus.com.au>
Message-ID: <3CFA914E.8573BAA2@engcorp.com>

Gold Fish wrote:
> 
> I like to create a list of file automatical based on series. For example
> user input the number of file say 5. So i want to generate the file
> according to this range , this mean the generated files could be
> filename1
> filename2
> filename3
> filename4
> filename5
> I using the for loop like that
> 
> for file in range(5):
>         lis = [file]
>         string.join('filename',lis[1:])
> how can it didn't work. I am confusing how to do it.

The 'for' statement is giving you five passes through the
loop, with 'file' set to the integer values 0, 1, ... 4.
You put them each time inside a list called 'lis', then
take a slice of the list *after* the only element you've
put in it.  Even if it worked, you would just be throwing
away the results of the string.join() since you don't
do anything with it.

Presumably you are getting an AttributeError as a result.
Please *always* post the trackback you get so we know what
the exact problem is.  Don't make us guess.

Anyway, what you want is more like this:

for fileIndex in range(5):
    fileName = 'filename%s' % fileIndex
    print fileName

Now if you want to actually create files with those
names, you'll need "file = open(fileName, 'w')" and
more.

Have you gone through the complete tutorial yet?  Something
tells me you're missing some of the basics still.

-Peter


From mark at mceahern.com  Wed Jun 12 14:48:40 2002
From: mark at mceahern.com (Mark McEahern)
Date: Wed, 12 Jun 2002 13:48:40 -0500
Subject: get a list of classes in this module
Message-ID: <NCBBLFCOHHDIKCAFGCFBCEEKLEAA.mark@mceahern.com>

Suppose I have a list of classes:

# validation.py
class CreditCardValidation:
  ...

class MasterCard(CreditCardValidation):
  ...

class Visa(CreditCardValidation):
  ...

and I want to make a helper method in the module to translate card type
strings into their appropriate class constructor.

# someother.py
def validate(card_number, card_type=None):
  import validation
  import inspect
  classes = inspect.getmembers(validation, inspect.isclass)
  class_dict = dict(classes)
  if not class_dict.has_key(card_type):
    # use generic validation
    c = validation.CreditCardValidation()
  else:
    c = class_dict[card_type]()
  c.card_number = card_number
  c.validate()

I know I can use inspect.getmembers() to do this, but what if I want to put
the helper method in the same file as the classes (i.e., validation.py)?
How do I refer to *this* module?  Why do I want the helper in the same
module?  Well, heck, why not?  ;-)

Should I use globals()?  I mean, that would work--so I guess the question is
whether there's a more Pythonic approach?

Cheers,

// m

-





From peter at engcorp.com  Fri Jun 21 09:30:36 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 21 Jun 2002 09:30:36 -0400
Subject: Dispatching a Win32 COM in the background?
References: <mailman.1024642348.12905.python-list@python.org>
Message-ID: <3D132A7C.94CE4670@engcorp.com>

David LeBlanc wrote:
> 
> Hmmm... if this is working for you, that's great - it just looks strange to
> me.
> 
> If Background_TTS is the background speacher thread, shouldn't it have the
> queue.get() in it and the foreground have the queue.put()?
> 
> It's also worth noting that self isn't used or needed at the global level,
> so self.speech is kind of different ;-)

I'm pretty sure from previous postings he was going to call this from
within another object, so the "self." used elsewhere might refer to it.

As near as I can tell, with that in mind, the .get() and .put() are in
exactly the right places.

-Peter


From aahz at pythoncraft.com  Fri Jun 21 14:45:34 2002
From: aahz at pythoncraft.com (Aahz)
Date: 21 Jun 2002 14:45:34 -0400
Subject: [newbie] Overwriting class definitions
References: <mailman.1024678349.29532.python-list@python.org>
Message-ID: <aevs8e$has$1@panix1.panix.com>

In article <mailman.1024678349.29532.python-list at python.org>,
Sean 'Shaleh' Perry <shalehperry at attbi.com> wrote:
>
>I find that existing the interpreter is the safest thing to do while doing
>interative coding like this.

s/existing/exiting/

And I generally work from the command line, running a script over and
over rather than going back into the interpreter and retyping stuff.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From jepler at unpythonic.net  Sat Jun  1 10:03:16 2002
From: jepler at unpythonic.net (jepler at unpythonic.net)
Date: Sat, 1 Jun 2002 09:03:16 -0500
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
In-Reply-To: <07d801c20970$cdebe050$6601a8c0@boostconsulting.com>
References: <06da01c208f4$d69003c0$6601a8c0@boostconsulting.com> <200206010121.g511LxX19223@pcp742651pcs.reston01.va.comcast.net> <073501c20960$23f367e0$6601a8c0@boostconsulting.com> <200206011334.g51DY1D21669@pcp742651pcs.reston01.va.comcast.net> <07d801c20970$cdebe050$6601a8c0@boostconsulting.com>
Message-ID: <20020601090316.A10396@unpythonic.net>

On Sat, Jun 01, 2002 at 09:32:39AM -0400, David Abrahams wrote:
> My question was about what a new type needs to do in order for things to
> work properly in Python. If, as I had incorrectly assumed, Python were
> checking a type's mutability before deciding whether it would be putting
> the result back into the sequence, I would need to know what criteria
> Python uses to decide mutability.

IMO the following is an odder behavior:
>>> t
([1], [2], [3])
>>> t[1] += [3]
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: object doesn't support item assignment
>>> t
([1], [2, 3], [3])

Jeff




From hfoffani at yahoo.com  Thu Jun 20 12:17:04 2002
From: hfoffani at yahoo.com (nntpserver.pppl.gov)
Date: Thu, 20 Jun 2002 18:17:04 +0200
Subject: Useful RE patterns (was: Variable Interpolation - status of PEP
 215)
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <4PfQ8.44245$n4.10307683@newsc.telia.net> <yciQ8.44268$n4.10312169@newsc.telia.net>
Message-ID: <3D120000.40905@yahoo.com>

Fredrik Lundh wrote:
> If I were to add a dozen (or so) patterns to the (S)RE module,
> what should I pick?  What patterns do you find yourself using
> over and over again?

Most of the time whenever I use regexes I'm trying to solve
specific problems.  But what comes to mind are lexers-token-like
patterns: identifiers, numbers (ints, floats), strings (yes,
know :-)
They don't have to be _perfect_ to be useful.  In fact, laziness
takes me to steal them from other places.

Regards,
-Hernan



From dswegen at software.plasmon.com  Thu Jun 27 10:23:08 2002
From: dswegen at software.plasmon.com (Dave Swegen)
Date: Thu, 27 Jun 2002 15:23:08 +0100
Subject: Difference between HTMLParser and htmllib.HTMLParser
In-Reply-To: <f5FS8.325922$352.35979@sccrnsc02>
References: <mailman.1025178934.31606.python-list@python.org> <f5FS8.325922$352.35979@sccrnsc02>
Message-ID: <20020627142308.GD15813@software.plasmon>

On Thu, Jun 27, 2002 at 02:14:35PM +0000, Emile van Sebille wrote:
> "Dave Swegen" <dswegen at software.plasmon.com> wrote in message
> news:mailman.1025178934.31606.python-list at python.org...
> > I've just started playing with these, and I'm curious which is the
> > better of the two (I noticed the HTMLParser module only came in in
> 2.2).
> > And is there any reason why the names are so confusingly similiar? I
> > tried searching in google for info on this, but the daft naming made
> it
> > rather hard ;)
> >

[ helpful explantion stuff removed ]

Cheers. That was exactly the sort of info I was looking for. I just
wasn't sure which one to use.

Cheers
    Dave




From peter at engcorp.com  Tue Jun 18 20:44:47 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 18 Jun 2002 20:44:47 -0400
Subject: NEWbie asks "is Python a good tool for automated file/folder 
 generation?"
References: <mailman.1024073951.24787.python-list@python.org> <3D0E8773.A1BA95ED@engcorp.com> <3D0F5967.2A9CEE86@noaa.gov>
Message-ID: <3D0FD3FF.49FBDFE0@engcorp.com>

Chris Barker wrote:
> 
> Peter Hansen wrote:
> > Better to learn to use raw-strings, with an "r" in front,
> > as in path = r"c:\this is a safe path" or better yet,
> > either forward slashes always: path = "c:/this works too".
> 
> Or even better, use os.path.join(), you never know when you will
> encounter a Mac. (of course, if it's OSX, the above will work as well)

I started writing that, then went to the interactive prompt
and concluded (after brief experimentation) that there was
no way to join the "c:" and the "this works too" portions
of the path without explicitly specifying a "/" to force it
to be an absolute path.  Doubtless I did something wrong,
but I decided to avoid the issue by skipping any mention of
os.path.join().

Of course, since drive letters are not portable anyway this
is all a moot point.  For any hardcoded path like that, 
you can't have a very good cross-platform solution I suppose.

-Peter


From bkc at Murkworks.com  Mon Jun  3 11:04:13 2002
From: bkc at Murkworks.com (Brad Clements)
Date: Mon, 3 Jun 2002 11:04:13 -0400
Subject: Interfacing to 1-Wire LAN, especially Dallas temperature sensors?
References: <a7iefuk650noamrqcmpvroqj60mtgbpf9v@4ax.com> <3CF82E8E.23C6FA10@engcorp.com>
Message-ID: <3cfb83e2$1_4@goliath.newsgroups.com>

There is a technique where you can use a UART at some whacky baud rate and
bits/byte setting to communicate. There's an APPNOTE somewhere, I don't have
it handy.

--
Novell DeveloperNet Sysop #5

_
"Peter Hansen" <peter at engcorp.com> wrote in message
news:3CF82E8E.23C6FA10 at engcorp.com...
> Christian Aastorp wrote:
> >
> > I'm considering doing some temperature control and logging, both
> > enviromental and inside computers. Did a google search, and found
> > references to the range of temperature sensors from Dallas
> > Semiconductors.
> >
> > I also found schematics for interfacing with the serial port, and
> > timing diagrams for implementing 1-Wire Lan using direct control of
> > the port.
> >
> > As I'm using  Python for my programming, these days, I would like some
> > pointers to modules or techniques to use Python to read the
> > temperature sensors.
>
> Although I have not tried it with Python, and I would never say
> something is impossible without more study, I am _fairly_ sure you
> cannot do this reliably with Python.
>
> The Dallas 1-Wire chips have protocol timing requirements somewhere
> down in the low _tens_ of microseconds.  This requires some kind of
> real-time code and not only is Python itself unlikely to have
> high enough performance on almost any current machine, but the
> OS you're running it on is even less likely to support the
> necessary latency.
>
> We've been struggling just to get a 16MHz 68HC12 working with
> these things, without having to disable interrupts for hundreds
> of microseconds at a time.  It hasn't been easy.
>
> It might be very instructive to try, however, and I'd be happy
> to hear somebody say they'd proven me wrong (and curious what
> was done to reach a solution).
>
> -Peter




-----------== Posted via Newsfeed.Com - Uncensored Usenet News ==----------
   http://www.newsfeed.com       The #1 Newsgroup Service in the World!
-----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers =-----


From aahz at pythoncraft.com  Wed Jun 19 15:19:52 2002
From: aahz at pythoncraft.com (Aahz)
Date: 19 Jun 2002 15:19:52 -0400
Subject: Pop a list from beginning ? and memory saving...
References: <aem8e8$4kq$1@news.idiom.com> <BIEJKCLHCIOIHAGOKOLHAELEDEAA.tim.one@comcast.net> <mailman.1024424930.10523.python-list@python.org> <aep4jb$1q8h$1@news.idiom.com>
Message-ID: <aeqlgo$gbk$1@panix1.panix.com>

In article <aep4jb$1q8h$1 at news.idiom.com>,
James T. Dennis <jadestar at idiom.com> wrote:
>
> It might be better to have a hack to the implementation of the
> list primitives --- so they defer the shift of remaining elements
> until GC.  The head of the list could point at the first none 
> deleted item.  Then pop(0) could be implemented in O(1) time and
> the garbage collector would eventually come along to do it's 
> job.

While this might be a worthwhile optimization, it has absolutely nothing
to do with GC.  Each time an element were popped off the front, its
object would be decreffed and possibly deleted.  GC runs completely
independently of individual objects.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From stefan1werner at hotmail.com  Tue Jun 11 16:34:52 2002
From: stefan1werner at hotmail.com (Stefan Werner)
Date: Tue, 11 Jun 2002 22:34:52 +0200
Subject: Embedding Python 2.2 in non-reentrant application
Message-ID: <ae5n1f$qtm$07$1@news.t-online.com>

Hi group,

I am working on embedding Python 2.2 in a large scale application. Now the
problem is, that Python 2.2 supports multithreading, where the functions we
want to expose to Python are not threadsafe. Is there something like a flag
to make my whole module single-threaded for Python access or do I have to
wrap every single function call in
Py_BEGIN_ALLOW_THREADS/Py_END_ALLOW_THREADS macros?

Thanks,
  Stefan




From benji_york at cknainc.com  Thu Jun 27 09:18:28 2002
From: benji_york at cknainc.com (Benji York)
Date: 27 Jun 2002 06:18:28 -0700
Subject: printing (again!!)
References: <afcusp$66r$1@newsg2.svr.pol.co.uk>
Message-ID: <b10fcc9a.0206270518.4224069a@posting.google.com>

"Geoff Tarrant" <geoff at galtar.co.uk> wrote in message news:<afcusp$66r$1 at newsg2.svr.pol.co.uk>...
> I want to use Python as the language in a Computing course that I teach.
> However, I can't find a way that is relatively straightforward to send data
> to a printer in a Windows environment.

If you have a recent win32all installed, then the function below will
send plain text (requires both carrage returns and line feeds) to the
printer.  You can also throw in some PCL if you like, but for more
complex stuff I would use some of the suggestions seen elsewhere in
this thread (PDF or wxPython (especially it's HTML printing
functions)).


import win32print

def rawPrint(data, jobName, printerName=None):
    if printerName == None:
        printerName = win32print.GetDefaultPrinter()

    printerHandle = win32print.OpenPrinter(printerName)
    printJob = win32print.StartDocPrinter(printerHandle, 1,
                                          (jobName, None, None))
    win32print.WritePrinter(printerHandle, data)
    win32print.EndDocPrinter(printerHandle)
    win32print.ClosePrinter(printerHandle)



happy-printing-ly y'rs,

Benji York
benji_york at cknainc.com


From Sascha.Ferley at infineon.net  Sun Jun 23 15:52:29 2002
From: Sascha.Ferley at infineon.net (Sascha Ferley)
Date: Sun, 23 Jun 2002 13:52:29 -0600
Subject: PIL, Python and CGI
Message-ID: <3D1626FD.DA934B09@infineon.net>

Hi,
I have a problem when writing a program of mine and moving it into a CGI
based form.
Basically if i run the python program in shell mode, it works fine. But
when i run it via cgi mode
it screws up horobly.
The problem is, I am using the function from the PIL library as follows.

import sys
import cgi
import Image
from urllib import urlretrieve

form = cgi.FieldStorage()
inputPath = form["inpP"].value
compPath = form["secP"].value

try:
        im1 = urlretrieve(inputPath)[0]
        im2 = urlretrieve(compPath)[0]
        print im1
        print im2
except:
        print "Error occured in grabbing files!"
        sys.exit(1)

try:
        Image1 = Image.open(im1)      # Tried
Image.open(im1).histogram()  still gives error
        Image2 = Image.open(im2)
except:
        print "Error opening files"
        sys.exit(1)

print image1.histogram()
print image2.histogram()


Looking at the error-log i see just following:

Traceback (most recent call last):
  File "/export/server/web/cgi-bin/compare.py", line 29, in ?
    print image1.histogram()
  File "/usr/local/lib/python2.2/site-packages/PIL/Image.py", line 577,
in histo
gram
    self.load()
  File "/usr/local/lib/python2.2/site-packages/PIL/ImageFile.py", line
126, in l
oad
    self.load_prepare()
  File "/usr/local/lib/python2.2/site-packages/PIL/ImageFile.py", line
180, in l
oad_prepare
    self.im = Image.core.new(self.mode, self.size)
  File "/usr/local/lib/python2.2/site-packages/PIL/Image.py", line 39,
in __geta
ttr__
    raise ImportError, "The _imaging C module is not installed"
ImportError: The _imaging C module is not installed


Anyone have any ideas?
Sascha




From fperez528 at yahoo.com  Thu Jun 20 17:51:35 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 20 Jun 2002 15:51:35 -0600
Subject: logical statements (in Python and Numeric)
References: <3D11F1B2.21587F31@lanl.gov> <aet2j4$h8c$1@peabody.colorado.edu> <3D122C5D.C6985686@noaa.gov> <aetb9n$ksn$1@peabody.colorado.edu> <3D124BEF.9EC71977@noaa.gov>
Message-ID: <aetipc$nma$1@peabody.colorado.edu>

Chris Barker wrote:

> oops, that's what I get for posting untested code. You need parentheses,
> the above was trying to bitwise-and -2.0 and x
> 
> (x >= -2.) & (x < 0.)
> works.
> 

Ah, thanks. That actually makes the code almost like the OP's (moudulo the 
necessary parens). An it looks nicer than the logical_and() variation.

Somethin new learned, thanks a lot.

f.


From aahz at pythoncraft.com  Sat Jun 22 15:14:16 2002
From: aahz at pythoncraft.com (Aahz)
Date: 22 Jun 2002 15:14:16 -0400
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <3d149f19$0$53133$edfadb0f@dspool01.news.tele.dk> <af28vp$j57$1@panix1.panix.com> <3D14AD7F.3D5D243F@engcorp.com>
Message-ID: <af2ia8$a92$1@panix1.panix.com>

In article <3D14AD7F.3D5D243F at engcorp.com>,
Peter Hansen  <peter at engcorp.com> wrote:
>Aahz wrote:
>> In article <3d149f19$0$53133$edfadb0f at dspool01.news.tele.dk>,
>> Frihtiof Andreas Jensen <nomailhere at lookitup.yourself.com> wrote:
>>>
>>>My bet is that once you cool off and sits down with some debugging tools
>>>like GDB, Boundschecker or Electric Fence you will indeed find the
>>>programming error and fix it.
>>>
>>>My bet is also that you will not tell anyone either.
>> 
>> Ya know, you could have saved yourself a lot of typing by simply saying
>> PBCAK.  ;-)
>
>To save us from those who can't Google:  Problem Between Chair And Keyboard.
>(It was new to me, too.)

<grin>  Ya coulda ast me, too.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From mark at mceahern.com  Tue Jun 25 09:26:53 2002
From: mark at mceahern.com (Mark McEahern)
Date: Tue, 25 Jun 2002 08:26:53 -0500
Subject: import precedence
Message-ID: <NCBBLFCOHHDIKCAFGCFBCEHALGAA.mark@mceahern.com>

Suppose I have a package like this:

  spam/
    __init__.py
    setup.py
    foo.py

For cygwin, setup.py creates:

  foo.dll

For linux2, setup.py creates:

  foo.so

But for win32, setup.py is basically a no-op because I can do what I need to
do in foo.py using pythoncom.

Here's the question:

Can I just assume that:

  from spam import foo

will only import the foo.py module if the foo.dll and foo.so are not
found--that is, if the import statement is executed on win32 (where those
extensions won't exist)?

I looked at the documentation for import and it's not clear to me that it
addresses this issue.  Perhaps because noone in their right mind would
consider this approach?  ;-)

Thanks,

// mark

-





From nadavh at envision.co.il  Tue Jun 18 11:03:46 2002
From: nadavh at envision.co.il (Nadav Horesh)
Date: Tue, 18 Jun 2002 17:03:46 +0200
Subject: Distutils for python 2.2.1 under win2000: util.py patch
Message-ID: <3D0F4BD2.5060100@envision.co.il>

Hi!
I had problems installing Gnuplot1.5 package under python2.2.1.  I got 
an error
which was a result of calling distutils.util.convert_path with an empty 
string argument.
To correct the problem I added

    if len(pathname ) == 0:
        return pathname

as the first executable statement in that function. After that the 
Gnuplot installation completed succesfully.

The question:
 Is it a legal path to keep (did I corrected a bug or just generated a 
new one)?

  Nadav.





From huaiyu at gauss.almadan.ibm.com  Fri Jun 21 13:25:25 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Fri, 21 Jun 2002 17:25:25 +0000 (UTC)
Subject: How to represent the infinite ?
References: <mailman.1024636889.6135.python-list@python.org>
Message-ID: <slrnah6oc4.amp.huaiyu@gauss.almadan.ibm.com>

>[Christophe Delord]
>> ...
>> >>> 1e200**2
>> Traceback (most recent call last):
>>   File "<pyshell#1>", line 1, in ?
>>     1e200**2
>> OverflowError: (34, 'Numerical result out of range')
>> >>> 1e400
>> inf
>>
>> When the overflow occurs, the OverflowError exception is thrown.
>> inf can't be the result of a computation (without patch and
>> without catching exception).

The patch is in fact very simple: just remove the extra code whose sole
purpose is to throw an exception.   No patch to catch the exception is
needed.  

Tim Peters <tim.one at comcast.net> wrote:
>
>It can, but it's a x-platform crapshoot as to exactly how.  For example,
>under current Windows CVS, this still "works":
>
>>>> 1e300 * 1e300
>1.#INF
>>>>
>
>Here's a clumsier way <wink>:
>
>>>> 1.6e308 + 1.6e308
>1.#INF
>>>>
>
>Both of those assume the user hasn't installed and enabled the fpectl
>module, whose purpose in life is to cause even this to complain.
>
>maybe-number-crunchers-will-agree-on-something-after-i'm-dead-
>    but-i'm-not-holding-my-breath-ly y'rs  - tim

Now I'm confused.  I thought from previous discussions that you intended to
disable inf and nan until full ieee floating point compliance is implemented
in Python.  And I infered that that's in order to make Windows work.

Since it could be made to work on Linux, I simply commented out the checking
in pyport.h.  Voila, instant ieee compliance!

Your example appears to show that Windows can also live with nan and inf.
So my question is, are there any significant platforms out there that still
really need such checkings which make nan and inf almost useless?

If, in reality, only some very niche platforms need such checkings, maybe it
is better to let those platforms without ieee compliance suffer (in terms of
having to add a small patch), rather than depriving most of the mainstream
platforms of a very useful feature?

I'm looking forward to the day when I don't need to patch Python when a new
version comes out.

Huaiyu



From john at johnnypops.demon.co.uk  Mon Jun 17 19:30:57 2002
From: john at johnnypops.demon.co.uk (John Popplewell)
Date: 17 Jun 2002 16:30:57 -0700
Subject: tray icon using tkinter
References: <mailman.1023950718.20009.python-list@python.org>
Message-ID: <e08b1d4.0206171530.563e67fd@posting.google.com>

Anoop P B <anoop79 at myrealbox.com> wrote in message news:<mailman.1023950718.20009.python-list at python.org>...
> hi,
> 
> 	is there a way that i can provide a tray icon ie. next to the clock (on 
> win32) for my program (using tkinter) ? i know that there is a way of 
> doing this in wxpython but i need to know whether its possible in tkinter.
> 
> thanks,
> anoop


Have you had a look at "Tray It!" ?

http://www.teamcti.com/trayit/index.html

cheers,
John.


From loewis at informatik.hu-berlin.de  Thu Jun  6 04:37:42 2002
From: loewis at informatik.hu-berlin.de (Martin v. =?iso-8859-1?q?L=F6wis?=)
Date: 06 Jun 2002 10:37:42 +0200
Subject: libpython as a shared library?
References: <20020604143402.GB22993@hal.mediasupervision.de> <Pine.OS2.4.32.0206051931120.1678-100000@tenring.andymac.org> <mailman.1023348398.27744.python-list@python.org>
Message-ID: <j4u1og6bvt.fsf@informatik.hu-berlin.de>

Gregor Hoffleit <gregor at hoffleit.de> writes:

> If I compile the sources with -fPIC, but put them in a static library,
> there's nearly no speed difference to non-PIC code in a static library.
> My best guess is that the speed penalty is caused by the indirection
> step that's necessary when calling routines from a shared lib; but then,
> that's no real explanation for a loss of 30% speed!

Can you elaborate a bit what you measured to get that speed difference?

If it involves startup time, the time for locating libpython surely
involves part of startup time. Also, it might be necessary to perform
many relocations to libpython at run time that might not be necessary
for the application.

Regards,
Martin



From emile at fcfw.fenx.com  Mon Jun 17 16:30:17 2002
From: emile at fcfw.fenx.com (Emile van Sebille)
Date: 17 Jun 2002 15:30:17 -0500
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Jun 17)
Message-ID: <52F61289899FD9D9.327A292D46CD5CB7.E0BCFB87CA7B8ABD@lp.airnews.net>

QOTW:  "People will inevitably associate me with my father, but I would
not have anyone believe that I am trading on the name Edison. I would
rather have you know me merely as the result of one of my father's
earlier experiments." - Charles Edison (son of the inventor), during
his campaign for governor of New Jersey in 1940
    http://theliterarylink.com/quotes.html

"Sprinting is intense and exhilarating. Everybody bounces questions and
ideas off of each other while sitting in front of real, evolving code.
You learn about coding practices such as unit tests and factoring
interface descriptions, as well as real Zope 3 component architecture
code. But far more interesting is the human interaction." -- Paul Everitt,
quoted from http://europython.zope.nl/interviews/entries/paul_everitt
  

    Happy Father's Day  -  Guido's First!  Watch for photos at:
       http://www.python.org/~guido/orlijn/#latest
    
    Collin Monahan asks about the nature of lists:
       http://groups.google.com/groups?threadm=mailman.1023833476.10249.python-list at python.org
    
    Fabio Corneti wants a way to create final methods, and Jeff Epler
    shows one way to do it:
       http://groups.google.com/groups?threadm=mailman.1023812475.3715.python-list at python.org
    
    Edward K Ream announces leo, a programmer's editor and a flexible
    browser for projects, programs, classes or data. Leo clarifies
    design, coding, debugging, testing and maintenance.
       http://groups.google.com/groups?selm=mailman.1023895323.9169.clpa-moderators at python.org
    
    Jerzy Karczmarczuk provides a different answer to the question of
    creating executables from python source by pointing us to an interview
    of Guido:
        http://groups.google.com/groups?selm=3CFDF148.AA2035FF at info.unicaen.fr

    Prabhu Ramachandran announces the availability of the MayaVi Data
    Visualizer version 1.2.  MayaVi is a free, easy to use, scientific
    data visualizer, written in Python and using VTK, the Visualization
    Toolkit.  One significant change is a license change from GPL/LGPL
    to BSD.
       http://groups.google.com/groups?selm=mailman.1023914707.18194.python-list at python.org
    
    "You knew the job was dangerous when you took it, Fred."
    James Rowe asks about documentation on __subclasses__...
       http://groups.google.com/groups?selm=a9bcagho74.22t.jimrowe at optilink.com
    ...a question still open in the in-progress Thinking in Python book...
       http://www.mindview.net/Books/TIPython/BackTalk/A_251
    ...but I did find some documentation (thanks google), including
    another new-class built in method, mro (method resolution order) at:
       http://cvs.astro-wise.org:4711/__builtin__.html#type
    
    Neil Schemenauer releases Version 0.1 of the Sancho unit testing
    framework, the unit test module in use at the MEMS Exchange:
       http://groups.google.com/groups?selm=mailman.1024086967.9147.clpa-moderators at python.org
    
    Anoop is looking for a way to create a systray icon on win32 using
    tkinter.  Interesting question -- any answers?
       http://groups.google.com/groups?selm=mailman.1023950718.20009.python-list at python.org
    
    Greg Ewing is working to speed things up with pyrex release 0.3. 
    Pyrex is a new language for writing Python extension modules.
        http://groups.google.com/groups?selm=mailman.1023502562.5896.clpa-moderators at python.org

    Jean-Sebastien Bolduc wants to understand about import * and namespaces...
       http://groups.google.com/groups?selm=56e1eff0.0206131437.4ca8ed9e at posting.google.com
    ...prompting Tim Delaney to point to section 1.1 at
       http://py-howto.sourceforge.net/doanddont/doanddont.html
    
    Aahz reminds us that a failed import doesn't get better with repeated
    imports...
       http://groups.google.com/groups?selm=aeat13$pli$1 at panix1.panix.com
    ... and questions a class/type unification bu^h^hfea^h^h^h... well, a
    'gotcha:
       http://groups.google.com/groups?threadm=aeau48$111$1 at panix1.panix.com
    
    Neal Josephson announces that Version 2.1 of the Hap python debugger
    and IDE has been made available for download as source and binary zip
    files at:
       http://sourceforge.net/projects/hapdebugger
    
    PythonCard, a GUI construction kit for building cross-platform desktop
    applications on Windows, Mac OS X, and Linux, moves to version 0.6.7.
    New additions include chat, webserver, pictureViewer, slideshow, and
    webgrabber.  See Kevin Atlis' announcement at:
       http://groups.google.com/groups?selm=gyoO8.45$YM3.33091 at news.uswest.net
    
    Wow.  Check out Dave's (age 12) request for help.  Note particularly
    the responses given... and more particularly the responses _not_ given.
    It's good to here...
       http://groups.google.com/groups?threadm=Xns922D5FE7B72D8duncanrcpcouk%40127.0.0.1
    ...and Duncan Booth provides a link to the LiveWires Python course,
    intended to teach the Python programming language to people who have
    never programmed before:
       http://www.livewires.org.uk/python/index.html
    

========================================================================

Everything you want is probably one or two clicks away in these pages:

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
	 http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Michael Hudson continued Andrew Kuchling's marvelous tradition
    of summarizing action on the python-dev mailing list once every
    other week, into July 2001.  Any volunteers to re-start this
    valuable series?
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

    Much of Python's real work takes place on Special-Interest Group
    mailing lists
        http://www.python.org/sigs/

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
	http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
	http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
	http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

    Links2Go is a new semi-automated link collection; it's impressive
    what AI can generate
        http://www.links2go.com/search?search=python

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning
(approximately), ask <claird at neosoft.com> to subscribe.  Mention
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

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


From dougz at cs.washington.edu  Wed Jun 19 22:57:01 2002
From: dougz at cs.washington.edu (Douglas Zongker)
Date: Thu, 20 Jun 2002 02:57:01 GMT
Subject: using os.chmod in windows
References: <3D110C99.5090607@portzer0.virtualave.net>
Message-ID: <GxzHJ1.24Ez@beaver.cs.washington.edu>

Locke <charter-news at portzer0.virtualave.net> wrote:
: I need to change a file to be 'hidden' in windows. [...] So how 
: do I do this?

If you have the win32 extensions that come with ActivePython
[aspn.activestate.com], then you can do:

    import win32api, win32con
    
    def hide_file( fn ):
        x = win32api.GetFileAttributes( fn )
        x |= win32con.FILE_ATTRIBUTE_HIDDEN
        win32api.SetFileAttributes( fn, x )
    
    def unhide_file( fn ):
        x = win32api.GetFileAttributes( fn )
        x &= ~win32con.FILE_ATTRIBUTE_HIDDEN
        win32api.SetFileAttributes( fn, x )

dz


From kragen at pobox.com  Mon Jun 24 18:00:55 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 24 Jun 2002 18:00:55 -0400
Subject: Generic Python
References: <3D16E465.F0D2509E@ifib.uni-karlsruhe.de>
Message-ID: <83ptygl4ko.fsf@panacea.canonical.org>

Uwe Mayer <Uwe.Mayer at ifib.uni-karlsruhe.de> writes:
> The problem is that I've got a base class and there should be many
> subclasses of it. Each subclass just overwriting an output method and
> serving as a template: 
> ...
> Is there another way to solve this problem?

Several people have offered ways to do this with metaclasses or
run-time code generation with exec, but I agree with Michael Chermside
that you should use a factory function or just override the output
method on individual instances instead.


From drifty at bigfoot.com  Mon Jun  3 17:54:10 2002
From: drifty at bigfoot.com (Brett C.)
Date: 3 Jun 2002 14:54:10 -0700
Subject: ANN: Pure Python strptime()
Message-ID: <f85b5ab1.0206031354.132f3148@posting.google.com>

Out of frustration of not always having strptime() available and not
liking something in the standard library saying something is only
available on "Most modern Unix systems" when it does not have to be
this way, I wrote strptime() in pure Python (requires Python 2.2 or
higher; if there is enough demand I could back-port to older versions
since most of the 2.2 capabilities used were for convenience of
performance issues).

The module can be found at the Python Cookbook at
http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/56036 .

I wrote strptime() before (over a year ago), but I got around to
rewriting it finally and removing any dependence on the programmer
knowing any locale-specific info ahead of time.  Now you do not need
to enter any locale-specific information; you can use the function
just like the version that is in the time module.  There is some extra
functionality, though, so look at the recipe to see what extras are
available (performance and locale abilities).

I personally would like to see this included in the standard library
since it is a nice function to have.  How should I go about this? 
Since the time module is a C extension, I realize that this can't just
be pasted into it.  Does this mean that I am going to have to write a
PEP to request that this be added as a module in the standard library?
 Do I just need to convince Guido to include it?  Do I need to push to
change the time module into a Python module with the current module
renamed _time and then just have a stub module to globally import the
_time module?

I have posted this as a patch at SF (patch #474274), but I am
wondering if I have to do anything that I just previously mentioned on
top of this.  Any guidance on this would be helpful.  Thanks.

-Brett C.


From tlirobot at tlirobotics.cc  Wed Jun 12 19:14:37 2002
From: tlirobot at tlirobotics.cc (Dylan Appleman)
Date: Wed, 12 Jun 2002 23:14:37 GMT
Subject: Best Hosting Package - 22810
References: <sGIN8.301359$t8_.51255@news01.bloor.is.net.cable.rogers.com>
Message-ID: <xBQN8.43543$zh2.13072602@twister.neo.rr.com>

HOSTING @ BSolution UNIX Servers
  "Best Solution" <bsolution at tut.by> wrote in message news:sGIN8.301359$t8_.51255 at news01.bloor.is.net.cable.rogers.com...
          
       
       
       
        HomePage Programming Web-Design Networking Careers About Us 
       
       



        Hosting Sign Up Form



        Sign Up Now!   
              HOSTING PRICES  
             
              Service Package Base Business Professional 
              Monthly $8 $15 $30 
              Annual $75 ($6.25/mo) $130 ($10.83/mo) $250 ($20.83/mo) 
              Setup Fee no no $20  
             
              Disk Space 30 MB 100 MB 500 MB 
              Traffic unlimited unlimited unlimited 
              Support ASP 3.0 / ASP .NET no no yes 
              PERL (CGI) yes yes yes 
              PHP 4.x Support yes yes yes 
              MS Access DataBases no no yes 
              MySQL 2000 yes yes yes 
              E-mail boxes 10 50 200 
              SSI Support yes yes yes 
              FTP yes yes yes 
              Statistics and .log access yes yes yes 
              SSL Support no yes yes 
              Password directories no yes yes 
             
              E-MAIL PRICES  
             
              E-Mail Setup Fee free 
              Monthly payment no 
             
              OTHER SERVICES 
             
              Domain name registration (.com, .net, .org, cc, .info, .biz) $20 
              Domain name: YOURNAME.bsolution.net free 
              On-Line Web Store Support $15/mo 
       
       
       
       

SPAMMER!!!!!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020612/d15137d6/attachment.html>

From huaiyu at gauss.almadan.ibm.com  Tue Jun  4 21:06:42 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 5 Jun 2002 01:06:42 +0000 (UTC)
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <3CFCF3E7.AE891EBE@noaa.gov> <adiumj$22g2$1@nntp6.u.washington.edu> <adjduf$92g$0@216.39.172.122>
Message-ID: <slrnafqp12.4ea.huaiyu@gauss.almadan.ibm.com>

[Huaiyu Zhu]
> This is not unlike the situation of the
> division operator with distinct behaviors on floats and ints.

[Chris Barker]
I suppose that is a pretty good parallel, but not quite the same, as the
behavior of += is much more similar for the different types ...

[Huaiyu Zhu]
In fact, the behavior of += is even more different in different situations.
Integer division sometimes equals real division, but modification in an
object is never the same as modification in namespace.  In-place
modification affects all names referencing the same object, as the following
example shows

[Steve Holden]
>>> a = b = (1, 2, 3)
>>> a += ('boo!',)
>>> a, b
((1, 2, 3, 'boo!'), (1, 2, 3))
>>> a = b = [1, 2, 3]
>>> a += ['boo!']
>>> a, b
([1, 2, 3, 'boo!'], [1, 2, 3, 'boo!'])

[Donn Cave]
Well, it is consistent.  Maybe not in a way that you can see
and still remain sane.  "Mutable" or not, it must always work
like "a = iadd(a, b)", where iadd attempts to perform some
computation like __iadd__() or eventually "+".  

[Huaiyu Zhu]
That "must" is only an artifact of trying to incorporate rebinding increment
for the sake of immutables.  It is similar to that / "must" return an integer
for integer oprerands.

[Huaiyu Zhu]
It is interesting to spell out the analogy with integer division:

- Real division is not well defined for integers (if result must be integer).
- Integer division is a useful approximation that works for intergers.
- Integer division is also meaningful for reals, but it is different from
  real division.
- Python uses the same symbol / for both operations depending on operarands,
  causing quite some inconsistency and confusion.
- The solution is to use / for real division only, while introducing another
  operator // for integer division.

- In-place increment is not well defined for immutables.
- Rebinding increment is a useful approximation that works for immutables.
- Rebinding increment is also meaningful for mutables, but it is different from
  in-place increment.
- Python uses the same symbol += for both operations depending on operands,
  causing quite some inconsistency and confusion.
- The solution might be to use += for inplace increament only, while
  introducing another operator (?) for rebinding increment.

Chris Barker gave details of several kinds of inconsistencies.

[Donn Cave]
Not really - the current semantics are actually better than the fixed
version, inasmuch as they are at least consistent.  As I interpret the
proposal, the fixed += would sometimes work without rebinding (if the
object supports it) and sometimes would be an assignment (falling back
to +.)  To me, to allow an object to dictate its relationship to its
computational environment like that is anathema.

[Huaiyu Zhu]
That is one proposal.  My proposal is to disallow += for immutables.
Another symbol would be used for rebinding operations.  (Or maybe swap the
role of these symbols.)  Thus for mutables both operations would be
available.  No, 'a=a+b' is not enough for rebinding, as the following shows:

[Bengt Richter]
 >>> foo()[0]+=1
 side effect
 >>> foo()[0] = foo()[0]+1
 side effect
 side effect


Huaiyu


From johnroth at ameritech.net  Mon Jun  3 18:32:34 2002
From: johnroth at ameritech.net (John Roth)
Date: Mon, 3 Jun 2002 18:32:34 -0400
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <06da01c208f4$d69003c0$6601a8c0@boostconsulting.com> <200206010121.g511LxX19223@pcp742651pcs.reston01.va.comcast.net> <073501c20960$23f367e0$6601a8c0@boostconsulting.com> <200206011334.g51DY1D21669@pcp742651pcs.reston01.va.comcast.net> <07d801c20970$cdebe050$6601a8c0@boostconsulting.com> <mailman.1022940122.3421.python-list@python.org> <Id5K8.171932$Po6.340972@rwcrnsc52.ops.asp.att.net> <oWIK8.179611$%u2.31494@atlpnn01.usenetserver.com>
Message-ID: <ufnrjqb4h0ub12@news.supernews.com>

"Steve Holden" <sholden at holdenweb.com> wrote in message
news:oWIK8.179611$%u2.31494 at atlpnn01.usenetserver.com...
> "Emile van Sebille" <emile at fenx.com> wrote ...
> > <jepler at unpythonic.net> wrote:
> > > IMO the following is an odder behavior:
> > > >>> t
> > > ([1], [2], [3])
> > > >>> t[1] += [3]
> > > Traceback (most recent call last):
> > >   File "<stdin>", line 1, in ?
> > > TypeError: object doesn't support item assignment
> > > >>> t
> > > ([1], [2, 3], [3])

What I believe is going on here is that the list t[1]
is being updated since it's mutable, and then the
attempt to insert the new version of the list into the
tuple is being rejected - hence the error.

The fact that it shows the change is most likely due
to the list not having moved in memory. I hope that
the reference count was handled properly!

AFIC, this is clearly wrong - operators should
be atomic - no side effects. Either they work, or
they don't.

> >
> > Here's another that's differently wrong:
> > >>> a = ('aaa',)
> > >>> a[0]+='bbb'
> > Traceback (most recent call last):
> >   File "<stdin>", line 1, in ?
> > TypeError: object doesn't support item assignment
> > >>> a
> > ('aaa',)
> > >>>

This is the exact same case, except that instead of
a list, the tuple contains a string - another immutable type.
Therefore, it fails on attempting to update the string,
rather than attempting to update the tuple.

Maybe. It might actually update the string, producing
'aaabbb' at a different location in memory,
and then (properly) fail to update the tuple.

John Roth





From gbien at intekom.co.za  Tue Jun 11 14:14:53 2002
From: gbien at intekom.co.za (Geoff Bien)
Date: Tue, 11 Jun 2002 20:14:53 +0200
Subject: unknown opcode Exception then AV
References: <ae546s$bg5$1@ctb-nnrp2.saix.net> <lkk7p5yh2w.fsf@pc150.maths.bris.ac.uk>
Message-ID: <ae5eov$m8i$1@ctb-nnrp1.saix.net>

Found the problem. There is a known bug
in version 2.1 ?? when "continue" is used
within a try...except block




From db3l at fitlinxx.com  Mon Jun 10 15:35:20 2002
From: db3l at fitlinxx.com (David Bolen)
Date: 10 Jun 2002 15:35:20 -0400
Subject: Python 2 Monolithic Executable
References: <mailman.1023724574.1452.python-list@python.org>
Message-ID: <u1ybfyljb.fsf@ctwd0143.fitlinxx.com>

Anoop P B <anoop79 at myrealbox.com> writes:

> 	I need to write a program which will take user input and mail it to a 
> predesignated email address. The constraint is that it needs to be a 
> single, small executable (without requiring installation and no 
> dependencies) (Windows). Is there a way to create such an executable 
> from a python program?

Gordan McMillan's installer package [1] supports creating a single
executable.  It's actually an archive in the form of an executable,
and when it runs, it unpacks temporary copies of external DLLs or
files that it needs, and the files are removed when the application
exits.  The end user need not know about this - so to them it can look
just like a single executable to run.

You're probably going to have a tough time with "small" though
depending on your definition.  A minimum application is likely to be
around 450-500KB, since it has to include at least the Python DLL and
whatever libraries you use.

[1] http://www.mcmillan-inc.com/install1.html

--
-- David
-- 
/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/


From calves at coelce.com.br  Wed Jun 19 07:30:14 2002
From: calves at coelce.com.br (Alves, Carlos Alberto - Coelce)
Date: Wed, 19 Jun 2002 08:30:14 -0300
Subject: =?ISO-8859-1?Q?ENC=3A_how_i_use_=27=E3=27_in_python=3F=3F=3F?=
Message-ID: <29A97D00F387D411AC7900902770E148058B2E65@lcoeexc01.coelce.net>

Python accept that, just represent it as hex. Look below:
>>> s='?'
>>> s
'\xe2'
>>> print s
?
Another work around, is to use the chr() function, just like that:
chr(226)=>?

> -----Mensagem original-----
> De: jubafre at zipmail.com.br [mailto:jubafre at zipmail.com.br] 
> Enviada em: quarta-feira, 19 de junho de 2002 01:39
> Para: python-list at python.org
> Assunto: how i use '?' in python???
> 
> 
> i?m brazilian, and in my laguage(Portguese) there are many 
> other caracters
> how ?, ?, ? and a string in python doesn?t support this
> how i can use??? have a import module for this???
> 
> BRASILEIROS DA LISTA COMO VCS CONSEGUEM USAR ACENTO NO PYTHON??
> 
> for exmple:
> s='?'
> doesn?t work?why??
> 
> 
> Juliano Freitas
> www.gebrasil.hpg.com.br
> 
> 
> 
> ------------------------------------------
> Use o melhor sistema de busca da Internet
> Radar UOL - http://www.radaruol.com.br
> 
> 
> 
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020619/c7fc5cda/attachment.html>

From threeseas at earthlink.net  Tue Jun 25 20:40:47 2002
From: threeseas at earthlink.net (Timothy Rue)
Date: Wed, 26 Jun 2002 00:40:47 GMT
Subject: Reminder of a python Project.
Message-ID: <2218.941T781T12224051threeseas@earthlink.net>

First off, it's ok that I got such a high level of resistance to my
initial posting here about this project. Thanks to some research that /.
had an article on, regarding the size of GPL project teams.

But anyway.... Remember the autocoding project I mentioned?


60 billion in cost to consumer/small business (shoddy software failure)
http://www.computerworld.com/managementtopics/management/itspending/story/0,10801,72245,00.html
(slashdot has an article today that links to this - and the comments...)

SCIENTIFIC AMERICAN - Sept. 1994
Article - "Software's Chronic Crisis"
Mass-produced PC products makes up less than 10% of the $92.8 billion
U.S. software market.


COMDEX SPRING and WINDOWS WORLD 95
Power Panel - "What's Wrong with Software Development"
** In The U.S. Only **
$81 Billion = 31% of software development gets cancelled before complete
$59 Billion = 53% of software development has cost over-runs of 189%
16% success - project success and failure ratio
61% customer requested features and functions make it in
Maintenance and repair is where most of the U.S. dollars are going,
instead of new, better, easier to use software.

That's 140 Billion before the cost of successful project is added in. And
a total cost increase over a 1 year period of: 47.2 + billion (I suppose
this can be cross checked with stock market or such.)

The Autocoding solution direction:
http://groups.google.com/groups?q=Autocoding&hl=en&lr=&ie=UTF-8&safe=off&selm=5512.785T1955T13695089threeseas%40earthlink.net&rnum=1

My 1997 estimate of software failures all around (total of industry and
consumer level development and use):  Low estimate of 300 Billion.
http://groups.google.com/groups?selm=2084.7123T966T2172%40mindspring.com&output=gplain

at 60 billion figured at 20% of total market ... = 300 billion.
Of course 10% means 600 billion.


---
*3 S.E.A.S - Virtual Interaction Configuration (VIC) - VISION OF VISIONS!*
   *~ ~ ~      Advancing How we Perceive and Use the Tool of Computers!*
Timothy Rue      What's *DONE* in all we do?  *AI PK OI IP OP SF IQ ID KE*
Email @ mailto:timrue at mindspring.com      >INPUT->(Processing)->OUTPUT>v
Web @ http://www.mindspring.com/~timrue/  ^<--------<----9----<--------<



From nhodgson at bigpond.net.au  Wed Jun 19 09:17:19 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Wed, 19 Jun 2002 13:17:19 GMT
Subject: imp.loadmodule creates compiled file
Message-ID: <zv%P8.332187$o66.856939@news-server.bigpond.net.au>

   I've been using imp.loadmodule with Python 2.2 (on both Windows and
Linux) to load changed versions of source code into a long running process
where they will run alongside objects already created from earlier versions
of the source code. An unexpected result of the call is to create a compiled
module on disk with a name based on the filename argument with "c" appended.

>>> f = file("x.py")
>>> import imp
>>> imp.load_module("y", f, "z", (".py", "r", 1))
<module 'y' from 'z'>

   At this point, a "zc" file is created. I would prefer no compiled file be
created. The documentation indicates that file and filename arguments can be
None and '' when the module is not being loaded from a file but this makes
litle sense as there is no other argument to provide the module contents
from other than a file. Trying '' as the filename argument leads to a
compiled file called "c". The compiled files can be unlinked but it would be
better to have a solution that avoided writing to disk at all as there may
be problems such as name clashes with other running instances or with being
in a protected directory.

   Any other techniques for creating mdoules from source code?

   Neil




From sonnenbu at informatik.uni-bonn.de  Sat Jun  8 13:16:02 2002
From: sonnenbu at informatik.uni-bonn.de (Frank Sonnenburg)
Date: Sat, 8 Jun 2002 19:16:02 +0200
Subject: Q: scope of module functions
Message-ID: <adte4c$nu8$1@f1node01.rhrz.uni-bonn.de>

Hello all,

a little tricky problem (at least from my point of view); given module
CALLER:

# i need "fct" and some other from module "UTILS"
from UTILS import *
fct()
# END of module CALLER


Now, inside fct(), how do i determine the module - in this case CALLER -
from where fct() was called?

Even concerning cases like (continuation of example):

# Now we are in __main__
import CALLER
fct_in_main = CALLER.fct
fct_in_main()


Appreciation for any help

    Frank Sonnenburg, BioSolveIT GmbH, Sankt Augustin, Germany




From eric.brunel at pragmadev.com  Wed Jun 19 11:25:31 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Wed, 19 Jun 2002 15:25:31 +0000
Subject: tkinter menu (newbie question)
References: <Xns9232778FCD335stefninnospamyahoofr@213.228.0.138> <aepps0$al$1@wanadoo.fr> <Xns9232967929C7Fstefninnospamyahoofr@213.228.0.138>
Message-ID: <aeq0cv$p5g$1@wanadoo.fr>

Stephane Ninin wrote:
> Thanks for the help.

No problem.

> If I want to use several of these windows,
> can still use for base class Tk, instead of Frame ?

Nope: the instance of Tk is your main window. There should only be one, and 
closing it will end your application. If you want to create other 
windows, you should create instances of Toplevel.

If you do not have a main window, there are means to work around this 
limitation, but they're not really natural. If you need to know how to do 
it, feel free to ask...

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From jason at tishler.net  Fri Jun 21 22:11:34 2002
From: jason at tishler.net (Jason Tishler)
Date: Fri, 21 Jun 2002 22:11:34 -0400
Subject: Python+Cygwin+XML SAX build failure of PyXML 0.7.1
In-Reply-To: <59d8a5f6.0206210804.25353aa3@posting.google.com>
References: <59d8a5f6.0206210804.25353aa3@posting.google.com>
Message-ID: <20020622021134.GA1792@tishler.net>

Dmitri,

On Fri, Jun 21, 2002 at 09:04:10AM -0700, Dmitri wrote:
> [snip]
> Searching the net:
> 1. I did not find the precompiled version of the PyXML for cygwin.
>    Does someone know where can I get it, it it does exist?

Sorry I don't, but see below...

> 2. The problem like "initializer element is not constant" is shown but
> [snip]

See the following for the solution to the above:

    http://cygwin.com/ml/cygwin/2002-06/msg00273.html

Would you like to contribute a pre-built Cygwin PyXML to the Cygwin
distribution?

Jason




From alf at logilab.fr  Tue Jun 11 08:20:00 2002
From: alf at logilab.fr (Alexandre Fayolle)
Date: Tue, 11 Jun 2002 12:20:00 +0000 (UTC)
Subject: [ANN] constraint-0.2
References: <mailman.1023468735.788.python-list@python.org> <slrnag50sd.gac.mlh@vier.idi.ntnu.no> <slrnag8to5.dr.alf@virgo.logilab.fr> <dl4N8.181687$Kp.17633467@bin7.nnrp.aus1.giganews.com>
Message-ID: <slrnagbqnf.ao.alf@virgo.logilab.fr>

Terry Reedy a ?crit :
> To paraphrase Magnus Lie Hetland, the general constraint-propagation
> system constraint-0.2 takes 100 times as long to do the 8-queens
> problem as the specific backtracking queens.py found in the standard
> Python distribution: 15 versus .1 sec on comparable machines.

Yup. But we do mean to make it faster.

> The is about the same comparison, for some computations, as between
> generic/dynamic Python and type-specific C.  In both cases, there are
> two more questions of practical interest:  which platform/method is
> faster for the *programmer*?  and, is the slow method fast enough?

This is a good point.
Using the constraint package for 8 queens is very easy, since all you
have to do is describe the problem in terms of variable and constraints,
and call Solver().solve(Repository(variables,domains,constraints))

Is it fast enough? Well, it really depends. The good point in the
situation at hand is that I have not yet started thinking about
optimizing the code. There are other things to deal with first (adding
other kind of domains, such as finite sets, and the basic constraints
that go with these domains), so my personal answer is that it is for 
now fast enough as far as I'm concerned. Which does not mean it's ready
for production use. Hey, this is version 0.2!

> So, what is the relative programming speed of the general constraint
> propagation system versus specific backtracking?  I suspect that some
> people could whip up something like queens.py as fast or faster that
> they could the (unpublished?) constraint input for the same problem.

O(N) algorithm available for N queens has been published and proven to
be of minimal complexity, and will always be faster than a constraint
propagation approach which needs to consider N*(N-1) constraints, and
therefore will always have a complexity > O(N^2)

> However, I also suspect the reverse would be true for the more
> realistic and useful scheduling example that Fayolle used to
> illustrate and motivate constraint-0.2.
> http://www.logilab.org/python-logic/documentation.html

N-Queens is a well known problem, and it can be used as a benchmark for
solvers. Same for SEND+MORE=MONEY, which is also in the examples
directory of the constraint module. Since these problems are both well
known and simple in their expression, they can give an idea of the
relative performance of solvers. 

The scheduling example on the web page is completely made up, and its
purpose is to give a more "real-life" flavour to the tutorial. There are
many other constraint satisfaction problems that can be solved with the
module, or the the constraint module will be able to solve in the near
future (how to select tunes from a CD to best fill up a tape, how to
arrange your furniture in the flat you're going to move in, finding
a new password that you'll be able to easily memorize while
passing through the devious filters set up by your sysadmin, choosing
which fruits and vegetables to buy in the market in order not to eat the
same same thing every week...)

> Real world problems may have no solution as given.  On the other hand,
> constraints may not be absolute.  In the scheduling example, two
> workshops *can* be scheduled simultaneously, if necessary, even though
> many people want to take both.  So a stepwise procedure may be needed.
> (Given multiple solutions, one might also want to add 'wishes' as
> pseudoconstraints to help select just one.)  So, is it equally easy,
> with the two approaches, to add or drop constraints and rerun?  This
> would be fairly simple with c-0.2 input but probably harder for
> backtracking code unless planned for from the beginning.  Is it
> equally easy to give constraints priorities and have the system
> automatically either add them in order from the highest or drop them
> from lowest?  Constraint-0.2 evaluates constraints in the order
> listed, but I do not know if  it properly backtracks from failure to
> report the most satisfying solutions.

The short answer is no, because solving overconstrained problems is a
tricky thing. 

The longer answer is that you could quite easily build on top of
the constraint package an engine which would try to relax constraints
when no solution can be found. When finite set domains are implemented
in the solver, building this engine will be possible with the constraint
package itself. But this approach will not be as efficient as a
dedicated overconstrained package solver. 

Now, a little annecdote: when writing the scheduling example, the
first set of constraints I had come up with gave 1152 possible solution,
which I found was too many. So I did exactly what you suggested, I added
more constraints (this was really easy because adding a consrtaint is a
one-liner), twiddled with the choice of the non-simultaneous
conferences, until I found a combination which gave "only" 64 possible
solutions. It took me about 20 runs to get this, which was fairly quick
because the conference.py script runs in about 2 seconds on my machine. 


Alexandre Fayolle
-- 
LOGILAB, Paris (France).
http://www.logilab.com   http://www.logilab.fr  http://www.logilab.org
Narval, the first software agent available as free software (GPL).


From marklists at mceahern.com  Fri Jun 28 08:33:31 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Fri, 28 Jun 2002 07:33:31 -0500
Subject: Is Python growing?
In-Reply-To: <slrnahnif7.1c6.matt@happy-hour.mondoinfo.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBIEPELGAA.marklists@mceahern.com>

[Matthew Dixon Cowles]
[...]
> Python is wonderfully easy to get started with. But someone who's new
> to programming doesn't need to learn everything about Python in order
> for it to be useful and fun. (That sounds obvious but it's less true
> in many other languages.) So if Guido adds a feature here and there,
> that doesn't necessarily make the learning curve much worse. I've been
> using Python for some years now and there are corners of the standard
> library that I haven't explored because I haven't needed to. Indeed,
> there are aspects of the core language that I haven't found out much
> about because I haven't needed them.
[...]

This is true for me as well.  For instance, I typically don't use map,
filter, zip, et al.  Paul Rubin replied to your post with an example that
used max and map to return the longest string in a list.  What's so
beautiful about Python is that all I had to do was fire up the interactive
interpreter and do this to remind myself of what map does:

  >>> print map.__doc__

So not only is the language incredibly expressive, clear, and pragmatic, but
it provides tools for exploration that make it easy to focus on what works.
Those tools allow you to expand your working subset of the language slowly
over time as needed.

// m

-





From ark at research.att.com  Fri Jun 14 12:50:01 2002
From: ark at research.att.com (Andrew Koenig)
Date: Fri, 14 Jun 2002 16:50:01 GMT
Subject: I feel stoopid ... this code is to verbose
References: <3D0A1791.3060307@mxm.dk>
Message-ID: <yu99ofedkdom.fsf@europa.research.att.com>

Max> I just cannot seem to find the nice 3-liner I expected from the
Max> beginning. Has anybody got a better solution ? I thought somebody
Max> might find it a fun exercise. Well I have...

Well, it's a little more than three lines, but how about this?

def stringSlicer(string, chunkSize = 3):
    suffix = ''
    while len(string) > chunkSize:
        suffix = '.' + string[-chunkSize:] + suffix
        string = string[:-chunkSize]
    return string + suffix

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From kragen at pobox.com  Thu Jun  6 12:58:59 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 06 Jun 2002 12:58:59 -0400
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet> <3CFDF148.AA2035FF@info.unicaen.fr> <3CFE49CA.46A5FEE1@tds.net>
Message-ID: <838z5sgx7w.fsf@panacea.canonical.org>

"Edward K. Ream" <edream at tds.net> writes:
> BTW, a compiled version of a Python program would be valuable
> commercially, as it would allow companies to protect their source code,
> as well as (possibly) improve performance.

I'm worried about this myself; perhaps the compiler could include the
source code in the resulting executable to prevent this.



From phd at phd.pp.ru  Thu Jun 27 12:50:42 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 27 Jun 2002 20:50:42 +0400
Subject: Python crashed on me
Message-ID: <20020627205042.A18986@phd.pp.ru>

Hello!

   Recently Python (one program that I am dbugging) started to crash.
FreeBSD kills it with "Bus error", Linux with "Segmentation fault".

   I think the program crashed in the cPickle.dump(file, 1), because after
the crash the file is corrupted and cannot be load back (upon loading
cPickle complain EOFile). The error is reproduceable.

   How can I debug the situation? May be my data grew too big? The pickle
file is about 800K.

   Details. The program is my URL robot
(http://phd.pp.ru/Software/Python/#bookmarks_db). The robot runs
urllib.urlretrive() on my bookmarks file and report the results. Between
runs data may be saved in different formats, including pickle.
   While trying to understand what is going on and make full run faster I
replaced the call to urlretrive with fake function that returns nothing.
The robot crashed with the same segfault.

   Python is

Python 2.2.1 (#2, Apr 10 2002, 15:15:56)
[GCC 2.95.2 20000220 (Debian GNU/Linux)] on linux2

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From kragen at pobox.com  Sat Jun  1 02:49:16 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 01 Jun 2002 02:49:16 -0400
Subject: Iterating through two lists
References: <3cee0ae6_1@news3.newsgroups.com> <mailman.1022234742.24014.python-list@python.org> <3cee2338_2@news3.newsgroups.com> <mailman.1022241402.21254.python-list@python.org> <1022297508.675568@yasure>
Message-ID: <831ybrwl1v.fsf@panacea.canonical.org>

"Donn Cave" <donn at drizzle.com> writes:
> Paul Prescod was right (as paraphrased on the followup page linked
> in the page above) - he really did choose an example that favored
> Lisp.

Sort of.

> The function that carries data is cool enough.  Lisp does it easier
> than Python, and in fact does "functional programming" better than
> Python by a long shot.

That's debatable.

> But then, Python does OOP better than Lisp!

Your certainty suggests deep ignorance.  How familiar are you with
CLOS?

> He seems vaguely mystified that Python programmers would favor the
> much more tedious class solution, but of course if you wanted to add
> some requirement like support two functions over that data instead
> of just one, then the mystery evaporates.

Like
(defun addergetter (x) (list (lambda (i) (incf x i)) (lambda () x)))
?

> When you take on a comparison between two decent languages, though,
> you have a lot of impossible questions about which big picture concepts
> are really going to pay off.  Does OOP live up to its promise on
> re-use of code, can ordinary people really "get" functional programming
> and become really productive in it, does one approach or another really
> reduce complexity in a meaningful way.  Jury's still out.

I think it's pretty clear that object orientation and functional
programming both reduce complexity in a meaningful way.



From kragen at pobox.com  Mon Jun 10 18:28:08 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 18:28:08 -0400
Subject: htmllib and parsing data sourrounded by HTML tags
References: <mailman.1023537792.4441.python-list@python.org>
Message-ID: <83elfeojk7.fsf@panacea.canonical.org>

Milos Prudek <milos.prudek at tiscali.cz> writes:
> I studied htmllib hard and I was able to subclass it and do some
> processing of HTML tags. I managed to extend it to process <TD>, for
> instance.
> 
> But what about the text that is surrounded by HTML tags?

I think you want to define a handle_data method.



From guido.goldstein at a-nugget.de  Fri Jun 28 01:17:18 2002
From: guido.goldstein at a-nugget.de (Guido Goldstein)
Date: 28 Jun 2002 07:17:18 +0200
Subject: How to get rid the new line
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org> <3D1AD534.A74EC9F4@ipm.fhg.de> <3D1B1A76.563E5394@engcorp.com> <pan.2002.06.28.00.14.48.789371.16627@linuxmail.org> <afg513$2426$1@news.idiom.com> <7xlm90cmxx.fsf@ruckus.brouhaha.com>
Message-ID: <m2lm90gext.fsf@nathan.mail.a-nugget.de>

Hi!

On 27 Jun 2002 16:38:02 -0700
  Paul Rubin <phr-n2002b at NOSPAMnightsong.com> wrote:
[...]
> If you don't mind removing all trailing whitespace, there's also
> 
>   line = line.strip()
> 
> or more portably
> 
>    import string
>    line = string.strip(line)

If you really would like to remove *trailing* whitespace you better
should use rstrip().
The strip() method strips from both sides!

Just my 2 cents.

HAND
  Guido


From eric.brunel at pragmadev.com  Thu Jun  6 14:57:31 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Thu, 6 Jun 2002 18:57:31 +0000
Subject: problem with isinstance and relative/absolute import
References: <87adq8jsq1.fsf@tux.ntw23.fr>
Message-ID: <ado3uu$sph$1@wanadoo.fr>

Sylvain Thenault wrote:
> does someone can explain to me the following test case, or should it be
> considered as a bug ?
> 
> [syt at tux syt]$ ls tests/
> __init__.py  module2.py  module.py
> [syt at tux syt]$ cat tests/__init__.py
> [syt at tux syt]$ cat tests/module.py
> class A: pass
> 
> a = A()
> 
> [syt at tux syt]$ cat tests/module2.py
> from module import A
> 
> b = A()
> 
> [syt at tux syt]$ export PYTHONPATH=$PWD
> [syt at tux syt]$ cd tests
> [syt at tux tests]$ python
> Python 2.1.3 (#1, Apr 20 2002, 10:14:34)
> [GCC 2.95.4 20011002 (Debian prerelease)] on linux2
> Type "copyright", "credits" or "license" for more information.
>>>> from module2 import b
>>>> from tests.module import A, a
>>>> isinstance(a, A)
> 1
>>>> isinstance(b, A)
> 0

I already noted this problem. The cause seems to be that in module2.py, you 
import A with:
from module import A
which is in turn imported in your main script via:
from module2 import b

But in your main script, you do:
from tests.module import A

This seems to confuse the interpreter which creates two classes named A: 
one in the module "module", and one in the module "tests.module", that are 
not identified as being the same.

However, when you do:
>>>> from tests.module import A, a
>>>> from tests.module2 import b
the line "from module import A" in module2.py is now executed from the 
import "from tests.module2 import b". So the interpreter knows that 
"module" is in the same directory than module2.py, and is therefore 
correctly identified as being the same than tests.module...

The solution is simple: try to avoid importing the same module in two 
different ways. First, it's quite ugly; second, it may confuse the 
interpreter...

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From nospam at bigfoot.com  Tue Jun 18 09:04:23 2002
From: nospam at bigfoot.com (Gillou)
Date: Tue, 18 Jun 2002 15:04:23 +0200
Subject: Listing all available (not just loaded) modules
References: <aem9om$888fb$1@ID-140151.news.dfncis.de>
Message-ID: <aenakn$a2j$1@norfair.nerim.net>

pydoc that's in the standard distro browses all that's in the python path.
Perhaps reading this module will give you the enlightenment.


--Gilles
glenfant AT bigfoot DOT com

"Michael Steuer" <mike at radiomag.com> a ?crit dans le message news:
aem9om$888fb$1 at ID-140151.news.dfncis.de...
> I've been searching the newsgroup for information on this but so far have
been
> unsuccessful:
>
> Is there a way to dynamically list all available modules (not just those
that
> are already loaded) in python? By available I mean those modules, which
are
> installed on the machine executing python (in the lib path(s)).
>
> Cheers,
>
> Mik.e
>
>



From python at rcn.com  Sat Jun 29 20:57:36 2002
From: python at rcn.com (Raymond Hettinger)
Date: Sat, 29 Jun 2002 20:57:36 -0400
Subject: Is this a __slot__ bug?
References: <pan.2002.06.30.00.21.54.133520.5016@vodeotron.ca>
Message-ID: <afll25$4rn$1@bob.news.rcn.net>

"Steve Menard" <steve.menard at vodeotron.ca> wrote in message
> However, I also love(need) private variables (those __X). It seems they
> dont work with slots.
 <snip>
> So if this is by design, I will not agree but I will do the workaround.
> If not, well then seems I founda bug :)
>
> Steve

Yes, you found a bug.
It takes a keen eye to spot one :-)

However, someone faster on the draw already reported it
and I've already fixed it (for Py2.2.2 and Py2.3).


Raymond Hettinger





From phr-n2002b at NOSPAMnightsong.com  Sun Jun 23 17:16:56 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 23 Jun 2002 14:16:56 -0700
Subject: Python hits the spot
References: <mailman.1024772248.9399.python-list@python.org> <yppfsn3dlpco.fsf@hotlips.cs.chalmers.se> <3D1634ED.B9ACF9EC@engcorp.com> <af5d0r$nlt$1@slb4.atl.mindspring.net>
Message-ID: <7xfzzdhf07.fsf@ruckus.brouhaha.com>

"Matt Gerrans" <mgerrans at mindspring.com> writes:
> > > for i in range(50*1000*1000):
> > >    pass
> 
> You call that a bug?   Did you consider trying xrange()?

Maybe not a bug, but a pretty serious misfeature IMO.


From sholden at holdenweb.com  Wed Jun  5 08:35:34 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Wed, 5 Jun 2002 08:35:34 -0400
Subject: semi-concatenated strings
References: <mailman.1022811221.4265.python-list@python.org> <3CF6F202.7274E43E@engcorp.com>
Message-ID: <nEnL8.268054$%u2.145302@atlpnn01.usenetserver.com>

"Peter Hansen" <peter at engcorp.com> wrote in message
news:3CF6F202.7274E43E at engcorp.com...
> "Delaney, Timothy" wrote:
> >
> > > From: Peter Hansen [mailto:peter at engcorp.com]
> > >
> > > And which has both embedded "\n" newlines and many more leading
> > > spaces before each line...
> >
> > That's why I would bind it to a name first - that allows me to
preprocess
> > the query before passing it in.
>
> > I would much rather write it so it is readable, then process it for the
> > database, than the other way around.
> >
> > sql = """
> >     select cities.city, state, country
> >     from   cities, venues, events, addresses
> >     where  cities.city like %s
> >            and events.active = 1
> >            and venues.address = addresses.id
> >            and addresses.city = cities.id
> >            and events.venue = venues.id
> > """
> >
# The next three lines are bogus
> > sql = string.split(sql)
> > sql = string.join(sql, ' ')
> > sql = string.strip(sql)
> >
> > rows = self.executesql(sql, (city,))
>
> What a lot of extra work to make something "readable"!
> I don't think that this is more readable than the simple, straightforward
> approach Skip presented.  It uses many more lines and a bunch of
> superfluous operations (without even a comment explaining why
> you are splitting and joining things this way instead of just using
> simple string concatenation).
>
Not only that, but for any reasonable database the whole schemozzle is
completely unnecessary. SQL is a free-format language, and newlines and
spaces embedded in statements make absolutely no difference.

> If I saw code like this in my team I'd ask that a comment be
> added to explain it, and if the comment said "# doing this so the
> code is more readable" I'd ask that it be refactored so it didn't
> need the comment at all, after I got back up off the floor.
> I expect the refactoring would take the form of the original
> example which needs no extra operations or commenting.
>
Or simply remove the three bogus lines and leave the SQL as it's generated.

> > Skip also asked why I would bind the statement to a name. For a query of
> > this size, I find it easier to read - the sql query is separate and (to
my
> > eye) cleaner than having the literal as a function parameter. Plus it
uses
> > much less horizontal white space.

I hadn't realised there was a shortage :-) [Seriously, I like the embedded
newline style, as I underline below...]
>
> Nothing wrong with separating things out like that if you think it
> looks more readable, but this is one case where I agree with Skip's
> approach even though it looks a lot like "hardcoded strings" which
> is generally a Code Smell.  Simpler is much better here.
>
> (I might agree with you more if I thought that maintaining the
> string-concatenated version would actually be harder.  Given the
> likely nature of the maintenance -- adding fields to the line with
> "from cities, ..." for example, or adding a new line to the
> "where" intersection -- I don't believe it would be *any* harder.)

As I said earlierrr, in another post: '''would personally much prefer to see
this written as

    ("""select cities.city, state, country
     from cities, venues, events, addresses
     where cities.city like %s
         and events.active = 1
         and venues.address = addresses.id
         and addresses.city = cities.id
         and events.venue = venues.id""",
    (city,))
'''

but it's largely a matter opf taate. What clealy *should* be avoided is
unnecessary programmed transformation of SQL statements.

regards
 Steve
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From mjais at web.de  Mon Jun  3 10:56:20 2002
From: mjais at web.de (Markus Jais)
Date: Mon, 03 Jun 2002 16:56:20 +0200
Subject: different gui-toolkits pros/cons...
References: <adfsoo$dmik4$1@hades.rz.uni-sb.de>
Message-ID: <_rLK8.12$fd2.979@news.ecrc.de>

with pyQT there might be problems with the license
if you want to make your programm freely available
(but I am not a lawyer, so I might be wrong)

pyGTK might be faster, because wxWindows is AFAIK an additional
layer on top of Gtk. but only in Unix/Linux platforms, AFAIK

I do not know if Gtk is really stable for Windows now. you 
should try to use Gtk2

pyGtk seems to be pretty fast, but I have made no tests

markus


Uwe Schmitt wrote:

> Hi,
> 
> I'm looking for an appropriate toolkit for GUI programming,
> which do you use ? what are pros and cons ???
> I'm specially interested in wxPython, pyQt and pyGTK...
> and I'm intersted in a windows-like look and feel on win platforms...
> Finally I'm intersted in execution speed.
> 
> So, I await your answers, greetings, Uwe
> 



From skin_pup-python-list at happypoo.com  Mon Jun 24 12:52:33 2002
From: skin_pup-python-list at happypoo.com (Skinny Puppy)
Date: Mon, 24 Jun 2002 12:52:33 -0400
Subject: My .py CGI script works in IE but not Netscape
In-Reply-To: <79b54d28.0205311837.731002aa@posting.google.com>
References: <79b54d28.0205311837.731002aa@posting.google.com>
Message-ID: <20020624165233.GF19586@shitbomb.com>

Chris [chouster at uclink4.berkeley.edu] wrote:
> My script doesnt work in Netscape for some reason. I'm developing on a
> UNIX envionrment with Python 2.2.1. I open up Netscape in windows and
> only the background picture of my webpage loads... nothing else.
> However in IE it works perfectly. Any general ideas as to what may be
> going on? Thanks.

I have run in to this before.  Make sure that your script has something
like this before it outputs other text.  

  print "Content-type: text/html"
  print ""

THe start of the issue is with the web server, most are set to default to
output "Content-type: text/plain" unless you tell it other wise (This is
correct BTW).  

Now when netscape receives this it renders your scripts output as a plain
text file, no HTML rendering. 

IE on the other hand starts to render the output as plain text, but it
notices the <HTML> tags and switches to rendering HTML.

Jeremy





From dreed at capital.edu  Tue Jun 11 20:26:44 2002
From: dreed at capital.edu (Dave Reed)
Date: Tue, 11 Jun 2002 20:26:44 -0400
Subject: lists and sequences
In-Reply-To: <LNBBLJKPBEHFEDALKOLCKEDOPMAA.tim.one@comcast.net> (message from
	Tim Peters on Tue, 11 Jun 2002 19:32:25 -0400)
References: <LNBBLJKPBEHFEDALKOLCKEDOPMAA.tim.one@comcast.net>
Message-ID: <200206120026.g5C0QiD24559@localhost.localdomain>

> From: Tim Peters <tim.one at comcast.net>
> Cc: cmonahan at yahoo.com
> 
> [Collin Monahan]
> > What is a python list? Is it a linked list or a random access structure
> > in contiguous memory?
> 
> A contiguous vector of pointers to objects.
> 
> > What type of situations should be avoided when using it?
> 
> Depends on the app and your judgment of appropriate tradeoffs.
> 
> > E.g. when does n^2 time happen using it?
> 
> alist.insert(0, newthing) physically moves len(alist) elements.  Ditto

Just to be clear, these are order n, not order n^2.

<other info snipped>

Dave




From jadestar at idiom.com  Tue Jun 18 02:13:55 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 18 Jun 2002 06:13:55 GMT
Subject: string generator
References: <mailman.1023127325.26015.python-list@python.org>
Message-ID: <aemj33$15dc$1@news.idiom.com>

Gustavo Cordova <gcordova at hebmex.com> wrote:

>> I like to create a list of file automatical based on series. 
>> For example
>> user input the number of file say 5. So i want to generate the file 
>> according to this range , this mean the generated files could be 
>> filename1
>> filename2
>> filename3
>> filename4
>> filename5
>> I using the for loop like that
>> 
>> for file in range(5):
>>         lis = [file]
>>         string.join('filename',lis[1:])
>> how can it didn't work. I am confusing how to do it.
>> 

> Yeah, but where do you wanna put your results?

> You could also:
> FNs = ["filename%d" % i for i in range(5)]
> or also you could:
> FNs = map(lambda i:"filename%d" % i, range(5))
> or many other variations thereof.

> Good luck!
> -gustavo

 Here's a variation:

>>> string.split(' filename'.join([str(x) for x in range(5)]))[1:]
['filename1', 'filename2', 'filename3', 'filename4']

>>> string.split(' filename'.join([str(x) for x in range(-1,5)]))[1:]
['filename0', 'filename1', 'filename2', 'filename3', 'filename4']

 although the map(lambda ...)) with the % operatator is even more
 clever.



From bokr at oz.net  Mon Jun 24 21:51:35 2002
From: bokr at oz.net (Bengt Richter)
Date: 25 Jun 2002 01:51:35 GMT
Subject: SOLVED (Re: Python hits the spot)
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <af2dim$sh6$0@216.39.172.122> <3D14CF95.4030501@kfunigraz.ac.at> <af3jk7$c4h$0@216.39.172.122> <3D158F3E.1010907@kfunigraz.ac.at> <af4vrl$ffe$0@216.39.172.122> <3D16DD5C.1050708@kfunigraz.ac.at> <uhdume2n1mhbba@corp.supernews.com> <3D171291.2060905@kfunigraz.ac.at>
Message-ID: <af8ib7$k51$0@216.39.172.122>

On Mon, 24 Jun 2002 14:37:37 +0200, Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote:

>RPM1 wrote:
>
>> Hey!  That's why I was asking if your laptop was a Hewlitt
>> Packard.  I had a Hewlitt Packard Pavillion laptop that would
>> die when the fan kicked on.
>> 
>> Any chess program will keep your CPU at 100%.
>> 
>> GnuChess is probably on your system now.  If not, it is a free
>> program.  Ask around on rec.games.chess.computer if you need
>> help.
>
>It was really the laptop processor. I tried the calculation on a 
>stationary machine (Pentium II 450 MHz and SuSE 8.0), and it works.
>
>Die Typen k?nnen jetzt was erleben.
>
Well, lucky guess on the heat. That _probably_ accounts for laptop
going to 87% and then shutting down, but how about the other huge
slowdowns in your big loop of 8? Did you get that diagnosed too?
Just curious ;-)

Regards,
Bengt Richter


From kragen at pobox.com  Sat Jun  1 02:11:55 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 01 Jun 2002 02:11:55 -0400
Subject: matching one time through a loop
References: <B43D149A9AB2D411971300B0D03D7E8BF0A335@natasha.auslabs.avaya.com> <mailman.1022806551.14232.python-list@python.org> <ad496f8.0205310435.4466841b@posting.google.com>
Message-ID: <837kljwms4.fsf@panacea.canonical.org>

vvainio at tp.spt.fi (Ville Vainio) writes:
> BTW, how does perl do re.findall()? Or re.sub using function return
> values to dynamically determine the replacement? I can't remember I
> ever used such features with perl.

re.findall():

my $string = 'a moon upon a lunar tune aroon';
my @list = ($string =~ /\w+n/g);
print map { "$_\n" } @list;

I'm afraid I always have to look up what // returns every time I use it.


re.sub using function return values is approximately the /e modifier,
but without the need to wrap a 'lambda:' around the replacement:

my %dict = qw(a the moon rock upon above lunar solar tune song aroon boon);
my $string = 'a moon upon a lunar tune aroon';
$string =~ s/\w+/$dict{$&}/ge;
print $string, "\n";

Compare Python:
import re
mydict = {'a': 'the', 'moon': 'rock', 'upon': 'above', 'lunar': 'solar',
          'tune': 'song', 'aroon': 'boon'}
astring = 'a moon upon a lunar tune aroon'
astring = re.sub(r'\w+', lambda mo: mydict[mo.group(0)], astring)
print astring

Not only is the Python harder to read (to me), but it took me four
tries to get it to work, because I made the following three mistakes:
- I got the order of re.sub arguments wrong (and got a *very* unhelpful
  traceback from line 177 in sre.py)
- I called the lambda argument 'aword' in one place and 'word' in the other
- I thought the lambda argument was the matched string, not the match object
  (if it had been the matched string, I could have written mydict.get instead
  of the lambda expression)

The Perl version worked the first time; you will note that none of the
three Python mistakes have a reasonable analogue in the Perl version.

> My experience is that Python regexps are better when you go beyond
> non-trivial, with clear group & matchobject semantics

I wholeheartedly agree.  But the vast majority of my regexps are
trivial.




From peter at engcorp.com  Mon Jun 24 21:58:33 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 24 Jun 2002 21:58:33 -0400
Subject: adding attributes to builtin functions
References: <Gy8Fr6.J7D@beaver.cs.washington.edu>
Message-ID: <3D17CE49.E3D16ABE@engcorp.com>

Douglas Zongker wrote:
> 
>    import mymodule
>    print mymodule.myfunction
>    mymodule.myfunction.special = 6
> 
> gives me:
> 
>    <built-in function myfunction>
>    Traceback (most recent call last):
>      File "<stdin>", line 1, in ?
>    TypeError: 'builtin_function_or_method' object has only read-only
>      attributes (assign to .special)
> 
> Is there a way to define a builtin function whose attributes aren't
> read-only?

Maybe something like this?

import mymodule

def myfunction(*argp, **argn):
    return mymodule.myfunction(*argp, **argn)  

mymodule.myfunction = myfunction
mymodule.myfunction.special = 6

Tada!  (Actually, I have *no* idea if this works. :-)

-Peter


From fperez528 at yahoo.com  Tue Jun 25 15:20:04 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Tue, 25 Jun 2002 13:20:04 -0600
Subject: Integrals with Python
References: <3D188551.6BA00DED@lanl.gov> <3D189307.1417ECD5@info.unicaen.fr> <3D18A010.CAF3FA9E@lanl.gov>
Message-ID: <afafp5$cun$1@peabody.colorado.edu>

Nathan Given wrote:

> Matlab has quad, quadl, and quad8...
> 
> 
> does Numeric have anything like those?

Go look at SciPy (http://scipy.org), it comes with a reasonable set of 
integration routines. But as Jerzy warned you, you first need to understand 
correctly the structure of your integrands. Black box integration is similar 
to russian roulette (just look at some of the disasters the famous 
Mathematica spits out).

But at the level you seem to be working, scipy would definitely be a good 
start. Then you can refine the routines therein if needed. Another option is 
to use C integrators wrapped with weave (part of Scipy). That's what I do 
because it's faster and I have tighter control over exactly what is going on 
under the hood. With weave, the wrapping job is very simple.

Cheers,

f


From pyth at devel.trillke.net  Fri Jun  7 08:17:49 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 7 Jun 2002 14:17:49 +0200
Subject: Efficient python programming...
In-Reply-To: <3D00A456.4315EDA3@engcorp.com>; from peter@engcorp.com on Fri, Jun 07, 2002 at 08:17:26AM -0400
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com>
Message-ID: <20020607141749.C16359@prim.han.de>

Peter Hansen wrote:
> Eddie Corns wrote:
> > 
> > A friend is fond of quoting this to me:
> > 
> >   Rules of Optimization:
> >     Rule 1: Don't do it.
> >     Rule 2 (for experts only): Don't do it yet.
> >   -- M.A. Jackson
> > 
> > I wasn't going to bother contributing since others have said the same thing
> > but the more people say the more you'll believe it :)
> > 
> > Especially for a beginner the important thing is to get clear and obvious
> > code.  Later, IF IT ISN'T FAST ENOUGH, then you think about ways to speed it
> > up.
> 
> You forgot the even more important first thing for a beginner: 
> get it correct!  The only good way to do this is to write unit
> tests that prove it.  If you don't have tests, you don't have 
> working code (or how can you prove it?).

yeah. writing tests is *the way* to go if you want other people use 
your code. Although successfull unittests theoretically don't prove 
anything<wink> they tend to do in reality<doublewink>. 

if-it's-not-proven-then-test-it-ly yours, holger




From shalehperry at attbi.com  Fri Jun  7 20:07:26 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 07 Jun 2002 17:07:26 -0700 (PDT)
Subject: Efficient python programming...
In-Reply-To: <3D013FB1.3D7477EA@engcorp.com>
Message-ID: <XFMail.20020607170726.shalehperry@attbi.com>

> 
> I disagree, if the tests are written first (and run, proving that
> they will fail when the code is wrong or missing).
> 
> The tests are executable requirement specifications.  Running
> them "proves" (yes, not 100%... nothing could) that the code
> meets the specs.  The specs could be wrong of course, but that
> doesn't mean the code is broken...
> 

this also assumes the test covers every possible failure.  this also assumes
one can write tests for that section of code.




From richardt at edshk.demon.co.uk  Tue Jun 18 08:16:56 2002
From: richardt at edshk.demon.co.uk (Richard Townsend)
Date: Tue, 18 Jun 2002 13:16:56 +0100
Subject: Embedding Python - time module problem
Message-ID: <cIgHcXA4SyD9swmz@edshk.demon.co.uk>

I have been having problems embedding Python in a small C program which
calls a Python script which, in turn, imports the time module.

This fails with the error:

"ImportError: dynamic module does not define init function (inittime)"

I have simplified this down to a minimal C program which just tries to
import the time module. This gives the same error.

  #include <stdio.h>
  #include <Python.h>

  int main(int argc, char **argv) 
  {
    PyObject *module;
    Py_Initialize();
    module = PyImport_ImportModule("time");
    if (module == NULL)
    {
      PyErr_Print();
    }
  }

If I substitute "os" or "sys" for "time" then no error occurs. However
"math" does give a similar error.  

If I run the interpreter on its own, I can import time and math with no
problems. 

I am using Python 2.2.1 on HP-UX 11i

Is there something special I need to do to be able to import time and
math modules when using embedded Python?




From maxm at mxm.dk  Mon Jun 24 11:39:14 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 24 Jun 2002 17:39:14 +0200
Subject: Python hits the spot
References: <mailman.1024928911.24460.python-list@python.org>
Message-ID: <3D173D22.9090304@mxm.dk>

Tim Peters wrote:

>>Don't some implementations of malloc offer a way to ask the operating
>>system to commit to the memory being available when you allocate it?
>>If so, I can imagine testing for that facility at config time and
>>using it if available.
> 
> SourceForge remains open 24 hours a day for patches.  My bet is that nobody
> cares enough to bother -- the barrier for complaining on Usenet is one an
> ant can hurdle with ease <wink>.

Oh... somebodys grumpy that he isn't at EuroPython ;-)

regards Max M



From shagshag13 at yahoo.fr  Sun Jun  2 08:15:21 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Sun, 2 Jun 2002 14:15:21 +0200
Subject: Pop a list from beginning ? and memory saving...
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de> <3CFA0676.68F80DE6@engcorp.com>
Message-ID: <add289$vud42$1@ID-146704.news.dfncis.de>

"Peter Hansen" <peter at engcorp.com> a ?crit dans le message de news: 3CFA0676.68F80DE6 at engcorp.com...
> Shagshag13 wrote:
> >
> > I had a huge list containing consuming data, as i only need a one pass in a
> > for statement, i'm wondering if by doing this with a pop i could save memory ?
>
> Maybe, maybe not.  The implementation _probably_ makes a copy of
> the entire list at some point as you shrink it.  That would mean that
> at least at some interval, you would temporarily consume almost twice
> the memory of the original list as the original list is copied, then
> freed.  You might end up with less memory used, but with double the
> memory needed.  Of course, the same would have held true as you built
> the list in the first place...
>
> Why are you trying to save memory?  Are you really working with such
> a large amount of data in one chunk that you will run out?  Maybe
> there is some other approach you haven't thought of.  This effort
> almost smells like premature optimization (the root of all evil).
>
> -Peter

in fact i had huge memory troubles !!! i'm a python newbie, but i use it because i needed a
language which allow fast development (and can't remind me of things like C or Java...)

and yes i'm working with large amount of data : i'm trying to index something like 1 go of
plain text by using conventionnal text retrieval structures (i mean without compression
tricks, because i *don't* have time to implement theses, but with stoplist and so on)
and wishing to do it in one pass without relying on hdd temp files (i'm starting to
think that i were dreaming...)

the most i think i could simply do is this : i use integer each time as possible, i free
everything i don't need anymore with = None, and what could i do more ? Advices
are ***greatly*** welcome !!!

s13.

ps : do you know if any python databases could handle without trouble more than
1 or 2 go of data ???







From sabine at textraeume.de  Mon Jun  3 06:48:52 2002
From: sabine at textraeume.de (Sabine Richter)
Date: Mon, 03 Jun 2002 12:48:52 +0200
Subject: word wrap on different OS
Message-ID: <3CFB4994.DD35AEA2@textraeume.de>

Hello,

I am not sure if I think clear on this subject:

A simple text file has been created for example on MS Dos or MacOS. So
the sign for a word wrap will be "\r\n" on MS Dos or "\r" on MacOS.
Then I save the file on a windows file system. The sign for the word
wrap will be converted to "\n". 
If I now read the file content into a string and search for a word wrap
in the string I have to search for "\n". For example with string.split
or with an regular expression. 
Is that all right?

And next question:
If I write a script, which uses MyODBC and the odbc-module from win32
extensions to connect to a MySQL-database, it can only be used on
windows. Isn't it?

So, in consequence I don't have to handle different signs for word wraps
in my script?
Or would it make any sense to do so? The purpose of the script is to
update a database structure based on a given dump file.

Thanks in advance for your answers
Sabine




From shalehperry at attbi.com  Sat Jun 29 11:58:00 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Sat, 29 Jun 2002 08:58:00 -0700 (PDT)
Subject: exception handing
In-Reply-To: <pan.2002.06.29.17.41.28.926443.440@myself.com>
Message-ID: <XFMail.20020629085800.shalehperry@attbi.com>

On 29-Jun-2002 Rhymes wrote:
> I'm reading "Learning Python" and in the section "gotchas"
> of the exception handling chapter the book itself states
> that the exception matching is made by identity not equality.
> It also write an example such as this:
> 
>>>> ex1 = "spam"
>>>> ex2 = "spam"
>>>>
>>>> ex1 == ex2, ex1 is ex2
> (1, 0)
> 
> <--- here i get (1, 1) ---->
> 

looks like the interpreter is now optimizing shared strings.

>>>> try:
> ...    raise ex1
> ... except ex1:
> ...    print 'got it'
> ...
> got it
>>>> try:
> ...    raise ex1
> ... except ex2:
> ...    print 'Got it'
> ...
> Traceback (innermost last):
>   File "<stdin>", line 2, in ?
> spam
> 
> but in my Python 2.2 i don't get back any error from the interpreter
> about the second example... why? What has changed (since 1.5 release)
> about objects identity?
> 

modern python uses exception classes and not strings.

>>> class BadThingsException: pass
... 
>>> try:
...   raise BadThingsException 
... except BadThingsException:
...   print 'got it'
... 
got it





From olc at ninti.com  Sat Jun  8 05:53:31 2002
From: olc at ninti.com (Michael Hall)
Date: Sat, 8 Jun 2002 19:23:31 +0930 (CST)
Subject: Creating Dynamic Web Pages (ZOPE)
In-Reply-To: <cb4ba455.0206070514.335e2cb2@posting.google.com>
Message-ID: <Pine.LNX.4.21.0206081919430.1782-100000@localhost.localdomain>

I notice no-one has suggested good ol' CGI. For a relatively simple job
such as this, a few basic Python CGI scripts, some templates and MySQLdb
could also be used ... oh, just noticed the MS-SQL database. There is
probably a module around to cope with MS-SQL, no?

Mick

On 7 Jun 2002, Bill Tate wrote:

> hwcowan at hotmail.com (Hugh Cowan) wrote in message news:<46ca81a0.0206062016.789d3af6 at posting.google.com>...
> 
> [snip]
> > My scenario is typical where I need to put together a company Intranet
> > with a few front end Web-Pages where users can select information,
> > various links, and enter search criteria.  The information is then
> > retrieved from the MS-SQL Server and the results are sent back to the
> > user.  It's nothing fancy, no shopping carts, or remembering sessions
> > states, etc.. just really a simple front-end interface for
> > non-technical people to easily retrieve information.
> >  
> > I don't know if ZOPE would fit my situation or not.  I know that it is
> > / can be used to create complex E-commerce sites, but I am wondering
> > if it is overkill for what I need done.  Would I be better off trying
> > to find some sort of RAD Web Tool instead (anyone have any good
> > suggestions) ?
> 
> I think its safe to say that there are any number of tools that could
> fit the bill here including Zope.  There are aspects of zope that can
> make your life much easier (e.g., zsql methods) but it does take some
> time to get used to it and everything it offers.  Several recent books
> on the subject I think would be very helpful to you in this regard. 
> If you go to zope.org, there is some information on using WYSIWYG
> tools such as Dreamweaver with zope that may also be of interest to
> you as well.
> 
> I think zope works bests for you over the long haul where it is
> important to keep things well organized and well structured.  The
> ability to collaborate amongst individuals is also a great strength. 
> If these issues are important to you, then I would say that Zope would
> be an excellent choice.
> 
> Good luck!!!
> 

-- 
--------------------------------
n   i   n   t   i  .   c   o   m
php-python-perl-mysql-postgresql
--------------------------------
Michael Hall     ninti at ninti.com
--------------------------------





From skip at pobox.com  Sun Jun 16 15:17:39 2002
From: skip at pobox.com (Skip Montanaro)
Date: Sun, 16 Jun 2002 14:17:39 -0500
Subject: [Extending] Py_BuildValue( "((OiiO))",...) creates empty string?
In-Reply-To: <3D0CDEEB.2070208@rogers.com>
References: <3D0CDEEB.2070208@rogers.com>
Message-ID: <15628.58451.320049.77386@12-248-41-177.client.attbi.com>

Does
    
    Py_BuildValue("(OiiO)", tagobj, childStart, childPosition, childResults);

create a valid tuple?  Do you have a simple test case you can post that
demonstrates the problem?

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html




From maxm at mxm.dk  Wed Jun 12 06:17:42 2002
From: maxm at mxm.dk (Max M)
Date: Wed, 12 Jun 2002 12:17:42 +0200
Subject: Reload an imported file
References: <ae76vb$4l6iq$1@ID-79019.news.dfncis.de>
Message-ID: <3D071FC6.5000309@mxm.dk>

Felix Seeger wrote:

> I have a config file. This is a normal python file.
> Can I reimport this file during runtime ?
> So changing the configuration will not need a restart of my program.


import myModule
reload(myModule)

regards Max M



From BPettersen at NAREX.com  Wed Jun 12 14:54:02 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Wed, 12 Jun 2002 12:54:02 -0600
Subject: Adding and running functions at runtime
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019221519E@admin56.narex.com>

> From: John Dunn [mailto:jhndnn at yahoo.com] 
> 
> Hello-
> 
> I am currently embedding Python in my C++ application and 
> would like to add function 'scriplets' defined by strings and 
> then call them from C. Let me explain what I am trying to do -
> 
> 1. load a string that defines a function ( or functions )
> 2. get the function object for a defined function
> 3. call function with arguments
> 4. get return value and modified argument values
> 
> so the code might look something like
> 
> const char * funcs = 
> "def foo( arg1, arg2 )\n"
> "    arg2 = arg2\n"
> "def foo2( arg1, arg2 )\n"
> "    return arg1+arg2\n"
> 
> Mythical_PyLoad( funcs );
> PyObject* foo = Mythical_PyGet_Function( "foo" );
> PyObject* foo2 = Mythical_PyGet_Function( "foo2" );
> PyOjbect* args = some code here to create arguments
> // this should return the sume of the args
> PyObject* result = PyEval_CallObject( foo2, args ); 
> // this should modify the second arg
> PyObject* result = PyEval_CallObject( foo, args ); 
> 
> I have been able to load code using Py_CompileString, but 
> that appears to run inline code, and I couldn't get it to 
> return a value or modify something in the dictionary.

I'm working on something similar. The interface looks like e.g.:

        py::Namespace ns;
        NDate d(1970, 5, 2);
        ns.set("tmp", d);

        py::stmts(
            "def two():          \n"
            "    return 2        \n",
            ns);

        int result = py::expr("two() + tmp.getYear()", ns);

i.e. it uses a namespace to keep track of your function definitions and
you get values by assigning the result of py::expr() to the appropriate
type. I can send you the code if you'd like...

oh, and it also deals with SIP wrapped libraries :-)

-- bjorn




From emile at fenx.com  Wed Jun 19 09:18:22 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 19 Jun 2002 13:18:22 GMT
Subject: Tkinter and the Menu Widget
References: <2259b0e2.0206190415.1f232447@posting.google.com>
Message-ID: <xw%P8.257292$cQ3.11997@sccrnsc01>

Michele Simionato
>     import Tkinter
>     def dosomething(s):
>         print s
>     root=Tkinter.Tk()
>     mn=Tkinter.Menu(root) #attach the menu mn to root
>     menuD={'File':['File1','File2'],'Help':['Help1','Help2']}
>     smn={} #dictionary keys->submenu-objects defined below
>     for key in menuD.keys():
>         smn[key]=Tkinter.Menu(mn) #define a submenu of mn
>         mn.add_cascade(label=key,menu=smn[key]) #attach smn[key] to mn
>         for submenu in menuD[key]:
>             smn[key].add_command(label=submenu,command=lambda :
>                                  dosomething(key+'='+submenu))

You've got a scope issue.  Try:

        smn[key].add_command(label=submenu,command=lambda key=key,
submenu=submenu:
                             dosomething(key+'='+submenu))

--

Emile van Sebille
emile at fenx.com

---------



From astavale at yahoo.co.uk  Thu Jun 27 13:48:26 2002
From: astavale at yahoo.co.uk (Alistair Thomas)
Date: Thu, 27 Jun 2002 18:48:26 +0100
Subject: sys.argv[0] - Returns path and filename.
References: <c07d0edc.0206270432.faf1fd3@posting.google.com>
Message-ID: <uhmk0q1aaa8e7e@corp.supernews.com>

In article <c07d0edc.0206270432.faf1fd3 at posting.google.com>, "Guy"
<guy at lightwork.co.uk> wrote:

> How do I get the current path of the script I am running, on the mac osx
> and the linux box ? 

Does 

sys.path[0]

make any difference?

--
Alistair


From jdhunter at nitace.bsd.uchicago.edu  Sat Jun 29 16:56:17 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Sat, 29 Jun 2002 15:56:17 -0500
Subject: can't import generators
References: <m2hejlc060.fsf@mother.paradise.lost> <afl5je$4iu$1@panix1.panix.com>
Message-ID: <m2d6u9by8e.fsf@mother.paradise.lost>

>>>>> "Aahz" == Aahz  <aahz at pythoncraft.com> writes:

    Aahz> Can't do this in interactive mode.  Use a script

Tried that first -- still a no go

> cat ./test.py
#!/usr/local/bin/python2.2
from __future__ import generators

> ./test.py
Traceback (most recent call last):
  File "./test.py", line 2, in ?
    from __future__ import generators
ImportError: cannot import name generators

> /usr/local/bin/python2.2 -V
Python 2.2.1

> /usr/local/bin/python2.2 /usr/local/lib/python2.2/test/test_generators.py
Traceback (most recent call last):
  File "/usr/local/lib/python2.2/test/test_generators.py", line 1, in ?
    from __future__ import generators
ImportError: cannot import name generators


Any other thoughts?  The only thing slightly non-standard about by
install is that I compiled python2.2 for large file support with 

OPT= -ggdb -O2 -Wall -Wstrict-prototypes -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64

but I can't see that making a difference.

John Hunter






From phd at phd.pp.ru  Thu Jun 13 12:36:36 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 13 Jun 2002 20:36:36 +0400
Subject: filter2
In-Reply-To: <20020613160557.GC25416@unpythonic.net>; from jepler@unpythonic.net on Thu, Jun 13, 2002 at 11:06:02AM -0500
References: <mailman.1023975618.6474.python-list@python.org> <lkptyvmdfz.fsf@pc150.maths.bris.ac.uk> <20020613160557.GC25416@unpythonic.net>
Message-ID: <20020613203636.H19424@phd.pp.ru>

On Thu, Jun 13, 2002 at 11:06:02AM -0500, Jeff Epler wrote:
>     funcs = [l2.append, l1.append]
>     map(lambda x: funcs[bool(test(x))](x), list)

   Wow! That's nice!

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From pimpao at bestway.com.br  Sun Jun 16 19:41:35 2002
From: pimpao at bestway.com.br (Thiago)
Date: 16 Jun 2002 16:41:35 -0700
Subject: How refresh a page trhough httplib
Message-ID: <6e458916.0206161541.100be8eb@posting.google.com>

hi,

 I need a problem, i was using python to build a application to verify
a status of web page, and i used proxy, to connect a internet, and i
would like to refresh the page on proxy, because it's cached on proxy!

Thiago


From gumuz at looze.net  Wed Jun 12 09:40:19 2002
From: gumuz at looze.net (Guyon Morée)
Date: Wed, 12 Jun 2002 15:40:19 +0200
Subject: Bits and bytes?
Message-ID: <3d074e6d$0$224$4d4ebb8e@news.nl.uu.net>

ok, this might be a strange question.... but i was wondering:

how can i read or write 'bits'?

if i have a string, how can i convert to it's basic 0's and 1's?

and offcourse, the other way around.... let's say i want to store some
bit-sequence(0011101110) to a file.


how do i do this?

thanx,

Guyon




From spam at melkor.dnp.fmph.uniba.sk  Wed Jun 19 02:38:59 2002
From: spam at melkor.dnp.fmph.uniba.sk (Radovan Garabik)
Date: Wed, 19 Jun 2002 08:38:59 +0200
Subject: procmail replacement in Python
References: <mailman.1024444823.19953.python-list@python.org>
Message-ID: <3u8pea.617.ln@127.0.0.1>

Gerhard H?ring <gh_pythonlist at gmx.de> wrote:
 : Does such a thing exist? I've recently looked into the procmail sources

<shameless advertising plug>
http://melkor.dnp.fmph.uniba.sk/~garabik/pycmail.html
</shameless advertising plug>


-- 
 -----------------------------------------------------------
| Radovan Garabik  http://melkor.dnp.fmph.uniba.sk/~garabik |
| __..--^^^--..__         garabik @ fmph . uniba . sk       |
 -----------------------------------------------------------
Antivirus alert: file .signature infected by signature virus.
Hi! I'm a signature virus! Copy me into your signature file to help me spread!


From kragen at pobox.com  Mon Jun 10 17:35:50 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 17:35:50 -0400
Subject: Correct code/unit testing (was Re: Efficient python programming...)
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <adq44b$obr$1@scotsman.ed.ac.uk> <3D00A456.4315EDA3@engcorp.com> <834rgd5isv.fsf@panacea.canonical.org> <3D02FB3B.CF75E925@engcorp.com>
Message-ID: <83u1oaq0jt.fsf@panacea.canonical.org>

Peter Hansen <peter at engcorp.com> writes:
> Kragen Sitaker wrote:
> > Peter Hansen <peter at engcorp.com> writes:
> > I also agree that unit tests are very useful for reducing bugginess.
> > 
> > But I do not think that they are either necessary or sufficient for
> > writing correct code.  
> 
> Without tests, how much confidence can you have in the correctness
> of your code?  It's been said "you don't know it works if you 
> don't test it".

Tests are one way to increase your confidence in the correctness of
your code, but they are not the only way.

> > I think (and extensive experimentation has
> > shown) that careful reasoning and code reviewing is necessary and
> > sufficient to write correct code.
> 
> Whoa!  "extensive experimentation has shown"??

Yes.  (Well, I guess "correct code" is probably overstating the case.
"Code with many fewer bugs per line than most code in use" is probably
more accurate.)  Are you familiar with Cleanroom programming?

> I would suggest that what extensive experience (I change the word
> deliberately) has shown many of us is that careful reasoning and
> code reviewing, no matter how extensive, will *always* leave bugs in
> the code.  Have you never spent a large amount of time carefully
> designing and coding a wonder little program, and checked it over
> very carefully, even with a peer, and later found a bug in it?

Of course!  People are fallible, it's much easier to do a bad code
review than a good one.

> Maybe you're just a better programmer than I, and many others, but I
> would say this inevitable outcome is the source of the term "egoless
> programming" -- if you have a big ego, you'll be devastated by the
> repeated proof that you will never, ever get it right the first
> time, or even after lots of reviewing.

Yes, one of the beautiful things about programming is how it impresses
upon each of us how fallible we are.

For what it's worth, the term "egoless programming" was invented by
Gerald Weinberg for his book _The Psychology of Computer Programming_
to describe a programming style in which group reviews focus on
finding bugs in code.  It comes from Weinberg's observation that code
reviews work very poorly if the author is defensive about bugs, and
much better if the author is trying to find bugs too.

Of course, it's true of testing too; testing works best if you're
trying to find bugs.

> > It is certainly possible to err too far in the direction of believing
> > your code correct simply because it passes its test suite, and this is
> > a common beginner's error.  I think it is also possible to err too far
> > in the direction of writing and running tests instead of desk-checking
> > code.
> 
> I can only suggest that until you have tried test-first development,
> in which the code is written in very, very small steps only to make
> small failing tests pass, and discovered the possibilities of 
> emergent design, and other aspects of this approach, then we will not
> find much common ground in this discussion.

Test-first development, exactly as you describe it above, has been my
preferred development style since mid-2000, at least when working on
programs of more than 100 lines, especially with other people.  I've
written a lot of code that way.

I think now we have each committed the same arrogant gaffe in this
discussion: "It is only possible to disagree with me if you do not
know what I am talking about."  I was wrong, and so are you.

> I simply disagree with your last sentence above, and I believe
> beginners _never_ write unit tests anywhere that I've seen, so I
> disagree with the first one too. :)

Beginners almost always test their code, but rarely by writing a
formal test suite --- usually by hand.

I think inspection can find bugs that are very difficult to find by
testing, and vice versa.  That is the reason for my second statement;
a balance of careful reasoning and testing can produce better code
than the same amount of time spent on either alone.



From gh_pythonlist at gmx.de  Mon Jun 17 12:06:22 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Mon, 17 Jun 2002 18:06:22 +0200
Subject: coding python by web?
In-Reply-To: <3eeda89d.0206170317.100549db@posting.google.com>
References: <3eeda89d.0206170317.100549db@posting.google.com>
Message-ID: <20020617160622.GA878@lilith.my-fqdn.de>

* Thinkit <aoeu at go.com> [2002-06-17 04:17 -0700]:
> Is there any way to test out simple python code with only a web
> browser (no input or files, just printing text)?  I'm guessing a
> simple cgi setup could do this, but has it been implemented anywhere?

Hi, "Thinkit",

dunno if it's implemented, yet, but it's not difficult to do. Below I've
put a quick, working draft, replace exec with print to see the code.

The trick is to redirect sys.stdout and use exec.

Things not implemented in the draft:
- using a restricted environment
- handling exceptions gracefully, displaying a traceback

import zlib,base64
exec(zlib.decompress(base64.decodestring("""
eNpdUstq3TAQ3esrplMu2HBjh0IpBNubQKCrLLItFNnWdURkSUjjJCbk3zvy49atVvM4OjNnZr5+
KacYSuM6acpW29LP9Oys0KN3gaAb9BniHHf/iYK2w89HIUiGQRHUgCWDbtLXVnYvvXOh8DMK8Uyj
SWnEKplN1bp+bgRsr7q4MMLKUuMpW6084gFC6p3DSoKVo6oxdkF7QgjuLdb47Rahc4atH98R3oL0
Nb7qQJM02JyyFZzHqtxZGvjLrK2fCGj2iXZqR03YVOUSPdQvU49H/z8JPqhrJTcRf071UrQqF80s
XohFaJ1GWTxoZfonckEOKsvF+pNzCVKw/FdpJpXtQs88vFwIF/TwO1LPFRjKyyhWR7TT5aICx/at
FLuRuK84BqxIIcZ5MK6VJnLs45PdZe+bJ9S76mDrSVu4gs+wA8W/tIfOtmaKqNRLdstde+6EAO+d
JWXpJo36DtIqlsn8srhDljs5wQeuF4B321Gw+uNkOb6V4FX2WX5Nc2I1WIH4A+u46F4=
""")))

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 29.2 ?C      Wind: 3.3 m/s




From maxm at mxm.dk  Tue Jun 25 05:04:31 2002
From: maxm at mxm.dk (Max M)
Date: Tue, 25 Jun 2002 11:04:31 +0200
Subject: ? and %s placeholders, help?
References: <af5oq7$sj1$1@news8.svr.pol.co.uk> <m2k7oolcrm.fsf@video.paradise.lost> <3D178542.604@mxm.dk> <af87cj$n5b$1@newsg3.svr.pol.co.uk>
Message-ID: <3D18321F.6010006@mxm.dk>

Duncan Smith wrote:

>>>>query = 'INSERT INTO %s (%s) VALUES (%s)' % (tblname, string.join(vars,
>>>
> ', '), string.join(['%s']*len(vars), ', '))
> 
>>>>query
>>>
> 'INSERT INTO tbl (var1, var2, var3) VALUES (%s, %s, %s)'
> 
>>>>curs.execute(query % ("'a'", "'b'", "'c'"))
>>>
> 1L
> #Hurray, but does this avoid parsing the statement on each INSERT?


No!


>>>>query = 'INSERT INTO %s (%s) VALUES (%s)' % (tblname, string.join(vars,
>>>
> ', '), string.join(['?']*len(vars), ', '))
> 
>>>>query
>>>
> 'INSERT INTO tbl (var1, var2, var3) VALUES (?, ?, ?)'
> 
>>>>curs.execute(query, ("'a'", "'b'", "'c'"))
>>>
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
>   File "C:\Python22\Lib\site-packages\MySQLdb\cursors.py", line 70, in
> execute
>     raise ProgrammingError, m.args[0]
> ProgrammingError: not all arguments converted


Here is where you misunderstand it. If you use the ? notation the dbi 
will automatically quote and escape as needed. No need to double quote a 
string like "'string'". Treat it as a plain string "string". This is one 
of the greatest features of th dbi.

You won't even have to escape 'weird' characters "this's also a good string"

So this will do instead::

     curs.execute(query, ("a", "b", "c"))


regards Max M



From peter at engcorp.com  Mon Jun 17 22:19:43 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 22:19:43 -0400
Subject: Import caches names (was Re: Problems with importing; must do it twice?)
References: <fgphgus0pv2c7hp8l27uaenjlerr97tdta@4ax.com> <aeat13$pli$1@panix1.panix.com>
Message-ID: <3D0E98BF.B2A9A6C2@engcorp.com>

Aahz wrote:
> 
> Python caches import names even when the import fails
> (protects against circular references)

I thought about this for a while and suppose I can see how it
protects against circular reference problems (I'm guessing it
creates a new module object and stores it in sys.modules 
immediately upon encountering the import statement, and only 
subsequently does it try to import and execute the contents),
but what I don't understand is why it doesn't clean up after
itself in the case of a failed import.  

Should it not be considered a bug that a failed import leaves 
a broken module around?  

Is it possible to redefine this so that it cleanly removes the 
newly created reference if an import fails?  

Or would it then be a problem that circularly imported modules
might now have references back to a module which appears never
to have been imported (because it failed, and was deleted)?

And have I now answered all my questions?  :-)

-Peter


From fredrik at pythonware.com  Wed Jun 26 04:10:23 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 26 Jun 2002 08:10:23 GMT
Subject: Reminder of a python Project.
References: <2218.941T781T12224051threeseas@earthlink.net> <mailman.1025059173.28468.python-list@python.org>
Message-ID: <PFeS8.45133$n4.10657214@newsc.telia.net>

Chris wrote:

> Gah.  I thought I had this guy in my message rules under "Delete from
> Server."
>
> I will remedy this immediately...

too late.  Guido has already added "True" to the
language...

</F>




From shagshag13 at yahoo.fr  Mon Jun  3 10:23:42 2002
From: shagshag13 at yahoo.fr (Shagshag13)
Date: Mon, 3 Jun 2002 16:23:42 +0200
Subject: Efficient posting-list
References: <adffdk$10mep9$1@ID-146704.news.dfncis.de> <3iKK8.99777$%y.10468246@bin4.nnrp.aus1.giganews.com>
Message-ID: <adfu52$10t9pp$1@ID-146704.news.dfncis.de>

> For English text, one would almost certainly reduce the number of text
> keys by deleting some suffixes, so that 'statistic', 'statistics',
> 'statistical', 'statistically', and maybe 'statistician' would be one
> key.  But I have no idea of word structure of non-Indo-European
> languages and what one should do to reduce key number.

I already do stemming via a pos tagger...

> By 'length of a posting list' do you mean number of text keys?  Ie,
> size of key dict?  The length of list attached to each key is at most
> number of documents.

you're right... key dict will be about 500,000 - 600,000, i currently have
140,000 documents but i'll reach 400,000 soon...

> You do not specify 'node object'.

By now i use a node with : 2 int and 1 float. Node objects are handled
in a python list.

> This is crucial since these take up
> most of memory.  (This a good reason not to index the 2000 or so most
> common keys.)

I already index only useful parts and can't *shrink* it more...

> For pure Python, tuple is most space efficient.  If
> each node really is (integer id, integer count), one could devise a
> system using a pair of array (module) objects or 2-dimensional
> numerical python arrays; which store actual integers rather than
> pointers to int PyObjects.  When filled, they would have to be copied
> to larger arrays in much the same manner as done automatically by
> Python lists.

So here you think that i should use only one list (i can't use a tuple, i
need to update values) containing :

[id_x, count_x, float_x, id_x+1, count_x+1, float_x+1, ...]

and so on ?

thanks,

s13.




From cliechti at gmx.net  Fri Jun 28 18:57:35 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 29 Jun 2002 00:57:35 +0200
Subject: Redirect stdout with createfilehandler
References: <3D1CE1FB.9060007@caramail.com>
Message-ID: <Xns923CA3702DD9cliechtigmxnet@62.2.16.82>

Fabien HENON <fabien.henon at caramail.com> wrote in 
news:3D1CE1FB.9060007 at caramail.com:
 
> With the script below I keep getting the error ' maximum recursion depth 
> exceeded'.
...
> 8<-------------snip--------------->8
> import sys,os
> import Tkinter
> from Tkinter import *
> 
> root = Tk()
> 
> T=Text(root)
> T.pack()
> 
> cmd = 'mega +i1.pov +w320 +h240 +dgt +v -f'
> c_in,c_out=os.popen4(cmd)
> 
> def stdout_handler():
>      T.insert('1.0',c_out)
>      root.tk.deletefilehandler(c_out)
>      root.tk.createfilehandler(c_out, Tkinter.READABLE, stdout_handler)
>      root.after(1000,stdout_handler())

don't know muchabout tkinter... but here your CALLING stdout_handler, 
you're sure you want this? from the 'root.after' i would expect that you 
should pass a callable as second parameter -> leave away the paranthereses.

  root.after(1000,stdout_handler)

> root.tk.createfilehandler(c_out, Tkinter.READABLE, stdout_handler)
here you pass a callable (a function object here), like i would expect it 
in the other cases too.

> root.after(1000,stdout_handler())
 same here                      ^^

> root.mainloop()
> 8<--------------snap--------------->8

chris

-- 
Chris <cliechti at gmx.net>



From dradul at yahoo.com  Fri Jun 21 21:40:57 2002
From: dradul at yahoo.com (P. Alejandro Lopez-Valencia)
Date: Fri, 21 Jun 2002 20:40:57 -0500
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <aer2fn$frc$1@peabody.colorado.edu> <3D1214B6.E9BB77D2@noaa.gov> <3D125B79.6090903@skippinet.com.au>
Message-ID: <af0lll$a7ntf$1@ID-99513.news.dfncis.de>

"Mark Hammond" <mhammond at skippinet.com.au> escribi? en el mensaje
news:3D125B79.6090903 at skippinet.com.au...
> Chris Barker wrote:
>
> > On that note: does anyone know if there is a way to turn a Python
script
> > into something that acts like an application on Windows...without
> > resorting to Py2exe and the like. What I want is something just
like:
>
> Add the .py extension to your PATHEXT environment variable, and you
can
> execute "foo.py" at the command prompt.  Like much of Windows, this is
> based on the extension rather than file attributes.

Ehem! :) This only works in WINNT class OSs that use cmd.exe as their
default command line processor. Lowly DOS32 OSs such as the Win98
mentioned by the OP, that only have command.com, just can't.



From benjl at cse.unsw.edu.au  Thu Jun 13 07:35:27 2002
From: benjl at cse.unsw.edu.au (Benno)
Date: 13 Jun 2002 04:35:27 -0700
Subject: How to call a function using apply with keyword args?
References: <ae9fud$u23$00$1@news.t-online.com>
Message-ID: <332c7c11.0206130335.6b97b7aa@posting.google.com>

"Achim Domma" <achim.domma at syynx.de> wrote in message news:<ae9fud$u23$00$1 at news.t-online.com>...
> Hi,
> 
> I have a function like this:
> 
> def MyFkt(A,B,C): pass
> 
> and some parameters from a config file:
> 
> params = {'A':5,'B':8,'C':9}
> 
> how can I call MyFkt with this parameters? I tried apply(MyFkt,[],params)
> but get this error:
> 
> TypeError: MyFkt() takes exactly 3 non-keyword arguments (0 given)

The code looks ok to me. This:

def pants(A, B, C):
    print A, B, C

dikt = {"A":1,"B":2, "C":3}

# We can use 2.1 syntax to avoid calling apply
pants(**dikt)

# Or we can use apply if we want to.
apply(pants, [], dikt)


This works for me, so i'm not sure what you have missed. 

Benno


From kseehof at neuralintegrator.com  Tue Jun 11 19:58:34 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Tue, 11 Jun 2002 16:58:34 -0700
Subject: Sorting list (maybe stupid)
In-Reply-To: <ae5uuo$4atsq$1@ID-71831.news.dfncis.de>
Message-ID: <EFEFIHAINIINNELLHFAIOEBLCJAA.kseehof@neuralintegrator.com>

> Is sort() stable -- that is, is the relative order of elements 
> which compare
> equal preserved? A quick shufti through the ActivePython docs wasn't
> illuminating.
> 
> --
> James Kew
> james.kew at btinternet.com

Documentation?  Real programmers don't read (or write)
documentation. :-)

>>> a = [1,1.0,3.0,3,2.0,2,0,0.0]
>>> a
[1, 1.0, 3.0, 3, 2.0, 2, 0, 0.0]
>>> a.sort()
>>> a
[0, 0.0, 1, 1.0, 2.0, 2, 3.0, 3]

So yes, sort() is stable.

I suppose it's theoretically possible for a different
python implementation to use a different sort algorithm,
but practically speaking, that is very unlikely.

- Ken Seehof






From quinn at upchuck.ugcs.caltech.edu  Thu Jun 27 01:53:29 2002
From: quinn at upchuck.ugcs.caltech.edu (Quinn Dunkan)
Date: 27 Jun 2002 05:53:29 GMT
Subject: Suggestions for good programming practices?
References: <mailman.1024954421.13028.python-list@python.org>
Message-ID: <slrnahla2o.fd3.quinn@upchuck.ugcs.caltech.edu>

On Mon, 24 Jun 2002 16:31:14 -0500, Mark McEahern <marklists at mceahern.com>
wrote:
>How do you choose between these:
>
>	if x is None:
>
>	if x:
>
>The latter is what you'd use if you wanted to work with instances that
>defined __nonzero__ and or __len__.  In other words, you don't really care
>what the type of x is, nor whether it's identity is zero, just whether it
>considers itself nonzero.
>
>Is one more or less polymorphic than the other?  I don't really know.  But
>it sure does sound good.

They both serve their own purposes and you have to know what you want.

For example, I recently tracked down an annoying bug where I said 'if x' when
I meant 'if x is not None'.  Everything worked until x's __len__ returned 0
and it was hard for me to see 'if x' as being buggy code.  I wasted a lot of
time trying to figure out why 'x' was occaisionally None.


From observer at NOSPAM.space.pl  Thu Jun 13 05:05:31 2002
From: observer at NOSPAM.space.pl (Johann)
Date: Thu, 13 Jun 2002 11:05:31 +0200
Subject: How to check version of Python under CGI
References: <rbmggus7ln1uiutktf5libf4bbngu29ji7@4ax.com> <873cvrin1q.fsf@tux.ntw23.fr> <cgnggu40ethjhpqcmltnfb06d5qvuf078k@4ax.com>
Message-ID: <t1ogguoua7b9ne9n3mb3n7rkp8mrqe2nee@4ax.com>

On Thu, 13 Jun 2002 10:57:21 +0200, Johann <observer at NOSPAM.space.pl>
wrote:

>Do you know how to check what modules are installed? E.g. I
>would like to check if PYANA or other xml modules are installed.

for x in dir(__builtins__): print x

Is it all i can get?

--
Johann


From dyoo at hkn.eecs.berkeley.edu  Sat Jun  1 04:00:36 2002
From: dyoo at hkn.eecs.berkeley.edu (Daniel Yoo)
Date: Sat, 1 Jun 2002 08:00:36 +0000 (UTC)
Subject: How to open a HTML file when the python cgi program is executing?
References: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de>
Message-ID: <ad9uv4$g2n$2@agate.berkeley.edu>

Ken <ken at hotmail.com> wrote:
: How do I open a HTML file when the cgi program is executing?

: The only method I can think of, which doesn't work is this:
: <meta HTTP-EQUIV="REFRESH" CONTENT="0; URL=main.html">

Not quite sure I understand yet.


Do you have a file called 'main.html' that you're trying to display
from your CGI program?  I just want to make sure we're work on the
same problem here before babbling.  *grin*


Talk to you later!


From irmen at NOSPAMREMOVETHISxs4all.nl  Tue Jun 11 17:41:11 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Tue, 11 Jun 2002 23:41:11 +0200
Subject: M2Crypto: select() behaves weird on SSL sockets
References: <3D04ED17.5070606@NOSPAMREMOVETHISxs4all.nl> <ae2str$b0q$1@nntp6.u.washington.edu>
Message-ID: <3D066E77.4070901@NOSPAMREMOVETHISxs4all.nl>

Donn Cave wrote:

> I'm no expert on this, but my theory is that you must call
> SSL_pending(self->ssl) to see if any data remains in its buffer.
> Then only when no data remains, you would call select.

I looked around a bit and oh boy, M2Crypto supplies a "pending"
method on SSL sockets, that apparently returns the remaining data.

I now have a workaround; if my socket is an SSL socket, and
there are still bytes pending (socket.pending()>0), I don't
call select. In all other cases it is 'safe' to call select.

My code (Pyro with SSL and timeouts) now works, I'm happy :-)


Thanks for the help!


Irmen de Jong



From tdelaney at avaya.com  Tue Jun 11 00:32:33 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 11 Jun 2002 14:32:33 +1000
Subject: Correct code/unit testing (was Re: Efficient python programmi
	ng...)
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A373@natasha.auslabs.avaya.com>

> From: Peter Hansen [mailto:peter at engcorp.com]
> 
> Kragen Sitaker wrote:
> > 
> > Peter Hansen <peter at engcorp.com> writes:
> > > Have you never spent a large amount of time carefully
> > > designing and coding a wonder little program, and checked it over
> > > very carefully, even with a peer, and later found a bug in it?
> > 
> > Of course!  People are fallible, it's much easier to do a bad code
> > review than a good one.
> 
> I have to confess I have never seen a successful code review, nor
> an environment in which they were used regularly and which had a
> high level of code quality.  Clearly as a result I'm fairly biased
> against them -- I should keep a more open mind on the issue perhaps.

I presume you mean a full, formal review by a group for a large piece of
code?

I'm certain that with your experience with XP, you have seen and taken part
in *many* successful reviews (where success means the code is improved as a
result).

Tim Delaney




From BPettersen at NAREX.com  Wed Jun 26 15:59:56 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Wed, 26 Jun 2002 13:59:56 -0600
Subject: tomorrow in yyyymmdd format
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E201922151B3@admin56.narex.com>

> From: Conrad [mailto:zneptune at zexcite.zcom] 
> 
> Years ago, Nostradamus predicted that on Tue, 18 Jun 2002 
> 20:52:25 -0500,
> Gerhard H?ring would write, saying:
> 
> > * Sean 'Shaleh' Perry <shalehperry at attbi.com> [2002-06-18 
> 18:38 -0700]:
> >> >> 
> >> >> Sure there is :-) time.mktime() can take a "malformed" 
> time tuple
> >> >> and do something sensible with it. You can therefore get a time
> >> >> tuple from gmtime, add one to the day position and call
> >> >> time.mktime() on the result:
> >> >> 
> >> >> >>> x = (2002, 2, 29, 0, 31, 42, 2, 170, 0)
> >> >> >>> time.gmtime(time.mktime(x))
> >> >> (2002, 3, 1, 7, 31, 42, 4, 60, 0)
> >> >> >>>
> >> >> >>>
> >> > Pure luck that this works. Don't complain if it stops workin in
> >> > Python 2.7 and your code breaks ;-)
> >> > 
> >> > 
> >> actually that is a function of the mktime() function in 
> the C library
> >> which the python module wraps.  Based on my reading of a 
> man page or
> >> two this behaviour seems guaranteed by POSIX.  Now, the 
> assumption made
> >> is the mktime() function on YOUR system follows POSIX.
> > 
> > Fine. I've just read the Python documentation which doesn't 
> make this
> > guarantee. I'll leave it to others to check if this works on win32,
> > MacOS, as I'm using mxDateTime for this kind of tasks, anyway.
> > 
> > Gerhard
> 
> 
> Hmm  - here's what I've done - a bit klutzy but it seems to work:
> 
> from time import *
> one_day = 60*60*24                       # number of seconds in day
>    # make a julian for tomorrow -
>    # note: depending on what you want be careful around midnight!
> tomorrow = mktime(localtime()) + one_day
> print 'tomorrow is',localtime(tomorrow)
> print 'today is',localtime()

The problem with this approach is that it doesn't account for DST. This may or may not be a problem for you...

00h15+24-1'ly y'rs
-- bjorn




From mjais at web.de  Thu Jun 13 10:55:20 2002
From: mjais at web.de (Markus Jais)
Date: Thu, 13 Jun 2002 16:55:20 +0200
Subject: Poll: Readable languages
References: <mailman.1023977423.23944.python-list@python.org>
Message-ID: <hn2O8.11$rY6.1476@news.ecrc.de>

Chris Gonnerman wrote:

> On Thu, 13 Jun 2002 09:36:06 -0400 (EDT), Attila Horvath wrote:
> 
>> In my years of experience I have come across VERY readable assembler and
>> very unreadable COBOL, PASCAL, FORTRAN, ADA, C/C++, JAVA, etc. So I
>> suggest that a langauge does not inherently make resultant code readable.
> 
> True, and false.  It is possible to write bad code in any language; but in
> my opinion, any programmer who has mastered Python (takes what, a couple
> of weeks? :-) *tends* to write more readable code than an equivalently
> accomplished programmer in any other language.
> 
> Those of us who have been here a while have learned from experience of the
> ease of writing Python code, and reading it again later.  I have to admit,
> most of the Python I write is little stuff to do jobs I would have done
> with Bourne shell in the past... but I tend to save them for reuse as I
> would not have done with a command-line pipeline.


I can confirm this. when I write Python or Ruby code the programms
tend to be much more readable that with Perl or C/C++.
and definitely much shorter than C/C++ which makes it also more readable

markus


From pyth at devel.trillke.net  Mon Jun 17 06:34:41 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Mon, 17 Jun 2002 12:34:41 +0200
Subject: a possibly foolish question about slices
In-Reply-To: <lkptyql02t.fsf@pc150.maths.bris.ac.uk>; from mwh@python.net on Mon, Jun 17, 2002 at 09:36:09AM +0000
References: <yu99sn3prbqm.fsf@europa.research.att.com> <eppstein-FCC2A4.12082514062002@news.service.uci.edu> <4c877253.0206162149.bff3ae3@posting.google.com> <lkptyql02t.fsf@pc150.maths.bris.ac.uk>
Message-ID: <20020617123441.O15079@prim.han.de>

Michael Hudson wrote:
> garth at deadlybloodyserious.com (Garth T Kidd) writes:
> 
> > > x[p:][:n]
> > 
> > Does the parser take the appropriate short-cut, here,
> 
> No.  Wouldn't be the parser anyway, would it?
> 
> > or do we end up with a temporary list?
> 
> It's very easy to answer questions about whether Python performs this
> sort of optimization: it doesn't.

:-)

[michael probably knows the following, nevertheless]

python's compiler can't possibly optimize unless it somehow had 
additional information about the *name* where an object is 
bound to. As things stand the names 'x','p','n' in the above expression
could point to almost any object with as many sideeffects as imaginable.
More specifically in 'x[p:]' the slice-operator of the object denoted by 'x' 
could have sideeffects which the compiler can't try to guess.

on the flip side you get the powerful dynamisms of python which spares
time in many other respects :-)

seriously though, the above expression is not very nice if 'x' is a
big list. It then executes *much much* slower compared to something like:

def numslice(x, start, maxnum):
    end=start+maxnum
    if start<0 and end>=0:
        return x[start:]
    return x[start:end]

Plus this also makes it possible to let

    numslice(x, -5, 10)

return the last 5 elements.

cheers,

    holger




From rowen at cesmail.net  Fri Jun 28 12:35:05 2002
From: rowen at cesmail.net (Russell E. Owen)
Date: Fri, 28 Jun 2002 09:35:05 -0700
Subject: Help with an algorithm wanted
References: <rowen-CE3C0F.13020226062002@nntp1.u.washington.edu> <7934d084.0206271628.665b03b3@posting.google.com>
Message-ID: <rowen-84A9E1.09350428062002@nntp4.u.washington.edu>

>So effectively you have a conversion graph, where each vertex
>represents a known representation, and each edge represents a
>conversion (presumably with a conversion function as an annotation to
>the edge).
>
>What you are asking for appears to be an algorthm to find the shortest
>path between to representations in the graph.  Fortunately for you,
>Dijkstra provided a solution to this problem a number of years ago,
>and implementing it is a standard algorthms assignment (I do hope
>you're not asking for help with a homework assignment?).
>
>You'll find an implementation in the Python Cookbook @
>http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/119466
>
>The recipie was submitted by David Eppstein.  It dosn't look
>optimised, but that's ok because it is legible (much more important)
>and the algorithm is efficient.

Thank you very much for the pointer. My problem is exactly one of 
converting data to different representations. In my case I am dealing 
with targets for a telescope that may be in FK5, ICRS, Galactic, 
Apparent Geocentric, Apparent Topocentric... coordinates. It's not a 
homework assignment, but is for a telescope control GUI I'm working on. 
The code will be available to others, probably being served at my web 
site <http://www.astro.washington.edu/owen/>. (I serve all my utility 
routines there, but the internals of the user interface I consider not 
of general interest, so I only send that by request. I suspect this code 
will end up in the former category, but I can't be sure until I finish 
modularizing and writing it.)

Finding the shortest path isn't an issue, because the path between any 
two representations will always be unique. Still, anything that can find 
a shortest path can find a unique path. I'll definitely have a look at 
the recipe.

The main thing I'm worried about is once I have a solution, how best to 
implement on object that implements it. I'm afraid that might not make 
much sense, but I hope it's clear from another posting I made with a 
solution to the problem. That solution is definitely not optimised but I 
hope it is legible.

Regards,

-- Russell


From ji at tarzan.it.jyu.fi  Fri Jun 14 04:50:39 2002
From: ji at tarzan.it.jyu.fi (Jonne Itkonen)
Date: Fri, 14 Jun 2002 08:50:39 +0000 (UTC)
Subject: Exheritance and Python
Message-ID: <aecaov$3ji$1@mordred.cc.jyu.fi>

ECOOP 2002 had an interesting workshop on inheritance -- whish
I could have attended. It had a particularly interesting paper
about exheritance -- well, I'm a bit tempted to read Mr Sakkinen's
papers :)

http://www.cs.jyu.fi/~sakkinen/inhws/papers/Sakkinen.ps

Exheritance is like inheritance, but other way around. As inheritance
specialises, exheritance generalises. If you, my humble reader, can't
think of any use for exheritance, please feel free to skip the rest
of the message.

My first idea after reading the paper was 'well, it should be possible
to exherit in Python', and after a bit of testing, I found out that
Python actually can do exheritance better than I thought...for old
type classes... :( (Old type class is class which does not inherit 
'object'.)

So, for an old type class C one can say:
    C.__bases__=C.__bases__+(A,)
and after that, even subclasses of C find methods of A.

(If you read Sakkinen's paper, hold on. I'm now just considering
using exheritance as 'interface exheritance' or even bare 'type
inheritance', that is, I'm not yet trying to exherit any methods
or attributes.)

For new type classes, those that inherit object, this doesn't work,
as __bases__ is read-only. Why? (Answering my own questions, PEP252
tells that "The introspection API is a read-only API", but continues
"A future PEP may define some semantics for some such assignments".
It seems I'll be to writing a PEP. PEP253 also has an innocent open
issue "assignment to __dict__, __bases__".)

I found an article posted here some time ago, which suggested using
type() to re-define class, like:
       C=type(C.__name__, C.__bases__+(A,), C.__dict__)
but for two reasons, this does not work:
    1. C.__dict__ is a proxy to dict, not a dict object.
    2. If there are classes using old class C, they hold
       references to that class, but new classes inherited 
       from C refer to new C. This is clear, type() doesn't 
       change the old class, it creates a new one 'overrides'
       the name of the old class.

So, is there a way to change the __bases__-attribute of a new type
class? Am I to write a PEP? Is anyone interested contributing? Or am
I just to use the old type of classes?

   Jonne


From ajw126NO at SPAMyork.ac.uk  Sat Jun 29 16:18:08 2002
From: ajw126NO at SPAMyork.ac.uk (Andrew Wilkinson)
Date: Sat, 29 Jun 2002 21:18:08 +0100
Subject: SSE With Swig
Message-ID: <uhs5g8hos5og51@corp.supernews.com>

Hi,

I'm trying to develop a Vector class that uses Intel's SSE extension. It
works very well when called using c or c++, but when I wrap the class with
Swig I get a privileged instruction exception error (this is under Windows
2000 using MSVC++ 6sp5 with the processor pack installed).

I've included the code below, any suggestions would be greatly appreciated
Andrew Wilkinson

swig.i
%module Vector
%{
#include "Vector.h"
%}
class Vector
{
public:
 Vector();
 virtual ~Vector();
 Vector* operator+(const Vector *other);
 float x;
 float y;
 float z;
 float w;
};
__
vector.h
class Vector
{
public:
 Vector();
 virtual ~Vector();

 Vector* operator+(const Vector *other);

 __declspec(align(16)) union {
  __m128 v;
  struct {
   float x;
   float y;
   float z;
   float w;
  };
 };
};
---
vector.cpp
#include "Vector.h"
#include <xmmintrin.h>

Vector::Vector()
{
 x = 0.0f; y = 0.0f; z = 0.0f; w = 0.0f;
}

Vector::~Vector()
{

}

Vector* Vector::operator +(const Vector *other)
{
 Vector* r = new Vector;

 r->v = _mm_add_ps(v, (other->v)); // The exception occurs here.

 return r;
}

--
---
Ditch The Decimal System!
Lets Use Hex - http://www.intuitor.com/hex/switch.html




From ngiven at lanl.gov  Wed Jun 26 17:56:56 2002
From: ngiven at lanl.gov (Nathan Given)
Date: Wed, 26 Jun 2002 15:56:56 -0600
Subject: Integrals with Python
References: <3D188551.6BA00DED@lanl.gov> <3D196758.3040604@kfunigraz.ac.at>
Message-ID: <3D1A38A8.9BD2CA47@lanl.gov>

> If you have got measurements only and want to numerically integrate that
> you can get a small Simpson function from me (a few lines of python
> code). Drop me a note. I use it heavily in combination with measurements
> (and the precision is okay; you know measurement values...).

that would be great...

you can email it to me...

thanks
--
nathan


From shalehperry at attbi.com  Sat Jun  8 01:20:21 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Fri, 07 Jun 2002 22:20:21 -0700 (PDT)
Subject: Efficient python programming...
In-Reply-To: <3D018306.D01A8C9B@engcorp.com>
Message-ID: <XFMail.20020607222021.shalehperry@attbi.com>

> 
> Automated testing of GUIs is a very underdeveloped area of software
> development, but it's a frontier that almost nobody is bothering to
> explore.  If more people check it out I think we'll make better 
> progress developing it.
> 

the problem is that extra level adds more code, more bloat and is one more
thing for new hackers to learn when they approach the project.

Yes I know tight coupling is usually bad.  But sometimes it is the best way to
solve the problem at hand.

and now I have walked this sufficiently off topic to bore most of the readers
of this list.




From hesterloli at hotmail.com  Tue Jun 18 16:48:14 2002
From: hesterloli at hotmail.com (George Hester)
Date: Tue, 18 Jun 2002 20:48:14 GMT
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com><aegfck$6tkre$1@ID-114079.news.dfncis.de><wEPO8.16066$uk2.8447171@twister.nyroc.rr.com><3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com><gYgP8.42356$GY.12361719@twister.nyroc.rr.com><aekcug$7js8n$1@ID-114079.news.dfncis.de><9YBP8.44418$GY.13865843@twister.nyroc.rr.com> <mailman.1024423525.24359.python-list@python.org>
Message-ID: <i0NP8.46300$GY.14481287@twister.nyroc.rr.com>

Bah humbug.

--
George Hester
_________________________________
"Cliff Wells" <logiplexsoftware at earthlink.net> wrote in message
news:mailman.1024423525.24359.python-list at python.org...
> On Tue, 18 Jun 2002 08:12:53 GMT
> George Hester wrote:
> > I think you guys don't how to do it.
>
> I think you wouldn't know an answer if it bit you:
>
> [You]
> What is a simple script that will tell me the version of python I am currently
> running?  IIS 5.0 Windows 2000.
>
> [Jeremy Yallop]
> Look up sys.version and sys.version_info.
>
> [You]
> Traceback (innermost last): File "<Script Block >", line 5, in ?
> Response.Write(sys.version) NameError: name 'sys' is not defined
>
> [Gerhard]
> You did not import the sys module.
>
> [You]
> I think you guys don't how to do it.
>
> [Duncan Booth, doing your job for you]
> Try this:
> <html>
> <head>
>   <title>Python Version</title>
> </head
> <body>
> <SCRIPT Language="Python" RUNAT=Server>
> import sys
> Response.Write(sys.version)
> </SCRIPT>
> </body>
> </html>
>
> Perhaps you'd like someone to come out and type it for you as well.
>
> Please remember that the people on this list aren't being paid to help you
> write your program.  They expect you to do at least /a little/ of the research
> yourself.  It's understandable if you don't know where to look, but when
> someone points you to the relevant portion of the docs, you should at least
> read it before asking again.  It's clear from your postings that you haven't
> even read the tutorial.
> If you can't do that then you should pay someone to write your program for
you.
>
> --
> Cliff Wells, Software Engineer
> Logiplex Corporation (www.logiplex.net)
> (503) 978-6726 x308  (800) 735-0555 x308
>
>




From donn at u.washington.edu  Mon Jun 24 19:04:45 2002
From: donn at u.washington.edu (Donn Cave)
Date: 24 Jun 2002 23:04:45 GMT
Subject: AF_UNIX + SOCK_DGRAM
References: <aes1he$gv2$1@news.idiom.com> <aet3ai$1l0o$2@nntp6.u.washington.edu> <af84af$1dkt$1@news.idiom.com>
Message-ID: <af88id$8oi$1@nntp6.u.washington.edu>

Quoth "James T. Dennis" <jadestar at idiom.com>:
[... re sockaddr_un->path fix ]
|  So, should this be reported to the Python dev list?  If so, how?
|  Can we get it fixed by 2.3?  How would my code work if it was fixed?
|  What would/should be returned for this case?

If Chris Liechti submits it to Sourceforge as he proposed to do,
that should eventually lead to a fix in the distributed version,
assuming everyone concerned agrees with the solution.  I don't
know where they are with 2.3 (heck, I've never even seen 2.2 yet),
but would think this kind of bug fix ought to be acceptable up
through alpha testing.

I can't guess at what your code does in any interesting detail.
Functions that return an address should return something that
will work as an address.  In the UNIX domain, that means a
string representing the file (path).

	Donn Cave, donn at u.washington.edu


From kragen at pobox.com  Sat Jun  1 15:52:50 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 01 Jun 2002 15:52:50 -0400
Subject: Which is fastest dict or list.index() ?
References: <adaih2$vtb8c$1@ID-146704.news.dfncis.de>
Message-ID: <83vg92vkrx.fsf@panacea.canonical.org>

"Shagshag13" <shagshag13 at yahoo.fr> writes:
> My question is which is the best way (best = quickest and memory safe) to
> handle this ?
> Should i subclasse from dict or list ?

I don't think so.

> Should i consider using only one list with toons (and use their index as id)
> ?

Yes, I think that's what you should do.  If you have a lot of toons,
that may be too slow, but unless you can quantitatively estimate the
amount of time list.index will take (and it's too much), try the
simpler method first, and then optimize by adding the dict if
necessary.

> Should i use 2 dicts ?

I don't think so.



From gerhard.haering at gmx.de  Fri Jun 21 19:15:14 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 22 Jun 2002 01:15:14 +0200
Subject: TimeoutQueue.py
In-Reply-To: <LNBBLJKPBEHFEDALKOLCGEDJPPAA.tim.one@comcast.net>
References: <Xns9234F249D7D85cliechtigmxnet@62.2.16.82> <LNBBLJKPBEHFEDALKOLCGEDJPPAA.tim.one@comcast.net>
Message-ID: <20020621231514.GA873@lilith.my-fqdn.de>

* Tim Peters <tim.one at comcast.net> [2002-06-21 18:32 -0400]:
> [Chris Liechti]
> > Anyone ever needed a Queue with timeout?
> 
> Yes.  I encourage you to submit a patch to SourceForge, [...]

I like it. It would certainly be a useful addition.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From quinn at upchuck.ugcs.caltech.edu  Thu Jun 27 01:11:06 2002
From: quinn at upchuck.ugcs.caltech.edu (Quinn Dunkan)
Date: 27 Jun 2002 05:11:06 GMT
Subject: Questiion on list
References: <pan.2002.06.27.14.36.03.151659.5728@linuxmail.org>
Message-ID: <slrnahl7j8.fd3.quinn@upchuck.ugcs.caltech.edu>

On Thu, 27 Jun 2002 14:36:03 +0000, SiverFish <occeanlinux at linuxmail.org> wrote:
>I got the list of words now i want to calculate how many word in that
>list which is the same word (insensitive case),can anyone show me how to do
>it for example ls = ['abc','cd','abc','adf',abc','dwc','cd']
>it will show abc = 3 ,cd = 2, adf = 1, dwc=1

import string
d = {}
for e in map(string.lower, ls):
    d[e] = d.get(e, 0) + 1
freqs = [ (count, w) for w, count in d.items() ]
freqs.sort()
freqs.reverse()
print ', '.join([ '%s = %d' %(count, w) for count, w in freqs ])


From safarzadeh at hotsheet.com  Thu Jun 13 00:21:09 2002
From: safarzadeh at hotsheet.com (H. Safarzadeh)
Date: 12 Jun 2002 21:21:09 -0700
Subject: Python GUI
Message-ID: <9ddd154e.0206122021.1b06f89a@posting.google.com>

Hi all,
I know that this subject is discussed in many other places, but I don't
think it is harmful to talk about it more!

 I want to what library I should use to make GUI for my Python apps?
 I have examined many of them, but I couldn't find what I need. Here is
the result of some of my examinations:

-Tkinter: A nice simple lib, but really simple! It dosn't have enough
widgets, and I do not like its look in addition!
-wxPython: Really perfect. It has many widgets, and do many works(not
only in GUI). But it is toooo big!(both on disk and in memory). I need a
smaller one.
-pyGTK and pyQt: I do not know enough about them. But I know that I do
not like GTK look on Windows and Qt for Windows has some problems with
its lisence(and it is tooo big, too!)
-Jython/Swing: ....it's Jython, after all!
-Win32 extensions: It is not portable!!!

Any comments? Other possibilities?

Thanks.


From maxm at mxm.dk  Mon Jun 24 16:46:58 2002
From: maxm at mxm.dk (Max M)
Date: Mon, 24 Jun 2002 22:46:58 +0200
Subject: ? and %s placeholders, help?
References: <af5oq7$sj1$1@news8.svr.pol.co.uk> <m2k7oolcrm.fsf@video.paradise.lost>
Message-ID: <3D178542.604@mxm.dk>

John Hunter wrote:


> As far as I know, the '?' syntax is used in the perl DBI but nowhere
> in the python MySQLdb.  So forget about that one.


It is also part of the Python DBI spec.


> To start with a simple example from the docs:
> 
> c.execute("""SELECT spam, eggs, sausage FROM breakfast
>             WHERE price < %s""", (max_price,))


c.execute("""SELECT spam, eggs, sausage FROM breakfast
             WHERE price < ?""", (max_price,))

Should work just as well.

> # use python string format to build the mysqldb query string
> # query == 'INSERT INTO mytable (var1,var2,var3) VALUES %s,%s,%s'
> query =  "INSERT INTO %s (%s) VALUES %s" % \
> 	(tbl, string.join(vars, ','), valsfmt) 

Or with the '?' as a placeholder

query =  "INSERT INTO %s (%s) VALUES %s" % \
	(tbl, string.join(vars, ','), ','.join(len(vars)*['?']))

Generally you would use the %s to insert tablenames, variablenames etc. 
And you will use ? to insert the variables.

you can do:

execute('select firstName, lastName from employess where age > 42')

or

execute('select firstName, lastName from employess where age > ?', (42,))

or

qs = 'select %s from employees where age > ?' % ('firstName, lastName', 
'employees')
execute(qs, (42,))


regards Max M



From aahz at pythoncraft.com  Mon Jun 17 16:34:46 2002
From: aahz at pythoncraft.com (Aahz)
Date: 17 Jun 2002 16:34:46 -0400
Subject: '__builtin__' vs. __builtins__
References: <Pine.GSO.4.33.0206171543060.8570-100000@techst02.technion.ac.il>
Message-ID: <aelh56$aq1$1@panix1.panix.com>

In article <Pine.GSO.4.33.0206171543060.8570-100000 at techst02.technion.ac.il>,
Beni Cherniavksy  <cben at tx.technion.ac.il> wrote:
>
>Why is the '__builtin__' module available under the __builtins__ name?
>By what's the logic one is supposed to remember it?  (If I do remember
>that it's different I would still confuse which is what...)

My paraphrase of what Tim Peters said the last time this came up:

    Main reason is to avoid getting flooded with output when calling
    ``vars()`` in interactive mode.

(From my upcoming OSCON tutorial.)
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From sarmstrong13 at mac.com  Thu Jun 13 13:36:17 2002
From: sarmstrong13 at mac.com (SA)
Date: Thu, 13 Jun 2002 12:36:17 -0500
Subject: Translating Tcl/Tk code into Python/Tkinter help please...
Message-ID: <B92E4241.6F6B%sarmstrong13@mac.com>

Hi Everyone-

    I am trying to translate a tcl program that use tk into python. I seem
to be going along fine until I get to the process that wants to "catch" the
results onto one text screen. Here is what I am looking at:

proc python {} {
    catch {exec /sw/bin/python -c [.t1 get 0.0 end]} output
    .t2 delete 0.0 end
    .t2 insert end $output
}

Basically, this process catches the output from running the code in the text
widget .t1 through python -c and places it into the variable ouput.
The text widget .t2 is then cleared and the contents of the variable $output
are displayed in the text widget .t2.

I can translate all of the widgets into python/tkinter code, but I'm not
sure about how to translate this process. Is there a function similar to
"catch" in Python?

Thanks.
SA



From BPettersen at NAREX.com  Wed Jun 12 14:35:35 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Wed, 12 Jun 2002 12:35:35 -0600
Subject: Can Distutils include .pyd files directly?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019221519D@admin56.narex.com>

> From: Thomas Heller [mailto:theller at python.net] 
> 
> "Bjorn Pettersen" <BPettersen at NAREX.com> wrote in message 
> news:mailman.1023837255.23228.python-list at python.org...
> > I've built my extension modules using the MS Devstudio IDE 
> and now I 
> > would like to use Distutils to create a Windows installer 
> for me that 
> > puts both the debug and release versions of the extension in the 
> > python22\DLLs directory. Is this possible? If not, is there 
> a way to 
> > have Distutils build my extension in both debug and release 
> mode and 
> > install both?
> >
> > -- bjorn
> 
> Sure. The 'build' command builds a 'pseudo installation tree' 
> in the directory build\lib.win32-2.2, all of this will later 
> be installed by the 'install' command. So you can
>   python setup.py build
> to build the release version,
>   python setup.py build -g
> to build the debug version, and

Is there any way I can skip having Distutils building the extension? If
there isn't, is there a way to provide a separate 'library_dirs' option
for debug libraries? (We have our release libraries in \active\release
and the debug libraries in \active\debug, and they have the same
names...)

>   python setup.py install
> will install all of this into Python22\lib\site-packages.
>   python setup.py bdist_wininst
> will create a windows installer containing the debug _and_
> the release version (everything that's in the build directory).
> 
> Thomas

-- bjorn




From gerhard.haering at gmx.de  Thu Jun 20 08:15:06 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Thu, 20 Jun 2002 14:15:06 +0200
Subject: module string.py
In-Reply-To: <3d11bd6a.11177520@news.polito.it>
References: <3d11bd6a.11177520@news.polito.it>
Message-ID: <20020620121505.GA1043@lilith.my-fqdn.de>

Hi "Nop",

* NOPollution <r_proietti at yahoo.com> [2002-06-20 11:38 +0000]:
> I have two trouble:
> 
> 1)  looking at the module "string.py" I have seen that are described
> some functions like "lower", "rjust", etc. Ther problem is that it
> seems not to describe exactly how this functions are made, but there
> is only a comment about their properties. So in which file are these
> functions described?

The behaviour of these functions is described in the documentation, for
example in http://www.python.org/doc/current/lib/module-string.html:

lower(s)
Return a copy of s, but with upper case letters converted to lower case. 

In this particular case, the docstring of the function has just as much
info:

>>> string.lower.__doc__
'lower(s) -> string\n\n    Return a copy of the string s converted to
lowercase.\n\n    '

If you're looking at additional info for _how_ exactactly these
functions are implemented, then the source code is the only info there
is. Btw. in current Python versions the string module just delegetages
to the string methods, which are the preferred option. So you could just
as well use.

> 2) I tried to cancel the file "string.py". After that, the Python
> Interpreter doesn't work anymore (it can't even start working). Why?

You're not supposed to delete modules from the standard library.  If you
don't want your Python to fail in obscure ways, you'd better get that
file back into place. Btw.  my Python 2.1.3 and 2.2.1 starts up fine
even without the string module.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From deckerben at freenet.de  Sun Jun 23 17:37:27 2002
From: deckerben at freenet.de (deckerben)
Date: Sun, 23 Jun 2002 23:37:27 +0200
Subject: Fork
References: <3d15c022$0$26988$9b622d9e@news.freenet.de> <3D15C638.EE5797F9@phekda.freeserve.co.uk>
Message-ID: <3d16405b$0$166$9b622d9e@news.freenet.de>

> Look in the libc docs:
>
>     info libc alpha fork

Thanks for the reference, it shed more light on the matter than the FAQ did.
I'm 100% sure now that its the fork that makes distutil builds using GCC
"run out of memory" :-)
I started working on on of the python library scripts that defines the make
sequence so that it no longer calls python funcion 'os.fork()'. It is
_possible_ that the next release of Python will support distutil builds
_to_a_limited_degree_.

For those who are interested, here is a new feature list of *currently*
running modules that exist in the latest port that were not present in the
previous test. Next release date is still undetermined.

socketmodule (using watterloo library - DOS packet driver is needed.)
expat module - James Clark's XML parser inside a Python wrapper
libxml wrap- the Python wrapper for the GNOME 'LibXML' library, latest CVS
release
Mathematical/Scientific visualization using NUMERIC (complete) and the
DISLIN display driver
Runar Tenfjord's Python adaptation of  Mark Spink's NURBS toolbox has been
incorperated as an example of the dislin/numpy facilities.
Others: termios, selectmodule, syslogmodule, parsermodule

The fate of cursesmodule depends on the hopefuly successful implementation
of the DXE2 dynamic linking library, as many symbols it defines/requires
interfere with DISLIN.

Ben






From fschaef at ces.clemson.edu  Wed Jun 19 17:18:44 2002
From: fschaef at ces.clemson.edu (Frank Schaefer)
Date: 19 Jun 2002 14:18:44 -0700
Subject: file .seek() under MS-Windows
Message-ID: <4169349c.0206191318.7f0d99d9@posting.google.com>

Hi,

The following code:


fh = open("myfile.pot", "r")


...
fh.seek(-1,1)
...

produces an error:
[IOError 22] Invalid Argument

under MS Windows (XP). It works fine under Unix, though (for years).

I appriciate any help.

Thanks,

Frank.


From campbells4 at ozemail.com.au  Fri Jun 28 07:56:09 2002
From: campbells4 at ozemail.com.au (Alistair Campbell)
Date: Fri, 28 Jun 2002 21:56:09 +1000
Subject: [NEWBIE]- Syntax error
Message-ID: <a7YS8.727$yY2.23825@ozemail.com.au>

Hi,
I imagine the answer to this post is very simple ... but I'm not proud.

Using Python2.2 on Windows OS

Anyway, when I try to run a script fom the command line in IDLE or Python

>>>python name_of_script.py

the interpreter complains that there is a syntax error and highlights the
"name_of _script" part.

I imagine that I have not initialised some PATH variable or other. I guess
that i need to put something in the autoexec.bat. Can anyone suggest what?

Alistair Campbell
Brisbane, Qld, Australia








From have.a.look at the-body.com  Tue Jun 18 18:25:16 2002
From: have.a.look at the-body.com (Giulio Cespuglio)
Date: Tue, 18 Jun 2002 22:25:16 +0000 (UTC)
Subject: tomorrow in yyyymmdd format
Message-ID: <qpbvguscmblone2r92o8sfc7evo5709mq8@4ax.com>

Hello everybody,

We all know that if I want to get the present GMT date in yyyymmdd
format I should say

	print time.strftime("%Y%m%d", time.gmtime() )

Now, what about tomorrow? Note that a simple string manipulation is
not the solution I'm looking for, as tomorrow might be next month,
next year, the 29th of February and such.
Do you confirm that there's nothing better than the following
silly-looking expression?

	time.strftime("%Y%m%d", time.gmtime(time.time() + 60*60*24))

Thanks,
Giulio
_______________________________________________

To reply, please replace ".pizza." with "."
giulio.pizza.agostini(at)libero.pizza.it


From TuxTrax at fortress.tuxnet.net  Wed Jun  5 06:31:19 2002
From: TuxTrax at fortress.tuxnet.net (TuxTrax)
Date: Wed, 05 Jun 2002 10:31:19 -0000
Subject: Compiling Python
Message-ID: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet>

Hello all

I am new to this forum, and to Python, please excuse the newbie question.

I have started reading the o'reilly book, "learning python" and find python
to be an amazing language so far. My first impression of it is a language that
is easier than basic to learn and use, and more powerful than many other
high level languages such as pascal.

My question is this; I realize that Python compiles to bytecode automatically
as part of the interpreters tokenization run, but is there a way to permanently
compile python code outputting to an executable binary?

Cheers,

Mathew

-- 
TuxTrax   (n.) An invincible, all terrain, Linux driven armored assault
vehicle that can take as much fire as it gives ;-)

<Carnivore activation string> 
ASSASINATION ANTHRAX PRESIDENT NUCLEAR TALIBAN AMMONIUM NITRATE
</Carnivore activation string>

Yes, I am a Penguin cult high priest. Fin readings upon request.
ROT13 this email address to mail me: uvtuqrfregzna at lnubb.pbz



From mwh at python.net  Wed Jun 12 09:59:45 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 12 Jun 2002 13:59:45 GMT
Subject: Zope Newbie query
References: <ugefre2980kj09@corp.supernews.com>
Message-ID: <lkbsagy4w6.fsf@pc150.maths.bris.ac.uk>

"Tony B" <tony at sentienc.co.yuk> writes:

> Apologies if this is an innaproprate newsgroup - it's the closest I could
> find.

http://lists.zope.org/mailman/listinfo/zope

is probably the best place to ask this question.

> I have been playing with zope and would like to use it properly now the
> problem I'm encountering is: How do I remotely install zope on a raq4 on
> which I am not the main raq4 administrator (i.e I have a subdirectory on
> which my site is hosted).? Do I have to have root priveliges to install and
> run zope?

Probably not.  Don't know any more than that though :)

Cheers,
M.

-- 
  Whaaat? That is the most retarded thing I have seen since, oh,
  yesterday                             -- Kaz Kylheku, comp.lang.lisp


From js_nntp at nospam_blackrocksoftware.com  Sat Jun  8 17:52:02 2002
From: js_nntp at nospam_blackrocksoftware.com (J Swartz)
Date: Sat, 08 Jun 2002 21:52:02 GMT
Subject: Comparison: HTTP with Java, Perl or Python
References: <3D009911.9000904@gmx.de> <h99M8.11172$p56.3361428@newsb.telia.net> <3D0127D2.2060604@gmx.de> <B9273737.4982%js_nntp@nospam_blackrocksoftware.com> <3D022634.9050304@gmx.de>
Message-ID: <B927CA91.49D3%js_nntp@nospam_blackrocksoftware.com>

This one:

public class URLTest {
    public static void main(String[] arguments) {
        String myurl = "http://www.uni-dortmund.de";
        try {
            java.io.BufferedReader reader = new java.io.BufferedReader
( new java.io.InputStreamReader((new java.net.URL(myurl)).openStream()) );
            for (String curLine = reader.readLine(); curLine != null;
curLine = reader.readLine())
                System.out.println( curLine );
            reader.close();
        } catch (java.io.IOException ex) { System.out.println(ex); }
    }
}

is 11 lines (text is wrapped here to make 13), took about 4 minutes to write
and test. It was quite straightforward.

BTW the Java 1.4 API is online here:
http://java.sun.com/j2se/1.4/docs/api/index.html

- JS



On 6/8/02 8:43 AM, in article 3D022634.9050304 at gmx.de, "Ingo Linkweiler"
<i.linkweiler at gmx.de> wrote:
> Thanks a lot for your versions.
> Which of ALL solutions is the shortest, and which the best ?
> Do you know good ways to measure the size? (LOC, irredundant or packed
> bytes....)
> But even without this:
> 
> The Python version is quite short, even after adding exception handling.
> TCL seems to be same.
> The Perl version seems to be the shortest, but can you read it without
> knowing Perl?
> 
> Can you estimate the time you needed for writing it?
> Did you have any difficulties?
> My Python version needed 2 Minutes, with Java I needed 20, because I did
> not unterstand the f******* manual.
> 
> 
> 
> 
> PS:
> This was my own Java solution:
> 
> package meintest;
> import java.lang.Exception;
> import java.net.*;
> import java.io.*;
> public class getURL {
> public static void main(String[] args) {
>   try {
>     URL url = new URL("http://www.uni-dortmund.de");
>     InputStream in = url.openStream();
>     for (;;)
>     {
>         int data = in.read();
>         if (data == -1)
>             break;
>         else
>             System.out.print ( (char) data);
>     }
>   }
>   catch (Exception e) { }
> }
> }
> 



From donn at drizzle.com  Thu Jun 13 01:13:28 2002
From: donn at drizzle.com (Donn Cave)
Date: Thu, 13 Jun 2002 05:13:28 -0000
Subject: socket module htonl/ntohl bug
References: <f2f390f9.0206121750.1fa384b2@posting.google.com>
Message-ID: <1023945194.166778@yasure>

Quoth davidma at eskimo.com (David Margrave):

| The following statement should work, but does not, on python 1.5.2 and
| python 2.2:
|
| >>> socket.htonl(pow(2L,32)-1)
| Traceback (innermost last):
|   File "<stdin>", line 1, in ?
| OverflowError: long int too long to convert
|
| 2^32-1 is a valid 32-bit value, equal to 32 1s in binary, and htonl
| should be able to handle it.  This is not related to the argument
| being a long int and htonl not accepting that type, because
| socket.htonl(4294967295) raises the same exception, and 4294967295 can
| be represented as a regular int in python2 (but not 1.5).

I think you may understand this, but 2^31-1 is not a valid 32-bit
signed integer, and that's what Python's int is (on platforms where
C int is 32 bits.)  Since there is no unsigned int, you would have
to represent the value as a negative number (v - pow(2L,32).)
Arguably the htonl function might convert to unsigned int, to match
the C variable, but it's going to return a signed int anyway, so
there's an element of consistency here.

On a DEC Alpha, incidentally, even though Python supports an int
of that size, I get
  >>> x = socket.htonl(y)
  Traceback (most recent call last):
    File "<stdin>", line 1, in ?
  OverflowError: signed integer is greater than maximum

while I get "long int too long to convert" with pow(2L, 64) - 1.
So in this particular context it looks like there's more going on
here than just an accidental collision with integer overflow.

	Donn Cave, donn at drizzle.com


From NineOfSix at gmx.de  Fri Jun 21 12:19:58 2002
From: NineOfSix at gmx.de (Axel Grune)
Date: Fri, 21 Jun 2002 18:19:58 +0200
Subject: enlightenment (Re: python cgi + IIS)
References: <3D0F1813.3010304@gmx.de> <x0NP8.165989$Fp1.91529@atlpnn01.usenetserver.com> <3D102AF2.5010204@gmx.de> <u4rfyc4gj.fsf@online.no>
Message-ID: <3D13522E.6010703@gmx.de>

> for, and so forth) give an account of all the steps you have performed

This steps: http://www.e-coli.net/pyiis.html

This doc tells you to change the 'App Mapping' in the properties of 
[IISAdmin] to enable the execution of python files for the entire server 
but for any reason it doesn't work.
In the end our admin suggested to change the properties of 'Default Web 
Site' (the node above [IISAdmin]) and then it works.

Thanks,
Axel


-- 

please check /dev/zero for further information



From mwh at python.net  Mon Jun 17 05:36:09 2002
From: mwh at python.net (Michael Hudson)
Date: Mon, 17 Jun 2002 09:36:09 GMT
Subject: a possibly foolish question about slices
References: <yu99sn3prbqm.fsf@europa.research.att.com> <eppstein-FCC2A4.12082514062002@news.service.uci.edu> <4c877253.0206162149.bff3ae3@posting.google.com>
Message-ID: <lkptyql02t.fsf@pc150.maths.bris.ac.uk>

garth at deadlybloodyserious.com (Garth T Kidd) writes:

> > x[p:][:n]
> 
> Does the parser take the appropriate short-cut, here,

No.  Wouldn't be the parser anyway, would it?

> or do we end up with a temporary list?

It's very easy to answer questions about whether Python performs this
sort of optimization: it doesn't.

Cheers,
M.

-- 
  I also feel it essential to note, [...], that Description Logics,
  non-Monotonic Logics, Default Logics and Circumscription Logics 
  can all collectively go suck a cow. Thank you.
              -- http://advogato.org/person/Johnath/diary.html?start=4


From mwh at python.net  Fri Jun  7 11:19:01 2002
From: mwh at python.net (Michael Hudson)
Date: Fri, 7 Jun 2002 15:19:01 GMT
Subject: Native code compiler for python?
References: <baf24bd4.0206070550.84f1376@posting.google.com>
Message-ID: <lkk7pbktg8.fsf@pc150.maths.bris.ac.uk>

d2002xx at softhome.net (nobody) writes:

> Is there any existing native code compiler for python? Or Python to C
> translator? Or someone is doing this?

http://www.google.com/search?q=psyco

Cheers,
M.

-- 
  Make this IDLE version 0.8.  (We have to skip 0.7 because that 
  was a CNRI release in a corner of the basement of a government 
  building on a planet circling Aldebaran.)
                             -- Guido Van Rossum, in a checkin comment


From salmonia at cf.ac.uk  Wed Jun 19 09:53:35 2002
From: salmonia at cf.ac.uk (Al)
Date: Wed, 19 Jun 2002 14:53:35 +0100
Subject: Newbie deluxe
References: <hdDO8.29984$Hn4.969427@news1.east.cox.net>
Message-ID: <3D108CDF.3010706@cf.ac.uk>

Yes Python can be used for graphical applications. I am writing one just 
now: http://salstat.sunsite.dk - it's for statistical analysis (a bit 
like a smaller, simpler version of SPSS). It has a spreadsheet like grid 
for data entry, the output is in html (in its own window, not a separate 
browser), and there is hypertext help, so I suppose fully fledged 
graphical applications can be developed.

I used wxPython for the GUI things (it's a separate download, but costs 
nothing) - this means my application now runs on Windows, Linux, Unix 
and MacIntosh OSX - without a single change of line code, something that 
I didn't find possible when I used Kylix and Delphi. The problem with 
wxPython is that you will probably need to learn how to program it - 
there are GUI designers if you want (Boa Constructor and PythonCard 
being 2 that spring to mind), but if you want the power, you will be 
best off learning how to write the code by hand. It really isn't that 
difficult once you get going - if I can do it, I am sure you can! ;)

To summarise - since using Python, I really haven't looked back once!

Alan.

btw - I found its functional language constructs to be superb and very 
time saving.

Zach wrote:
> I've just started taking a look at Python (experienced in Delphi) and so far
> it looks very clean and comprehensible.  My question is, what are the
> fundamental differences and similarities between it and Perl?  I hear so
> much about these two languages, especially in Linux circles and just
> wondered what sets them apart.  Can full fledged graphical applications be
> built with Python or does it lean more towards scripting?  I'm aware that it
> has OOP capabilities so I assume it reaches beyond just scripting but like I
> said, I've no experience with it.  I would like to broaden my coding skills
> with the addition of another language so any general opinions with regards
> to its usefulness are appreciated.
> 
> TIA.
> 
> Zach
> 
> 




From mcherm at destiny.com  Fri Jun 14 09:08:38 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Fri, 14 Jun 2002 09:08:38 -0400
Subject: Medium to Large Scale Python Deployments
Message-ID: <3D09EAD6.10701@destiny.com>

> I'm working on a large web project whose name I hafta jump thru
> permission hoops to reveal

No problem... if it's OK with you, I'll list it with your name but no 
project name. Will that satisfy the hoopmasters?

-- Michael Chermside







From phd at phd.pp.ru  Fri Jun 21 17:24:52 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Sat, 22 Jun 2002 01:24:52 +0400
Subject: do...while
In-Reply-To: <3D13841D.F3C9EB5B@noaa.gov>; from Chris.Barker@noaa.gov on Fri, Jun 21, 2002 at 12:53:02PM -0700
References: <mailman.1024613916.27915.python-list@python.org> <3D13841D.F3C9EB5B@noaa.gov>
Message-ID: <20020622012452.A24417@phd.pp.ru>

On Fri, Jun 21, 2002 at 12:53:02PM -0700, Chris Barker wrote:
> while line = file.readline():

   In Python it is spelled

for line in file.xreadlines():
   ...

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From logiplexsoftware at earthlink.net  Fri Jun 21 01:09:04 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: 20 Jun 2002 22:09:04 -0700
Subject: Inhibiting duplicate apps in Windows?
In-Reply-To: <3D1283EC.A28A45CB@engcorp.com>
References: <3D11F9E8.DFFFC8E4@tds.net>
	<Xns9233F02DB217Ccliechtigmxnet@62.2.16.82>  <3D1283EC.A28A45CB@engcorp.com>
Message-ID: <1024636145.10524.2.camel@localhost.localdomain>

On Thu, 2002-06-20 at 18:39, Peter Hansen wrote:
> Chris Liechti wrote:
> > 
> > "Edward K. Ream" <edream at tds.net> wrote in news:3D11F9E8.DFFFC8E4 at tds.net:
> > > What I would like is for the second copy of Leo to detect that another
> > > copy of leo.py is already running, send a message to the first copy and
> > > then exit. I wonder if anyone knows how this might be done?
> 
> > dumb solution: create a "lock file" and remove it on programm exit. the
> > other problem is that you have to delete it manually when the app crashes
> > (i deleted may lock files with older Netscapes ;-)
> 
> I wonder whether there's not a simple way to avoid manual deletion.
> If one application creates the lock file, it can hold it open as 
> long as it is running.  A second invocation can detect the first
> by its failure to open the same file for writing.  If the file
> already exists, it can try to delete it with os.remove(). That
> should fail if the first application still has it open, but if 
> the first application actually crashed, the OS should allow the
> file to be removed, at which point the new invocation can safely
> recreate it, knowing it now has the lock.

A fairly fool-proof/portable method is to open a socket. If the app is
run a second time it will fail on the attempt.  Even if the app crashes,
the socket will be closed by the OS (although it may take a couple of
seconds on occasion).

Regards,

Cliff







From gh_pythonlist at gmx.de  Tue Jun 11 16:54:23 2002
From: gh_pythonlist at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Tue, 11 Jun 2002 22:54:23 +0200
Subject: RfD: Automagically making a class instance 'synchronized' (draft impl.)
Message-ID: <20020611205423.GA28113@lilith.my-fqdn.de>

For one of my projects, I'll need something not quite unlike the
following code. It's not tested, yet, but it looks ok and complete to
me. I thought that it could be useful to other developers who need to
use multithreading, so I'm posting it here. Maybe something like this
would even be a useful addition to the Standard Library?

Gerhard


from threading import RLock

class Connection:
    def begin(self):
        print "begin"

    def commit(self):
        print "commit"

class LockingWrapper:
    def __init__(self, proxy, obj):
        self.proxy = proxy
        self.obj = obj

    def __call__(self, *args, **kwargs):
        self.proxy.__dict__["rlock"].acquire(1)
        try:
            self.obj(*args, **kwargs)
        except:
            self.proxy.__dict__["rlock"].release()
            raise
            return
        self.proxy.__dict__["rlock"].release()

class LockingProxy:
    """LockingProxy proxies an instance of a class, so that in that instance,
    only one method can be active at any given point in time. In Java-speak, it
    automatically makes all methods of this class 'synchronized'."""
    def __init__(self, obj, methodnames):
        """obj -> a callable
           methodnames -> a list of method names"""
        self.__dict__["obj"] = obj
        self.__dict__["rlock"] = RLock()
        self.__dict__["methodnames"] = methodnames

    def __getattr__(self, attr):
        val = getattr(self.__dict__["obj"], attr)
        if attr not in self.__dict__["methodnames"]:
            return val
        else:
            return LockingWrapper(self, val)

conn = Connection()
lp = LockingProxy(conn, ["begin", "commit"])
lp.begin()
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 16.6 ?C      Wind: 0.6 m/s




From quinn at vomit.ugcs.caltech.edu  Sun Jun 23 22:00:17 2002
From: quinn at vomit.ugcs.caltech.edu (Quinn Dunkan)
Date: 24 Jun 2002 02:00:17 GMT
Subject: re.rsearch?
References: <slrnahcg6u.382.quinn@baht.ugcs.caltech.edu> <ZGrR8.44834$n4.10519185@newsc.telia.net>
Message-ID: <slrnahcv9h.r27.quinn@vomit.ugcs.caltech.edu>

On Sun, 23 Jun 2002 22:10:01 GMT, Fredrik Lundh <fredrik at pythonware.com> wrote:
>Quinn Dunkan wrote:
>> Occaisionally I find myself wanting an re.search that will find the rightmost
>> match.
>
>just prepend ".*" to the pattern...

Oh, I should have thought of that :)  Of course, that changes the match, but
I can get my span with ()s and m.span(1)

>> Unfortunately, AFAIK, re.search also lacks a 'start_searching_at'
>> type option.
>
>if you use a compiled pattern, you can pass in start
>and end offsets:
>
>    p = re.compile(pattern)
>    p.search(string, start, end)

Phoey... don't know how I missed that one in the docs.  Guess I looked at
the modules search() and assumed the method was the same...

thanks!


From jepler at unpythonic.net  Fri Jun  7 11:17:52 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Fri, 7 Jun 2002 10:17:52 -0500
Subject: positional modifiers in python?
In-Reply-To: <20020607143021.GA3038@foof.i3.cz>
References: <20020607143021.GA3038@foof.i3.cz>
Message-ID: <20020607151751.GB12501@unpythonic.net>

On Fri, Jun 07, 2002 at 04:30:21PM +0200, fuf wrote:
>  hello everyone,
> 
>  does python support positional parameters in the print method? ie.
>  something like:
>     print "there %2$s something %1$s" % ("there", "is")

print "there %(verb)s something %(where)s" % {'verb': 'is', 'where': 'there'}

Jeff




From huaiyu at gauss.almadan.ibm.com  Tue Jun 18 15:02:39 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Tue, 18 Jun 2002 19:02:39 +0000 (UTC)
Subject: What If.....    Strong Types
References: <3D0C0F0D.4030900@bgb.cc>
Message-ID: <slrnagv0ue.46u.huaiyu@gauss.almadan.ibm.com>

Don Garrett <garrett at bgb.cc> wrote:
>   I'm not suggesting any real changes to any, only proposing a thought 
>experiment.
>
>   What if Python had been built from the start with strong types? And I mean 
>strong types, not strong typing. The distinction I am trying to draw is that 
>variables would not be typed, but that all types would have rigid interfaces.
>
>   Primarily, what if classes always had rigidly defined interfaces. I mean 
>that public members had to be declared to exist, and that methods couldn't be 
>modified on an instance. Private members could exist but would be really private.
>
>   All types would be classes (including ints and such). Amoung other things, 
>I would add the concepts of Interfaces, instanceof operators.
>
>   Local variables would not require declaration, and would work just like 
>today. Introspection would work, but be read-only.
>
>   My belief is that almost all the convenience of Python would be maintained, 
>but that compilation, optimization, and automated code analysis would be simpler.
>
>   I'm just wondering if the change would be good or bad. Would it really 
>break anything important in Python today? Would it prevent any common errors? 
>Would it help with automated code analysis (compilation and optimization 
>included) as much as I think it would?

What you are talking about is not strong types.  Python already has strong
types.  You are talking about interfaces, which is quite different from
classes.

Classes define a hierarchy of implementation, so that attributes of base
classes can be used in derived classes.

Interfaces define a hierarchy of requirements, so that a sub-"type" can be
used in places that require a base-"type".

They cannot be combined easily because the two hierarchies do not share the
same inheritance structure.

Example 1: You might want to define a class of file-like objects that shares
no implementation with builtin files.  You want just the interface.

Example 2: Mathematically a square is a kind of rectangle.  So anywhere a
rectangle is required, a square should do.  But for graphical classes it's
often better to code Rectangle as a subclass of Square, because it uses most
of the methods of Square, in addition to its own.  This is implementation.

Example 3: It is often necessary to require that an object is one of several
types, AFTER the types have been defined.  For example, you might code up a
Time class, whose __init__ accepts numbers (as seconds), strings (in the
common time format), and tuples (as hours, minutes and seconds).  If you
want to label all these as TimeData, it cannot be a base class for strings,
numbers and tuples, as these have already been defined.  

The relations of interfaces are often built from specific to general, in the
opposite direction of class inheritances.

We can conclude from these considerations that mixing the implementation
hierarchy with the requirement hierarchy is not a good general solution,
although it is sometime useful.

Java's interfaces and C++'s abstract classes re-uses the same inheritance
structure of implementation for interfaces.   Such interfaces are not as
useful as it could have been, while creating some additional problems.

To do it right, interfaces must have an inheritance hierarchy that is
separate from that of implementation.  It is possible to re-use the existing
class hierarchy for this purpose if done carefully in special cases, such as
Python's exception classes.  But a general type-hierarchy (serving as
interfaces) has to be separate from class-heirarchy.

Will it be beneficial to Python?  At least, it will allow many formal
correctness guarantees.  It will help compiling.  It will reduce much of
current type checking against implementational classes.  It will reduce the
computational burden of much of run time type checking.

How difficult would it be?  Well, it has to be optional.  It has to have an
inheritance syntax that is not confused with classes.  There need to be
syntax to specify requirements (such as with function arguments).  There
need to be syntax to specify invariance assertions.  The checking should
best be done at C-level, possibly in the compile stage, for the sake of
efficiency.  Last but not least, there need to be a good name beside the
word 'type', which is becoming similar to 'class' in Python now.

I think that once it is decided what it should look like, there are many
capable people who can actually implement it.

Huaiyu


From phd at phd.pp.ru  Fri Jun 28 04:52:51 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 28 Jun 2002 12:52:51 +0400
Subject: Python crashed on me
In-Reply-To: <afftgs$5p7$0@216.39.172.122>; from bokr@oz.net on Thu, Jun 27, 2002 at 08:45:16PM +0000
References: <mailman.1025196737.29740.python-list@python.org> <afftgs$5p7$0@216.39.172.122>
Message-ID: <20020628125251.A27385@phd.pp.ru>

Hi!

On Thu, Jun 27, 2002 at 08:45:16PM +0000, Bengt Richter wrote:
> >   Recently Python (one program that I am dbugging) started to crash.
> >FreeBSD kills it with "Bus error", Linux with "Segmentation fault".
> >
> >   I think the program crashed in the cPickle.dump(file, 1)

   I replaced cPikle.dump with pikle.dump and got infinite rcursion. The
traceback is below.

   What's that? Are there any limits that an object to be pikled must
follow? Could it be a tree with loops? (I am pretty sure it could - I used
the program for years, and data structures was not changed much). Could it
be "new" Python class? (Recently I changed one of my classes to be derived
from builtin list instead of UserList).

   Well (or not so well), the traceback:

Traceback (most recent call last):
  File "/home/phd/lib/bookmarks_db/check_urls.py", line 158, in ?
    run()
  File "/home/phd/lib/bookmarks_db/check_urls.py", line 145, in run
    storage.store(root_folder)
  File "bkmk_stpickle.py", line 23, in store
  File "/usr/local/lib/python2.2/pickle.py", line 973, in dump
    Pickler(file, bin).dump(object)
  File "/usr/local/lib/python2.2/pickle.py", line 115, in dump
    self.save(object)
  File "/usr/local/lib/python2.2/pickle.py", line 219, in save
    self.save_reduce(callable, arg_tup, state)
  File "/usr/local/lib/python2.2/pickle.py", line 245, in save_reduce
    save(arg_tup)
  File "/usr/local/lib/python2.2/pickle.py", line 225, in save
    f(self, object)
  File "/usr/local/lib/python2.2/pickle.py", line 374, in save_tuple
    save(element)
  File "/usr/local/lib/python2.2/pickle.py", line 225, in save
    f(self, object)

[about 1000 lines skipped - they are all the same]

  File "/usr/local/lib/python2.2/pickle.py", line 498, in save_inst
    save(stuff)
  File "/usr/local/lib/python2.2/pickle.py", line 225, in save
    f(self, object)
  File "/usr/local/lib/python2.2/pickle.py", line 447, in save_dict
    save(value)
  File "/usr/local/lib/python2.2/pickle.py", line 219, in save
    self.save_reduce(callable, arg_tup, state)
  File "/usr/local/lib/python2.2/pickle.py", line 245, in save_reduce
    save(arg_tup)
  File "/usr/local/lib/python2.2/pickle.py", line 225, in save
    f(self, object)
  File "/usr/local/lib/python2.2/pickle.py", line 374, in save_tuple
    save(element)
  File "/usr/local/lib/python2.2/pickle.py", line 225, in save
    f(self, object)
  File "/usr/local/lib/python2.2/pickle.py", line 414, in save_list
    save(element)
  File "/usr/local/lib/python2.2/pickle.py", line 143, in save
    pid = self.persistent_id(object)
RuntimeError: maximum recursion depth exceeded

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From christian.aastorp.killspam at siemens.no  Thu Jun  6 08:24:15 2002
From: christian.aastorp.killspam at siemens.no (Christian Aastorp)
Date: Thu, 06 Jun 2002 14:24:15 +0200
Subject: Interfacing to 1-Wire LAN, especially Dallas temperature sensors?
References: <a7iefuk650noamrqcmpvroqj60mtgbpf9v@4ax.com> <1bf96e7f.0206060356.417a7394@posting.google.com>
Message-ID: <fqkufuoljmafu6si319bhmt3f83pr587uc@4ax.com>

On 6 Jun 2002 04:56:02 -0700, python at tykebsd.net (Dave Moor) wrote:

>Hi
>
>I have just started producing a PyOneWire module it currently can
>search the one wire network for families of devices and read the
>temperatures from DS1820 devices.  What platform are you wanting to
>run it on, I could send you a prerelease versiob to try.

That is great!
I'm running on Windows 2000, and XP.
Remove the obvious from my mail adress.


From pde at cs.mu.oz.au  Sat Jun 29 10:15:08 2002
From: pde at cs.mu.oz.au (Peter Eckersley)
Date: Sun, 30 Jun 2002 00:15:08 +1000
Subject: Partial Parsing of Python
Message-ID: <20020629141507.GA6507@luminance.reworld.org>

Hi,

I'm currently working on a python-based alternative to a standard *nix
shell.  I imagine lots of other people have built such animals, but this
one is supposed to be graphical and include nifty features like smart
auto-completion and running commentary on one's commands (which should be
handy for newbies).

Anyway, having implemented some of the completion features the simple 
way-- using tokenize.tokenize(), string analysis, and the odd call to
parser.expr() to guess what the user might want to type next, I've
reached the conclusion that doing too much more will lead to Madness
and Spaghetti Code.

So the alternative is of course to use a proper python parser, which 
can include hooks suggesting how bits of the grammar can be completed.
Basically, what this parser needs to do is attempt to accept its input,
noting along the way when it meets the user's cursor.  If it fails,
it needs to preserve an incomplete parse tree, record which production
rule it was trying to use, and which symbol in that rule caused the
problem.

It's not immediately obvious how to achieve this with a minimum of fuss.
A few different strategies might be:

* hack the cPython parser to provide this extra stuff (not having emersed
  myself in the parser, I imagine this could be a lot of work, but would
  love to hear information to the contrary, or receive few pointers as
  to how it might be done)

* use bison to implement a new parser from scratch (ugly,
  perilous, and I'd have to learn yacc/bison along the way :( )

* write a minimalist, brute-force recursive descent parser in python,
  which would be a more-or-less direct translation of the grammar, but
  would almost certainly be damnably slow

Suggestions about which strategy to pursue would be gratefully
appreciated!

(CCs off-list, IINTMT)

-- 
Peter Eckersley
Department of Computer Science   &                  mailto:pde at cs.mu.oz.au 
IP Research Institute of Australia             http://www.cs.mu.oz.au/~pde
The University of Melbourne               




From gnpowell at powerup.com.au  Sat Jun  1 23:27:54 2002
From: gnpowell at powerup.com.au (Daniel Powell)
Date: Sun, 2 Jun 2002 13:27:54 +1000
Subject: system variables
Message-ID: <WhgK8.3670$3t6.160209@ozemail.com.au>

hi guys,
I'm new to python and I'm teaching myself from a website. I'm stumped on one
thing this is the exact wording

"You will need to cause the directory containing the file named python.exe
to be listed in your system environment variable named path."
the website is
http://softwaredev.earthweb.com/sdopen/article/0,,12077_626311,00.html
it says when you do this if you type something like
"python junk.py" it will run the "junk.py" scipt
if anyone can help me do this i would grealty appreciate it.
My OS is Win 98SE

thanks in advance
Daniel Powell
thecubical at hotmail.com




From Grant_member at newsguy.com  Mon Jun  3 13:52:31 2002
From: Grant_member at newsguy.com (Grant Griffin)
Date: 3 Jun 2002 10:52:31 -0700
Subject: getting windows processes list
Message-ID: <adgacv01l41@drn.newsguy.com>

Hi Gang,

Does anybody know offhand how to use Python to get the list of processes that
are currently running on Windows?  (I'm essentially trying to get the
information that appears in Window's Task Manager's "Processes" tab.)

not-gifted-at-finding-needles-in-haystacks-ly y'rs,

=g2

_________________________________________________________________________

Grant R. Griffin                                           g2 at dspguru.com
Publisher of dspGuru                               http://www.dspguru.com
Iowegian International Corporation                http://www.iowegian.com



From dom at edgereport.put_a_c_o_m_here  Fri Jun  7 16:45:36 2002
From: dom at edgereport.put_a_c_o_m_here (Domenic R. Merenda)
Date: Fri, 07 Jun 2002 20:45:36 GMT
Subject: Medium to Large Scale Python Deployments
References: <mailman.1023460091.14067.python-list@python.org>
Message-ID: <QX8M8.21882$EP.2548532334@newssvr10.news.prodigy.com>

"Michael Chermside" <mcherm at destiny.com> wrote in message
news:mailman.1023460091.14067.python-list at python.org...
> Folks, I think this is a pretty useful thing to collect.

> In fact, it might be really nice to collect a list of large (successful)
> Python projects and post it somewhere so the Python community could use
> it for advocacy.

   Michael, I agree with you.  I was formerly the Vice President of Business
Development for BeOpen.com, while we had PythonLabs there, and I can tell
you that we were compiling such a list for that very reason.  Using the
WayBack Machine, I was able to grab the list.  Here ya go:


Digital Creations
http://web.archive.org/web/20001119052700/http://www.digicool.com/
Digital Creations built its flagship product, Zope, in Python. Zope is an
advanced, highly adaptable web application server and content management
system in production use across the globe. Zope enables web developers and
content managers to rapidly deploy interactive, dynamic, scalable,
community-oriented web sites. Zope is available at
http://web.archive.org/web/20001119052700/http://www.zope.org/.


 eGroups (acquired by Yahoo!)
http://web.archive.org/web/20001119052700/http://www.egroups.com/
eGroups empowers people to share interests and ideas by delivering the
world's best web-based email group service, a technology built in Python.
Email groups facilitate group communication through the ease and convenience
of a user's email account. Python has allowed eGroups to rapidly evolve its
service offering as well as enable it to scale to unprecedented levels.

eGroups decision to use Python helped them achieve rapid growth, leading to
their acquisition by another market leader, Yahoo!.


 Google.com
http://web.archive.org/web/20001119052700/http://www.google.com/
Google's mission is to organize the world's online information and to make
it universally accessible and useful.

Google rocketed to the top of the web portal industry by making extensive
use of Python throughout its web crawler and search engine technologies
(later optimized for further performance in C). Google's web crawler now
spans the entire public Internet, with over 1 billion indexed URL's. The
total number of searches served by Google exceeds 13 million per day.


 Helix Code
http://web.archive.org/web/20001119052700/http://www.helixcode.com/
Helix Code is a startup building a free Internet desktop based on the GNOME
desktop platform, for Linux and Unix. The company is headed by the leading
devevelopers of the GNOME project. Helix Code ships a variety of Python
programming interfaces for the GNOME libraries and its associated graphical
widget set GTK+.


 Hewlett Packard e-speak
http://web.archive.org/web/20001119052700/http://www.e-speak.net/
HP used Python to build a truly open Market Making Broker with XML/HTTP
interfaces at all major interface points. This broker was far ahead of its
time, working across a large space of market makers. Initially, Python was
used only to produce the first prototype, with plans for a production
version to be done in Java. After the Java-intensive product proved less
stable with even fewer features than the Python prototype, Java work was
discontinued and the Python prototype was reworked in order to meet critical
deadlines.


 IBM
http://web.archive.org/web/20001119052700/http://www.ibm.com/
IBM uses Python on OS/390 UNIX System Services to facilitate automated
testing of Internet tools.

IBM is using Python to create the business practice logic for the factory
tool control applications that manage a semiconductor plant. Code that had
been previously written in an internal, proprietary scripting language has
been given new life via its port to Python.


 Infoseek (acquired by Go Network)
http://web.archive.org/web/20001119052700/http://www.infoseek.com/
Ultraseek Server, Infoseek's commercial search engine product, is
implemented as an elaborate multi-threaded Python program, with the
primitive indexing and search operations performed by a built-in module.
Most of the program is written in Python, and both a built-in spider and
HTTP server can be customized with additional Python code. The program
contains over 11,000 lines of Python code while the user interface contains
over 17,000 lines of Python-scripted HTML templates.


 ITI
http://web.archive.org/web/20001119052700/http://www.iti-oh.com/pdi
ITI's Product Data Interoperability (PDI) business provides software tools
and services enabling companies to effectively exchange, reuse, and share
CAD/CAM/CAE/PDM and related engineering data among diverse software
applications used throughout a manufacturing enterprise and it's supply
chain. ITI has had amazing success with Python.


 Lawrence Livermore National Laboratories
Los Alamos National Laboratory
http://web.archive.org/web/20001119052700/http://www.llnl.gov/
http://web.archive.org/web/20001119052700/http://www.lanl.gov/
A group at the Lawrence Livermore National Laboratories based its new
numerical engineering environment, CDAT, on Python, replacing a home-grown
scripting language of ten-year standing. CDAT used Python's exceptionally
versatile design to integrate C, C++, and FORTRAN subroutines, affording
climate research scientists an effortless analysis of global climate model
data sets.


 MCI Worldcom
http://web.archive.org/web/20001119052700/http://www.mciworldcom.com/
The Data Subnet Manager, MCI Worldcom's internal frame relay configuration
and provisioning suite, uses JPython for regression testing, interactive
debugging, and rapid prototyping of its Java Client. They also use Python in
their automated build processes and CGI scripts.


 Microsoft eShop
http://web.archive.org/web/20001119052700/http://www.eshop.com/
eShop, Inc. was originally formed in December 1993 to produce desktop
electronic commerce software. By November 1995, Pandora's Box had been
opened -- the Internet was something that had to be addressed directly,
rather than simply accounted for, and eShop had to move quickly or lose its
competitive advantages. By the time eShop was acquired by Microsoft, less
than a half-dozen engineers had built an entire Web-based shopping system
with Python in about four months. The resulting server was stable,
maintainable, robust, and scalable showing no compromise in quality for such
a brief development period.


 NASA
http://web.archive.org/web/20001119052700/http://www.nasa.gov/
Johnson Space Center uses Python as the standard scripting language in its
Integrated Planning System. Efforts are underway to develop a modular
collection of tools for assisting shuttle pre-mission planning and replacing
older tools written in PERL and shell dialects. Python will also be
installed in the new Mission Control Center to perform auxiliary processing
integrated with a user interface shell. Ongoing developments include an
automated grammar based system, whereby C++ libraries may be interfaced
directly to Python via compiler techniques. This technology can be extended
to other languages in the future.


 Penguin Computing
http://web.archive.org/web/20001119052700/http://www.penguincomputing.com/
Penguin Computing is a leading manufacturer of Linux servers for web,
database, and other enterprise uses. Much of the software developed at
Penguin for internal use and for shipping with their servers is developed in
Python -- after other languages proved too limiting and difficult to
maintain.


 PyBiz
http://web.archive.org/web/20001119052700/http://www.pybiz.com/
Python enabled a small team of senior architects at PyBiz to produce and
deliver both XDisect and eContentMgr on short schedule. It was estimated
that it would have required 3 times the build time and 1.5 times the number
of developers to deliver comparable functionality in Java.

Building in Python allowed delivery of a working product early in the life
cycle, providing more time to evolve the product towards customers' needs as
well as optimize performance of critical areas.


 Real Networks
http://web.archive.org/web/20001119052700/http://www.real.com/
RealNetworks has developed a Python binding for their RealMedia client. It
is used extensively in load testing and feature testing of both their server
and client on all of their supported platforms. Additionally, RealNetworks'
build system and bug tracking system have very significant Python
components.


 Red Hat
http://web.archive.org/web/20001119052700/http://www.redhat.com/
In the highly competitive Linux market, easy-to-use and flexible
installation technologies have become a key business differentiator. Red
Hat, a leader in Linux and Open Source technologies, identified Python as a
unique and powerful tool for building their installer, Anaconda.


 Sapient
http://web.archive.org/web/20001119052700/http://www.shn.net/
Sapient Information Systems built the Sapient Health Network for WebMD with
Python. The service provides chronically and terminally ill people with
information related to their illness. All of the backend web programming
uses Python, as do a large number of internal tools.


 Sonics
http://web.archive.org/web/20001119052700/http://www.sonicsinc.com/
Sonics engineers use Python exclusively to construct graphical and
configuration tools (the FastForward Development Environment) which provide
the layout of micro-network-based system-on-chips. Their experiences over
the years with many languages and systems led them to conclude that, with
the available time and resources, they could never have developed the
technology without Python.


 Yahoo! Mail and Four11
http://web.archive.org/web/20001119052700/http://www.yahoo.com/
Yahoo! acquired a company called Four11, a technology providing both an
Internet White Pages and a web-based e-mail reading system (renamed from
Rocketmail to Yahoo! Mail). Both of these services were written in Python to
obtain strong performance and flexibility. The Internet White Pages holds a
large number of email address records and the complete US residential
telephone directory -- so searching performance is key. Having acquired both
411 and eGroups, Yahoo! now relies on a wide range of Python-driven
technologies to provide the highly interactive experience its users expect.


 80-20 Software
http://web.archive.org/web/20001119052700/http://www.80-20.com/
80-20 Software used Python to rapidly develop a scalable and extensible
document management system that sits on top of the popular Microsoft
Exchange messaging technology. By using Python, they were able to develop
and release to market a feature rich client/server solution in under a year
with only 4 developers. Python provided an excellent prototyping tool for
the server technology, which originally was supposed to be converted to C++
for production use. 80-20 found during its testing that only a few
components needed to be in C++ for greater speed and that more than 90% of
the prototype code could be used unchanged without any significant
performance impact.


> Alternatively, if such a list already exists, let me know where, and
> I'll volunteer to help update and promote it.
>
> Domenic's
>    ERP system   >100      --

   Let's not call it that. :-)  And I will check upstairs when we get back
from our vacation to see if it's okay to release the number of lines in the
code.  Don't be surprised if it's treated like a National Security issue.
(Shh!  If they know how many lines are in the code, they can CRACK US!)


--
Domenic R. Merenda
Editor, The Edge Report
http://www.edgereport.com




From graham_guttocks at yahoo.co.nz  Tue Jun  4 14:54:13 2002
From: graham_guttocks at yahoo.co.nz (Graham Guttocks)
Date: Tue, 4 Jun 2002 11:54:13 -0700 (PDT)
Subject: os.path.expanduser ignores os.seteuid
In-Reply-To: <XFMail.20020604113833.shalehperry@attbi.com>
Message-ID: <20020604185413.26385.qmail@web10303.mail.yahoo.com>

Sean 'Shaleh' Perry <shalehperry at attbi.com> wrote:
>
> It uses short circuit eval for the '~' case and simply
> reads the environment $HOME.  If you could pass '~user/' 
> it would give you the answer you want.

Well, I also need to support lots of existing ~/foo cases.

Looks like I can just set $HOME after the os.seteuid() call
with the following:

os.environ['HOME'] = pwd.getpwuid(os.geteuid())[5]

Then subsequent os.path.expanduser calls should work as I want.

Regards,
Graham

__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com




From hesterloli at hotmail.com  Sat Jun 15 18:23:38 2002
From: hesterloli at hotmail.com (George Hester)
Date: Sat, 15 Jun 2002 22:23:38 GMT
Subject: python version?
Message-ID: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com>

What is a simple script that will tell me the version of python I am currently
running?  IIS 5.0 Windows 2000.

--
George Hester
_________________________________




From jeremy at jdyallop.freeserve.co.uk  Wed Jun 19 06:34:34 2002
From: jeremy at jdyallop.freeserve.co.uk (Jeremy Yallop)
Date: 19 Jun 2002 10:34:34 GMT
Subject: zlib/gzip + anydbm or ...
References: <aeoufa$rkk$1@news.idiom.com>
Message-ID: <aepmnq$90217$1@ID-114079.news.dfncis.de>

* James T. Dennis
| Is there an obvious and simple way to use the existing standard
| libraries to support transparently compressed dbm files?
| 
| (Any combination of zlib/gzip etc with any of the dbm files would
| be considered).

I've done something similar to this before using dbhash/gdbm and zlib.

The code was something like:

import dbhash, zlib, tempfile

   if file_exists(filename):
       dbfile = open(filename)
       tmpfilename = tempfile.mktemp()
       temporary_file = open(tmpfilename, 'w')
       
       temporary_file.write(zlib.decompress(dbfile.read()))
       temporary_file.flush()
       dbfile.close()

       db = dbhash.open(tmpfilename, mode)

with code to recompress the file when the database is closed.

It worked fairly well, but I was only using it for fairly small files
(20 megabytes or less).

If you'd like the full code (around 180 lines) contact me by email.

Jeremy.



From fcorneti at libero.it  Tue Jun 11 07:35:30 2002
From: fcorneti at libero.it (Fabio Corneti)
Date: Tue, 11 Jun 2002 11:35:30 GMT
Subject: Defining a method final
Message-ID: <6glN8.35376$86.884025@twister1.libero.it>

Is there any way to define a method final in Python?
I've searched through the docs and through the web, but
it seems there's no way to to such a thing using standard
language statements. Some suggestions?

-- 
Fabio Corneti
Network Administrator
fcorneti AT libero DOT it


From aleax at aleax.it  Mon Jun 24 13:10:01 2002
From: aleax at aleax.it (Alex Martelli)
Date: Mon, 24 Jun 2002 19:10:01 +0200
Subject: [Python-Help] File Copying
In-Reply-To: <1024934961.9bd8aff8rchopra@myrealbox.com>
References: <1024934961.9bd8aff8rchopra@myrealbox.com>
Message-ID: <E17MXMM-0002me-00@mail.python.org>

On Monday 24 June 2002 06:09 pm, Rajat Chopra wrote:
> How do you copy a file from one directory to another. I have looked all
> over and haven't found any documentation on it. Essentially, I want to copy
> a file from $SRC_DIRECTORY to $TARGET_DIRECTORY.

Python does not support variable names starting with $, so I can only 
guess at what you mean.  Maybe you want environment variables: you'll
find those in dictionary os.environ (obviously you need to import os).

You don't tell us what is the filename you want to copy; assuming you
have it in a string, you can join that string to each directory name with
os.path.join.  Finally, shutil.copyfile does the copy for you (you need to
import shutil).  Also look at shutil.copy, shutil.copy2, etc etc, depending
on exactly what you want to do.

Please don't post to both python-help AND python-list.  Choose one or the
other.  python-list has a lot more traffic; python-help is for those cases in
which you'd rather NOT let everybody in python-list know about your
problems.  Thanks for your cooperation on this issue.

Alex




From whisper at oz.net  Tue Jun 11 09:52:27 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 11 Jun 2002 06:52:27 -0700
Subject: Defining a method final
In-Reply-To: <6glN8.35376$86.884025@twister1.libero.it>
Message-ID: <GCEDKONBLEFPPADDJCOEGECFDIAA.whisper@oz.net>

"final" is a Java concept not supported in Python.

You could possibly simulate it with an 'init' method called after an
instance of a subclass is created to adjust the dictionary so that the base
classes' method is always called in preference to any same-named method in
the derived class(es). Potentially this could even be done in the derived
classes' __init__ method that calls the base class __init__ and thus not
need an additional explict initialization step.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Fabio Corneti
> Sent: Tuesday, June 11, 2002 4:36
> To: python-list at python.org
> Subject: Defining a method final
>
>
> Is there any way to define a method final in Python?
> I've searched through the docs and through the web, but
> it seems there's no way to to such a thing using standard
> language statements. Some suggestions?
>
> --
> Fabio Corneti
> Network Administrator
> fcorneti AT libero DOT it
> --
> http://mail.python.org/mailman/listinfo/python-list





From marklists at mceahern.com  Thu Jun 13 12:00:24 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 13 Jun 2002 11:00:24 -0500
Subject: modifying string objects
In-Reply-To: <MPG.1772db35d89f7b9c989685@news.rz.uni-karlsruhe.de>
Message-ID: <NCBBLFCOHHDIKCAFGCFBIEHHLEAA.marklists@mceahern.com>

Short answer:  Stick the string in something that is mutable:

s = "foobar"

def change_it(x):
    x[0] = x[0] + "foo"

x = [s]
change_it(x)
s = x[0]

print s

// m
-





From sholden at holdenweb.com  Tue Jun 25 13:26:35 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 25 Jun 2002 13:26:35 -0400
Subject: HTML Error from Python CGI Script
References: <79b54d28.0206241618.6c78dbde@posting.google.com> <slrnahh3d3.sar.jon+usenet@snowy.squish.net>
Message-ID: <nI1S8.23098$mF.7005@atlpnn01.usenetserver.com>

"Jon Ribbens" <jon+usenet at unequivocal.co.uk> wrote ...
> In article <79b54d28.0206241618.6c78dbde at posting.google.com>, Chris wrote:
> > I copy this exact text to my own computer (versus operating over
> > network) and open it and it looks fine. Exactly what I need. However,
> > when I put a link to it on another webpage on the network, an internal
> > server error occurs and complains about the following:
> >
> > syntax error at line 2: `newline or ;' unexpected
>
> Obviously your server is trying to use your static page as some kind
> of script. You haven't given enough information to work out why this
> might be though.

Look at your server's configuration, and see whether the extension you are
using for the temporary files is treated as executable on the server (in
Apache you'll be looking at httpd.conf or .htaccess files).

Since you don't say whether the error appears on the client or the server, I
suppose it's also possible you are delivering content that the client-side
(i.e. the machine you run your browser on) is trying to execute your
content.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From xah at xahlee.org  Sun Jun 16 07:58:23 2002
From: xah at xahlee.org (Xah Lee)
Date: 16 Jun 2002 04:58:23 -0700
Subject: python install through fink, mac os x
References: <7fe97cc4.0206142251.164d9ec5@posting.google.com> <mailman.1024175660.17023.python-list@python.org>
Message-ID: <7fe97cc4.0206160358.6dbd1696@posting.google.com>

problem solved by doing a
fink selfupdate
first.

 Xah
 xah at xahlee.org
 http://xahlee.org/PageTwo_dir/more.html


> > [xah at localhost ~][Fri Jun 14,23:48:52]
> > fink install python
> > sudo /sw/bin/fink install python
> > Password:
> > Reading package info...
> > Information about 435 packages read in 5 seconds.
> > The following 9 additional packages will be installed:
> >  db3 db3-shlibs expat gdbm gdbm-shlibs gmp gmp-shlibs manconf tcltk
> > Do you want to continue? [Y/n] y
> > curl -f -L -s -S -O http://www.sleepycat.com/update/3.3.11/db-3.3.11.tar.gz
> > curl: (22) The requested file was not found
> > ### curl failed, exit code 22
> > Downloading the file "db-3.3.11.tar.gz" failed.
> >
> > (1)  Give up
> > (2)  Retry
> >
> > How do you want to proceed? [2]


From loredo at astro.cornell.edu  Wed Jun 19 18:53:09 2002
From: loredo at astro.cornell.edu (Tom Loredo)
Date: Wed, 19 Jun 2002 18:53:09 -0400
Subject: Windows versions of Python---pros and cons?
Message-ID: <3D110B55.4D8AD7FA@astro.cornell.edu>

Hi folks-

I've been using Python heavily on a Solaris box and a Mac for a
few years now.  For reasons I'll spare you from, I now am venturing
into Windows territory (a Dell Inspiron 7500 laptop, if that matters).
I see at least two Windows versions of Python---that at Python.org,
and the ActiveState version.  The obvious "pro" for the ActiveState
version is simplicity of the install (batteries included, etc.).
But what about after that?  Are their substantial differences between
the versions that I will bump into down the line?  In particular,
I write many C extensions; can one use distutils and the Borland
or gcc free compilers to build C extensions in both versions?

Thanks for your patience and advice for this PythonWin newbie!

-Tom Loredo


From jknapka at earthlink.net  Tue Jun 25 01:09:53 2002
From: jknapka at earthlink.net (Joseph A Knapka)
Date: Tue, 25 Jun 2002 05:09:53 GMT
Subject: Python and Eclipse
Message-ID: <3D17FAF4.35CC9F1B@earthlink.net>

Hi folks,

I am interested in having Python support in
Eclipse <URL: http://www.eclipse.org> - interested
enough to contribute code, or to tackle the
task myself, if need be. Of course, I'd like
the Python support to be at least as good as the
Java support, which is a pretty tall order,
'cause the Java support in Eclipse is awfully
damn good --- it nearly allows me to *enjoy*
writing Java code :-)

Anyway, I know some other folks have expressed
interest in this as well, and I'd like to hook
up with them. If you're working on Python+Eclipse
and you've got code in any state of completion,
I'm willing to alpha- or beta-test, fix stuff,
add new stuff, etc.

Cheers,

-- Joe
   "It was me, drove off the offramp / of the
    sweetheart highway..."
    -- Joe Strummer & The Mescaleros, "Bummed-Out City"


From phd at phd.pp.ru  Thu Jun 27 04:47:21 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Thu, 27 Jun 2002 12:47:21 +0400
Subject: How to get rid the new line
In-Reply-To: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org>; from occeanlinux@linuxmail.org on Thu, Jun 27, 2002 at 06:34:27PM +0000
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org>
Message-ID: <20020627124721.C12145@phd.pp.ru>

On Thu, Jun 27, 2002 at 06:34:27PM +0000, SiverFish wrote:
> Anyone tell me how to skip the empty line when you read through the file
> and copy the information in to the list(no empty line )

   It depends on how you have opened the file - in text or binary mode. In
text mode it is simple:

infile = open(filename, 'r')

for line in file: # require python 2.2
   if line != '\n': # end-of-line in the beginning of the line - the line is empty
      process(line)

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From whisper at oz.net  Sun Jun  9 16:45:19 2002
From: whisper at oz.net (David LeBlanc)
Date: Sun, 9 Jun 2002 13:45:19 -0700
Subject: Socket error problem
In-Reply-To: <m33cvwb42k.fsf@mira.informatik.hu-berlin.de>
Message-ID: <GCEDKONBLEFPPADDJCOEOELGDHAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Martin v. Loewis
> Sent: Sunday, June 09, 2002 13:17
> To: python-list at python.org
> Subject: Re: Socket error problem
>
>
> "David LeBlanc" <whisper at oz.net> writes:
>
> > The general.* constant references are to colorize the output
> which is turned
> > off but still results in the "31" in the connection error line in the
> > program output below. NOTE: I didn't write this code, so I have
> no idea why
> > it's returning the username in the socket.error instead of the
> host. FWIW,
> > this code is running in a thread, but there is only one thread
> trying to get
> > mail.
>
> That deserves some investigation since this, again, sounds impossible
> (how can it possibly know what the user name as, when this piece of is
> not being passed the user name). The only possible explanation is that
> self.error_msg prints unrelated information; I recommend to remove all
> except clauses and let the exception pass through.
>
> >                 self.sock = socket.socket(af, socktype, proto)
>
> Please put a print statement before this socket call, to print all
> three arguments.
>
> >                 self.sock.connect(sa)
> >             except socket.error, msg:
>
> And put one here, printing socket.error.
>
> Regards,
> Martin
> --
> http://mail.python.org/mailman/listinfo/python-list

It knows the username because it got it out of a config file and supplied it
to the initializtion of POP3Server, which is what "serv" is.

I've put print statements around all this before I first tryed the ml, but
here they are again:
    def __init__(self, host, port = POP3_PORT):
        self.host = host
        self.port = port
        msg = "getaddrinfo returns an empty list"
        self.sock = None
        print 'poplib._init__: self.host = %s ' % self.host, 'self.port =
%s' % self.port
        for res in socket.getaddrinfo(self.host, self.port, 0,
socket.SOCK_STREAM):
            af, socktype, proto, canonname, sa = res
            try:
                self.sock = socket.socket(af, socktype, proto)
                self.sock.connect(sa)
            except socket.error, msg:
                print 'poplib.__init__ socket.error msg = %s' % msg
                if self.sock:
                    self.sock.close()
                self.sock = None
                continue
            break
        if not self.sock:
            raise socket.error, msg
        self.file = self.sock.makefile('rb')
        self._debugging = 0
        self.welcome = self._getresp()

And the output is:

K:\tmp\animail>animail -v
Opening Config file: c:/user/dave/.animail/animailrc

Trying to connect (mail.oz.net)
poplib._init__: self.host = mail.oz.net  self.port = 110
poplib.__init__ socket.error msg = (10022, 'Invalid argument')
("Connection error (whisper):  (10022, 'Invalid argument')", 31)


David LeBlanc
Seattle, WA USA





From peter at engcorp.com  Tue Jun  4 23:34:44 2002
From: peter at engcorp.com (Peter Hansen)
Date: Tue, 04 Jun 2002 23:34:44 -0400
Subject: Pop a list from beginning ? and memory saving...
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de>	<3CFA0676.68F80DE6@engcorp.com>	<add289$vud42$1@ID-146704.news.dfncis.de>	<Xns9221CE2AF76D0cliechtigmxnet@62.2.16.82>	<adhrsl$118kvn$1@ID-146704.news.dfncis.de> <mailman.1023194347.14669.python-list@python.org> <3CFD5B60.90831F21@engcorp.com> <aIeL8.120588$Kp.12235125@bin7.nnrp.aus1.giganews.com>
Message-ID: <3CFD86D4.C4F81404@engcorp.com>

Terry Reedy wrote:
> 
> "Peter Hansen" <peter at engcorp.com> wrote in message
> news:3CFD5B60.90831F21 at engcorp.com...
> > >     self._v[:] = []
> > I can't parse that.  What does it do?
> 
> Try the wonerful interactive interpreter:
> 
> >>> l = range(10)
> >>> l
> [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
> >>> l[3:7] = []
> >>> l
> [0, 1, 2, 7, 8, 9]
> >>> l[:] = []
> >>> l
> []

Okay, so it's modifying the list in place to be empty, 
equivalent to the del self._v[:] thing someone else posted?

I must have missed the _why_ part...  ah!  enlightenment:
you are removing the contents of the list, which might
have other references to it, rather than rebinding the name
self._v to a new empty list but leaving those other 
references alone.

Still have to think about times when that would be desirable,
but thanks for the nudge. :)

-Peter


From fredrik at pythonware.com  Sat Jun 29 13:16:45 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sat, 29 Jun 2002 17:16:45 GMT
Subject: Newbie that don't understand
References: <AIlT8.13914$p56.4478653@newsb.telia.net>
Message-ID: <1YlT8.45606$n4.10833103@newsc.telia.net>

Martin St?hl wrote:

> if a == b:
>     print "The absolute values of", a,"and",b,"are equal"
> else:
>     print "The absolute values of a and b are different"
>
> with the output being:
> The absolute values of 23 and 23 are equal"
>
> how can 23 and -23 be equal?

"value of" != "absolute value of"

in ordinary math notation, the absolute value of a number
x is written |x|, and is simply what you get if you remove
the sign.

in python, it's usually written:

    abs(x)

also see:
    http://mathworld.wolfram.com/AbsoluteValue.html

</F>




From tim.one at comcast.net  Sat Jun 29 21:22:07 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 29 Jun 2002 21:22:07 -0400
Subject: can't import generators
In-Reply-To: <m2hejlc060.fsf@mother.paradise.lost>
Message-ID: <LNBBLJKPBEHFEDALKOLCOEAEABAB.tim.one@comcast.net>

[John Hunter]
> I am having trouble working with generators.  Anyone know what might
> cause this behavior:
>
> > /usr/local/bin/python2.2
> Python 2.2.1 (#1, Apr 22 2002, 21:20:54)
> [GCC 3.0.4] on linux2
> Type "help", "copyright", "credits" or "license" for more information.
> >>> from __future__ import generators
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
> ImportError: cannot import name generators
> >>>

Start Python with the -v switch to see how imports are getting resolved.
__future__.py is an actual module that actually gets imported.  If you're
picking up a wrong version of the Python libraries, that would explain it
(maybe due to a bad PYTHONPATH setting, or some local module named
__future__.{py,pyc,pyo}).

Note that you *are* successfully importing *some* module named __future__,
else the error msg would have been different.

Here's what's in the correct __future__.py for 2.2.1:

>>> import __future__
>>> dir(__future__)
['CO_FUTURE_DIVISION', 'CO_GENERATOR_ALLOWED', 'CO_NESTED',
 '_Feature', '__all__', '__builtins__', '__doc__', '__file__',
 '__name__', 'all_feature_names', 'division', 'generators',
 'nested_scopes']
>>> __future__.generators
_Feature((2, 2, 0, 'alpha', 1), (2, 3, 0, 'final', 0), 4096)
>>>





From a.schmolck at gmx.net  Mon Jun 10 09:47:06 2002
From: a.schmolck at gmx.net (Alexander Schmolck)
Date: 10 Jun 2002 14:47:06 +0100
Subject: __imul__ broken for 'objects' in 2.2.1
References: <yfsd6v3qa86.fsf@black132.ex.ac.uk> <3n79guskvoaeqtr4srg11kaqse7jcqutt4@4ax.com>
Message-ID: <yfswut76yat.fsf@black132.ex.ac.uk>

Christos Georgiou <DLNXPEGFQVEB at spammotel.com> writes:

> On 07 Jun 2002 18:17:45 +0100, rumours say that Alexander Schmolck
> <a.schmolck at gmx.net> might have written:
> 
> >class Breaky(object):
> >    def __imul__(self, other):
> >        print "imuling"
> >        return self
> >sq = Breaky()
> >sq *=1.
> >
> >gives:
> >
> >Traceback (most recent call last):[...] line 10, in ?
> >    sq *=1.
> >TypeError: can't multiply sequence to non-int
> 
> Just in case: it is indeed your intention to multiply the object with a
> float, right?
> 

Indeed it is and this nice little bug basically means that the matrix class
I've been working on for quite some time is now somewhat screwed (I need a
new-style class *and* inplace operations).

> PS It seems that a Breaky instance is not (incorrectly) considered a
> sequence in 2.3a0 (20020603).

Nice to hear, but only a limited consolation since it quite some time will
pass before 2.3 comes out and is sufficiently widely adopted.

I really wonder where this bizzarre behavior comes from -- it is really none
of the interpreter's business to take guesses whether my class is a sequence
or not -- finding out if there is a problem with the supplied arguments should
be left to the discretion of the method itself.

alex


From mcfletch at rogers.com  Tue Jun  4 11:40:12 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Tue, 04 Jun 2002 11:40:12 -0400
Subject: Creating explorer link on Win32
References: <3CFCB72E.4010405@haftmann-online.de> <20020604115516.GB704@lilith.my-fqdn.de>
Message-ID: <3CFCDF5C.8070903@rogers.com>

There's a demo of creating a shell link in
	site-packages\win32comext\shell\test\link.py
(obviously only if you've installed win32all).  Just so you don't have 
to do a lot of searching for it.

Enjoy,
Mike

Gerhard H?ring wrote:
> * Florian Fredegar Haftmann [2002-06-04 13:48 +0100]:
> 
>>Hi!
>>
>>I'm searching for a possibility to create an explorer link (*.lnk) to a 
>>file on Win32. Has anyne heard of a python module able to do that in a 
>>simple way?
> 
> 
> Mark Hammond's win32 extensions. On my homepage, under VitaminP, there
> is a pure-C solution, which you can live with software explictely marked
> as 'experimental'. IOW, I'd go with the win32 extensions.
> 
> Gerhard
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From emile at fenx.com  Tue Jun 11 10:30:38 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 11 Jun 2002 14:30:38 GMT
Subject: if number is 1-2, 4 or 6-8 or 12-28, 30, 32...
References: <Xns9229EE722467Fgustafl@195.100.94.182> <c76ff6fc.0206110601.6680e899@posting.google.com>
Message-ID: <iQnN8.153796$352.7562@sccrnsc02>

John Machin
> === bot bait ================================
> bignum, long, fast, speed, optimize snot out of
> 


Hook many bots this way?  ;-))

My favorites would include:

impossible, floatingpoint, pre-compile, 

and depending on what I was fishing for, I might also try:

Free, QT, PSU, Tkinter, 


If-it-was-easy-they'd-call-it-catching ly y'rs

--

Emile van Sebille
emile at fenx.com

---------



From dfackrell at DELETETHIS.linuxmail.org  Fri Jun 21 12:41:10 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Fri, 21 Jun 2002 10:41:10 -0600
Subject: [newbie] Overwriting class definitions
References: <3D13563D.AB80D324@athensgroup.com>
Message-ID: <3d135729$1_2@hpb10302.boi.hp.com>

Mike,

You'll need to reload(circle), as import will simply reuse a version that
has already been loaded.

--
Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.


"Mike Christie" <mike.christie at athensgroup.com> wrote in message
news:3D13563D.AB80D324 at athensgroup.com...
> I'm a complete newcomer to Python; I'm teaching myself Python and this is
just
> about the only Python code I've ever typed in, so my apologies for posting
a
> dumb beginner question.  Here's my problem.
>
> I tried a slight variation on the Circle class example from "The Quick
Python
> Book" by Harms and McDonald.  I figured I'd put the class definitions in
> script files so I could edit and reload them quickly, rather than
retyping.  I
> entered this in a file called circle.py:
>
> class Circle:
>     def __init__(self):
>         print "Hello, version 1"
>
> Then I did this:
>
> import circle
> circle.Circle()
>
> and got
>
> Hello, version 1
>
> as expected.  Then I edited the file to say "version 2" and saved it, and
then
> typed in the import statement and circle.Circle() statement again.  It
says
> version 1 again.
>
> So what's happening?  I assumed that when I imported the file again, it
would
> read in the class definition again and overwrite the old definition.  But
that
> doesn't appear to be happening.  When I exit Pythonwin and reenter, it
runs
> just fine.
>
> Do I need to destroy the existing class def first?
>
> Mike




From BPettersen at NAREX.com  Wed Jun 26 17:13:32 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Wed, 26 Jun 2002 15:13:32 -0600
Subject: Help with an algorithm wanted
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158F04@admin56.narex.com>

> From: Russell E. Owen [mailto:rowen at cesmail.net] 
> 
> I'd like some help solving a problem in Python.
> 
> The basic idea is I have an item of data that can be 
> represented in any 
> of four ways: A,B,C and D. I also have conversion functions between 
> certain pairs of representations, such that there's always a path 
> between any two data representations. In this case suppose I 
> know A<->C, 
> B<->C and C<->D, or graphically:
> 
> A <--+
>      +--> C <---> D
> B <--+
> 
> I'd like to create objects that have one of these items 
> (A,B,C,D) as a 
> given, then be able to query any such object for any data 
> representation.
> 
> For instance, an object with A as the given would return D by solving 
> A->C->D.
> 
> What I'm trying to figure out is a simple and reasonably 
> elegant way to 
> code such objects.
> 
> I am pretty sure recursion is the heart of the solution. An 
> object with 
> A as the given could have a method getD that returned D<-C 
> and a method 
> getC that returned C<-A.
> 
> The real issue is creating such objects without hand coding each one 
> (since in the real world I have more representations and the 
> number of 
> casese proliferates badly).
> 
> I suspect I should use recursion to create the object, too, but it 
> sounds like it'll need a lot of introspection. For instance 
> to create an 
> object with A as the given, one might:
> - start by defining method getA to return A
> - look through the set of conversion functions that output A; 
> the only 
> one is A<-C, so method getC returns C<-A.
> - repeat with all methods that return C and don't have 
> methods already 
> defined; this gives method getB returns B<-C and method getD returns 
> D<-C.
> 
> Alternatively, I can imagine making each object basically identical 
> except for some kind of meta-data that tells each method 
> which data item 
> to query; the appropriate function can then be looked up as 
> needed (e.g. 
> via a dictionary that takes as a key the output and input 
> representation). This would slow down the methods but avoid the 
> introspection.
> 
> Any suggestions?

Here's some lightly tested code to get you started. It does a depth
first search of the graph. Detecting cycles in the graph, and populating
the mapping is left as an exercise to the reader <wink>

-- bjorn

def fromAtoB(): pass
def fromAtoC(): pass
def fromCtoD(): pass

mapping = {
	'A' : [('B', fromAtoB), ('C', fromAtoC)],
	'C' : [('D', fromCtoD)]
	}
	
def findShortestPath(mapping, source, dest):
	bestPath = [[]] # ref list
	
	def _findSP(source, dest, res):
		try:
			next = mapping[source]
		except KeyError:
			return # dead end
			
		for item in next:
			if item[0] == dest:
				res += [item[1]]
				if bestPath[0] == [] or len(res) <
len(bestPath[0]):
					bestPath[0] = res
			else:
				_findSP(item[0], dest, res + [item[1]])
		
	_findSP(source, dest, [])
	return bestPath[0]

print findShortestPath(mapping, 'A', 'D')




From erno-news at erno.iki.fi  Tue Jun 11 10:36:52 2002
From: erno-news at erno.iki.fi (Erno Kuusela)
Date: 11 Jun 2002 17:36:52 +0300
Subject: How could i get execution times ?
References: <mailman.1023581654.26137.python-list@python.org>
Message-ID: <kuwut5napn.fsf@lasipalatsi.fi>

In article <mailman.1023581654.26137.python-list at python.org>, Ken
Seehof <kseehof at neuralintegrator.com> writes:

| In other words, time.clock() is far more precise for elapsed
| time (the above example demonstrates precision on the order
| of microseconds -- more than adequate for profiling python
| code).  But time.time() apparently has a granularity on the
| order of a tenth of a second (on windows), which is fine for
| profiling slower applications like Microsoft Outlook :-)

the above is correct on windows. the situation is approximately the
reverse on unix. on unix, time.time() is very accurate in measuring
elapsed wall clock time, whereas time.clock() has poor resulution and
poor accuracy but tries to measure consumed cpu time instead of wall
clock time.

  -- erno



From irmen at NOSPAMREMOVETHISxs4all.nl  Wed Jun 12 14:20:17 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Wed, 12 Jun 2002 20:20:17 +0200
Subject: M2Crypto: select() behaves weird on SSL sockets
References: <3D04ED17.5070606@NOSPAMREMOVETHISxs4all.nl> <ae2str$b0q$1@nntp6.u.washington.edu> <3D066E77.4070901@NOSPAMREMOVETHISxs4all.nl> <ae6bmb$frc$1@dahlia.singnet.com.sg>
Message-ID: <3D0790E1.9080904@NOSPAMREMOVETHISxs4all.nl>

Ng Pheng Siong wrote:

 > Just wondering: Does your code look contrived structured the way you 
have?
 > IOW, will any (simple) change(s) to M2Crypto's interfaces make it more
 > natural to express what you wish in this case?

Well, I would appreciate a wrapper around select.select, one that
acts on SSL sockets the way I expected (from the man page of select).
That means: a select() wrapper that
1) works on regular files and sockets
2) works on SSL sockets
3) in the case of SSL sockets, honors the pending bytes that might still
    be available after a recv() (or send): don't block but return the
    socket in the 'read' or 'write' list if bytes are pending...

This would greatly simplify my socket reading/writing functions...


Irmen



From ftobin at neverending.org  Mon Jun 10 23:05:44 2002
From: ftobin at neverending.org (Frank Tobin)
Date: Mon, 10 Jun 2002 23:05:44 -0400
Subject: Iterators vs. Generators
In-Reply-To: <ae3a6f$9d9$1@panix1.panix.com>
References: <ae3a6f$9d9$1@panix1.panix.com>
Message-ID: <Pine.LNX.4.44.0206102301580.4174-100000@palanthas.neverending.org>

On 10 Jun 2002, Aahz wrote:

> So when would one actually write an iterator instead of a generator?
> I've been trying to think of an example and failing.

Probably any time you want to have the thing 'do more' by being a full
object.  I personally haven't written any generators yet, but I do use
iterators a lot, just because I use OOP a lot.

-- 
Frank Tobin			http://www.neverending.org/~ftobin/



From db3l at fitlinxx.com  Wed Jun  5 20:19:05 2002
From: db3l at fitlinxx.com (David Bolen)
Date: Wed, 5 Jun 2002 20:19:05 -0400 
Subject: self
Message-ID: <926F937512224245A616323693D3F16B1C01EF@ctmsg01.corp.fitlinxx.com>

> everything is a google-search away if you know the right codewords :-)
> I don't in this case, though ...

Sorry, I must admit to being temporarily overcome with a fit of laziness.  I
had actually fumbled around a bit before finding the posts I mentioned, so I
should really have tried to include a reference.

Lest the codewords remain secret, I managed to wind my way there with a
search on "eval dictionary instance" in comp.lang.python (which referenced a
post of Tim's on the fourth page of results).

The messages I had found occurred deep into a "Dictionary from list?" thread
from last October which crossed over into the subject of subclassing
dictionaries:

http://groups.google.com/groups?th=2338421ec8397bec&seekm=mailman.1004338951
.27778.python-list%40python.org

is a pointer to the relevant portion including a response by both Tim and
Guido.

-- David

/-----------------------------------------------------------------------\
 \               David Bolen            \   E-mail: db3l at fitlinxx.com  /
  |             FitLinxx, Inc.            \  Phone: (203) 708-5192    |
 /  860 Canal Street, Stamford, CT  06902   \  Fax: (203) 316-5150     \
\-----------------------------------------------------------------------/




From kseehof at neuralintegrator.com  Sat Jun  1 16:04:50 2002
From: kseehof at neuralintegrator.com (Ken Seehof)
Date: Sat, 01 Jun 2002 13:04:50 -0700
Subject: prototyping good OOdesign in Python?
In-Reply-To: <Pine.LNX.4.33.0206012225290.16212-100000@rnd.onego.ru>
Message-ID: <EFEFIHAINIINNELLHFAIKEHPCHAA.kseehof@neuralintegrator.com>

> Maybe this was already discussed, but I'd liked to raise this topic again.
>
> Python has superb OOProgramming support. Lets not dispute this for a
> moment. Now let's consider that we need to make well-designed OOP solution
> first prototyped in Python.
>
> What troubles are there? Python is too dynamic and all that. And all
> those nifty features need to be translated properly into, say, C++.
>
> Recently I encountered the following piece of code
> (from Arts++ project):
>
> //================================================================
> ===========
> //  Copyright Notice
> //
> //  By accessing this software, arts++, you are duly informed
> //  of and agree to be bound by the conditions described below in this
> //  notice:
> //
> //  This software product, arts++, is developed by Daniel W. McRobb, and
> //  copyrighted(C) 1998 by the University of California, San Diego
> //  (UCSD), with all rights reserved.  UCSD administers the CAIDA grant,
> //  NCR-9711092, under which part of this code was developed.
> //
> //  There is no charge for arts++ software.  You can redistribute it
> //  and/or modify it under the terms of the GNU Lesser General Public
> //  License, Version 2.1, February 1999, which is incorporated by
> //  reference herein.
> //
> //  arts++ is distributed WITHOUT ANY WARRANTY, IMPLIED OR EXPRESS, OF
> //  MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE or that the use
> //  of it will not infringe on any third party's intellectual
> //  property rights.
> //
> //  You should have received a copy of the GNU Lesser General Public
> //  License along with arts++.  Copies can also be obtained from:
> //
> //    http://www.gnu.org/copyleft/lesser.html
> //
> //  or by writing to:
> //
> //  Free Software Foundation, Inc.
> //  59 Temple Place, Suite 330
> //  Boston, MA 02111-1307
> //  USA
> //
> //  Or contact:
> //
> //    info at caida.org
> //================================================================
> ===========
>
> //  constructor
>
> ............skip
>
> Arts::Arts()
> {
>   this->_data._ipPath = (ArtsIpPathData *)0;
>   this->_data._asMatrix = (ArtsAsMatrixData *)0;
>   this->_data._netMatrix = (ArtsNetMatrixData *)0;
>   this->_data._portTable = (ArtsPortTableData *)0;
>   this->_data._portMatrix = (ArtsPortMatrixData *)0;
>   this->_data._protocolTable = (ArtsProtocolTableData *)0;
>   this->_data._selectedPortTable = (ArtsSelectedPortTableData *)0;
>   this->_data._interfaceMatrix = (ArtsInterfaceMatrixData *)0;
>   this->_data._nextHopTable = (ArtsNextHopTableData *)0;
>   this->_data._bgp4RouteTable = (ArtsBgp4RouteTableData *)0;
>   this->_data._rttTimeSeriesTable = (ArtsRttTimeSeriesTableData *)0;
>   this->_data._tosTable = (ArtsTosTableData *)0;
>
>   #ifndef NDEBUG
>     ++_numObjects;
>   #endif
> }
>
> Arts::Arts(const Arts & arts)
> {
>   this->_header = arts.Header();
>   this->_attributes = arts.Attributes();
>
>   switch (this->_header.Identifier()) {
>     case artsC_OBJECT_IP_PATH:
>       this->_data._ipPath = new ArtsIpPathData;
>       assert(this->_data._ipPath != (ArtsIpPathData *)0);
>       *(this->_data._ipPath) = *(arts.IpPathData());
>       break;
>
>     case artsC_OBJECT_AS_MATRIX:
>       this->_data._asMatrix = new ArtsAsMatrixData;
>       assert(this->_data._asMatrix != (ArtsAsMatrixData *)0);
>       *(this->_data._asMatrix) = *(arts.AsMatrixData());
>       break;
>
>     case artsC_OBJECT_NET:
>       this->_data._netMatrix = new ArtsNetMatrixData;
>       assert(this->_data._netMatrix != (ArtsNetMatrixData *)0);
>       *(this->_data._netMatrix) = *(arts.NetMatrixData());
>       break;
>
>     case artsC_OBJECT_PORT:
>       this->_data._portTable = new ArtsPortTableData;
>       assert(this->_data._portTable != (ArtsPortTableData *)0);
>       *(this->_data._portTable) = *(arts.PortTableData());
>       break;
>
>     case artsC_OBJECT_SELECTED_PORT:
>       this->_data._selectedPortTable = new ArtsSelectedPortTableData;
>       assert(this->_data._selectedPortTable !=
> (ArtsSelectedPortTableData *)0);
>       *(this->_data._selectedPortTable) = *(arts.SelectedPortTableData());
>       break;
>
>     case artsC_OBJECT_PORT_MATRIX:
>       this->_data._portMatrix = new ArtsPortMatrixData;
>       assert(this->_data._portMatrix != (ArtsPortMatrixData *)0);
>       *(this->_data._portMatrix) = *(arts.PortMatrixData());
>       break;
>
>     case artsC_OBJECT_PROTO:
>       this->_data._protocolTable = new ArtsProtocolTableData;
>       assert(this->_data._protocolTable != (ArtsProtocolTableData *)0);
>       *(this->_data._protocolTable) = *(arts.ProtocolTableData());
>       break;
>
>     case artsC_OBJECT_TOS:
>       this->_data._tosTable = new ArtsTosTableData;
>       assert(this->_data._tosTable != (ArtsTosTableData *)0);
>       *(this->_data._tosTable) = *(arts.TosTableData());
>       break;
>
>     case artsC_OBJECT_INTERFACE_MATRIX:
>       this->_data._interfaceMatrix = new ArtsInterfaceMatrixData;
>       assert(this->_data._interfaceMatrix !=
> (ArtsInterfaceMatrixData *)0);
>       *(this->_data._interfaceMatrix) = *(arts.InterfaceMatrixData());
>       break;
>
>     case artsC_OBJECT_NEXT_HOP:
>       this->_data._nextHopTable = new ArtsNextHopTableData;
>       assert(this->_data._nextHopTable != (ArtsNextHopTableData *)0);
>       *(this->_data._nextHopTable) = *(arts.NextHopTableData());
>       break;
>
>     case artsC_OBJECT_BGP4:
>       this->_data._bgp4RouteTable = new ArtsBgp4RouteTableData;
>       assert(this->_data._bgp4RouteTable != (ArtsBgp4RouteTableData *)0);
>       *(this->_data._bgp4RouteTable) = *(arts.Bgp4RouteTableData());
>       break;
>
>     case artsC_OBJECT_RTT_TIME_SERIES:
>       this->_data._rttTimeSeriesTable = new ArtsRttTimeSeriesTableData;
>       assert(this->_data._rttTimeSeriesTable !=
> (ArtsRttTimeSeriesTableData *)0);
>       *(this->_data._rttTimeSeriesTable) =
> *(arts.RttTimeSeriesTableData());
>       break;
>
>     default:
>       break;
>   }
>
>   #ifndef NDEBUG
>     ++_numObjects;
>   #endif
> }
>
> ..........skip the rest of file - there are 2-3 more similar
> .......... switch-constructs.
>
> ------------------------------------------------------------
>
> This code looks like... automatically generated. Rather than manually
> written. I do not know why it was written in C++ at all. Of course, I
> imagine 2-3 variants of how to refactor it if I were writing the code
> in Python. I am pretty sure there were no select statement (switches)
> in Python code, because that is what polymorphism is about.
>
> But then again, translating back to C++ could be difficult. The
> mentioned example is all about reading a specially formatted binary
> file to produce certain slices of information. And for each
> class it uses separate .cc/.hh file! While all the difference
> between different classes - structure of corresponding data entries.
>
> My approach in this case could be one resource file, with format
> specification (like DTD). Plus one univarsal class to handle big
> binary file according to "DTD"). Plus maybe separate resource file
> with data query specifications. And it could be at most 80 Kb Python
> project, not that monster Arts++ is in both source and binary forms...
>
> Well, if you still follow this thread, the question is - how well
> Python could serve as an ARCHITECTURAL prototype if it has too rich
> OOP facilities? That is, is Python of help when trying to
> prototype certain design? Developing in C++ looks so unnecessary hard
> after things were done in Python...
> Or, it may be put this way: what discipline a Python programmer must
> obey to allow it's prototype to be conveniently rewriten in C++?
> The above thoughts aren't probably well-formed. But I hope
> my concern is understood.
>
>
> Sincerely yours, Roman Suzi

I agree with what you are saying.  Here's my approach :-)

0. Write a project schedule for management; something like this:
   1. Prototype in python
   2. Convert to C++
1. Prototype in python
2. Give a nice demo, and convince management to skip step 2
3. Re-implement a few cpu intensive classes in C++

Since most of the code never /actually/ gets rewritten in C++,
the need for special C++ "discipline" in the python code goes
away.  Weeks are slashed from the schedule, and the program is
more maintainable.

Another nice advantage is that the python programmer gets to ask
for a raise, and has excellent job security!

There may be a few oddball examples of projects that actually
need to end up written in C++ after prototyping in python, but
for the most part, the most common reason to target C++ is to
alleviate fear of the unknown.  (Hmm, is this also why people,
(including myself), still buy from Microsoft? :-)

- Ken Seehof





From zopestoller at thomas-guettler.de  Fri Jun 28 07:40:48 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Fri, 28 Jun 2002 13:40:48 +0200
Subject: GUI-Control for HTML
Message-ID: <3D1C4B40.6090803@thomas-guettler.de>

Hi!

Is there are GUI-control which can display HTML?

The HTML that will be used will be simple, no javascript or
form support is needed.

Is there are cross-plattform solution? (Unix + Win32)

The tags will be:
-li
-ol
-br
-h?
-table, td, tr, th,
-a
-img

  thomas



From fabien.henon at caramail.com  Fri Jun 28 18:23:55 2002
From: fabien.henon at caramail.com (Fabien HENON)
Date: Sat, 29 Jun 2002 00:23:55 +0200
Subject: Redirect stdout with createfilehandler
Message-ID: <3D1CE1FB.9060007@caramail.com>

I am trying to redirect stdout from a raytracer back to a Tkinter Text 
with createfilehandler.

With the script below I keep getting the error ' maximum recursion depth 
exceeded'.
I have been stuck with this redirection problem for 2 months. This is 
driving really nuts now.

If I turn off the verbose in the command line '-v', the image is 
raytraced till the end with the recursion error.
If I turn the verbose on, the raytrace stops after roughly a second.

What the raytracer does when the verbose is turned on, is to spit out 
the line being rendered while raytracing.
I get the image, not the stdout to the Tkinter text.

Could one Guru tell me what I am doing wrong here.
Or at least have one working example

One thing :
You can download the executable and the POV script (the raytracer) at

http://pyvon.sourceforge.net/executable_file.tgz

Fabien HENON

8<-------------snip--------------->8
import sys,os
import Tkinter
from Tkinter import *

root = Tk()

T=Text(root)
T.pack()

cmd = 'mega +i1.pov +w320 +h240 +dgt +v -f'
c_in,c_out=os.popen4(cmd)

def stdout_handler():
     T.insert('1.0',c_out)
     root.tk.deletefilehandler(c_out)
     root.tk.createfilehandler(c_out, Tkinter.READABLE, stdout_handler)
     root.after(1000,stdout_handler())


root.tk.createfilehandler(c_out, Tkinter.READABLE, stdout_handler)
root.after(1000,stdout_handler())
root.mainloop()
8<--------------snap--------------->8



From rdacker at pacbell.net  Fri Jun 21 13:00:44 2002
From: rdacker at pacbell.net (rdack)
Date: 21 Jun 2002 10:00:44 -0700
Subject: urllib POST question
References: <644f6688.0206200930.48110837@posting.google.com> <slrnah4dbq.r77.insyte-clp@petra.bitstream.net>
Message-ID: <644f6688.0206210900.68dd930c@posting.google.com>

Ben Beuchler <insyte-clp at emt-p.org> wrote in message news:<slrnah4dbq.r77.insyte-clp at petra.bitstream.net>...
> In article <644f6688.0206200930.48110837 at posting.google.com>, rdack wrote:
> > i am doing a POST to a server using urllib and it works.
> > is there some way i can see exactly what is being posted?
> > a way to get the post object out of urllib - the exact lines it is sending?
> 
> I'm a big fan of Ethereal...
> "Sniffing the glue that holds the internet together!"
> 
> http://www.ethereal.com/

i am looking into it. seems overkill for wanting to see a few bytes
that python is generating.
No way within python to see what urllib intends to send out for the
POST?


From tim.one at comcast.net  Mon Jun  3 17:27:05 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 03 Jun 2002 17:27:05 -0400
Subject: python tool: finding duplicate code
In-Reply-To: <Pine.LNX.4.44.0205301241220.4084-100000@mercury.sabren.com>
Message-ID: <LNBBLJKPBEHFEDALKOLCAECPPKAA.tim.one@comcast.net>

[Michal Wallace]
> Thanks for the link! I found a javascript version of a
> suffix tree algorithm online. I ported it to python and it
> seems to work... Unfortunately the original code is very
> hard to understand. I did a straight port and then tried to
> clean it up and make it a little more object oriented, but
> when I started looking into the algorithm, I just couldn't
> track what was going on.

Don't feel bad!  Suffix trees are very surprising.  In fact, they caught
everyone by surprise when they were invented, allowing linear-time solution
of several problems that were conjectured to be impossible to solve in
linear time.  Ukkonen's algorithm (on which the JavaScript version seems to
be based) is considered much clearer than its predecessors, but it's also
subtle to the point of madness <wink>.

A much more obvious approach is to materialize all the suffixes explicitly
into a list, and then just sort the list.  Stop right there and you're close
to a "suffix array".  Building a tree from the sorted list is simple then.
But this doesn't work in linear time, so misses part of the point.

In any case, suffix tree algorithms probably need to be coded in C to be of
practical use (lots of low-level fiddling involved).

> ...
> There's also a suffix tree module written in C, but with a
> python binding here:
>
>     http://www-hkn.eecs.berkeley.edu/~dyoo/python/suffix_trees/

Wow!  Danny sure gets around -- if I can weasle a few free hours I'll
definitely play with this.  Thanks for the link!





From shalehperry at attbi.com  Thu Jun  6 12:01:33 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Thu, 06 Jun 2002 09:01:33 -0700 (PDT)
Subject: Program to an interface, not an implementation
In-Reply-To: <20020606121159.A975@bork.demon.nl>
Message-ID: <XFMail.20020606090133.shalehperry@attbi.com>

On 06-Jun-2002 Egbert Bouwman wrote:
> Hello,
> Can someone shed some light on the GoF adagium:
> 'program to an interface, not an implementation'
> and especially what it means for programming in Python ?
> GoF is not easy.

Sure we can even use python as the example (-:

Any object can act like a list or a dict if they implement the right methods. 
This is an interface.

l = [1,2,3] # use a python list

uses the implementation.

To further this, if your class included code which made decisions based on how
you knew say dict was implemented rather than just using ['string'] or
.has_key() that would be violating the interface.




From Chris.Barker at noaa.gov  Fri Jun 21 16:01:55 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Fri, 21 Jun 2002 13:01:55 -0700
Subject: How to represent the infinite ?
References: <mailman.1024636889.6135.python-list@python.org> <slrnah6oc4.amp.huaiyu@gauss.almadan.ibm.com>
Message-ID: <3D138632.8B40F968@noaa.gov>


Huaiyu Zhu wrote:

> If, in reality, only some very niche platforms need such checkings, maybe it
> is better to let those platforms without ieee compliance suffer (in terms of
> having to add a small patch), rather than depriving most of the mainstream
> platforms of a very useful feature?

hear hear!! (or is that here here!)

-Chris
-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From davegaramond at icqmail.com  Fri Jun 21 06:34:51 2002
From: davegaramond at icqmail.com (David Garamond)
Date: Fri, 21 Jun 2002 17:34:51 +0700
Subject: I'd give up Perl tomorrow if only...
References: <pan.2002.06.21.09.27.15.276676.4527@NOSPAM.optushome.com.au> <3D12F6FC.3000608@wedgetail.com>
Message-ID: <3D13014B.8070203@icqmail.com>

Derek Thomson wrote:
>> .. there was something equivalent in scope and breadth to CPAN.
> I (almost) agree, and sympathise :)

whom is it that you sympathize on? perl? python? the original poster? i 
pick python to sympathize on. :-)

--
dave





From peter at engcorp.com  Sat Jun 22 13:01:51 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 22 Jun 2002 13:01:51 -0400
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <3d149f19$0$53133$edfadb0f@dspool01.news.tele.dk> <af28vp$j57$1@panix1.panix.com>
Message-ID: <3D14AD7F.3D5D243F@engcorp.com>

Aahz wrote:
> 
> In article <3d149f19$0$53133$edfadb0f at dspool01.news.tele.dk>,
> Frihtiof Andreas Jensen <nomailhere at lookitup.yourself.com> wrote:
> >
> >My bet is that once you cool off and sits down with some debugging tools
> >like GDB, Boundschecker or Electric Fence you will indeed find the
> >programming error and fix it.
> >
> >My bet is also that you will not tell anyone either.
> 
> Ya know, you could have saved yourself a lot of typing by simply saying
> PBCAK.  ;-)

To save us from those who can't Google:  Problem Between Chair And Keyboard.

(It was new to me, too.)

-Peter


From donn at u.washington.edu  Fri Jun  7 12:16:52 2002
From: donn at u.washington.edu (Donn Cave)
Date: 7 Jun 2002 16:16:52 GMT
Subject: Using popen interactively
References: <slrnag0fhu.6u3.joe@zork.plover.net>
Message-ID: <adqm9k$22u6$1@nntp6.u.washington.edu>

Quoth Joe Mason <joe at notcharles.ca>:
...
| So I'm trying to use popen2.Popen3("executable-name") to get handles to
| stdin and stdout.  Works fine, except that I'm not sure how to detect
| when the app has stopped printing and is waiting for input.  readline()
| gets the first lines ok, but then the ">" line has no newline on the
| end, so it blocks.  Apparently read() is supposed to return an empty
| string when it hits EOF, but it blocks too - so apparently on an
| interactive file stream like this EOF isn't set when its waiting for
| input.

Right.  I expect you'd do better to throw away the file objects
popen2 makes for you, and use the pipe file descriptors directly.
Instead of
    infile = app.fromchild
    ...
        line = infile.readline()
say
    infile = app.fromchild.fileno()
    ...
        data = posix.read(infile, 8192)
        #  split into lines if you care

The stdio buffering in a file object is only trouble for you.
Part of this trouble is on the other end, if your external
process buffers output - in that case you'll have to figure
out how to use ptys.

	Donn Cave, donn at u.washington.edu


From mhammond at skippinet.com.au  Mon Jun 17 18:50:23 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Mon, 17 Jun 2002 22:50:23 GMT
Subject: Help: win32gui & threads
References: <33803989.0206170318.9d3a225@posting.google.com>
Message-ID: <3D0E6849.3050909@skippinet.com.au>

Miki Tebeka wrote:
> Hello All,
> 
> I've written a little logger daemon with UI in Windows taskbar.
> (See code below)
> It seems to work fine but when I choose 'Quit' the windown closes but
> the pythonw is still alive and I need to kill it from the TaskManager.
> 
> Any idea how to make it close?

You will need to do a little more work to give us more information.

* Is the DestroyWindow() call being made?
* Does the main thread ever appear to end - ie, if you put a print 
statement after the PumpMessages(), will it appear.

Mark.



From emile at fenx.com  Tue Jun 18 21:48:58 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 19 Jun 2002 01:48:58 GMT
Subject: I have .py, I want .exe, how??
References: <mailman.1024448546.1172.python-list@python.org>
Message-ID: <eqRP8.2009$EP.54@sccrnsc03>

<jubafre at zipmail.com.br> wrote in message
news:mailman.1024448546.1172.python-list at python.org...
i have a file.py
how i make e executable file .exe for run in windows???

http://www.python.org/doc/FAQ.html#4.28

--

Emile van Sebille
emile at fenx.com

---------



From quinn at groat.ugcs.caltech.edu  Sat Jun 15 01:49:33 2002
From: quinn at groat.ugcs.caltech.edu (Quinn Dunkan)
Date: 15 Jun 2002 05:49:33 GMT
Subject: filter2
References: <mailman.1023975618.6474.python-list@python.org> <aeaava$56l8p$1@ID-59885.news.dfncis.de> <mailman.1023979895.13512.python-list@python.org>
Message-ID: <slrnagllbc.5mj.quinn@groat.ugcs.caltech.edu>

On Thu, 13 Jun 2002 18:50:23 +0400, Oleg Broytmann <phd at phd.pp.ru> wrote:
>On Thu, Jun 13, 2002 at 04:39:51PM +0200, Thomas Heller wrote:
>> 
>> "Oleg Broytmann" <phd at phd.pp.ru> wrote in message
news:mailman.1023975618.6474.python-list at python.org...
>> >    I want to have new python builtin - filter2. It is exactly like filter,
>> > but it returns also a list of items that didn't pass the test.
>> 
>> IIRC, in Smalltalk they were named 'select' (like filter), and 'reject' (like your filter2).
>> IMO these are much better names then filter and filter2.
>
>   It is too late to rename builtin filter, so I modelled fiter2 after it.

In haskell they call it "partition" which I think is better since the
emphasis is that it returns both selected and rejeted.

I wrote mine in python with a plain for loop.  I'm not sure C would really get
you that much efficiency, but it would be easy enough to find out.

Not sure it's worth being in the core (maybe there should be a module for
list utils or something), but I like partition better than filter2.


From peter at engcorp.com  Mon Jun 17 20:28:22 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 20:28:22 -0400
Subject: Embedded systems
References: <aeaajv$5db01$1@ID-17980.news.dfncis.de>
Message-ID: <3D0E7EA6.1E96C53F@engcorp.com>

Stuart wrote:
> 
> Is Python practical for embedded systems?   I am working on a project at the
> moment that uses a GSM/GPRS modem and a GPS receiver.  I have prototyped the
> whole thing in Visual Basic, and now I need to port it over to a suitable
> processor (possibly on a 68000 device using Linux).
> 
> I don't really want  to get into C for this project, and was wondering if
> Python was suitable?  Are memory requirements an issue?

We are building embedded Linux systems using Python as the primary
language (with bits of C for some custom drivers).  Our processors
are x86 but a standard port to 68K should work fine I suppose.
Memory is about what you might expect, which is to say that it
requires more than C by several times, but any system capable of
running Linux and a few apps can handle Python on top of it
without much trouble, IMHO.

For what it's worth, we had a 386-based proprietary system 
running with a compacted Python running in 1MB RAM and 1MB
flash, but it was way too tight for our needs.  We now use
486/100 processors with 32MB RAM and 64MB CompactFlash cards,
and have _plenty_ of room to spare.  We could probably get
by with as little as 4-8MB RAM and 16-32MB flash for what we
currently do.

-Peter


From dw133 at cs.york.ac.uk.go.away  Fri Jun  7 19:45:08 2002
From: dw133 at cs.york.ac.uk.go.away (Darren Winsper)
Date: Sat, 08 Jun 2002 00:45:08 +0100
Subject: Detecting OS and Window Manager
References: <mailman.1023304388.9659.python-list@python.org> <ufvsco8mlhf213@corp.supernews.com> <3d00bfb5$1_2@hpb10302.boi.hp.com> <ug1h5q29aulga9@corp.supernews.com> <3d00d248$1_1@hpb10302.boi.hp.com>
Message-ID: <ug2hcuoq7egm6d@corp.supernews.com>

In <3d00d248$1_1 at hpb10302.boi.hp.com>, Daniel Fackrell wrote:

>> Well, I could attempt to see if something like "kdesktop" is running,
>> since if it is, the chances of running KDE are very high.  It's kludgy,
>> but I can't really think of anything else, since KDEINIT would be running
>> if a single kde application is running IIRC.
> 
> If you're thinking of scanning the currently running processes for the
> existence of a particular process, then you won't get the results you want
> there, either.  The reason for this is that there can be multiple logins for
> a single user or multiple users, and KDE-related processes will exist if any
> one of those logins is running KDE, but not necessarily the current one.

Good point.

> Maybe import your kde module(s), and then try to do something KDE specific
> without KDE running and see what kind of error this causes so you will know
> what to trap for?

I don't believe there really is anything there you can do.  I certainly
don't know of anything in KDE that requires you to be running the KDE DE
to actually perform the command.  If there is anything, I'm fairly certain
it'll be a kludge and won't really be portable.

> Hopefully some other newbie (for a much less significant
> value of new, AKA expert) is still following this thread to give you a
> better answer.

Alas, I don't think there is an easy answer to this.  I think we'll just
have to either:
1) Settle for a default
2) Provide different commands such as gfunky and kfunky and just provide
menu entries to point to the appropriate one depending upon which DE you
are running.

-- 
ICQ: 8899775 Jabber ID: darren at rooter.dyndns.org
Stop the EUCD before it's too late:
http://uk.eurorights.org/issues/eucd/
http://www.tomchance.uklinux.net/articles/darkages.shtm



From emile at fenx.com  Tue Jun 18 22:30:00 2002
From: emile at fenx.com (Emile van Sebille)
Date: Wed, 19 Jun 2002 02:30:00 GMT
Subject: I have .py, I want .exe, how??
References: <3D0F6623000012D8@www.zipmail.com.br> <mailman.1024451845.24394.python-list@python.org>
Message-ID: <I0SP8.249962$cQ3.11582@sccrnsc01>

Gerhard H?ring
>
> It's a FAQ and answered here:
>
> http://www.python.org/cgi-bin/faqw.py?req=show&file=faq04.028.htp
>
> Gerhard, who considers making an editor macro for this question

When you get enough macros you can apply for bot status.

Of course, you'll have to find out where to apply from the PS






From skip at pobox.com  Tue Jun 25 10:12:42 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 25 Jun 2002 09:12:42 -0500
Subject: Thread Memory Leak
In-Reply-To: <3D18745D.554A7C44@engcorp.com>
References: <pan.2002.06.25.12.37.12.535288.17223@pobox.com>
        <mailman.1025010754.9180.python-list@python.org>
        <3D18745D.554A7C44@engcorp.com>
Message-ID: <15640.31322.834546.849385@12-248-8-148.client.attbi.com>

    >> You never remove any thread objects from your listThread list.  It
    >> fills up with completed threads.

    Peter> Actually, the list is rebound to a new empty list each pass
    Peter> through the while loop...

Duh...  That's what I get for answering questions so early in the morning...

I do see some strange behavior, now that I actually run the code.  Running
it as

    python threadgrowth.py | egrep -v RSS | uniq

yields output like

     3288 1984
     3292 1992
     3296 1996
    13536 2016
     3296 1996
    15584 2020
     3296 1996
    21728 2044
    19680 2028
    15584 2020
     3296 1996
    19680 2028
     3296 1996
    17632 2032
     3296 1996

I don't know why it bounces around like that, but the minimum VSZ does go
up:

     3296 1996
    19680 2032
     7392 2008
     3296 1996
    20704 2032
     4320 2000
    16608 2028
    20704 2040
    20704 2036
     4320 2000
    20704 2036
     4320 2000
    16608 2024
     4320 2000
    14560 2020
     4320 2000
    20704 2032
     4320 2000
    16608 2024
    18656 2032
     4320 2000
    18656 2028
     4320 2000

This was after just a couple minutes of running.

Skip




From logiplexsoftware at earthlink.net  Thu Jun 20 13:47:15 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Thu, 20 Jun 2002 10:47:15 -0700
Subject: How to represent the infinite ?
In-Reply-To: <3d11ee9b$1@news.swissonline.ch>
References: <3d11ee9b$1@news.swissonline.ch>
Message-ID: <20020620104715.5d4f11ae.logiplexsoftware@earthlink.net>

On Thu, 20 Jun 2002 17:02:33 +0200
erreur wrote:

> Will somebody have an idea, to represent the infinite one?
> 
> I have variables to initialize with is +inf (or - inf).  To be sure that
> later, all will be smaller (or larger) than my variables.
> I tried to redefine the operators on an object.  It goes for Inf>10 but I do
> not arrive for 10>Inf (because that takes > of Int).
> 
> --------------------
> 
> def __gt__(self, val):
> 
>     return 1
> 
> --------------------

Well, here's a start:

class inf(int):
    def __init__(self):
        int.__init__(self)
        self.sign = 1

    def __str__(self):
        return "<infinity>"

    def __repr__(self):
        return "<infinity>"

    def __cmp__(self, n):
        if isinstance(n, inf):
            return cmp(self.sign, n.sign)
        return self.sign
    
    def __neg__(self):
        retval = inf()
        retval.sign = self.sign * -1
        return retval

if __name__ == '__main__':
    import sys
    
    m = -inf()
    n = inf()
    assert(m < n)
    assert(n > m)
    assert(m == m)
    assert(n == n)
    assert(m != n)
    assert(m < 0)
    assert(n > 0)
    assert(m < -sys.maxint)
    assert(n > sys.maxint)
    assert(m < -m)
    assert(n == -m)
    



-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From irmen at NOSPAMREMOVETHISxs4all.nl  Fri Jun 28 13:33:44 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Fri, 28 Jun 2002 19:33:44 +0200
Subject: M2Crypto: select() behaves weird on SSL sockets
References: <3D04ED17.5070606@NOSPAMREMOVETHISxs4all.nl> <ae2str$b0q$1@nntp6.u.washington.edu> <3D066E77.4070901@NOSPAMREMOVETHISxs4all.nl> <ae6bmb$frc$1@dahlia.singnet.com.sg> <3D0790E1.9080904@NOSPAMREMOVETHISxs4all.nl> <831ybavprx.fsf@panacea.canonical.org> <3D0919AF.7050903@NOSPAMREMOVETHISxs4all.nl> <83it45lxfb.fsf@panacea.canonical.org> <3D1B92A2.7030605@NOSPAMREMOVETHISxs4all.nl> <q0ZS8.106386$mF.41854@atlpnn01.usenetserver.com>
Message-ID: <3D1C9DF8.3010607@NOSPAMREMOVETHISxs4all.nl>

Steve Holden wrote:

> asyncore/asynchat really repay the effort of study, so I would advise you to
> take a fresh look when you get time.

I will. Am I correct to conclude that the benefits of using asyncore as a
basis for Pyro are:
1. uses a well-tested stable I/O module (though the current implementation
    in Pyro isn't bad at all)
2. able to process multiple requests without using threads, because
    the I/O is multiplexed
3. able to handle very many concurrent connections (no thread overhead
    per connection -- but is a thread really that heavy?)


> To assist you, I plan to add documentation for asynchat for 2.3.

That's nice because I was looking for that one just yet :)
But glancing at the code of asynchat makes me think that this
one isn't suited for Pyro because Pyro is no line-based command-response
protocol.

> when-sourceforge-stops-playing-silly-buggers-ly y'rs  - steve
What's wrong with SF?

Cya
Irmen de Jong



From pyth at devel.trillke.net  Sun Jun 16 06:23:29 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 16 Jun 2002 12:23:29 +0200
Subject: What If.....    Strong Types
In-Reply-To: <3D0C0F0D.4030900@bgb.cc>; from garrett@bgb.cc on Sun, Jun 16, 2002 at 04:08:26AM +0000
References: <3D0C0F0D.4030900@bgb.cc>
Message-ID: <20020616122329.H15079@prim.han.de>

Hi Don,

You wrote:
>    What if Python had been built from the start with strong types? And I mean 
> strong types, not strong typing. The distinction I am trying to draw is that 
> variables would not be typed, but that all types would have rigid interfaces.

your wording doesn't really make sense to me. There is no such thing as "variables" 
in python. EVERYWHERE in modules, classes, instances, globals(), locals() ... 
there are only 

        name=object  "name-object-bindings"

Objects *have* strong types (or strong typing if you like). The difference
to e.g. Java is that python does DYNAMIC typing instead of java's STATIC
typing. This is the main hinderance for doing certain optimizations 
on compilation time. See the types-SIG for further and very interesting
discussion:

    http://www.python.org/sigs/types-sig/

>    Primarily, what if classes always had rigidly defined interfaces. I mean 
> that public members had to be declared to exist, and that methods couldn't be 
> modified on an instance. Private members could exist but would be really private.

OK. There is lots of discussion about 'interface' concepts. Today interfaces 
only exist implicitely in python. There have been lots of efforts to
introduce interface concepts like the one for Zope2/Zope3's components:

http://www.zope.org/Members/jim/PythonInterfaces/Interface (maybe outdated)

I think there is a reason, though, why there is no great rush to introduce
any current interface concept to python. None of the current proposals have 
really convinced a large enough portion of python developers to become part of
the core language. I am sure that when python gets explicit interfaces 
it's going to be a great thing (tm). But IMO one of the main questions is 

    how do you mix the declarative nature of interfaces with pythons
    powerful dynamic object and namespace model?

My personal fantasies focus on connecting IDL-specified interfaces 
(OMG's *I*nterface *D*efinition *L*anguage which is used in many domains) 
with python namespaces. Something basically like:

    mymodule.py    # a python module
    mymodule.idl   # interface definitions 

python's compiler would use *both* files to make sure that they match. 
For a to-be-discussed definition of 'match' :-)

regards,

    holger




From gerson.kurz at t-online.de  Fri Jun  7 12:40:27 2002
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Fri, 07 Jun 2002 16:40:27 GMT
Subject: Newby: How do I strip HTML tags?
References: <Xns92278CC566FEAusYourStandingInIt@130.133.1.4>
Message-ID: <3d00e1b1.47853937@news.t-online.de>

This is a quite straight forward function: 

def StripTags(text):
    finished = 0
    while not finished:
        finished = 1
        # check if there is an open tag left
        start = text.find("<")
        if start >= 0:
            # if there is, check if the tag gets closed
            stop = text[start:].find(">")
            if stop >= 0:
                # if it does, strip it, and continue loop
                text = text[:start] + text[start+stop+1:]
                finished = 0
    return text

Or, if you feel lucky, you could use this more complicated solution

def StripTags(text):
    flag = [1]
    def stripfunc(c):
        if not flag[0]:
            if c == '>':
                flag[0] = 1
                return 0
        elif c == '<':
            flag[0] = 0
        return flag[0]               
    return filter(stripfunc,text)




From baas at ira.uka.de  Mon Jun 24 16:52:46 2002
From: baas at ira.uka.de (Matthias Baas)
Date: Mon, 24 Jun 2002 22:52:46 +0200
Subject: PYREX
References: <3d1749d3.19107439@news1.news.adelphia.net>
Message-ID: <ev0fhuojju9fhr2bepqnhno79qo7ir6vv1@4ax.com>

On Mon, 24 Jun 2002 16:34:57 GMT, edlsoft at mindspring.com (Burt
Leavenworth) wrote:
>Has anyone installed Pyrex on Windows? If so, can you share the
>details?

- Download Pyrex and Plex.
- Unpack Pyrex (you can use filzip (www.filzip.de), winzip
(www.winzip.com) or the cygwin tools (www.cygwin.com))
- Unpack Plex 
- Copy the "Plex" directory into the Pyrex directory (the other files
aren't necessary for executing pyrex)

Now you can already call pyrexc, but probably you also want to add the
Pyrex directory to your path and rename "pyrexc" into "pyrexc.py".

If you want to use Plex yourself, it'd be better to install it as a
regular module, but if you just want to get pyrex running you can get
away with the above procedure.

- Matthias -



From peter at engcorp.com  Thu Jun 20 21:05:45 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 20 Jun 2002 21:05:45 -0400
Subject: Flame fests (was Re: python version?)
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <3D107948.B3FEE4BB@engcorp.com> <aer53s$7is$1@panix1.panix.com> <3D112BA6.4CA5F584@engcorp.com> <aerosm$p18$1@panix1.panix.com>
Message-ID: <3D127BE9.9BD93FC3@engcorp.com>

Aahz wrote:
> 
> In article <3D112BA6.4CA5F584 at engcorp.com>,
> Peter Hansen  <peter at engcorp.com> wrote:
> >Aahz wrote:
> >>
> >> Everybody have fun playing "kick the newbie"?  Maybe it's time to quit?
> >
> >I think we'd all object strongly to the characterization that we were
> >"kicking a newbie".
> 
> Well, then, if you're not kicking a newbie, you're swallowing trolled
> bait.  Which makes you feel better?

Swallowing trolled bait, no contest.  I'm not yet convinced it was such
though, although similar further postings would prove your point.

-Peter


From aahz at pythoncraft.com  Thu Jun 20 20:20:16 2002
From: aahz at pythoncraft.com (Aahz)
Date: 20 Jun 2002 20:20:16 -0400
Subject: Raw String Output
References: <aetnkf$tot$1@news.umbc.edu> <Xns9234F35131B7cliechtigmxnet@62.2.16.82>
Message-ID: <aetrg0$47e$1@panix1.panix.com>

In article <Xns9234F35131B7cliechtigmxnet at 62.2.16.82>,
Chris Liechti  <cliechti at gmx.net> wrote:
>"mike" <gringro1 at umbc.edu> wrote in news:aetnkf$tot$1 at news.umbc.edu:
>>
>> If I have a string variable with the value '\n' how do I output it to
>> a file with the value of '\012' instead ?
>
>i assume that its clear to you that '\n' == '\012' so i guess you want the 
>string r'\012' in the file and not the newline character?!?
>
>then you must do it by hand, i think. this one replaces all characters < 32 
>with their octal value:
>>>> print ''.join([ c>=' ' and c or '\\%03o' % ord(c) for c in 'hello\n'])
>hello\012

Actually, I'd guess the problem is more that Mike is using a text file
on a platform where \n is *not* \012 (e.g. Windows).  In that case,
he'll have to stick chr(012) into the files directly, preferably opening
files in binary mode first.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

Project Vote Smart: http://www.vote-smart.org/


From deltapigz at telocity.com  Sat Jun 29 21:29:51 2002
From: deltapigz at telocity.com (Adonis)
Date: Sat, 29 Jun 2002 21:29:51 -0400
Subject: possibly a dumb question
Message-ID: <3d1e5eff$1_7@nopics.sjc>

can a class return a value? i have treid to google this, but have found
myself short of an answer, i know i could place it in a member and retrieve
it or through a function, but im trying to simplify my life.. just a
little..

any help would greatly be appreciated.

Adonis





From pollosgbr at netscape.net  Sat Jun 29 17:51:04 2002
From: pollosgbr at netscape.net (Just a dude)
Date: Sat, 29 Jun 2002 21:51:04 GMT
Subject: Newbie that don't understand
References: <AIlT8.13914$p56.4478653@newsb.telia.net>
Message-ID: <3d1e262f.3755054@news.clara.net>

On Sat, 29 Jun 2002 17:00:16 GMT, "Martin St?hl"
<martin.stahl at telia.com> wrote:

>I have never been programming before and now I am reading the

<snip you all know what he said>

I have just [this week] decided to try Python myself. I haven't done
any programming since I had a 48K Specky [except for a little dabble
in Rexx some years back]
I too have started with the "Non-Programmers Tutorial For Python" but
understand [I think] absolute values.

But if I didn't I don't think I'd understand your reply's. I'm sorry I
know the value of these groups and really appreciate the help you all
give freely and I'm very grateful, but for 'Newbies' like myself I
would ask for simple answers if that's possible. Please don't take
this the wrong way but as a beginner what seems easy and obvious to
you is sometimes way over our heads.

>with the output being:
>The absolute values of 23 and 23 are equal"
>
>how can 23 and -23 be equal?
>
To some this can be a hard concept to grasp. My son is an electronics
design engineer, and I still can't get my head around negative
voltages, how the hell do you get a -12 V supply... to him it's basic
but to me..............

Cheers and thanx for you help Steve.



From rs at onsitetech.com  Fri Jun 28 16:07:46 2002
From: rs at onsitetech.com (Robb Shecter)
Date: 28 Jun 2002 20:07:46 GMT
Subject: Introspection: determining parent classes?
Message-ID: <3D1CC1CA.70004@onsitetech.com>

Hi,

Is there any way to determine the parent classes given an instance?

I'm writing unit test code, and I want to assert that a particular 
result is a subclass of some other class.  In Java, you can do something 
like:

anInstance.getClass().isSubclassOf( aClass )

I looked through the type(), __class__, etc. info, but these seem to 
just return string-like objects that don't allow further introspection.

Is there an equivalent in Python?

Thanks,
Robb



From dm_top at hotmail.com  Mon Jun  3 11:30:01 2002
From: dm_top at hotmail.com (Dmitri)
Date: 3 Jun 2002 08:30:01 -0700
Subject: thread module under Windows
Message-ID: <59d8a5f6.0206030730.1be14ffb@posting.google.com>

Hi All,

I have been searching the conference for the simple question like "
where the thread module is " and got the answer that it's inside the
python if the threading support is on.
Ok. 

Reading the conference I see that many people use the python in
multi-threading. How they get the python version for windows with
treading support on ( I am not talking about other optional services
now... )???

Now the question - I understand, that I could and must rebuild the
python under the unix/linux/cygwin environment to have the treading
support be on.
What about the windows? Could I download the python with optional
services installed, i.e.  should I build the python for the windows
myself if I want to have some optional services be on? If yes, I think
that the only way to do it is to build the python under cygwin. Would
it be compatible enough to run the python from windows command prompt
after that?


With the best regards,
  Dmitri Top


From dave5774 at hotmail.com  Mon Jun 17 17:44:36 2002
From: dave5774 at hotmail.com (Dave)
Date: Mon, 17 Jun 2002 17:44:36 -0400
Subject: Please help with Python Script/MS-Access (DAO)
Message-ID: <aell87$3uv$1@netnews.upenn.edu>

Hey all,
I'm having trouble writing a Python script relating to Microsoft Access.  If
anyone could help me I would greatly appreciate it.  Here's the problem :

I built an MS-Access table with 5 columns(fields) : OrderID, OrderDate,
PlantID, ProductID, OrderQty, with OrderID as the primary key.  All fields
are numeric, except for OrderDate(which is a date field).

Now I want to populate this table with data from a web page
(http://opim.wharton.upenn.edu/~opim101/spring02/Plant1.html).  please check
the site.

My object is to read the the data and insert it record by record into my
Access table.

Here's what I have so far :

import urllib, string
Plant1 =
urllib.urlopen('http://opim.wharton.upenn.edu/~opim101/spring02/Plant1.html'
)
lines1 = Plant1.readlines()
import win32com.client
engine = win32com.client.Dispatch("DAO.DBEngine.35")
db=engine.OpenDatabase(r'd:\Plants.mdb')

I have several problems :
Firstly, when I read the lines, I notice that not all of them are relevant.
I need to ignore <html>, <body>, etc. at the beginning of the file, as well
as the column headers, so that the action really begins from line 4 of the
file and ends with </body>
Also, I need to eliminate <p> from every line in the file in order to get to
the five comma-delimited fields that follow, and then split the rest of the
string to get to each field.  Here's what I think part of the code should
look like :

for line in lines1.readlines():
    line=line.rstrip()
    inOrderID,inOrderDate,inRegionID,inProductID,inOrderQty =
line.split(",")

Something like that, I'm not sure...
Lastly, now that I've read all the fields, I need to insert them each as a
single record into my table of my access database using Python DAO using the
Insert statement.  Also, the OrderDate field that I read in, looks like a
text type to Access, so I need to pad it with the # sign on either side
before inserting it into the Access database, so that Access interprets it
as a Date/Time type.

Finally, I need to get the same information from 2 other plants and tag them
onto the same table...from the web page
http://opim.wharton.upenn.edu/~opim101/spring02/PlantPortal.html
where plant1's data shows up first, plant2's data shows up below it, and
plant3's data shows up last...so all plants' data in the same table.

If anyone could help me by writing the script I would greatly appreciate
it...
Thank you so much.
Sincerely,
Dave





From cliechti at gmx.net  Sun Jun  2 14:19:05 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 2 Jun 2002 20:19:05 +0200
Subject: Python Turing machine
References: <d9akfu0c7ro0sebio1vd9c6ik2i0fg59le@4ax.com>
Message-ID: <Xns9221CF0CC998Dcliechtigmxnet@62.2.16.82>

Tobias Brandvik <tmagna at online.no> wrote in 
news:d9akfu0c7ro0sebio1vd9c6ik2i0fg59le at 4ax.com:

> Has anyone seen a model of a Universal Turing Machine coded in Python?
> I've seen examples in Scheme, Java and Perl, but I'd really like to
> look at one made in Python... Any pointers?

hint: post some URLs of those implementations to the list. that way we 
don't need to look up what it is and people here have fun trying to solve 
things pythonic...

chris

-- 
Chris <cliechti at gmx.net>



From hwcowan at hotmail.com  Fri Jun 14 16:01:00 2002
From: hwcowan at hotmail.com (Hugh Cowan)
Date: 14 Jun 2002 13:01:00 -0700
Subject: Creating Dynamic Web Pages (ZOPE)
References: <46ca81a0.0206062016.789d3af6@posting.google.com> <J29M8.358696$%u2.156726@atlpnn01.usenetserver.com>
Message-ID: <46ca81a0.0206141201.23db4a8c@posting.google.com>

Steve,

Thanks for the suggestion and sample web-site, I will definetly
check-it out.  Does DreamWaver support ASP then, or rather provide
support?  I mean, I know that you could just add the ASP code mixed in
with the HTML manually, but does it offer any RAD type tools to help
make using ASP easier?

Also, I thought that Dreamweaver was more of just a HTML Editor, while
something like Zope provided extra functionality, tools, wizards,
etc.. to allow you to create Dynamic Web-Sites quicker and easier?  I
am just trying to figure out where to concentrate my search for
software and tools.

I originally looked through the various *Freeware* web-sites for HTML
editors -- but there are so many of them out there (hard to pick one
from the other -- almost makes you want to go back to a plain old text
editor).  While some offer advantages over others, they all seem to
basically assist you in generating the HTML code for web-pages (either
by coding or using a WYSIWYG type interface)?

Thanks,

Hugh

> 
> At the risk of sounding like a heretic...
> 
> I recently started to use DreamWeaver UltraDev (one of my clients mandated
> its use) and I've been relatively impressed. For a really small
> database-driven web created in DreamWeaver take a look at
> 
>     http://www.holdenweb.com/TinySite/
> 
> This web has precisely four pages, and I built it to explain to beginners
> how database contents can affect a site's navigation as well as its readable
> content.
> 
> DreamWeaver MX is now out, and targets several more backends. It would be
> nice if it had one that drove some suitable Python framework -- possibly
> WebWare.
> 
> If you are already familiar with ASP, this will get you up to speed on your
> tasks much more quickly than learning Zope. Won't be as much fun, though.
> 
> regards
>  Steve


From osantana at netbank.com.br  Fri Jun  7 21:31:55 2002
From: osantana at netbank.com.br (Osvaldo Santana Neto)
Date: Fri, 7 Jun 2002 22:31:55 -0300
Subject: How to install Python without root priviledges ?
In-Reply-To: <acvmin$svjfb$1@ID-146704.news.dfncis.de>
References: <acvmin$svjfb$1@ID-146704.news.dfncis.de>
Message-ID: <20020608013155.GD28708@netbank.com.br>

On Tue, May 28, 2002 at 12:36:23PM +0200, Shagshag13 wrote:
> 
> I'm a real newbie in unix's world and i would like to install python on a
> machine where *i can't have* any root priviledges, how should i do that ?
> 
> (it's on a HP Unix, but don't even know how to check the exact os name or
> version...)
> 
> i do this :
> 
> mkdir /home/shagshag/local
> download Python-2.2.1.tgz
> gunzip
> tar -xvf Python-2.2.1.tgz
> cd Python-2.2.1
> ./configure --prefix=/home/shagshag/local
> make

make install

> 
> And i get nothing !!!!
> 
> Is it the good way to install ???
> 
> Thanks in advance,
> 
> S13.
> 
> please, don't forget i'm not a skilled unix guy...
> 
> 
> -- 
> http://mail.python.org/mailman/listinfo/python-list

-- 
Osvaldo Santana Neto    -    aCiDBaSe
ICQ#: 11287184 - www.osantana.f2s.com




From gerson.kurz at t-online.de  Sat Jun  1 10:15:19 2002
From: gerson.kurz at t-online.de (Gerson Kurz)
Date: Sat, 01 Jun 2002 14:15:19 GMT
Subject: Python and HTML frames
References: <ada6hf$vico3$1@ID-49758.news.dfncis.de>
Message-ID: <3cf8d6d3.14806203@news.t-online.de>

I'm assuming you're writing a CGI, so:

#! /usr/bin/python
import cgi, sys, os

print "Content-type: text/html\n"

def GenerateMainFrame():
    print """<html><frameset cols="529,*">
<frame name="Inhalt" target="Hauptframe"
src="/cgi-bin/frames.py?cmd=frame1">
<frame name="Hauptframe" src="/cgi-bin/frames.py?cmd=frame2">
<noframes><body><p>blablabla</p>
</body></noframes></frameset></html>"""

def GenerateFrame1():
    print "<html><body><h1>TODO: add code for frame
#1</h1></body></html>"

def GenerateFrame2():
    print "<html><body><h1>TODO: add code for frame
#2</h1></body></html>"

page, form = None, cgi.FieldStorage()
if form.has_key('cmd'):
    page = form['cmd'].value

if page == "frame1":
    GenerateFrame1()
elif page == "frame2":
    GenerateFrame2()
else:
    GenerateMainFrame()



From gumuz at looze.net  Tue Jun 25 04:30:38 2002
From: gumuz at looze.net (Guyon Morée)
Date: Tue, 25 Jun 2002 10:30:38 +0200
Subject: 'parent object'?
References: <3d18228c$0$221$4d4ebb8e@news.nl.uu.net> <3D182662.CBC925DB@alcyone.com>
Message-ID: <3d182943$0$227$4d4ebb8e@news.nl.uu.net>

thanx a lot for your quick response!


"Erik Max Francis" <max at alcyone.com> wrote in message
news:3D182662.CBC925DB at alcyone.com...
> "Guyon Mor?e" wrote:
>
> > Is there, like 'self, also a 'parent' object of some sort?
> >
> > The problem i'm having is that i have a class, which contains a list
> > of
> > instances of another class. How can i let this 'childclass' call a
> > function
> > which is on the 'parent class'?
>
> An instance has access to its class (.__class__ attribute), and a class
> has access to its parent classes (.__bases__ attribute).
>
> --
>  Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
>  __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
> /  \ See the son in your bad day / Smell the flowers in the valley
> \__/ Chante Moore
>     Bosskey.net: Aliens vs. Predator 2 / http://www.bosskey.net/avp2/
>  A personal guide to Aliens vs. Predator 2.




From kragen at pobox.com  Thu Jun 20 11:20:39 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: Thu, 20 Jun 2002 11:20:39 -0400
Subject: Fw: Inter Process Communication with Numeric arrays
Message-ID: <20020620152039.GA16770@canonical.org>

The problem with http://pobox.com/~kragen/sw/arrayfrombuffer/ has been
fixed.  Thanks for letting me know.




From js_nntp at nospam_blackrocksoftware.com  Sat Jun  8 07:23:20 2002
From: js_nntp at nospam_blackrocksoftware.com (J Swartz)
Date: Sat, 08 Jun 2002 11:23:20 GMT
Subject: Comparison: HTTP with Java, Perl or Python
References: <3D009911.9000904@gmx.de> <h99M8.11172$p56.3361428@newsb.telia.net> <3D0127D2.2060604@gmx.de>
Message-ID: <B9273737.4982%js_nntp@nospam_blackrocksoftware.com>

On 6/7/02 2:38 PM, in article 3D0127D2.2060604 at gmx.de, "Ingo Linkweiler"
<i.linkweiler at gmx.de> wrote:
> thanks a lot for your help.
> 
> Do you know SHORTER solutions in Java?

Here's one:

public class URLTest {
    public static void main(String[] arguments) {
        String myurl = "http://www.uni-dortmund.de";
        try {
            java.io.BufferedReader reader = new java.io.BufferedReader
( new java.io.InputStreamReader((new java.net.URL(myurl)).openStream()) );
            for (String curLine = reader.readLine(); curLine != null;
curLine = reader.readLine())
                System.out.println( curLine );
            reader.close();
        } catch (java.io.IOException ex) { System.out.println(ex); }
    }
}


Instead of reading the url completely before printing, this version prints
line by line. Still, it's pretty close. The java.io package could definitely
use a simple method to download the complete contents of a stream at once.

Below is a similar version:

public class URLTest {
    public static void main(String[] arguments) {
        String myurl = ( "http://www.uni-dortmund.de" );
        try {
            java.io.BufferedInputStream input = new
java.io.BufferedInputStream( (new java.net.URL(myurl)).openStream() );
            java.io.ByteArrayOutputStream output = new
java.io.ByteArrayOutputStream();
            for (int val = input.read(); val != -1; val = input.read())
                output.write( val );
            input.close();
            System.out.println( output );
        } catch (java.io.IOException ex) { System.out.println(ex); }
    }
}

This one is 2 lines longer but it is closer in function to your sample
program.



- JS









From tjreedy at udel.edu  Tue Jun  4 22:23:40 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Wed, 05 Jun 2002 02:23:40 GMT
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <slrnafps1c.4ea.huaiyu@gauss.almadan.ibm.com> <ufqm0voskv86cc@news.supernews.com> <slrnafqppg.4ea.huaiyu@gauss.almadan.ibm.com>
Message-ID: <MCeL8.114006$%y.11961913@bin4.nnrp.aus1.giganews.com>

"Huaiyu Zhu" <huaiyu at gauss.almadan.ibm.com> wrote in message
news:slrnafqppg.4ea.huaiyu at gauss.almadan.ibm.com...
> Objects are referenced by their ids (or equivalent).  Their memory
location
> may or may not change when it is modified, but if it is the same
object,
> every name that was refering to it will remain referencing to it
after the
> modification, regardless of its memory location.

Conceptually, PyObjects do not even *have* a memory location.  In the
current CPython implementation for computers, they have a *fixed*
location, which happens to be used for the unique integer id.  A
(fixed-location) list object contains a pointer to an auxiliary array
which may be moved when resized.  Same for dicts.

Terry J. Reedy





From fperez528 at yahoo.com  Thu Jun 20 20:36:50 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 20 Jun 2002 18:36:50 -0600
Subject: Passing inner functions as arguments
References: <d72779c7.0206201628.33d6a657@posting.google.com>
Message-ID: <aetsf4$sam$1@peabody.colorado.edu>

Pablo Ripolles wrote:

> Hello all!
> 
> I am new to python, so sorry if i am asking the obvious. Here it
> comes. Is there anything wrong with this code?
> 
> # Module to define math tools.
> def momentum(f, k, a, b):
> # Momentum of kth order of the given function f in [a, b].
>     from Quadrature import quadrature
>     def integrand(x):
>     # Momentum's integrand in terms of the passed function f and k.
>         return (f(x)) ** k
>     return quadrature(integrand, a, b)
> 
> I am not sure if it is correct. It runs OK in Python 2.2.1 but not at
> all in Jython 2.1 (java1.4.0). I am not sure about the scoping rules.
> I there's something strange, please let me know!

Nested scopes are not on by default in 2.1 You need AT THE TOP of your file to 
put 'from __future__ import nested_scopes'.

Cheers,

f.

ps. De paso, en ingl?s se usa 'moment/moments' para describir los momentos de 
una funci?n/distribuci?n. 'momentum/momenta' es la palabra cuyo significado 
es el momento mec?nico (masa*velocidad en su descripci?n m?s simple, o 
derivada de la Lagrangiana en general). El problema es que en espa?ol usamos 
'momento' para ambas cosas :)



From phoengeist38259 at arcor.de  Mon Jun  3 13:25:34 2002
From: phoengeist38259 at arcor.de (phoengeist38259 at arcor.de)
Date: Mon, 3 Jun 2002 19:25:34 +0200
Subject: Entschuldigen Sie bitte die Störung!
Message-ID: <E17EvaN-0002FU-00@mail.python.org>

Entschuldigen Sie bitte die St?rung!

Mir ist etwas zu Ohren gekommen.
Eine relativ aussergew?hnliche Ger?chtek?che,
aus der man mir ein schwerverdauliches S?ppchen vorgesetzt hat,
ist der Grund meiner Mail.
Unappetitlich ist gar kein Ausdruck!
Ist es m?glich auf  funktechnischem Wege(in welchen Frequenzbereichen?)
 jemanden zu beeinflussen oder zu manipulieren?
Oder sogar zu schikanieren und terrorisieren?
Unter dem Motto:"Einen am Sender?Nich ganz alleine?
Kleine Mannim Ohr?Falsche Wellenl?nge?Bohnen in den Ohren?
Auf den Zahn gef?hlt(Amalgam)?Mal unverbindlich reinh?ren?
Der Pullacher Wanzentanz?
Ist das Spinnerei?Das geht doch gar nicht,oder?
Und wenn wie sieht das ethisch moralisch aus?
Zur technischen Seite der Sache gibt es zwar Berichte und Webseiten:
Totalitaer,de - Die Waffe gegen die Kritik
http://www.raum-und-zeit.com/Aktuell/Brummton.htm
http://www.fosar-bludorf.com/Tempelhof/
http://jya.com/haarp.htm
http://www.zeitenschrift.at/magazin/zs_24_15/1_mikrowaffen.htm
http://www.bse-plus.de/d/doc/lbrief/lbmincontr.htm
http://home.nexgo.de/kraven/bigb/big3.html
http://w3.nrl.navy.mil/projects/haarp/index.html
http://cryptome.org/
http://www.raven1.net/ravindex.htm
http://www.calweb.com/~welsh/
http://www.cahe.org/
http://www.parascope.com/ds/mkultra0.htm
http://www.trufax.org/menu/mind.html
http://www.trufax.org/menu/elect.html
http://mindcontrolforum.com/
http://www.trufax.org/menu/elect.html
usw.
usw.
usw.
,aber,das kann doch nicht sein,das soetwas gemacht wird,oder?
Eine Menschenrechtsverletzung sonder gleichen!?!
Ist es m?glich,durch Pr?paration,der
Ohren und im Zusammenspiel mit eventuell vorhandenem Zahnersatz?
Mit relativ einfacher Funktechnik??
In diesem Land?Hier und heute???
Unter welchen Motiven?
Wo ist eigentlich die Abteilung  5 des BND und des Verfassungsschutzes?
Kann es sein,da? es Leute gibt,die dem BND/Verfassungsschutz,auf
funktechnischem Wege 
permanent einen Situationsbericht abliefern,ohne es selbst zu merken,im
Kindesalter machbar gemacht??
Werden durch solche inoffiziellen Mitarbeiter,beim BND und
Verfassungsschutz,nach Stasimanier,
Informationen von und ?ber,rein theoretisch, jeden Bundesb?rger,gesammelt?
Gibt es dann noch ein Recht auf  Privatsphere? Wer kontrolliert eigentlich
den BND,MAD und Verfassungsschutz auf Unterwanderung???
In der Mail geht es mir eigentlich um die Frage,ob es kriminellen Elementen,
aus dem Motiv der Bereicherung,oder Gruppierungen aus ideologischen Motiven,
m?glich ist ,sich Wissen und Technik anzueignen,die zu anderen Zeiten,
aus anderen Motiven(Westfernsehen?),entwickelt wurde.
Und stellt der technische Wissensstand,
der der Allgemeinheit bekannt ist  wirklich das Ende der Fahnenstange dar?
Ist es denn nicht kriminellen Elementen genauso m?glich,
ich sage das jetzt mal verharmlost und verniedlichend,
einzelne Personen oder Gruppen mit relativ einfachen Mitteln,
aus welchen Motiven auch immer, auszuspionieren?
Und stellt diese "Ausspioniererei" nicht einen erheblichen Eingriff in die
Privatsph?re dar? 
Ist es m?glich einzelne Personen oder Gruppen,
eine Akzeptans einer gewissen ?ffentlichkeit(suggeriert?),
die z.B. mit Hilfe von Internetseiten,wie zum Beispiel dem
"Pranger"geschaffen werden k?nnte,
mal vorausgestzt,zu terroriesieren und oder zu schikanieren,
und das in aller (suggerierten)?ffentlichkeit?Haben die Leute die da am
Pranger,
oder auf irgendeiner anderen Seite verunglimpft,oder gar Verleumdet werden,
eigentlich eine Chance zur Gegen?ffentlichkeit?Ist das nicht Rufmord?
Vor einigen Jahren bin ich per Zufall auf die Seite "Der Pranger" gesto?en,
damals lief das noch nicht unter dem Deckmantel der Partnervermittlung.
K?nnen sich einzelne Personen,oder Interessengemeinschaften,
aus reinem Selbstzweck,solcher Seiten bedienen,
um unter dem Deckmantel einer fragw?rdigen Zivilkourage,
durch anzetteln irgendwelcher Hetzkampagnen,eigene,
ganz pers?hnliche Interessen durchsetzen?
K?nnen solche Seiten zur Koordination von kriminellen machenschaften dienen?
Die Frage,ist es  M?glichkeit oder Unm?glichkeit,technisch und
gesellschaftlich,
einzelne Personen,oder auch Gruppierungen,aus einer
kriminellen/ideologischen
Energei heraus,zu manipulieren oder zu beeinflussen,terrorisieren oder zu 
schickanieren,und zwar gezielt.
Zielgruppenmanipulation durch Massenmedien sind allt?gliche Manipulation,
der mansich,mehr oder weniger,entziehen kann.
Wird das Recht auf Privatsph?re,schleichend,tiefenpsychologisch,
durch Sendungen,wie,zum Beispiel "Big brother",untergraben?
Sollte bei einem der Angemailten ein gewisser Wissensstand zum Thema
vorhanden sein,
w?re ich ?ber Hinweise zum Thema froh.
Auf der Suche nach Antworten auf meine Fragen
maile ich verschiedene Adressen aus dem Internet an,
und hoffe aufkonstruktive Antworten und Kritiken.
?ber einen Besuch auf der Seite
<http://hometown.aol.de/reinerhohn38259/homepage/index.html>
w?rde ich mich freuen.
Sollten Sie von mir mehrfach angeschrieben worden
sein,so bitte ich Sie,mir dies zu entschuldigen,
das war nicht beabsichtigt.
Der Grund f?r meine Anonymit?t ist die Tatsache,
da? bei derlei Fragenstellerei,
verst?ndlicherweise,schnell der Ruf nach der Psychatrie laut wird.
Was auch Methode hat(ist).
Sollten Sie die Mail als Bel?stigung empfinden,
m?chte ich mich hiermit daf?r entschuldigen!
Big brother is watching you?


Excuse please the disturbance!

Me something came to ears.
A relatively unusual rumor kitchen,
from which one put forward to me a heavydigestible soup,
is the reason of my Mail.
Unappetizing is no printout!
Is it possible on radio Wege(in for which frequency ranges?)  to
influence or manipulate someone?
Terrorize or to even chicane and?
Under the Motto:"Einen at the Sender?Nich quite alone?
Small Mannim Ohr?Fal Wellenlaenge?Bohnen in the ears?
On the tooth clean-hear gefuehlt(Amalgam)?Mal witthout obligation?
The Pullacher bug wanzentanz?
Isn't the Spinnerei?Das goes nevertheless at all, or?
And if as looks ethicalally morally?
For the technical page of the thing there is to report and web page:
Totalitaer,de - Die Waffe gegen die Kritik
http://www.raum-und-zeit.com/Aktuell/Brummton.htm
http://www.fosar-bludorf.com/Tempelhof/
http://jya.com/haarp.htm
http://www.zeitenschrift.at/magazin/zs_24_15/1_mikrowaffen.htm
http://www.bse-plus.de/d/doc/lbrief/lbmincontr.htm
http://home.nexgo.de/kraven/bigb/big3.html
http://w3.nrl.navy.mil/projects/haarp/index.html
http://cryptome.org/
http://www.raven1.net/ravindex.htm
http://www.calweb.com/~welsh/
http://www.cahe.org/
http://www.parascope.com/ds/mkultra0.htm
http://www.trufax.org/menu/mind.html
http://www.trufax.org/menu/elect.html
http://mindcontrolforum.com/
http://www.trufax.org/menu/elect.html
usw.
usw.
usw.
but, that cannot be nevertheless, which is made soetwas, or?
A violation of human rights resemble special!?!
Is it possible, by preparation, the ears and in interaction with
possibly available artificial dentures?
With relatively simple radio engineering??
In this Land?Hier and today???
Under which motives?
Where is the department actually 5 of the BND and the protection of the
constitution?
Can it be that there are people, which deliver the Federal
Intelligence Service/protection of the constitution, on radio way
permanently a situation report, without noticing it, in the infancy
feasiblly made?
By such unofficial coworkers, with the BND and protection of the
constitution, after Stasimanier, is information collected of and
over,purely theoretically, each Federal citizen?
Is there then still another right to Privatsphere?
Who actually checks the BND, WAD and protection of the constitution for
infiltration???
Into the Mail actually concerns it to me the question whether it
criminal items, from which motive of enriching, or groupings from
ideological motives is possible, to acquire itself knowledge and
technique which were developed at other times, from other
Motiven(Westfernsehen?).And does the technical knowledge status place, to
that the public admits is really the end of the flag bar?
Is it not to criminal items just as possible, I legend that now times
played down and does nice-end, individual persons or groups with
relatively simple means, to spy from whatever motives always?
And doesn't this " Ausspioniererei " represent a substantial
intervention into the privatsphaere?
It is possible individual persons or groups, one acceptance to of a
certain Oeffentlichkeit(suggeriert?),  e.g. by Internet pages, how for
example the " Pranger"geschaffen could become, times vorausgestzt, to
terroriesieren and or chicane, and in everything (the people
suggerierten)Oeffentlichkeit?Haben there at the Pranger, or on any
other page to be reviled, or slandered, actually a chance to the
Gegenoeffentlichkeit?Ist that not character assassination?
Some years ago I am by coincidence the page " the Pranger "
encountered, at that time ran not yet under the cover of the partner
switching.Itself can individual persons, or communities of interests, from
pure self purpose, such pages to serve, over under the cover of a doubtful
Zivilkourage, through plot any rushing campaigns, own, quite
persoehnliche interests to intersperse?
Can such pages serve for the co-ordination of criminal machinations?
The question, is it possibility or impossibility, technically and
socially, individual persons, or also groupings of manipulating or of
influencing from an criminal/ideological Energei, terrorizes or to
schickanieren, directed.Target group manipulation by mass media are
everyday manipulation, from which, more or less, can extract itself.
Does the right to privatsphaere, creeping, by transmissions become
deep psychological, how, for example " Big undermine brother"?
If the Angemailten should be available a certain knowledge status to
the topic with one, I would be glad over notes to the topic
On the search for responses to my questions maile I different
addresses from the Internet on, and hope up-constructional responses
and criticisms.Over an attendance on the page
<http://hometown.aol.de/reinerhohn38259/homepage/index.html>
wuerde I are pleased.If you should have been written down by me several
times, then please
I you to excuse me this that was not intended.
The reason for my anonymity is the fact that with such
Fragenstellerei, understandably, fast after the call the Psychatrie
loud becomes.  Which also method hat(ist).
If you should feel the Mail as annoyance, I would like to apologize
hereby for it!  Big is watching you?


Veuillez excuser le d?rangement!

Moi quelque chose concernant des oreilles est venu.
   Une cuisine de bruit relativement inhabituelle, dont on m'a plac? un
Sueppchen schwerverdauliches devant, est la raison de mes Mail.Aucune
expression n'est  peu app?tissante!
   Il est possible sur un Wege(in funktechnischem pour quelles r?ponses
fr?quentielles?)  quelqu'un influencer ou manipuler?
Ou m?me  schikanieren et terroriser?
   Sous le Motto:"Einen au Sender?Nich tout ? fait seulement?
   Petits Mannim Ohr?Falsche Wellenlaenge?Bohnen dans les oreilles?
   Sur la dent gefuehlt(Amalgam)?Mal non contraignant reinhoeren?
   Le Pullacher Wanzentanz?
Le Spinnerei?Das n'est-il quand m?me pas du tout va, ou?
   Et si comme cela para?t ?thiquement moralement?
   Au c?t? technique de la chose, il y a certes des rapports et des
Webseiten:
Totalitaer,de - Die Waffe gegen die Kritik
http://www.raum-und-zeit.com/Aktuell/Brummton.htm
http://www.fosar-bludorf.com/Tempelhof/
http://jya.com/haarp.htm
http://www.zeitenschrift.at/magazin/zs_24_15/1_mikrowaffen.htm
http://www.bse-plus.de/d/doc/lbrief/lbmincontr.htm
http://home.nexgo.de/kraven/bigb/big3.html
http://w3.nrl.navy.mil/projects/haarp/index.html
http://cryptome.org/
http://www.raven1.net/ravindex.htm
http://www.calweb.com/~welsh/
http://www.cahe.org/
http://www.parascope.com/ds/mkultra0.htm
http://www.trufax.org/menu/mind.html
http://www.trufax.org/menu/elect.html
http://mindcontrolforum.com/
http://www.trufax.org/menu/elect.html
usw.
usw.
usw.
toutefois qui ne peut quand m?me pas ?tre qui on fait soetwas, ou?
   Une violation des droits de l'homme s?parer ressembler!?!
   Il est possible, par la pr?paration, des oreilles et dans l'effet avec
la proth?se dentaire ?ventuellement existante?
Avec la technique de radio relativement simple??
   Dans ce Land?Hier et aujourd'hui
   Sous quels motifs?
   O? le d?partement est-il en r?alit? 5 du BND et de la protection
d'constitution?
peut il ?tre qu'il y a les personnes qui livrent en permanence le
BND/Verfassungsschutz, de mani?re funktechnischem un rapport de situation,
sans le remarquer le -m?me , dans l'enfance rendu possible??
Par de tels collaborateurs officieux, avec le BND et la protection
d'constitution, apr?s mani?re, des informations sont-elles rassembl?es et
plus de, purement th?oriquement, chaque citoyen allemand?
   Il y a alors encore un droit ? des Privatsphere?  Qui contr?le en
r?alit? le BND, mad et protection d'constitution sur une infiltration???
Il s'agit en r?alit? dans le Mail me la question de savoir si lui ?l?ments
criminels, dont le motif de l'enrichissement, ou de groupements des motifs
id?ologiques, possible de s'acqu?rir le savoir et la technique qui ?
d'autres temps, est autre MotivenEt place-t-il le savoir technique dont le
public vraiment la fin la barre de drapeau a connaissance ?
   Il n'est pas donc exactement la m?me chose possible pour des ?l?ments
criminels, moi cela maintenant fois verharmlost et minimisant une l?gende,
personnes ou groupes particuliers avec des moyens relativement simples, de
quels motifs aussi toujours, auszuspionieren?(Westfernsehen?), a ?t?
d?velopp?.
Et ce "Ausspioniererei" ne repr?sente-t-il pas une intervention
consid?rable dans la vie priv?e?
   Il est possible personnes ou groupes particuliers, pour certain
Oeffentlichkeit(suggeriert?),  celui p. ex. ? l'aide des c?t?s Internet,
comme par exemple "le Pranger"geschaffen pourrait, fois vorausgestzt
schikanieren  terroriesieren et ou ,
et qui toute (suggerierten)Oeffentlichkeit?Haben les personnes ceux l?, ou
d'un autre c?t? verunglimpft, ou  on ne pas calomnie, en r?alit? une
chance au Gegenoeffentlichkeit?Ist qui meurtre d'appel?
Il y a quelques ann?es, je ne suis pas encore par hasard sur le c?t?
"celui" pouss?, fonctionnais alors cela sous la couche de pont de
l'entremise partenaire.
   Des personnes particuli?res, ou des communaut?s d'int?r?ts le
peuventelles, d'un autobut pur, de tels c?t?s servent, sous la couche de
pont d'un Zivilkourage douteux, tracent plus de  des campagnes de
pr?cipitation, propres int?r?ts tout ? fait persoehnliche entrem?lent?
De tels c?t?s peuvent-ils servir ? la coordination des manoeuvres
criminelles?
Question, est lui possibilit? ou impossibilit? de manipuler ou
d'influencer  techniquement et socialement, particuli?re personnes, ou
aussi groupements, criminelle/ponctuel id?ologique Energei dehors, ,
terroriser ou  schickanieren, et ce.Une manipulation de groupe cible par
des masse-m?dias ?tre la manipulation quotidienne qui peut extraire
mansich, plus ou moins.
   Le droit ? la vie priv?e est-il min?, ramment, tiefenpsychologisch, par
des envois, comme, par exemple "des Big brother"?
   Avec un les Angemailten si un certain savoir devait exister sur le
th?me, je serais heureux sur des indications sur le th?me.Sur la recherche
des r?ponses ? mes questions je diff?rentes adresses maile d'Internet
dessus, et esp?re r?ponses et critiques aufkonstruktive.
   Sur une visite du c?t?
http://hometown.aol.de/reinerhohn38259/homepage/index.html>
je me r?jouirais.
   Si vous deviez avoir ?t? ?crit ? diff?rentes reprises par moi, je vous
demande de m'excuser cela  qui n'?tait pas envisag?.
La raison de mon anonymat est le fait qu'avec telle des Fragenstellerei,
l'appel devient ce qui est bien compr?hensible, rapidement bruyant apr?s
le Psychatrie.
   Ce que la m?thode a ?galement (ist).
   Si vous deviez ressentir les Mail comme un ennui, je voudrais m'excuser
par ceci pour cela!
   Big brother is watching you?





From hammond at U.Arizona.EDU  Fri Jun 21 11:17:26 2002
From: hammond at U.Arizona.EDU (Michael T Hammond)
Date: Fri, 21 Jun 2002 08:17:26 -0700
Subject: python, xml, windows, sunos
Message-ID: <Pine.SOL.4.44.0206210813510.27754-100000@bast.u.arizona.edu>

Hi.

I've got a program that parses an xml file. It runs just fine under the
most recent python (from www.python.org) on windows, but I can't get it to
run on my sun machine. I have version 2.2.1 installed there. I'm not sure
it's necessary, but I have Pyxml installed there as well (as on the
windows machine).

The error I get is:

xml.sax._exceptions.SAXReaderNotAvailable: No parsers found.

Does anybody have any idea what to do here?

Mike H.



From irmen at NOSPAMREMOVETHISxs4all.nl  Mon Jun 10 15:29:51 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Mon, 10 Jun 2002 21:29:51 +0200
Subject: M2Crypto: select() behaves weird on SSL socket
References: <3D04ECED.2030202@NOSPAMREMOVETHISxs4all.nl> <slrnag9rie.enn.akuchlin@ute.mems-exchange.org> <3D04FBF5.8090407@NOSPAMREMOVETHISxs4all.nl>
Message-ID: <3D04FE2F.6070302@NOSPAMREMOVETHISxs4all.nl>

To add some more info...

> A.M. Kuchling wrote:
> 
>> Perhaps there really is no more data on the socket because OpenSSL has
>> read it all, but the decrypted data is now sitting in a buffer

The manual page of select(2) says:
"select  waits  for  a number of file descriptors to change status.
  Three independent sets of descriptors are watched.   Those
  listed  in  readfds  will  be watched to see if characters
  become available for reading (more precisely, to see if  a
  read  will not block - in particular, a file descriptor is
  also ready on end-of-file),  those  in  writefds  will  be
  watched  to  see  if  a write will not block, and those in
  exceptfds will be watched for exceptions."

The way I interpret this is that -buffering or not- select must
return a FD if there is more data available. In my case, the SSL
socket indeed has more data available, and a recv() on the socket
will not block. I therefore think that either M2Crypto or OpenSSL
is doing the wrong thing because select() doesn't behave as it should?
The recv() may read from a buffer, I don't care, but it doesn't
block on my next call and thus I expect select() to return the socket
as being ready for reading... am I wrong here?


irmen



From smst at bigfoot.com  Mon Jun  3 15:18:49 2002
From: smst at bigfoot.com (Steve Tregidgo)
Date: 3 Jun 2002 12:18:49 -0700
Subject: Help for Tkinter and python's programming
References: <mailman.1022985123.22656.python-list@python.org> <adc8kj$1ovk$1@nntp6.u.washington.edu>
Message-ID: <fdffd27c.0206031118.2c2a9301@posting.google.com>

"Russell E. Owen" <owen at nojunk.invalid> wrote in message news:<adc8kj$1ovk$1 at nntp6.u.washington.edu>...
[snip to code fragment of GC class]
>     def __call__(self, *lastArgs, **kwArgs):
>         if kwArgs:
>             netKWArgs = self.__firstKWArgs.copy()
>             netKWArgs.update(self.__kwArgs)
>         else:
>             netKWArgs = self.__firstKWArgs 
>         return self.__callback (*(self.__firstArgs + lastArgs), 
> **netKWArgs)

I think this code contains a mistake.  'self.__kwArgs' should probably
be simply 'kwArgs' (4th line above).  Convenient class, anyhow!

Cheers,
Steve


From marco.stagnoNOSPAM at libero.it  Mon Jun 10 16:31:58 2002
From: marco.stagnoNOSPAM at libero.it (Marco Stagno)
Date: Mon, 10 Jun 2002 20:31:58 GMT
Subject: How convert all list/tuple...
Message-ID: <523aguoljk3qceh4kkki7emn6283d6sqpt@4ax.com>

Hi
There is a fast way to convert all item in a tuple/list into a string type,
without do a loop?

I wish to translate something like:
 [1,2,"string"]
into:
 ['1','2','string']

in the fastest way

thank you in advance

                                  MAS!



From kragen at pobox.com  Mon Jun 10 18:08:27 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 18:08:27 -0400
Subject: What is a classmethod for?
References: <addgv30f4d@enews4.newsguy.com> <Pine.LNX.4.44.0206031559001.8474-100000@palanthas.neverending.org>
Message-ID: <83y9dmokh0.fsf@panacea.canonical.org>

Frank Tobin <ftobin at neverending.org> writes:
> On Sun, 2 Jun 2002, Dennis Peterson wrote:
> > That's nice, but I can't think of a use for it, given that we have
> > inheritance and isinstance(). Can anyone enlighten me?
> 
> Personally, I like using classmethods as secondary constructors, generally
> gones that are able to construct 'from' something.
> 
> e.g.:
> 
> class HostPortAddress(tuple):
> 	def from_str(self, s):
> 		return self(s.split(':', 1))
> 	from_str = classmethod(from_str)

That's so you don't have to write this?

class HostPortAddress(tuple): pass
def HostPortAddress_from_str(s): 
    return HostPortAddress(s.split(':', 1))

Are there times when it's more compelling?



From mcfletch at rogers.com  Thu Jun  6 22:45:51 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Thu, 06 Jun 2002 22:45:51 -0400
Subject: Medium to Large Scale Python Deployments
References: <mailman.1023405789.30972.python-list@python.org> <aJTL8.21666$ZF5.2505752733@newssvr10.news.prodigy.com> <3D000B36.5C7BCEBF@engcorp.com>
Message-ID: <3D001E5F.3000808@rogers.com>

Hey, that's a fun little tool (pycount)...

I seem to create projects in libraries of around 20KLOC.  Do projects 
accumulate the LOCs of their used libraries though?  Is it only 
libraries you have created that count?  If so, I've got 2 projects of 
~50KLOC (the top-level projects, and one of the ~20KLOC libraries were 
closed-sourced, the rest were my open-source work).

If used libraries don't count, most of my "projects" are ~ 5 thousand 
lines of glue code.

Ah well, back to work, thanks for the diversion,
Mike

Peter Hansen wrote:
> "Domenic R. Merenda" wrote:
> 
>>"Sam Penrose" <spenrose at intersight.com> wrote:
>>
>>>I'm also curious about large amounts of Python, period. Who knows of
>>>~50K Python lines-of-code (LOC) systems? 100K LOC? More?
...
> Running pycount (http://starship.python.net/crew/gherman/playground/pycount/ )
> as "pycount.py -R *.py" in the top level of your source tree would tell
> you something more accurate if you wished.
> 
> -Peter
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From sandeep182 at hotmail.com  Fri Jun 14 19:03:21 2002
From: sandeep182 at hotmail.com (Sandeep Gupta)
Date: Fri, 14 Jun 2002 23:03:21 GMT
Subject: Python and SQL Server 2000
Message-ID: <ZCuO8.10006$9b.678135@typhoon.austin.rr.com>

I'm running python 2.2 on Redhat 7.2 and I need to access SQL Server 2000.
The command are simple SQL queries and updates.  We don't even use joins.

There seem to be several common methods for doing so:
1. Use the python application at
http://www.object-craft.com.au/projects/mssql/ with Sybase ASE 11.9.2
2. Use the python application at
http://www.object-craft.com.au/projects/mssql/ with FreeTDS.
3. Use mxODBC with iODBC
4. Use mxODBC with unixODBC

It is not clear to me if I will still need to use Sybase ASE or FreeTDS with
iODBC/unixODBC.

Which option do people recommend?  Are there any other options?

Thanks
Sandeep




From aahz at pythoncraft.com  Thu Jun 13 16:08:40 2002
From: aahz at pythoncraft.com (Aahz)
Date: 13 Jun 2002 16:08:40 -0400
Subject: Type subclassing: bug or feature
Message-ID: <aeau48$111$1@panix1.panix.com>

Consider the following code:

class MyStr(str):
    def contains(self, value):
        return self.find(value) >= 0

s = MyStr("hello, world!")
s = s.capitalize()
if s.contains('Hello'):
    print "Found it!"

It fails with an AttributeError when it calls s.contains(), because
s.capitalize() returned a str instead of a MyStr.  Anyone want to take a
whack at defending this as the correct behavior?
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From johnroth at ameritech.net  Sat Jun 29 18:34:38 2002
From: johnroth at ameritech.net (John Roth)
Date: Sat, 29 Jun 2002 18:34:38 -0400
Subject: private
References: <pan.2002.06.29.17.38.51.789527.440@myself.com> <3D1DE132.4A890D9D@engcorp.com> <afl3e4$f8abc$1@ID-71831.news.dfncis.de>
Message-ID: <uhsddip50prt0b@news.supernews.com>

"James Kew" <james.kew at btinternet.com> wrote in message
news:afl3e4$f8abc$1 at ID-71831.news.dfncis.de...
>
> I don't -- in my admittedly very limited experience -- find Python
classes
> as immediately self-documenting. It's not obvious from the class
definition
> what is intended to be called or modfied by the client and what is
internal
> to the implementation.

The fact is, there is no such thing as a 'self-documenting' language.
COBOL demonstrated that to me a long time ago. If you want your
code to say something to someone, it's up to the two (or however many)
of you to set up coding conventions that communicate the intent - and
stick to them.

There are a number of such conventions within the Python community,
including the underscore for internal methods.

As a number of posters have mentioned, the issue has been
discussed many times. The general feeling is that access protection
attributes (and similar things) are not what Python is about.

Python does not support declarations at all. Some people feel this
is a weakness. Some people feel it is a strength. In general, the kinds
of problems which declarations are intended to solve are less visible
in Python than in compiled languages.

Hope this helps give you some background.

John Roth




From stever at gate.net  Fri Jun 21 23:44:46 2002
From: stever at gate.net (Steve Reeves)
Date: 22 Jun 2002 03:44:46 GMT
Subject: telnetlib option negotiation callback should be a method
Message-ID: <slrnah7sle.4pu.stever@debian.skree.com>

I have an application using telnet that requires option subnegotiation
(specifically, it needs to set a terminal type in the manner of RFC
1091).  The standard telnetlib doesn't currently support this.  I
would like to extend it to do so, but the current design of the option
negotiation callback makes this difficult.

Currently, the callback is a function taking a socket as an argument.
This is adequate for sending data to the server, but not for
receiving.  It bypasses the Telnet object's internal queue, which may
already contain the data you're waiting for.  It also requires you to
reimplement the special processing the protocol requires (undoubling
IAC characters), and the default negotiation behavior.

The callback really needs to be a method of the Telnet object, with
access to the internal queue.  It can implement the default DONT/WONT
negotiation.  Subclasses can override it to handle the specific
options they want and forward the rest to the base class.  The
set_option_negotiation_callback() method would no longer be needed.

(This question of function vs. method was briefly raised when the
telnetlib patch was first submitted, <http://www.python.org/sf/416079>,
but it was left as a function.)

Opinions?

-- 
Steve Reeves
stever at gate.net


From logiplexsoftware at earthlink.net  Fri Jun 21 14:17:33 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Fri, 21 Jun 2002 11:17:33 -0700
Subject: ssl on windows
In-Reply-To: <uh6qucjo5omi5e@corp.supernews.com>
References: <uh6qucjo5omi5e@corp.supernews.com>
Message-ID: <20020621111733.2cbf3463.logiplexsoftware@earthlink.net>

On Fri, 21 Jun 2002 13:09:33 -0500
Scott Hathaway wrote:

> I would like to use ssl on Windows with Python 2.2.  Is this available yet?
> If so, how can I go about getting it installed?  When I try to get an https
> address with urllib, it tells me that https is an unknown protocol.

Google turned up the following for python+ssl+windows:

http://mail.python.org/pipermail/python-announce-list/2002-March/001312.html

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From uwe at rocksport.de  Sat Jun  1 19:31:28 2002
From: uwe at rocksport.de (Uwe Schmitt)
Date: 1 Jun 2002 23:31:28 GMT
Subject: wxPython performance
References: <slrnafhmoa.389.ak@ak.silmarill.org>
Message-ID: <adblgg$dmc37$1@hades.rz.uni-sb.de>

Andrei Kulakov <ak at silmarill.org> wrote:
| Hello,

|     "hello world" type program from wxPython tutorial takes 6
| seconds to start on my cel 366 128mb system (running Debian). Is
| this typical? Why is it so slow? I'm looking for a graphic
| toolkit that'd be fairly quick. I don't want to use Tk because of
| the looks.

I had the same problem which appeared when I used py2exe.
The generated exe just neads one second instead of four...

Greetings, Uwe.


-- 

Dr. rer. nat. Uwe Schmitt      ICQ# 159647634   Uwe.Schmitt at num.uni-sb.de


From gcordova at hebmex.com  Thu Jun  6 18:39:21 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 6 Jun 2002 17:39:21 -0500 
Subject: SNIP IT: Object gatekeeper for multithread apps.
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F062330116F332@mail.hebmex.com>

Hi y'all.

I've been toying around with multithreading apps,
and one of the (small) deficiencies I've seen is
that, besides Queue, there's no other object gatekeeper
to regulate access to shared objects.

Now; Queue is quite cool, very nice. But, because of
it's design (I mean, it *is* a queue), only the objects
at the top can be accessed (or the bottom... you know).

Also, they have no name. I didn't like that.

So, I began writing my own gatekeeper. I've been
through many iterations, each one getting killed because
of my inherent developer paranoia ("And, what if one
thread does this... *damn*, another deadlock posibility").

Until this version. It seems as clean as it's ever gonna
get, so I'm sending it out to the wild, in hopes that
it may prove useful to somebody else.

Please, enjoy.

Also, I'm taking bug reports, *possible* bugs,
design deficiencies (or improvements!), and also,
please please please, a better name :-) "Holder"
just doesn't ring right.

So, please, enjoy.

-gustavo

pd: Ah! This is 2.2+ specific, because it uses properties.
    Sorry for all you pre-2.2 programmers. This is my
    properties self-test-exam. I Passed!! ;-)

-------------- next part --------------
A non-text attachment was scrubbed...
Name: Holder.py
Type: application/octet-stream
Size: 6394 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020606/e0b33d94/attachment.obj>

From jarvin24 at tutor.cc.tut.fi  Fri Jun 28 04:05:30 2002
From: jarvin24 at tutor.cc.tut.fi (=?iso-8859-1?Q?J=E4rvinen?= Petri)
Date: Fri, 28 Jun 2002 08:05:30 +0000 (UTC)
Subject: Packages +reload()
Message-ID: <slrnaho66b.3of.jarvin24@tutor.cc.tut.fi>

Hey, 

I feel like complite lost. I have created nice package structure,
 but I encountered lot's of troubles when using it with IIS and ASP.

Structure is:

	eNAct\
		documents\
			listDocuments\
				showPage()
		enactXML()

-> main package eNAct and sub-packages eg. documents and it's
 sub-packages eg. listDocuments

The problem is that I can't reload the modules. 

I have tried various ways of importing listDocuments but IT doesn't 
get reloaded.

eg.
from eNAct.documents import listDocuments
reload(listDocuments)

listDocuments.showPage()

or

import eNAct
reload (enact)

eNAct.documents.listDocuments.showPage()

Questions:

1. How should the structure be?
2. Is it possible that I only reload the eNAct and all sub-modules
   get reloaded.
3. What does reload(eNAct.documents) reload?


- Petri J?rvinen


From dkuhlman at rexx.com  Sat Jun 29 14:45:32 2002
From: dkuhlman at rexx.com (Dave Kuhlman)
Date: Sat, 29 Jun 2002 11:45:32 -0700
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com> <mailman.1025336262.12245.python-list@python.org> <3d1dc547.468619140@netnews.attbi.com>
Message-ID: <afkv73$f2274$1@ID-139865.news.dfncis.de>

candiazoo at attbi.com wrote:

[snip]

Thank you for posting this.  Very helpful.

Could we have a bit of evaluation and explanation please.  I'm 
suspicious about whether some of these changes actually made a 
difference.  I'd like a little understanding before I start 
changing my Python coding style.

I'd also like some enlightenment about *why* they would make an 
improvement.

>  my aforementined optimizations...
> 
> 1) "import sys" became "import exit from sys"
> 2) "import os" became "import makedires from os"

Does importing a function from a module (as opposed to importing 
the whole module) save a look-up each time the function is called?

> 3) "import time" bacame "import asctime from time" (this one is
> used frequently) 
> 4) I localized constants I was using in boolean
> tests (there were declared at the beginning of the module, I moved
> them to the function call just before the loop).

Is finding things in globals slower than finding them in locals?  
Did this change in Python 2.2?

> 5) I localized self.__source_id...  source_id = self.__source_id
> before using it in every loop iteration.
> 6) I localized self.fix_date_range()... fix_date =
> self.fix_date_range before using it in every loop iteration/
> 7) I changed my method that built the output string by changing...
> 
> return self.val_a + self.value_b + etc..
> 
> ...to...
> 
> return "%-10.10s%-200.200s (etc)" % (self.val_a, self.val_b, etc.)
> 
> ...and those were basically the changes I made!  They made a huge
> difference!
> 
> Mike J.

Thanks in advance.

  - Dave

-- 
Dave Kuhlman
dkuhlman at rexx.com
http://www.rexx.com/~dkuhlman



From eric.brunel at pragmadev.com  Thu Jun 13 10:40:41 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Thu, 13 Jun 2002 14:40:41 +0000
Subject: Question about asksaveasfilename
References: <ed0O8.23891$nZ3.2215@rwcrnsc53>
Message-ID: <aea3ha$lv$1@wanadoo.fr>

Mike Callahan wrote:

> I am using asksaveasfilename from tkFileDialog. I want to restrict my
> users to save filenames with these extensions .unl or .sql. This is my
> call:
> 
> from tkFileDialog import asksaveasfilename
> fn = asksaveasfilename(filetypes=[('Unloaded', '*.unl'),
> ('Queries','*.sql')])
> 
> This gives me the correct dialog box, but how do I force the correct
> extension on fn?  Thanks.

There's a defaultextension option, but it does not depend on the selected 
file type, i.e. you'll have to do either:

fn = asksaveasfilename(filetypes=[('Unloaded', '*.unl'),
                                  ('Queries',  '*.sql')],
                       defaultextension='.unl')

or:

fn = asksaveasfilename(filetypes=[('Unloaded', '*.unl'),
                                  ('Queries',  '*.sql')],
                       defaultextension='.sql')

which isn't what you're looking for, I fear...

I don't know any means to force an extension depending on the selecetd file 
type with asksaveasfilename, so my advice would be to split the two save 
options: one "Save unloaded" and one "Save query"... But if there's another 
solution, I'm interested too ;-).

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From marklists at mceahern.com  Sat Jun 29 12:00:30 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Sat, 29 Jun 2002 11:00:30 -0500
Subject: exception handing
In-Reply-To: <pan.2002.06.29.17.41.28.926443.440@myself.com>
Message-ID: <NCBBLFCOHHDIKCAFGCFBMECHLHAA.marklists@mceahern.com>

> >>> ex1 = "spam"
> >>> try:
> ...    raise ex1
> ... except ex1:
> ...    print 'got it'

The problem is that you are using strings as exceptions--that is
discouraged.  Instead, raise a built-in exception or subclass the Exception
object.

class ex1(Exception):pass
class ex2(ex1):pass

try:
	raise ex2("some ex2 error")
except ex1, e:
	print e

// m

-





From peter at engcorp.com  Sun Jun  9 13:38:13 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 09 Jun 2002 13:38:13 -0400
Subject: How to capture stdout into a string?
References: <2YLM8.905$Yw3.119357@news20.bellglobal.com>
Message-ID: <3D039285.72DFC71D@engcorp.com>

Michael Davis wrote:
> 
> Hi,
> 
> I'm using libftp. The FTP.dir() function gets the remote directory listing
> and prints it to standard out. How can I capture that in a string?
> 
> If I were using an external process, I know I could open a pipe to it. But
> it's not an external process.
> 
> I'm sure there's a tricky way of doing it by getting the file descriptor for
> stdout and reading from it directly, but I was hoping to avoid going that
> low-level. I will if I must though.

The best option by far (and really the easiest) to figure out
stuff like this is to check the source.

In this case, you can see that dir() calls retrlines() with
the LIST argument and no callback function.  As the notes
say, this directs the output to stdout.

You must specify a callback function of some kind.  The best
is probably a small class with a __call__ method, but this
little baby seems to work too, although it may not be quite
enough to work in the general case yet.  It grabs lines that
are passed to it into a list, until it is called with an
empty argument, at which point it returns the saved list
and deletes it, ready for another bunch of lines:

>>> def grablines(line='', lines=[]):
...    if line:
...       lines.append(line)
...    else:
...       result = lines[:]
...       del lines[:]
...       return result
...
>>> grablines()
[]
>>> grablines('test')
>>> grablines('test')
>>> grablines('test')
>>> grablines()
['test', 'test', 'test']
>>> ftp.dir(grablines)
>>> grablines()
['total 62', '-rw-r--r--  1 engcorp  netdial      0 Jan 11  2000 .addressbook',
'-rw-------  1 engcorp  netdial   2285 Jan 11  2000 .addressbook.lu', '-rw------
-  1 engcorp  netdial   6459 May 21 23:42 .bash_history', '-rw-r--r--  1 engcorp
  netdial     29 Jun 17  2001 .forward', '-rw-------  1 engcorp  netdial  12221
.............]


-Peter


From dom at edgereport.put_a_c_o_m_here  Wed Jun  5 05:18:34 2002
From: dom at edgereport.put_a_c_o_m_here (Domenic R. Merenda)
Date: Wed, 05 Jun 2002 09:18:34 GMT
Subject: Medium to Large Scale Python Deployments
Message-ID: <KHkL8.21383$eD2.2399923271@newssvr10.news.prodigy.com>

Greetings,

   I am running a home-grown Enterprise Resource Planning system, written
entirely in Python, to manage the operations and manufacturing aspects of a
medium ($100,000,000 annual sales) manufacturing organization.  I would be
interested in hearing from other individuals who are likewise "betting the
farm" on Python as an ERP or MRP solution, and some of the experiences they
may have had.

   We have found Python to be a robust solution for the problems we've
encountered to date, and pass approximately half a gigabyte of daily data
through our custom systems.  We have merged the AS/400 (Python 2.1, soon to
go to 2.2.1) with midrange (PIII700) PC's to create an interesting
heterogenous environment.


Cheers!
--
Domenic R. Merenda
Editor, The Edge Report
http://www.edgereport.com




From never at mind.info  Sun Jun  2 16:54:29 2002
From: never at mind.info (G. Willoughby)
Date: Sun, 2 Jun 2002 21:54:29 +0100
Subject: Python Turing machine
References: <d9akfu0c7ro0sebio1vd9c6ik2i0fg59le@4ax.com> <Xns9221CF0CC998Dcliechtigmxnet@62.2.16.82>
Message-ID: <ade0mc$j4r$1@newsg3.svr.pol.co.uk>

> hint: post some URLs of those implementations to the list. that way we
> don't need to look up what it is and people here have fun trying to solve
> things pythonic...

I belive he means a chat bot that tries to fool humans into beliving its a
real person.

--G. Willoughby




From jadestar at idiom.com  Tue Jun 11 05:28:58 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 11 Jun 2002 09:28:58 GMT
Subject: python, shell, environment variable
References: <a7c2c93c.0206102219.28e2d2f7@posting.google.com>
Message-ID: <ae4fsq$mtt$1@news.idiom.com>

oliver <fwang2 at yahoo.com> wrote:

> hi, folks,

> I am trying to get a set of python and bash scripts to work, and I
> constantly run into trouble with environment variables:

> First, my python script need to modify environment variable, say,
> PATH, and want to keep the modification even after the script is done.
> os.environ["PATH"]= ... doesn't seem to work, any idea?

 Yes.  You seem to have a fundamental misunderstanding about how
 environment variables work (under UNIX and UNIX-like OSes, at least).

 The environment is a region of your process' memory that is not
 overwritten by an exec*() system call.  Thus it is an efficient way
 for a parent process to pass some (textual or textually encoded) data
 to its children.  The environment is *NOT* a share memory mechanism.
 
 To execute Python your shell performed a fork() and then an exec*()
 (some form of exec; it's not important which one).  fork() creates a
 new private address space for your new process, and address space that
 is initially an (almost) exact copy of the parent's memory (except for
 the portion of the stack or heap that's holding the return value to 
 the fork() call itself).  [Please understand that this is a new
 "virtual" address space; the actual physical memory pages are usually
 shared through page table aliasing on modern processors/architectures
 running modern kernels --- using a technique called CoW: copy on write].

 So your assignments to os.environ["PATH"]=something are just happening
 to your address space.  They can't affect your parent's memory (that
 would be a SEGV, segmentation violation; or it would require some form
 of shared memory).  The assignments would be visible in any processes
 you created *under* your python process (os.popen('printenv') and search
 for the PATH entry for one way to demonstrate that), and they should
 remain set in that process (that was python) if you finish your script
 with one of the os.exec*() functions (except for the exec*e() which
 require you to explicitly prepare and pass a new environment to them).

> Second, it would be better if my python script can call bash shell and
> still keep the environment variable modification done by bash. I tried
> os.popen("source some_shell"), it doesn't work. ?

 os.popen is creating yet another process; and the source command
 therein is modifying popen's memory.

> Thanks for help.

> -oliver

 This is a fundamental consequence of UNIX' fork()/exec*() and 
 memory management/protection model.  It is a basic principle that
 leads to several FAQs.  It is something I have to emphasize at 
 considerable length in my shell scripting classes.  It is also one
 of the best examples of how important it is to understand the big
 picture, the "model" of an architecture, an OS, a language, or any
 other technology that one must control and use in any but the most
 trivial ways.  Once you understand how fork(), exec*() and *NIX
 memory management works then you'd never have this question --- because
 the answer would be so obvious.  (When my son tattoos his arm, why
 can't my grandpa read it off *his* arm?  Because it's a different skin
 that's being modified!).
 



From mcherm at destiny.com  Mon Jun  3 15:29:39 2002
From: mcherm at destiny.com (Michael Chermside)
Date: Mon, 03 Jun 2002 15:29:39 -0400
Subject: Is there a Text to HTML conversion module?
Message-ID: <3CFBC3A3.8090806@destiny.com>

> Does anyone know of a module that converts a text file to html?

Sure!

def textToHtml(text):
     return '<html><body><pre>%s</pre></body></head>' % text


<wink>

(sorry... couldn't resist)

-- Michael Chermside








From martin at strakt.com  Thu Jun 13 05:40:26 2002
From: martin at strakt.com (Martin =?iso-8859-1?Q?Sj=F6gren?=)
Date: Thu, 13 Jun 2002 11:40:26 +0200
Subject: Extension objects in Python 1.5 and 2.2
In-Reply-To: <lk3cvry1pi.fsf@pc150.maths.bris.ac.uk>
References: <20020613071813.GA16764@strakt.com> <15624.19045.700125.215173@12-248-41-177.client.attbi.com> <mailman.1023953777.18976.python-list@python.org> <lk3cvry1pi.fsf@pc150.maths.bris.ac.uk>
Message-ID: <20020613094026.GA22933@strakt.com>

On Thu, Jun 13, 2002 at 09:20:53AM +0000, Michael Hudson wrote:
> Martin Sj?gren <martin at strakt.com> writes:
> 
> > On Thu, Jun 13, 2002 at 02:31:49AM -0500, Skip Montanaro wrote:
> > > 
> > >     Martin> How do I write extension objects so that they work in both
> > >     Martin> Python 1.5 and the 2.x series? 
> > > 
> > > Michael Hudson has been working on a new include file named pymemcompat.h.
> > > Since it's only used for backward compatibility it's in the Misc directory.
> > > You can grab it from CVS:
> > > 
> > >     http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/python/python/dist/src/Misc/pymemcompat.h
> > > 
> > > The idea is that you should be able to include it and program to the 2.3
> > > memory api in code that runs under 1.5.2 or newer versions of the
> > > interpreter.
> > 
> > Cool. Thanks a lot!
> 
> Pleasure!  Be sure to complain if it doesn't work, won't you?

So far it's looking good. I still had to ifdef on the Python version since
I'd used PyModule_AddObject, but that was easily fixed.

Regards,
Martin

-- 
Martin Sj?gren
  martin at strakt.com              ICQ : 41245059
  Phone: +46 (0)31 7710870       Cell: +46 (0)739 169191
  GPG key: http://www.strakt.com/~martin/gpg.html




From johnroth at ameritech.net  Tue Jun  4 20:15:49 2002
From: johnroth at ameritech.net (John Roth)
Date: Tue, 4 Jun 2002 20:15:49 -0400
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <slrnafps1c.4ea.huaiyu@gauss.almadan.ibm.com>
Message-ID: <ufqm0voskv86cc@news.supernews.com>

"Huaiyu Zhu" <huaiyu at gauss.almadan.ibm.com> wrote in message
news:slrnafps1c.4ea.huaiyu at gauss.almadan.ibm.com...
> John Roth <johnroth at ameritech.net> wrote:
> >
> >"Huaiyu Zhu" <huaiyu at gauss.almadan.ibm.com> wrote in message
> >news:slrnafo7qv.37c.huaiyu at gauss.almadan.ibm.com...
> >> Gustavo Cordova <gcordova at hebmex.com> wrote:
> >
> >>
> >> Great explanation!  The problem stems from the fact that the
current
> >+=
> >> operator implements two semantically distinct operations, chosen
> >according
> >> to some properties of the operands.  This is not unlike the
situation
> >of the
> >> division operator with distinct behaviors on floats and ints.
> >
> >This is not an issue with '+='. It's an issue with assignment,
> >regardless
> >of whether it is standalone or combined with some other operation.
>
> The point is that += does not have to contain any assignment at all.
In
> most cases for mutable objects, the result is the same object
(mutated).
> This was the motivation for introducing += into Python in the first
place.
> The need for assignment only comes from immutable objects, but its
> implementational implication extends to mutable objects as well,
producing
> this spurious exception.  It is spurious because the last assignment
step is
> not necessary.
>
> >
> >And I disagree that assignment implements two distinct operations.
> >Assignment always rebinds. Always. In the examples cited, it
> >attempted to rebind into a tuple, because that's what was asked
> >for when the left hand side had a subscript.
>
> This missed the point.  In-place operations should not contain
assigments.

I think you're mistaken. To consider the example at the top of
the thread in more detail, you would argue that the only operation
is the append operation against the second list which is the second
operand of the tuple.

The assignment occurs because the pointer to the list has to
be changed to point to the (possibly) new location of the list
in memory. In other words, the second element of the tuple
has to be rebound.

There is no way of avoiding this, since objects are known
internally by their location in memory, and if you change the
object, you may (not necessarilly will) change that object's
location. Therefore, the pointer to the object must also be
changed in this instance. That is what causes the exception.

John Roth







From dreed at capital.edu  Sat Jun 29 16:21:41 2002
From: dreed at capital.edu (Dave Reed)
Date: Sat, 29 Jun 2002 16:21:41 -0400
Subject: private
In-Reply-To: <3D1DE132.4A890D9D@engcorp.com> (message from Peter Hansen on
	Sat, 29 Jun 2002 12:32:50 -0400)
References: <pan.2002.06.29.17.38.51.789527.440@myself.com> <3D1DE132.4A890D9D@engcorp.com>
Message-ID: <200206292021.g5TKLf928740@localhost.localdomain>

> From: Peter Hansen <peter at engcorp.com>
> X-Accept-Language: en
> Newsgroups: comp.lang.python
> Xref: news.baymountain.net comp.lang.python:170296
> Sender: python-list-admin at python.org
> Date: Sat, 29 Jun 2002 12:32:50 -0400
> 
> Rhymes wrote:
> > 
> > Is there any possibility to build up a _real_ private attribute?
> 
> Please define what "_real_" means to you.  It's not apparent.
> 
> > Such as C++ private data members...
> 
> You realize that it's quite possible to get at private data members
> in C++, don't you?

Kind of off-topic, but this got me curious. The only way I could think
of is to assume (which I think is always true) that the private data
members are in order from the starting address so based on the offset
from the address of the object you could access each one.

Is there another way?

Dave




From nookieNO_SP_AM at online.no  Wed Jun  5 12:02:26 2002
From: nookieNO_SP_AM at online.no (Erlend J. Leiknes)
Date: Wed, 05 Jun 2002 16:02:26 GMT
Subject: distutils and distributing "scripts"
References: <adl4e4$19ep$1@news.adamastor.ac.za> <nAqL8.10168$fG3.343408@news2.ulv.nextra.no>
Message-ID: <mCqL8.10171$fG3.343358@news2.ulv.nextra.no>

oh, sorry, didnt read your question good enough...

=(

"Erlend J. Leiknes" <nookieNO_SP_AM at online.no> wrote in message
news:nAqL8.10168$fG3.343408 at news2.ulv.nextra.no...
> do not call the files .pyw, just use py
>
> when you then use py2exe, do the following
> python setup.py py2exe -w
>
> the -w means --window or --windows (dont remember), anyway, this will use
> the pythonw.exe.
>
>
> "Hugo van der Merwe % str (33) >" <<s1%s61562 at bach.sun.ac.za> wrote in
> message news:adl4e4$19ep$1 at news.adamastor.ac.za...
> > I have recently started using distutils. I have a .pyw file for the
> windows
> > users, how do I get this to be a part of my ... sdist, for example?
> >
> > I have added this to the scripts list passed to "setup()", this seems to
> me
> > the "correct" place for it (I want to eventually play with py2exe).
> >
> > It seems "python setup.py sdist" does not add "scripts" by default
though,
> > just the py_modules. The docs seem to confirm this. How do I get my .pyw
> > file distributed, do I really have to go and do all that MANIFEST.in
> stuff,
> > just for this one file?
> >
> > Thanks,
> > Hugo van der Merwe
>
>




From robin at jessikat.fsnet.co.uk  Sat Jun 29 14:12:46 2002
From: robin at jessikat.fsnet.co.uk (Robin Becker)
Date: Sat, 29 Jun 2002 19:12:46 +0100
Subject: yield equivalent in C/JavaScript?
Message-ID: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk>

Is there any way to do yielding easily in C? I have a bunch of programs
which are normally connected together using unix pipes. I would like to
merge them and the problem is then how to get them to communicate nicely
using a single process.

the typical process currently looks like

        initialise();
        while((c=getc(stdin))>=0){
                process(c);
                putc(c,stdout);
                }

so if the inputs are reasonable I could just run each to completion
before proceeding with the next by replacing the file streams with
memory ones. That could get messy if the inputs are large.

Is there a neat way to simulate the yield idea? In C I can just about
consider doing some longjump madness, but that wont wash in the other
possible environment which has JavaScript for implementation.
-- 
Robin Becker


From SSchukat at dspace.de  Tue Jun  4 10:43:19 2002
From: SSchukat at dspace.de (Stefan Schukat)
Date: Tue, 4 Jun 2002 15:43:19 +0100 
Subject: Creating explorer link on Win32
Message-ID: <84257D042AA7D411B3F700D0B7DB9B7C13B17F@PDC-DSPACE>

Use the standard windows functions or look at the Python Cookbook
http://aspn.ActiveState.com/ASPN/Python/Cookbook/

	Stefan

-----Original Message-----
From: Florian Fredegar Haftmann [mailto:fredegar at haftmann-online.de]
Sent: Tuesday, June 04, 2002 1:49 PM
To: python-list at python.org
Subject: Creating explorer link on Win32


Hi!

I'm searching for a possibility to create an explorer link (*.lnk) to a 
file on Win32. Has anyne heard of a python module able to do that in a 
simple way?

Thanks,
FFH

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

-------------- next part --------------
A non-text attachment was scrubbed...
Name: CreateShortCut.py
Type: application/octet-stream
Size: 5310 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020604/bc6dfecd/attachment.obj>

From ark at research.att.com  Thu Jun 13 15:33:34 2002
From: ark at research.att.com (Andrew Koenig)
Date: Thu, 13 Jun 2002 19:33:34 GMT
Subject: dictionary: keys() and values()
References: <m2elfbuhn1.fsf@mother.paradise.lost>
Message-ID: <yu991ybbdldd.fsf@europa.research.att.com>

John> keys = m.keys()
John> values = m.values()

John> Do we know if the values come out in the same order, that is does
John> m[keys[i]] necessarily equal values[i] ?

>From the library reference, 2.2.7, footnote 3:

If keys() and values() are called with no intervening modifications to the
dictionary, the two lists will directly correspond.  This allows the creation
of (value, key) pairs using zip():  "pairs = zip(a.values(), a.keys())".

-- 
Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark


From logiplexsoftware at earthlink.net  Thu Jun 20 15:47:15 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Thu, 20 Jun 2002 19:47:15 GMT
Subject: How to represent the infinite ?
References: <3d11ee9b$1@news.swissonline.ch> <mailman.1024600723.6989.python-list@python.org>
Message-ID: <20020620.124436.1266235189.32516@software1.logiplex.internal>

In article <mailman.1024600723.6989.python-list at python.org>, "Cliff Wells"
<logiplexsoftware at earthlink.net> wrote:



> On Thu, 20 Jun 2002 17:02:33 +0200
> erreur wrote:
> 
>> Will somebody have an idea, to represent the infinite one?  I have
>> variables to initialize with is +inf (or - inf).  To be sure that
>> later, all will be smaller (or larger) than my variables. I tried to
>> redefine the operators on an object.  It goes for Inf>10 but I do not
>> arrive for 10>Inf (because that takes > of Int).  --------------------
>> def __gt__(self, val):
>> 
>>     return 1
>> --------------------


> Hm, for some reason, my last two posts on this thread didn't seem to go
> through...  one more try (with a couple of little enhancements):

Odd, the newsgroup -> mailing list transfer must be really slow (or down)
today, I pulled up Pan and lo, my posts are there, they just haven't made
it to the mailing list yet (after almost 2 hours).

Or maybe I accidentally killfiled myself  ;)


From mhammond at skippinet.com.au  Thu Jun  6 18:55:56 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Thu, 06 Jun 2002 22:55:56 GMT
Subject: win32com problem: No Such Interface Supported
References: <adoed4012g2@enews3.newsguy.com>
Message-ID: <3CFFE8F8.1010404@skippinet.com.au>

Unfortunately, win32com does not support calling arbitary vtable 
interfaces - only IDispatch is guaranteed.

This may be fixed sometime this year, but is not at the moment - just 
one person offering to help may tip this over ;)

Mark.

Lindsey Smith wrote:
> I'm having trouble using a COM object from Python 2.2.1 using the
> win32all-143 extensions. I want to access ISubscriptionMgr, which is
> implemented by webcheck.dll, a part of Internet Explorer 5+.
> ISubscriptionMgr is defined by subsmgr.idl and subsmgr.h in the MS Visual C
> vc98\include directory.
> 
> What I did was:
> ] midl subsmgr.idl
> 
> which yields subsmgr.tlb, then
> ] python makepy.py -i path\subsmgr.tlb
> 
> which yields:
> {C54FD88A-FFA1-11D0-BC5E-00C04FD929DB}, lcid=0, major=1, minor=0
> 
>>>># Use these commands in Python code to auto generate .py support
>>>>from win32com.client import gencache
>>>>gencache.EnsureModule('{C54FD88A-FFA1-11D0-BC5E-00C04FD929DB}', 0, 1, 0)
>>>
> 
> 
> then making test.py:
> 
> from win32com.client import gencache
> mod = gencache.EnsureModule('{C54FD88A-FFA1-11D0-BC5E-00C04FD929DB}', 0, 1,
> 0)
> 
> o = mod.SubscriptionMgr()
> print o
> 
> 
> but test.py bombs out in this way:
> 
> Traceback (most recent call last):
>   File "test.py", line 5, in ?
>     o = mod.SubscriptionMgr()
>   File
> "D:\DEV\PYTHON22\lib\site-packages\win32com\gen_py\C54FD88A-FFA1-11D0-BC5E-0
> 0C04FD929DBx0x1x0.py", line 37, in __init__
>     if oobj is None: oobj = pythoncom.new(self.CLSID)
> pywintypes.com_error: (-2147467262, 'No such interface supported', None,
> None)
> 
> Any ideas on what I'm doing wrong?
> 
> thanks,
> Lindsey
> 
> 
> 




From observer at NOSPAM.space.pl  Thu Jun 13 06:41:03 2002
From: observer at NOSPAM.space.pl (Johann)
Date: Thu, 13 Jun 2002 12:41:03 +0200
Subject: How to check version of Python under CGI
References: <rbmggus7ln1uiutktf5libf4bbngu29ji7@4ax.com> <873cvrin1q.fsf@tux.ntw23.fr> <cgnggu40ethjhpqcmltnfb06d5qvuf078k@4ax.com> <t1ogguoua7b9ne9n3mb3n7rkp8mrqe2nee@4ax.com> <87y9djh6tg.fsf@tux.ntw23.fr> <ndpggu8uqsmi6ppumkem4c16rt7p5i6k9o@4ax.com> <87ptyvh6g7.fsf@tux.ntw23.fr>
Message-ID: <mktggu4ufogo65e48t5sgnvpnf1thl0g67@4ax.com>

On 13 Jun 2002 11:30:00 +0200, Sylvain Thenault <thenault at nerim.net>
wrote:

>I didn't see the solution on the newsgroup, could you share it ?

source: http://snakefarm.org/snakecharmer.tar.gz

#!/usr/bin/env python

#
# SnakeCharmer v1.3, Carsten Gaebler (cg at snakefarm.org)
#
# SnakeCharmer gives you the location of your Python interpreter,
# your path to Sendmail, your environment variables and a list of the
# Python modules that are installed on your web server.
#

import glob, os, string, sys, traceback
from stat import *

def is_executable(path):
    return (os.stat(path)[ST_MODE] & (S_IXUSR | S_IXGRP | S_IXOTH)) !=
0

def get_toplevel_libdirs():
    libdirs = sys.path[1:]
    libdirs.sort()

    tlds = {}
    for i in range(len(libdirs) - 1):
        cp = os.path.commonprefix(libdirs[i:i+2])
        if cp == "/":
            tlds[libdirs[i]] = None
            continue
        if cp[-1] == "/":
            cp = cp[:-1]
        if os.path.isdir(cp):
            tlds[cp] = None
        else:
            tlds[libdirs[i]] = None

    libdirs = tlds.keys()
    libdirs.sort()
    return libdirs
	

def listmodules(arg, dirname, names):	
    has_init_py = '__init__.py' in names
    if has_init_py:
	for p in sys_path:
            if dirname[:len(p)] == p:
                dirname = dirname[len(p):]
                break
        if dirname[:1] == '/':
            dirname = dirname[1:]
        if dirname[:4] == 'test':
            return
	dirname = string.replace(dirname, "/", ".")
	installedmodules.append(dirname)
        
    for mod in names:
        if (mod[0] != "_") and (mod[:4] != 'test'):
            ext = mod[-3:]
            if ext == '.py':
                mod = mod[:-3]
            elif ext == '.so':
                if mod[-9:-3] == 'module':
                    mod = mod[:-9]
                else:
                    mod = mod[:-3]
            else:
                continue
            if has_init_py:
                installedmodules.append(dirname + "." + mod)
            else:
                installedmodules.append(mod)


print "Content-Type: text/html"
print "X-Powered-By: Python %s" %
string.join(string.split(sys.version), " ")
print '''
<html><body bgcolor="#FFFFFF" alink="#FDB900" link="#BF0425"
vlink="#1200FD">
<h3><center><a
href="http://snakefarm.org/">snakefarm.org</a></center></h3>
<h1><center>Snake Charmer 1.3</center></h1>
'''

try:
    sys_path = sys.path[1:]
    sys_path.sort(lambda a, b: cmp(len(b), len(a)))

    installedmodules = list(sys.builtin_module_names)

    for dir in get_toplevel_libdirs():
        os.path.walk(dir, listmodules, None)

    installedmodules.sort()
    for i in range(len(installedmodules) % 3):
        installedmodules.append('')

    py = {}
    py["version"] = string.split(sys.version)[0]
    py["platform"]= sys.platform

    location = string.split(os.popen("whereis python", "r").read())
    location = filter(lambda p: p[0] == '/', location)
    location.sort()

    for i in range(len(location)):
        if os.path.islink(location[i]):
            location[i] = location[i] + " -> " +
os.readlink(location[i])

    py["location"] = string.join(location, "<br>\n")

    location = string.split(os.popen("whereis sendmail", "r").read())
    location = filter(lambda p: p[0] == '/', location)
    location.sort()
    location = filter(is_executable, location)
    py["sendmail"] = string.join(location, "<br>\n")

    libdirs = sys.path
    if libdirs[0] == '':
        libdirs[0] = './'

    py["libdirs"] = string.join(libdirs, "<br>\n")
    font = '<font face="Verdana, Arial, Sans Serif">'
    py["font"] = font

    print '<table width="100%">'
    print """
    <tr><td BGCOLOR='#99ccff' colspan="2"
align="center">%(font)s<b>Program Paths</b></td></tr>
    <tr><td bgcolor="GHOSTWHITE" width="35%%"
valign="top">%(font)s<b>Python version</b></td><td
width="65%%">%(font)s%(version)s</td></tr>
    <tr><td bgcolor="GHOSTWHITE" width="35%%"
valign="top">%(font)s<b>Python OS platform</b></td><td
width="65%%">%(font)s%(platform)s</td></tr>
    <tr><td bgcolor="GHOSTWHITE" width="35%%"
valign="top">%(font)s<b>Location of Python</b></td><td
width="65%%">%(font)s%(location)s</td></tr>
    <tr><td bgcolor="GHOSTWHITE"width="35%%"
valign="top">%(font)s<b>Location of Sendmail</b></td><td
width="65%%">%(font)s%(sendmail)s</td></tr>
    <tr><td bgcolor="GHOSTWHITE"width="35%%"
valign="top">%(font)s<b>Directories searched for Python
modules<b></td><td width="65%%">%(font)s%(libdirs)s</td></tr>
    """ % py
    print "</table>"

    print '''<table width="100%%">
    <tr><th BGCOLOR="#99ccff" colspan="2" align="center">%sEnvironment
Variables</td></tr>
    ''' % font

    envvars = os.environ.keys()
    envvars.sort()
    for envvar in envvars:
        value = os.environ[envvar]
        print '<tr><td bgcolor="GHOSTWHITE" width="35%%"
valign="top">%s<b>%s</b></td><td width="65%%">%s%s</td></tr>' % (font,
envvar, font, value)
    print "</table>"

    print '''<table width="100%%">
    <tr><th BGCOLOR="#99ccff" align="center">%sInstalled
Modules</th></tr>
    <tr><td><pre>
    ''' % font

    rows = len(installedmodules) / 3
    mods = [ [], [], [] ]
    maxlen = max(map(len, installedmodules))

    for i in range(rows):
            s = "%%-%ds %%-%ds %%s" % (maxlen, maxlen)
            print s % (installedmodules[i], installedmodules[rows +
i], installedmodules[2*rows + i])

    print """
    </pre></td></tr>
    </table>
    </body></html>
    """

except:
    tb = traceback.format_exception(sys.exc_type, sys.exc_value,
sys.exc_traceback)
    tb = string.join(tb, "")
    print '''
    <pre>
    %s
    </pre>
    </body>
    </html>
    ''' % tb
    



From eric.brunel at pragmadev.com  Fri Jun  7 13:52:19 2002
From: eric.brunel at pragmadev.com (Eric Brunel)
Date: Fri, 7 Jun 2002 17:52:19 +0000
Subject: positional modifiers in python?
References: <mailman.1023460302.20078.python-list@python.org>
Message-ID: <adqkgl$5ob$1@wanadoo.fr>

fuf wrote:
>  hello everyone,
> 
>  does python support positional parameters in the print method? ie.
>  something like:
>     print "there %2$s something %1$s" % ("there", "is")
>  it doesn't work but maybe there's some other way?

There's an utterly ugly way to do this:

def toPosDict(l):
  d = {}
  map(lambda k,v,d=d: d.update({str(k):v}), range(len(l)), l)
  return d
print "there %(1)s something %(0)s" % toPosDict(("there", "is"))

For post-Python 2.2 users (I'm still 2.1, sorry), the "map" line may 
certainly be replaced by:

map(d.__setitem__, [str(i) for i in range(len(l))], l)

but it may be less efficient than the code above (one map + one list 
comprehension instead of a single map in the first solution). Maybe the 
second's a little more readable (or a little less unreadable ;-).

HTH
-- 
- Eric Brunel <eric.brunel at pragmadev.com> -
PragmaDev : Real Time Software Development Tools - http://www.pragmadev.com


From lhudson at geminidataloggers.com  Thu Jun 27 11:22:49 2002
From: lhudson at geminidataloggers.com (Lawrence Hudson)
Date: Thu, 27 Jun 2002 16:22:49 +0100
Subject: How to find out operating system
References: <mailman.1025159195.9033.python-list@python.org>
Message-ID: <3D1B2DC9.30504@geminidataloggers.com>

If you have PyWin32, try this:

d:\>python
Python 2.0 (#8, Jun 12 2002, 10:20:16) [MSC 32 bit (Intel)] on win32
Type "copyright", "credits" or "license" for more information.
 >>> import win32api
 >>> win32api.GetVersionEx()
(5, 0, 2195, 2, 'Service Pack 2')
 >>>

majorVersion:   Identifies the major version number of the operating system.
minorVersion:   Identifies the minor version number of the operating system.
buildNumber:    Identifies the build number of the operating system in 
the low-order word. (The high-order word contains the major and minor 
version numbers.)
platformId:     Identifies the platform supported by the operating 
system.  May be one of VER_PLATFORM_WIN32s, VER_PLATFORM_WIN32_WINDOWS 
or VER_PLATFORM_WIN32_NT
version:        Contains arbitrary additional information about the 
operating system.

Thanks,
Lawrence


A wrote:
 > Hi,
 > What is the best way of  finding out the kind operating system?
 > I can use os.name but if the system is Windows I would like also
 > know if the system is Windows98 or Windows ME or W2K or
 > Windows XP.
 > Thanks for help.
 > Ladislav
 >
 >
 >
 >




From saint at ghost.lt  Thu Jun 20 12:08:52 2002
From: saint at ghost.lt (martynas sklizmantas)
Date: 20 Jun 2002 18:08:52 +0200
Subject: Function to add commas in numbers?
In-Reply-To: <Xns923372EFBBC5Dmdkmdkmdkmdk@130.133.1.4>
References: <Xns923372EFBBC5Dmdkmdkmdkmdk@130.133.1.4>
Message-ID: <1024589333.15471.82.camel@naked>

hello, 


On Thu, 2002-06-20 at 17:17, mdk wrote:
> Hello,
> 
> Is there a function that will take, for example, 1234567 and return 
> 1,234,567?


you could use locale module
(http://www.python.org/doc/current/lib/module-locale.html )

for more interesting solutions look -
http://mail.python.org/pipermail/python-list/2002-June/thread.html#108787


btw, 
i have some problems with locale module on debian sid:


Python 2.1.3 (#1, Apr 20 2002, 10:14:34) 
[GCC 2.95.4 20011002 (Debian prerelease)] on linux2
Type "copyright", "credits" or "license" for more information.
>>> import locale
>>> locale.setlocale(locale.LC_NUMERIC, 'da_DK')
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "/usr/lib/python2.1/locale.py", line 374, in setlocale
    return _setlocale(category, locale)
locale.Error: locale setting not supported


am i missing something in my installation?


best regards, 

ms

-- 
You fill a much-needed gap.





From cjw at sympatico.ca  Sat Jun 15 08:25:59 2002
From: cjw at sympatico.ca (Colin J. Williams)
Date: Sat, 15 Jun 2002 08:25:59 -0400
Subject: Integer conversion
Message-ID: <3D0B3257.E8572AE5@sympatico.ca>

int(string, base) => integer

For int('127',  8), I had expected this function to deliver 87.

The example below correctly gives errors at the top and bottom of the
range and
for int('127', 10) delivers 127.  The other cases are not what I
expected.

I would appreciate any suggestions.

Colin W.

for b in range(2,38):
  try:
    print b, '[', int('127', b), ']'
  except:
    pass
    print 'error'



From wilson at visi.com  Thu Jun 20 21:41:04 2002
From: wilson at visi.com (Tim Wilson)
Date: Thu, 20 Jun 2002 20:41:04 -0500
Subject: Generating SWFs using Ming via CGI
Message-ID: <2wvQ8.19$eH2.59629@ruti.visi.com>

Hi everyone,

I'm finishing up a class on building Flash animations. For my final 
project, I thought it would be cool to build a SWF (the Flash output file) 
using input from a Web form and processed using the Ming library 
(http://ming.sourceforge.net/) through CGI.

Ming is "an SWF output library and PHP module" that also provides a Python 
wrapper. There is a function reference at 
http://ming.sourceforge.net/docs/index.php?mode=py and examples at 
http://ming.sourceforge.net/examples/index.html

I'm having trouble getting a simple example to work via cgi on my Web 
server. (I know it's not a cgi configuration problem. There are other CGI 
scripts that work fine.) I'm not sure how to set the Content-type header 
and then output the SWF movie.

Here's a short PHP script followed by my Python translation.

<?
  $s = new SWFShape();
  $s->setLine(4, 0x7f, 0, 0);
  $s->setRightFill($s->addFill(0xff, 0, 0));
  $s->movePenTo(10, 10);
  $s->drawLineTo(310, 10);
  $s->drawLineTo(310, 230);
  $s->drawCurveTo(10, 230, 10, 10);

  $m = new SWFMovie();
  $m->setDimension(320, 240);
  $m->setRate(12.0);
  $m->add($s);
  $m->nextFrame();

  header('Content-type: application/x-shockwave-flash');
  $m->output();
?>

This script is found at 
http://ming.sourceforge.net/examples/example.php?name=shape

Here's my Python translation.

#!/usr/bin/python

from ming import *

s = SWFShape()
s.setLine(4, 0x7f, 0, 0)
s.setRightFill(s.addFill(0xff, 0, 0))
s.movePenTo(10, 10)
s.drawLineTo(310, 10)
s.drawLineTo(310, 230)
s.drawCurveTo(10, 230, 10, 10)

m = SWFMovie()
m.setDimension(320, 240)
m.setRate(12.0)
m.add(s)
m.nextFrame()

print "Content-type: applicaton/x-shockwave.flash\n"
m.output()

Note, I can substitute m.save('shape.swf') for m.output() and the script 
works fine. But I want the swf to be viewed immediately in the browser.

BTW, running the Python version listed above produces the following in my 
Web server's error log:

[Thu Jun 20 20:08:15 2002] [error] [client 208.42.140.222] malformed header 
from script. Bad header=FWSQ: /home/wilson/public_html/cgi-bin/shape.py

Running the shape.py script from the server's command line produces:

wilson at beethoven:~/public_html/cgi-bin$ python shape.py
Content-type: applicaton/x-shockwave.flash

FWSQp
     ??
       C???p
            0@???P?5,?G????&Q h?@wilson at beethoven:~/public_html/cgi-bin$

Can anyone provide a hint about how I can make this work?

-Tim

-- 
Tim Wilson      | Visit Sibley online:         | Check out:
Henry Sibley HS | http://www.isd197.k12.mn.us/ | http://www.zope.org/
W. St. Paul, MN |                              | http://slashdot.org/
wilson at visi.com |   <dtml-var pithy_quote>     | http://linux.com/


From edream at tds.net  Thu Jun 20 19:49:36 2002
From: edream at tds.net (Edward K. Ream)
Date: Thu, 20 Jun 2002 23:49:36 GMT
Subject: tkinter font picker?
References: <3D122BB6.92593026@tds.net> <aetkg6$1qbo$1@nntp6.u.washington.edu>
Message-ID: <3D126A11.172D2565@tds.net>

> >Does anyone know of tkinter code that will put up a font chooser
> >dialog?  There are a number of such packages written in tk/tcl.  I'm
> >wondering if anyone knows of a port to tkinter.
> 
> I have a simple one that is part of a Tkinter-based preferences package.
> You can find it at <http://www.astro.washington.edu/owen/> in the
> Util.zip package; unpack and look in the Prefs folder.

Thanks for this link :-)

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


From nookieNO_SP_AM at online.no  Sun Jun  9 17:56:28 2002
From: nookieNO_SP_AM at online.no (Erlend J. Leiknes)
Date: Sun, 09 Jun 2002 21:56:28 GMT
Subject: import problem
References: <2se7guc3kkmm3bgooj776j55igpkmgpst4@4ax.com>
Message-ID: <gaQM8.1178$g65.19246@news4.ulv.nextra.no>

if you have the debugging application strace (its most likely you have it)

write a small application who only does: from qt import *

save it, run it like this: strace python your_app.py 1> log1.txt 2>log2.txt
wait for it to finnish, and edit log2 (or log1).txt, scroll to the bottom,
and look for io signales that tries to open libsub.so.*
now you should find out what your filename should be.

"Michael Eager" <eager at eagercon.com> wrote in message
news:2se7guc3kkmm3bgooj776j55igpkmgpst4 at 4ax.com...
> I'm trying to use Python 2.2 and PyQt on a RedHat-7.2 system.
>
> I get an error when I execute "from qt import *" -- there is a failure
> in qt.py when it executes "import libsup".  The error is "No module
> named libsup".
>
> libsup.so.9 is a link to libsup.so.9.0.3 and is in
> /usr/lib/python2.2/site-packages.  If I enter "import libsup", I get
> the same error message.
>
> Any suggestions?
>
> -- Mike Eager, eager at eagercon.com




From gerhard at bigfoot.de  Fri Jun 28 11:00:13 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 28 Jun 2002 15:00:13 GMT
Subject: distutils/boost question
References: <afhp5p$fqf0d$1@hades.rz.uni-sb.de>
Message-ID: <slrnahou7h.140.gerhard@lilith.my-fqdn.de>

In article <afhp5p$fqf0d$1 at hades.rz.uni-sb.de>, Uwe Schmitt wrote:
> Hi,
> 
> I used the Boost.Python library to write some python-extension.
> How can I distribute the generated (binary) .so-file with distutils ?
> I don't want to distribute the  sources, only the binaries.
> In detail: I have to distribute the libbpl.so from boost and
> my own .so-file. One problem is, that I have to add the
> path of libbpl.so to $LD_LIBRARY_PATH. When I only copy it
> to the extension-folder of python, it does not work...

The easiest solution is to statically link in the Boost.Python library.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From reason5000 at yahoo.com  Wed Jun 19 10:35:44 2002
From: reason5000 at yahoo.com (Rod G)
Date: 19 Jun 2002 07:35:44 -0700
Subject: web based e-mail
Message-ID: <83e21f2.0206190635.14617b0a@posting.google.com>

Hello. Does anyone know of a tutorial or how-to on creating
a web based e-mail system using Python? I'm wanting to 
develop my own system as a learning project and need 
a starting point.

Thanks,
Rod


From Chris.Barker at noaa.gov  Thu Jun 20 13:46:12 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Thu, 20 Jun 2002 10:46:12 -0700
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <aer2fn$frc$1@peabody.colorado.edu>
Message-ID: <3D1214B6.E9BB77D2@noaa.gov>

Fernando P?rez wrote:
> If you're an old unix hand and want to preserve your sanity, you have two
> options:
> 
> 1- put linux on that laptop and be happy.

I have to concur. I am mostly a Linux user, but when I use Python on
Windows, I really have a hard time with it. It's nothing to do with
Python, and everything to do with Windows. Windows simply does not
provide a programing enviroment that will feel at all productive to
someone used to *nix. I run Linux on a Dell laptop, and am very happy
with it.

On that note: does anyone know if there is a way to turn a Python script
into something that acts like an application on Windows...without
resorting to Py2exe and the like. What I want is something just like:

chmod +x scriptname
and a #! line at the top 

on Unix. I want it to work on the windows command line.

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From gerhard.haering at gmx.de  Wed Jun 19 21:18:33 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 20 Jun 2002 01:18:33 GMT
Subject: My Python/MySQL code is too slow
References: <aer2s4$mp7$1@news7.svr.pol.co.uk>
Message-ID: <slrnah2bbb.cpt.gerhard.haering@lilith.my-fqdn.de>

Duncan Smith wrote in comp.lang.python:
> I have already posted this problem on mailing.database.mysql as I see it as
> a problem with my lack of knowledge of SQL.  But no responses so far, and
> maybe Python could do more of the work.

Doing work in Python that the db can do is not generally advisabele
for performance reasons. MySQL, however, is a rather limited RDBMS. If
you want better SQL support, I'd recommend to take a look at
PostgreSQL. Outer joins and subselects are missing from MySQL, but
available in PostgreSQL and it looks like they could be useful in your
case.

> I repost the original question below (SQL easier to follow), and the
> Python code below that.  Any help appreciated.  TIA.

> 
> ---------------------------------------------------------------------
> I have two tables (say dbA and dbB) and need to find the number of rows in
> dbB which have exactly 2 matching rows in dbA (on the selected columns).
> I'm using Python/MySQL, but the relevant SQL is below.  This is slow, and
> I'm sure it's possible without creating the temporary table and without the
> join.  But everything I try fails (because of my limited knowledge of SQL).
> Can anyone show me an efficient way of getting the answer?  All I need is
> the number of rows in dbB which have exactly 2 matching rows in dbA.  Thanks
> in advance.
> 
> CREATE TEMPORARY TABLE tmp SELECT var1, var2, var3 FROM dbA GROUP BY var1,
> var2, var3 HAVING COUNT(*)=2;
> 
> SELECT dbB.var1, dbB.var2, dbB.var3, tmp.var1, tmp.var2, tmp.var3 FROM dbB,
> tmp WHERE dbB.var1=tmp.var1 AND dbB.var2=tmp.var2 AND dbB.var3=tmp.var3;

I have no idea what dbA and dbB contain, but it looks like they
contain the same sort of data. This is not good db design, but that's
normal for SQL newbies. I'm not very good at explaining the theory,
but you could take a look at one of these links:

http://www.google.com/search?hl=en&lr=&q=relational+database+normalization

("normalization" is the basic concept of good relational db design)

It's very likely that reorganizing your db schema into a one-to-many
relationship between two tables, and thus unifiying dbA and dbB will
simpify your problem.

> [...]
>         conn = MySQLdb.connect()
>         curs = conn.cursor()
>         curs.execute('USE %s' % (dbname,))

conn = MySQLdb.connect(db=dbname)

HTH,

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From gcordova at hebmex.com  Fri Jun  7 17:44:38 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Fri, 7 Jun 2002 16:44:38 -0500 
Subject: direct output to printer
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F0623301194786@mail.hebmex.com>

> 
> Hi,
> 
> I'd like to print barcodes with a special barcode printer.
> Therefore I have to send raw ASCII data to the printer.
> How can I do this using Python under Windows ???
> 
> Greetings, Uwe.
> 

Same way you'd do it in MSDOS
(if you have it connected to a parallel port):

>>> lpt = open("lpt1:", "w")
>>> lpt.write(BarcodeDataString)
>>> lpt.close()

You can also use "prn:" instead of "lpt1:"

:-)

-gustavo

pd: Win's just MSDOS with icing on top.  ;-)




From metaliu at yahoo.com  Fri Jun  7 20:01:21 2002
From: metaliu at yahoo.com (Bill)
Date: Fri, 7 Jun 2002 17:01:21 -0700
Subject: Getting pixel values under cursor
Message-ID: <adrhgi$mh$1@nntp1.jpl.nasa.gov>

I'm using Tkinter right now for a GUI.  Is there a way for me to get, in
real time or when the user clicks the mouse button on a pixel, the pixel
intensity value?  Currently, I have a tkinter label which is showing a large
grey scale image.  I would like the user to move the mouse cursor over the
image and show the pixel intensity under the cursor.  Is there a way to do
this?  Perhaps some kind of event binding with the mouse and the image?
Would putting the image in a canvas instead provide more flexibility?

Thanks in advance,
Bill





From mertz at gnosis.cx  Wed Jun  5 15:54:44 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Wed, 05 Jun 2002 15:54:44 -0400
Subject: ANN: Gnosis (XML) Utils 1.0.2
Message-ID: <Eym/8kKkXwOd092yn@gnosis.cx>                

This release contains one major bugfix, and one new XML module:

   * Added module gnosis.xml.validity.  This module contains custom
     classes for creating a Python object that is constrained by
     XML validity rules (DTD or Schema).  Taking its inspiration from the
     Haskell module HaXml, a programmer using gnosis.xml.validity can
     guarantee that no operation on an object will result in something
     that would be serialized as invalid XML.  See
     gnosis/doc/xml_matters_20.txt for details.

   * Fixed bug with integer parsing in gnosis.xml.pickle (a single "0"
     was being treated as a failed octal number, rather than as...well,
     zero).

It may be obtained at:
   http://gnosis.cx/download/Gnosis_Utils-1.0.2.tar.gz

The current release is always available as:
   http://gnosis.cx/download/Gnosis_Utils-current.tar.gz

Try it out, have fun, send feedback!

David Mertz (mertz at gnosis.cx)
Frank McIngvale (frankm at hiwaay.net)

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

BACKGROUND:  Gnosis Utilites contains a number of Python libraries, most
(but not all) related to working with XML.  These include:

    gnosis.xml.pickle       (XML pickling of Python objects)
    gnosis.xml.objectify    (Any XML to "native" Python objects)
    gnosis.xml.validity     (Enforce validity constraints)
    gnosis.xml.indexer      (XPATH indexing of XML documents)
    gnosis.indexer          (Full-text indexing/searching)
    [...].convert.txt2html  (Convert ASCII source files to HTML)
    gnosis.util.dtd2sql     (DTD -> SQL 'CREATE TABLE' statements)
    gnosis.util.sql2dtd     (SQL query -> DTD for query results)
    gnosis.util.xml2sql     (XML -> SQL 'INSERT INTO' statements)
    gnosis.util.combinators (Combinatorial higher-order functions)
    gnosis.util.introspect  (Introspect Python objects)

...and so much more! :-)


SUMMARY:
<P><A HREF="http://gnosis.cx/download/Gnosis_XML_Util.ANNOUNCE">
Gnosis_XMLUtil 1.0.2</A>  (6-June-02)


--
    _/_/_/ THIS MESSAGE WAS BROUGHT TO YOU BY: Postmodern Enterprises _/_/_/
   _/_/    ~~~~~~~~~~~~~~~~~~~~[mertz at gnosis.cx]~~~~~~~~~~~~~~~~~~~~~  _/_/
  _/_/  The opinions expressed here must be those of my employer...   _/_/
 _/_/_/_/_/_/_/_/_/_/ Surely you don't think that *I* believe them!  _/_/





From peter at engcorp.com  Sat Jun 22 13:05:01 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 22 Jun 2002 13:05:01 -0400
Subject: urllib POST question
References: <644f6688.0206200930.48110837@posting.google.com> <slrnah4dbq.r77.insyte-clp@petra.bitstream.net> <644f6688.0206210900.68dd930c@posting.google.com> <3D13D62B.D7AE6554@engcorp.com> <644f6688.0206220830.46583e8f@posting.google.com>
Message-ID: <3D14AE3D.18BFBE55@engcorp.com>

rdack wrote:
> 
> Peter Hansen <peter at engcorp.com> wrote in message news:<3D13D62B.D7AE6554 at engcorp.com>...
> > > > In article <644f6688.0206200930.48110837 at posting.google.com>, rdack wrote:
> > > > > i am doing a POST to a server using urllib and it works.
> > > > > is there some way i can see exactly what is being posted?
> > > > > a way to get the post object out of urllib - the exact lines it is sending?
> > Howzabout this?
> >
> > >>> import httplib
> > >>> httplib.HTTPConnection.debuglevel = 1
> > >>> import urllib
> > >>> page = urllib.urlopen('yoururlhere').read()
> 
> perfect! thanks. although i was able to install ethereal and see the
> packet-it was the 100 megabyte solution.
> this was the zero byte solution.
> of course, now i have the roll royce of packet sniffers.
> anyway - you are right - python-rocks - when you know which button to
> push. :-).

The right button was "Use the Source, Luke!".  I found in urllib
that it was doing an h = httplib.HTTP() so I went to httplib and
found that the class had a debuglevel field which was used to 
control that feature.  This button is available for most other
Python modules, too.  ;-)

-Peter


From cpbotha at i_triple_e.org  Thu Jun 13 04:40:25 2002
From: cpbotha at i_triple_e.org (Charl P. Botha)
Date: Thu, 13 Jun 2002 08:40:25 +0000 (UTC)
Subject: py2exe 4 linux?
References: <3D08563C.33D041E3@gol.ge>
Message-ID: <ae9lpp$h4g$1@news.tudelft.nl>

In article <3D08563C.33D041E3 at gol.ge>, Giorgi Lekishvili wrote:
> Is there anything like py2exe for Linux?

Try Gordon McMillan's Installer at http://www.mcmillan-inc.com/install1.html

This works incredibly well and is cross-platform to boot!  I've used it to
make standalones of quite an extensive application that makes use of
wxPython and VTK and was amazed at how little I had to do to get it working
perfectly.

HTH,
Charl

-- 
charl p. botha http://cpbotha.net/ http://visualisation.tudelft.nl/


From merkosh at hadiko.de  Sun Jun 16 19:52:52 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Mon, 17 Jun 2002 01:52:52 +0200
Subject: what *is* a class?
Message-ID: <MPG.17774343cc5a589098968a@news.rz.uni-karlsruhe.de>

hi

i've got an object 'Group'. i instanciate it by calling, f.e.
g = Group()

what actually does Group() return and is there any way to get hold of 
that again?
i want to write a method which returns the instance, so that f.e. the 
following is possible:

g = Group()
g2 = g.foobar()
g2 is g

so that g2 is identical to g.
any idea?

thanks in advance
Uwe


From BPettersen at NAREX.com  Mon Jun 24 12:31:37 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Mon, 24 Jun 2002 10:31:37 -0600
Subject: File Copying
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192158EF6@admin56.narex.com>

> From: Rajat Chopra [mailto:rchopra at myrealbox.com] 
> 
> How do you copy a file from one directory to another. I have 
> looked all over and haven't found any documentation on it. 
> Essentially, I want to copy a file from $SRC_DIRECTORY to 
> $TARGET_DIRECTORY.
> 
> Does anyone know which modules are used? Or does anyone have 
> some sample scripts?

Look in the shutil module.

-- bjorn




From peter at engcorp.com  Mon Jun 17 21:28:18 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 21:28:18 -0400
Subject: symbolic python
References: <3CEFAE4C.38A8022B@doc.ic.ac.uk> <3CEFD881.D7992095@doc.ic.ac.uk> <3cf8ce86$1@giga.realtime.net> <83sn46vkgi.fsf@panacea.canonical.org> <3CF942C7.2E4AA4CF@engcorp.com> <838z5xtcla.fsf@panacea.canonical.org> <6a625eea.0206161221.50d351ad@posting.google.com>
Message-ID: <3D0E8CB2.E831F8AB@engcorp.com>

Ira Baxter wrote:
> 
> > > no support for Python is (yet) mentioned in the list of supported
> > > languages.
> 
> That's because our tools are designed to accept langauge definitions
> easily.   A Python grammar would take only a few days to develop.
> 
> If you want to transform Python symbolically, DMS will do the
> job very well, and I don't see many alternatives out there.

"Will"?  When?  Or did you mean "could"?

-Peter


From mwh at python.net  Wed Jun 19 11:47:11 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 19 Jun 2002 15:47:11 GMT
Subject: procmail replacement in Python
References: <oqelf3s7n4.fsf@carouge.sram.qc.ca> <NCBBLFCOHHDIKCAFGCFBAEGILFAA.marklists@mceahern.com> <20020619181953.E4127@phd.pp.ru> <mailman.1024499148.5977.python-list@python.org>
Message-ID: <lk4rfzl1a3.fsf@pc150.maths.bris.ac.uk>

Gerhard H?ring <gerhard at bigfoot.de> writes:

> > Fetchmail also has good GUI configuration program, written in Python
> > and Tkinter.
> 
> Pointless. fetchmail (and getmail even more) configuration is very easy.
> If there were a GUI tool to write procmail rules, now that would make
> sense. But even then only if it included a regex-wizard ;-)

What *I* want is a tool that will subscribe you to a mailman mailing
list, write an appropriate procmail rule to shove the messages into a
new mail folder, send the confirmation and tell gnus about the folder.

That would be neat.

Cheers,
M.

-- 
  The Internet is full.  Go away.
                      -- http://www.disobey.com/devilshat/ds011101.htm


From jepler at unpythonic.net  Tue Jun  4 12:20:26 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Tue, 4 Jun 2002 11:20:26 -0500
Subject: Python command line?
In-Reply-To: <B9224E06.628B%sarmstrong13@mac.com>
References: <B9224E06.628B%sarmstrong13@mac.com>
Message-ID: <20020604112025.A24006@unpythonic.net>

On Tue, Jun 04, 2002 at 10:59:18AM -0500, SA wrote:
> Hi Everyone-
> 
> I would like to run python code line by line using:
> 
> python -c some line of code
> 
> Will this work?

Probably not.  For instance,
    python -c 'def f(): return 3'
    python -c 'print f()'
won't do the same thing as
    >>> def f(x): return 3
    ...
    >>> print f(x)
    3
You probably need to use 'eval' or 'exec':
    >>> gns = {'__builtins__': __builtins__} 
    >>> lns = {}
    >>> exec "def f(): return 3" in gns, lns
    >>> exec "print f()" in gns, lns
    3
    >>> print lns['f']()  # Or access items created in the local namespace
    3

Jeff
PS you escape all sorts of quoting and length-limit problems by avoiding
the likes of os.system("python -c '%s'", cmdstr) too -- for instance,
this won't work at all on NT (since ''-quoting doesn't exist in the
shell afaik) and the cmdstr of "'; echo '0wn3d" will not do what you
want on Unix.




From logiplexsoftware at earthlink.net  Tue Jun 18 14:07:16 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: Tue, 18 Jun 2002 11:07:16 -0700
Subject: python version?
In-Reply-To: <9YBP8.44418$GY.13865843@twister.nyroc.rr.com>
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com>
	<aegfck$6tkre$1@ID-114079.news.dfncis.de>
	<wEPO8.16066$uk2.8447171@twister.nyroc.rr.com>
	<3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com>
	<gYgP8.42356$GY.12361719@twister.nyroc.rr.com>
	<aekcug$7js8n$1@ID-114079.news.dfncis.de>
	<9YBP8.44418$GY.13865843@twister.nyroc.rr.com>
Message-ID: <20020618110716.67cbf88f.logiplexsoftware@earthlink.net>

On Tue, 18 Jun 2002 08:12:53 GMT
George Hester wrote:

> I think you guys don't how to do it.
> 
> --
> George Hester

You may in fact be correct.  If I were you I'd email "Timothy Rue"
<threeseas at earthlink.net> as he would probably be able to give you a more
direct answer.

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726 x308  (800) 735-0555 x308




From max at alcyone.com  Thu Jun 20 00:59:18 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 19 Jun 2002 21:59:18 -0700
Subject: can I call operator overloader in superclass?
References: <H2dQ8.1805$CZ4.569945@news1.news.adelphia.net>
Message-ID: <3D116126.433AF0E4@alcyone.com>

"Jon J. Morin" wrote:

> TypeError: repr not string
> 
> For one, I don't understand the error.  For seconds, can I call
> __repr__ in
> the superclass from __repr__ in the subclass?

I'm guessing you have an older version of Python.  In Python 2.2 this is
the error you'll get:

	TypeError: __repr__ returned non-string (type NoneType)

which I can only presume was changed to be a little more descriptive. 
The calling superclass methods in subclasses is actually leading you
astray -- all the error is telling you is that your __repr__ method must
return and yours isn't.  Change the __repr__ method in your subclass so
that it returns a string and you're off and running.

To answer your other question (which really wasn't the issue here), you
can certainly explicitly call a superclass method in a subclass, and
you'd had the syntax right:  Specify the superclass method is an unbound
method and then explicitly pass in the self object:

	class Superclass:
	    ...

	    def method(self):
	        ...

	class Subclass(Superclass):
	    ...

	    def method(self):
	        ...
	        Superclass.method(self)
	        ...


-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Who'd ever think it / Such a squalid little ending
\__/ The American and Florence, _Chess_
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From ylee12 at uiuc.edu  Fri Jun  7 14:22:54 2002
From: ylee12 at uiuc.edu (Young-Jin Lee)
Date: Fri, 7 Jun 2002 13:22:54 -0500
Subject: About writing Jython tutorial...
References: <c427d639.0206070612.4ea8ce0c@posting.google.com>
Message-ID: <%N6M8.12540$U7.145594@vixen.cso.uiuc.edu>

It would be great to have a nice Jython tutorial.
I have been looking for it quite a while, but had no luck.

YJ.
"Robert Hanlin" <findler_lambda at yahoo.com> wrote in message
news:c427d639.0206070612.4ea8ce0c at posting.google.com...
> Hi all,
>
> I'm indoctrinating friends into the dark, beautiful world of Jython.
> But I find there is no good online documentation, and I only have an
> easy time because I happen to own the O'Reilly Jython Essentials book.
>  (For example, when you use jythonc to compile classes, if something
> is missing on the classpath you're more likely to get a null pointer
> exception than an appropriate class not found exception.)
>
> If I happened to write a walkthrough, is there a website where it
> would be welcome?  If not, maybe I should make a weblog...
>
>
> Thanks!
> findler




From john_taylor_1973 at yahoo.com  Sat Jun 22 01:35:22 2002
From: john_taylor_1973 at yahoo.com (John Taylor)
Date: 21 Jun 2002 22:35:22 -0700
Subject: urllib POST question
References: <644f6688.0206200930.48110837@posting.google.com> <slrnah4dbq.r77.insyte-clp@petra.bitstream.net> <644f6688.0206210900.68dd930c@posting.google.com>
Message-ID: <fc30bff.0206212135.21af0bb3@posting.google.com>

rdacker at pacbell.net (rdack) wrote in message news:<644f6688.0206210900.68dd930c at posting.google.com>...
> Ben Beuchler <insyte-clp at emt-p.org> wrote in message news:<slrnah4dbq.r77.insyte-clp at petra.bitstream.net>...
> > In article <644f6688.0206200930.48110837 at posting.google.com>, rdack wrote:
> > > i am doing a POST to a server using urllib and it works.
> > > is there some way i can see exactly what is being posted?
> > > a way to get the post object out of urllib - the exact lines it is sending?
> > 
> > I'm a big fan of Ethereal...
> > "Sniffing the glue that holds the internet together!"
> > 
> > http://www.ethereal.com/
> 
> i am looking into it. seems overkill for wanting to see a few bytes
> that python is generating.
> No way within python to see what urllib intends to send out for the
> POST?

If you have the tcpdump program available (most likely on unix), then
you can use the script below.  The program is a wee bit on the slow
side
when processing large files, so any suggested speed ups would be
appreciated.

Also available at http://www.terry.uga.edu/~jft/ptd.py


#!/usr/bin/env python

# ptd - Print TcpDump
# John Taylor 2002-06-03

# Requires Python 2, will not work on Python 1.5.2

# Example:
# tcpdump -n -i eth0 -x -vv -t -s 4096 ip and not port 22 | ptd.py

import sys

# if set to 1, you will see a line count status on stderr
VERBOSE = 1
VERBOSE_PERIOD = 2000

#
# Main
#

def main():
        line_count = 0
        for line in sys.stdin.xreadlines():
                line_count += 1
                newline = " "
                hexdata = line.split("\t")
                if 4 == len(hexdata) :
                        hexline = hexdata[3].split()
                        for w in hexline:
                                h1 = w[:2]
                                h2 = w[2:]
                                if( len(h1) ):
                                        i = int("0x"+h1,16)
                                        if i >= 32 and i <= 127:
                                                newline += chr(i)
                                        else:
                                                if " " == newline[-1]
:
                                                        newline += "%s
" % (h1)
                                                else:
                                                        newline += "
%s " % (h1)

                                if( len(h2) ):
                                        i = int("0x"+h2,16)
                                        if i >= 32 and i <= 127:
                                                newline += chr(i)
                                        else:
                                                if " " == newline[-1]
:
                                                        newline += "%s
" % (h2)
                                                else:
                                                        newline += "
%s " % (h2)

                        print newline

                else:
                        one = line[0:1]
                        if " " == line[1:2] and (">" == one or "<" ==
one or "B" == one):
                                print
                        print line,

                if 1 == VERBOSE and (0 == line_count %
VERBOSE_PERIOD):
                        sys.stderr.write("%d\n" % line_count)

#
# Program execution begins here
#

main()


From whisper at oz.net  Tue Jun 11 12:04:17 2002
From: whisper at oz.net (David LeBlanc)
Date: Tue, 11 Jun 2002 09:04:17 -0700
Subject: compiling error: undefined symbol: __gxx_personality_v0
In-Reply-To: <93523ebb.0206110732.791e745@posting.google.com>
Message-ID: <GCEDKONBLEFPPADDJCOEIEDFDIAA.whisper@oz.net>

You might get more help with a boost problem like this on
http://mail.python.org/mailman/listinfo/c++-sig which is the mailing list
for boost & python.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of mitsch
> Sent: Tuesday, June 11, 2002 8:32
> To: python-list at python.org
> Subject: compiling error: undefined symbol: __gxx_personality_v0
>
>
> Hi,
> I try to wrap C functions in Python.
> when I compile, here's the result:
> Traceback (most recent call last):
>   File "./boost.py", line 2, in ?
>     import getting_started1
> ImportError: ./getting_started1.so: undefined symbol:
> __gxx_personality_v0
>
>
> can somebody tell why this error?
>
> here's my files:
> getting_stated!.cpp :
> #include <boost/python/class_builder.hpp>
> namespace python = boost::python;
> #include <string>
>
> namespace { // Avoid cluttering the global namespace.
>
>   // A couple of simple C++ functions that we want to expose to
> Python.
>   std::string greet() { return "hello, world"; }
>   int square(int number) { return number * number; }
> }
>
> BOOST_PYTHON_MODULE_INIT(getting_started1)
> {
>     // Create an object representing this extension module.
>     python::module_builder this_module("getting_started1");
>
>     // Add regular functions to the module.
>     this_module.def(greet, "greet");
>     this_module.def(square, "square");
> }
>
>
>
>
>
> my compile command:
> g++3 -fpic  -g  -I/usr/include/python2.1
> -I/usr/local/lib/python/boost_1_28_0/  -c ./getting_started1.cpp
> gcc -shared -lc  getting_started1.o  -o getting_started1.so
> --
> http://mail.python.org/mailman/listinfo/python-list





From zopestoller at thomas-guettler.de  Tue Jun 25 10:36:54 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Tue, 25 Jun 2002 16:36:54 +0200
Subject: Parsing strings (\n and \\)
Message-ID: <3D188006.7050202@thomas-guettler.de>

Hi!

Is there a function for parsing c-like
strings? I need to parse strings containing
'\n' and '\\' '\"'.

thomas



From gmcm at hypernet.com  Sat Jun  1 11:17:30 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 01 Jun 2002 15:17:30 GMT
Subject: Installer vs py2exe: Windows DLLs
References: <u1ooxxn8.fsf@morpheus.demon.co.uk>
Message-ID: <Xns9220730326E0Fgmcmhypernetcom@199.171.54.215>

Paul Moore wrote:

> When building a distribution for the same script using Gordon
> McMillan's Installer and py2exe, I notice that Installer includes
> PyWinTypes22.dll and win32api.pyd, whereas py2exe doesn't. 

[snip]

> More to the point, in some ways, why does Installer think these files
> are needed? Is there some situation in which py2exe generated
> executables will fail because these files aren't there?

David's answer was 100% correct.

The specific call here is os.path.abspath(...). Before 2.2 and 
without win32api, this simply examined its argument. In some
rare corner cases, it would yield an incorrect answer.

-- Gordon
http://www.mcmillan-inc.com/


From amckay at merlintechnologies.com  Mon Jun 10 13:08:48 2002
From: amckay at merlintechnologies.com (Andy McKay)
Date: Mon, 10 Jun 2002 10:08:48 -0700
Subject: Newby: How do I strip HTML tags?
In-Reply-To: <3d0190e7.5774640@news.t-online.de>
References: <Xns92278CC566FEAusYourStandingInIt@130.133.1.4> <3d00e1b1.47853937@news.t-online.de> <mailman.1023469751.21489.python-list@python.org> <3d0190e7.5774640@news.t-online.de>
Message-ID: <20020610170848.GA2228@mckay.merlintechnologies.com>

> Standalone "<" and ">" indicate invalid HTML code, one should use &lt;
> and &gt; instead. You are of course right, in the end a use of
> predefined classes is almost always better than reinventing the wheel
> yourself.

Of course the main problem is there is a lot of invalid HTML out there ;)

I cant remember where the article was, but I found a good one a while ago
that showed a list of valid, but admittedly rare situations that can crop up
a simple html parser. Should make set of unit tests for any parser
actually...

-- 
  Andy McKay
  Merlin Technologies




From tdelaney at avaya.com  Thu Jun 13 20:40:33 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 14 Jun 2002 10:40:33 +1000
Subject: newbie: from .. import *
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A388@natasha.auslabs.avaya.com>

> From: jseb at cs.mcgill.ca [mailto:jseb at cs.mcgill.ca]
>
> In the interactive mode, if I import the code above as:
> >>> from test import *
> 1
> 0

http://py-howto.sourceforge.net/doanddont/doanddont.html

Look at 1.1 ...

Tim Delaney




From rnd at onego.ru  Tue Jun 11 16:20:00 2002
From: rnd at onego.ru (Roman Suzi)
Date: Wed, 12 Jun 2002 00:20:00 +0400 (MSD)
Subject: Why does Python mix OO concepts and non OO concepts for operation
  s  on basic types?
In-Reply-To: <3D054A98.3090708@bgb.cc>
Message-ID: <Pine.LNX.4.33.0206120008180.9668-100000@rnd.onego.ru>

On Tue, 11 Jun 2002, Don Garrett wrote:

>Hans Nowak wrote:
>> Besides, while Python is object-oriented, that isn't the only
>> paradigm in the language. People coming from a functional
>> background may find the len() function more natural than a
>> method.
>> 
>> YMMV,
>> 
>
>   That's one thing I don't get. Why wasn't len() added as a member to the 
>various types during the recent rework that added so many other members to the 
>native types?

...and what for do we need all those funny slices,
attribute assignments, ..?

a.set(b, c)
print a.get(b)
a.set_b(c)

is much nicer than:

a[b] = c
print a[b]
a.b = c

I also wonder why to contaminate a wonderful language with
all those <, >, &, |, ..., +, - ? Look, how nice these examples are:

a.add(b.multiply(c))

instead of ugly, Perlish, unOOPish:

a + b * c

...

I only forgot how will I do this:

map(string.split, s)

if GvR will deprecate string module...

Sincerely yours, Roman Suzi
-- 
\_ Russia \_ Karelia \_ Petrozavodsk \_ rnd at onego.ru \_
\_ Tuesday, June 11, 2002 \_ Powered by Linux RedHat 7.2 \_
\_ "I can't use Windows. The cat ate my mouse." \_





From jepler at unpythonic.net  Fri Jun 14 09:48:58 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Fri, 14 Jun 2002 08:48:58 -0500
Subject: Why does Python mix OO concepts and non OO concepts for opera tion s  on basic types?
In-Reply-To: <B43D149A9AB2D411971300B0D03D7E8BF0A387@natasha.auslabs.avaya.com>
References: <B43D149A9AB2D411971300B0D03D7E8BF0A387@natasha.auslabs.avaya.com>
Message-ID: <20020614134857.GB30070@unpythonic.net>

On Fri, Jun 14, 2002 at 09:57:12AM +1000, Delaney, Timothy wrote:
> > From: holger krekel [mailto:pyth at devel.trillke.net]
> > 
> > probably. With python 1.6 many methods of the string *module* were
> > included included directly with the *type* string ('str'). Most 
> > python developers think that the string module should be deprecated 
> 
> I think "most" is a little bit strong. The general feeling that I get seems
> to be pretty much split down the middle.
> 
> I personally *don't* favour getting rid of the string module - there are
> many things which lend themselves better to a non-object-oriented style -
> and many of the string functions are included.
> 
> Now, if the str type included *all* of this functionality, *and* was aliased
> to string, then you may have a case ... although 'import string' would still
> need to work (for backwards compatibility) until 3.0.

There are other problems with using "str" instead of "string", even when
the methods exist:
>>> u = U"  hi there   " 
>>> str.rstrip(u)
TypeError: descriptor 'rstrip' requires a 'str' object but received a 'unicode'
>>> string.rstrip(u)
u'  hi there'

Jeff




From gerhard at bigfoot.de  Wed Jun 12 10:52:11 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 12 Jun 2002 14:52:11 GMT
Subject: (no subject)
References: <mailman.1023640273.18210.python-list@python.org> <m37kl8b4tp.fsf@mira.informatik.hu-berlin.de>
Message-ID: <slrnageo1j.1bg.gerhard@lilith.my-fqdn.de>

In article <m37kl8b4tp.fsf at mira.informatik.hu-berlin.de>, Martin v. Loewis wrote:
> [cgi programming with Python]
> There are several strategies, but I recommend to use plain print
> statements.

I would recommend to write to a StringIO.StringIO instance for the http body
instead, then, at the end, when you finally know that nothing exceptional
happened, you know which HTTP header to write, and then just print the body.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From donn at u.washington.edu  Mon Jun 24 18:52:52 2002
From: donn at u.washington.edu (Donn Cave)
Date: 24 Jun 2002 22:52:52 GMT
Subject: AF_UNIX + SOCK_DGRAM
References: <aes1he$gv2$1@news.idiom.com> <aet3ai$1l0o$2@nntp6.u.washington.edu> <af84af$1dkt$1@news.idiom.com> <Xns92386002B71cliechtigmxnet@62.2.16.82>
Message-ID: <af87s4$e8e$1@nntp6.u.washington.edu>

Quoth Chris Liechti <cliechti at gmx.net>:
[quoting me]
| >> Actually think you have run into a bug in Python's socket module.
| >> If it's convenient for you to build your own socket module, change
| >> it like this and see what happens (I'm looking at version 2.1 here.)
| > 
| >> *** socketmodule.c.dist Sun Apr 15 17:21:33 2001
| >> --- socketmodule.c      Thu Jun 20 10:10:12 2002
| >> ***************
| >> *** 597,602 ****
| >> --- 597,603 ----
| >>         case AF_UNIX:
| >>         {
| >>                 struct sockaddr_un *a = (struct sockaddr_un *) addr;
| >> +               a->sun_path[addrlen - (sizeof(*a) -
| >> sizeof(a->sun_path))] = 0; 
| >>                 return PyString_FromString(a->sun_path);
| >>         }
| >>   #endif
|
| why not simply use PyString_FromStringAndSize(const char *v,  int len)?
|
| couln't you just file a patch on SF? should i do it?

PyString_FromStringAndSize is OK by me, and it's also fine with me if
you would like to submit a patch to Sourceforge.

	Donn Cave, donn at u.washington.edu


From talon34 at hotmail.com  Wed Jun 26 10:49:34 2002
From: talon34 at hotmail.com (Talon)
Date: Wed, 26 Jun 2002 14:49:34 GMT
Subject: Installation problems DCOracle2 & W2000
References: <talon34-C09E81.19332125062002@netnews.attbi.com> <3D1934D1.3000707@nospam.majid.fm>
Message-ID: <talon34-431D67.07480126062002@netnews.attbi.com>

In article <3D1934D1.3000707 at nospam.majid.fm>,
 Fazal Majid <news at nospam.majid.fm> wrote:

> Talon wrote:
> > I am having a heck of a time getting DCOracle2 to funcion properly.  I 
> > am running on a W2K machine, python2.1.3 and Zope 2.5.1, both of which 
> > run just fine.  I installed DCOracle2 and the ZOracleDA under Zope and 
> > it works fine.  But standalone python scripts don't see the dco2.py 
> > module.  So I tried to install DCOracle2 in the python21/lib folder.  I 
> > got that latest version from Zope.org and it said it installed properly.  
> > But it doesn't run.  If you try to import it into a script, you get 
> > complaints that it can't find this or that module, all of which are part 
> > of the Zope path.
> 
> I have gotten DCOracle2 to work correctly on my Windows 2000 setup with 
> 2.1.1 (but I haven't tried it with Zope). You do need to set your 
> Pythonpath environment variable correctly, as well as copy the correct 
> dco2.pyd (DLL) version for your Python and Oracle versions.
> 
> -- Fazal Majid
> 
> If you want to respond by email, take the nospam out of my email address.
> 

Thanks for the response Fazal.  I believe the python path is correct, as 
other python scripts run fine.  The dco2.pyd file is copied to the 
correct directory by the latest version of install.py that comes with 
it.  But when a script tries to import DCOracle2, I get an error saying 
it is looking for Buffer or DateTime, all of which are part of the Zope 
path.

Mark


From bokr at oz.net  Wed Jun 12 14:56:38 2002
From: bokr at oz.net (Bengt Richter)
Date: 12 Jun 2002 18:56:38 GMT
Subject: Negative long literals (was Re: Does Python need a '>>>' operator?)
References: <mailman.1023711733.6056.python-list@python.org>
Message-ID: <ae85h6$t0d$0@216.39.172.122>

On Mon, 10 Jun 2002 08:19:09 -0400, Tim Peters <tim.one at comcast.net> wrote:

>[Beni Cherniavksy]
>> I just got another idea: use 0x1234 for 0-filled numbers and 1xABCD for
>> 1-filled ones.  That way you impose no restrictions on what follows the
>> prefix and keep backward compatibility.  0xFFFFFFFF stays a 2^n-1
>> _positive_ number, as it should be.  The look of 1x is weird at first but
>> it is very logical...
>
>Indeed, on both counts <wink>.  I like it a lot!
>
>For the other points related to getting rid of machine long-size
>assumptions, see PEP 237:  <http://www.python.org/peps/pep-0237.html>.
>
It was too easy for me to grasp at first ;-) Now that I know that
1x7 == 1xf7 == 1xff7, I can relate it to my (with Greg mod) suggested
format 0hf7. Mostly they're interchageable by substituting 0hf for 1x
and 0h0 for 0x as prefixes, but I like Beni's better. (They'd be fully
interchangable if 1x == 1xf ;-)

In a private email Paul Foley joked about writing 042 for 42 and 942
for -58 (i.e., why limit it to radix 16), and protested that machine
representation was being exposed.

It was not my purpose to expose machine representation, and neither
format (mine or Beni's) should be thought of as such IMO. I was
actually going to suggest an analogous binary notation (0b...), so
I'm happy not to limit it to radix 16.

The following decoding routine shows that the formats are not
about exposing machine representation, IMO:

 >>> def benidecode(s):
 ...     if s[1].lower()=='x': return long(s[2:],16)-long(s[0])*16L**(len(s)-2)
 ...     elif s[1].lower()=='b': return long(s[2:],2)-long(s[0])*2L**(len(s)-2)
 ...     elif s[1].lower()=='d': return long(s[2:],10)-long(s[0])*10L**(len(s)-2)
 ...     elif s[1].lower()=='o': return long(s[2:],8)-long(s[0])*8L**(len(s)-2)
 ...     raise NotImplementedError, 'radix "%s" not implemented"' % s[1]
 ...
 >>> benidecode('1x7')
 -9L
 >>> benidecode('1b0111')
 -9L
 >>> benidecode('1o67')
 -9L
 >>> benidecode('0x7')
 7L
 >>> benidecode('0b0111')
 7L
 >>> benidecode('0o07')
 7L
 >>> benidecode('0z07')
 Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "<stdin>", line 6, in benidecode
 NotImplementedError: radix "z" not implemented"

And for Paul ;-)
 >>> benidecode('0d42')
 42L
 >>> benidecode('0d042')
 42L
 >>> benidecode('1d42')
 -58L
 >>> benidecode('1d942')
 -58L
 >>> benidecode('1d9942')
 -58L
 
Regards,
Bengt Richter


From gina02122000 at yahoo.com  Fri Jun 28 13:44:36 2002
From: gina02122000 at yahoo.com (ginak)
Date: Fri, 28 Jun 2002 17:44:36 +0000 (UTC)
Subject: Is Grail alive?
Message-ID: <afi7a4$vv$1@reader2.panix.com>

The links from http://grail.sourceforge.net/source/license.html seem
to be invalid...

Thanks,

G.



From BPettersen at NAREX.com  Thu Jun 13 17:06:01 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Thu, 13 Jun 2002 15:06:01 -0600
Subject: Type subclassing: bug or feature
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E2019221519F@admin56.narex.com>

> From: Aahz [mailto:aahz at pythoncraft.com] 
> 
> Consider the following code:
> 
> class MyStr(str):
>     def contains(self, value):
>         return self.find(value) >= 0
> 
> s = MyStr("hello, world!")
> s = s.capitalize()
> if s.contains('Hello'):
>     print "Found it!"
> 
> It fails with an AttributeError when it calls s.contains(), because
> s.capitalize() returned a str instead of a MyStr.  Anyone 
> want to take a whack at defending this as the correct behavior?

Well, in e.g. C++ this would make perfect sense, since a base class
method can't know the real type of self/this. In Python we can do better
of course:

>>> class A:
...   def foo(self):
...     return self.__class__()
...
>>> class B(A):
...   pass
...
>>> b = B()
>>> b.foo()
<__main__.B instance at 0x00864CC0>

which version is correct is mostly a religious question, but personally
I think that if we _can_ do better than C++ we should <wink>.

File a bug and see what Guido says...

-- bjorn




From bokr at oz.net  Tue Jun 25 21:09:39 2002
From: bokr at oz.net (Bengt Richter)
Date: 26 Jun 2002 01:09:39 GMT
Subject: SOLVED (Re: Python hits the spot)
References: <mailman.1025043213.28349.python-list@python.org>
Message-ID: <afb48j$vl2$0@216.39.172.122>

On Tue, 25 Jun 2002 15:12:47 -0700, "Jimmy Retzlaff" <jimmy at retzlaff.com> wrote:

>Bengt Richter wrote:
>>>> Well, lucky guess on the heat. That _probably_ accounts for laptop
>>>> going to 87% and then shutting down, but how about the other huge
>>>> slowdowns in your big loop of 8? Did you get that diagnosed too?
>>>> Just curious ;-)
>>>
>>>On my stationary machine definitely not. The 8 steps remain constant.
>I=20
>>
>>Still curious ;-)
>
>I believe that newer mobile CPUs will throttle back their clock speed in
>response to heat. This could explain the slow down. Apparently the
                                      ^^^-- or "some of the"
>lowest clock speed wasn't low enough to keep the CPU happy...
>
Yeah, I understood that (having introduced the concept into this thread, I think ;-)
But it was not clear to me whether that explained _everything_ (see other post re two
readings of "On my stationary machine definitely not. The 8 steps remain constant." ;-)

Regards,
Bengt Richter


From pedro_rodriguez at club-internet.fr  Sat Jun 29 15:27:52 2002
From: pedro_rodriguez at club-internet.fr (Pedro RODRIGUEZ)
Date: Sat, 29 Jun 2002 21:27:52 +0200
Subject: aspect-oriented demo using metaclasses
References: <mailman.1025278420.26105.python-list@python.org>
Message-ID: <pan.2002.06.29.19.23.48.841533.2265@club-internet.fr>

On Fri, 28 Jun 2002 17:31:06 +0200, Mark McEahern wrote:

<snip>

> I owe much to previous threads on AOP, in particular Pedro Rodriguez's
> non-metaclass implementation:
> 

Mark,

If you really want something more consistent than my posting, I will
recommend you to also check:

    - Pythius by Juergen Hermann
        where Frank J. Tobin has introduced an aop module
        http://pythius.sourceforge.net/ by Juergen Hermann

    - Transwarp by Phillip J. Eby
        this a more complex framework using in some way concepts of aop
        http://telecommunity.com/TransWarp/
        http://www.zope.org/Members/pje/Wikis/TransWarp/HomePage


<snipping introduction on metaclass>

> 
> We want to iterate through the class's dict and do something for each
> method (I'm not worried about distinguishing staticmethod and
> classmethod type methods for now):
> 

You should <wink>. At least for the purpose of the exercise ;)
(Honestly I didn't try either, it may end up with a two liner...
but I wonder if they will take two minutes or two hours to be written ;)


<sniping>

> Rather than importing types and using types.FunctionType, I just compare
> the type of the item to the type of an anonymous function (via lambda).
> Is that cool or what?  ;-)
> 

Even if deprecation of the types module have been discussed on the devel
list, I think it is preferable.


<sniping>

> Now we're starting to get somewhere!  The next step is to create a
> framework where multiple observers can plug into the before and after
> events for any given method call.  Rather than taking more baby steps to
> get there, this is the complete implementation I have so far:
> 

You went even further than that. In the next step, you start separating
(at least ;) the aspect and the 'aspectified' object in a less intrusive 
way.


> <code>
> import sys
> 
> class wrapped_method(object):
> 
>     def __init__(self, cls, method):
>         self.class_name = cls.__name__
>         self.method = method
>         self.method_name = method.func_name
>         self.observers = []
> 
>     def __call__(self, *args, **kwargs):
>         self.before(*args, **kwargs)
>         self.method(self, *args, **kwargs)
>         self.after(*args, **kwargs)
> 

Oops... Didn't I make the same mistake... 
... forgetting about the returned value ;)

Something like could be better:
     def __call__(self, *args, **kwargs):
         self.before(*args, **kwargs)
         ret = self.method(self, *args, **kwargs)
         self.after(*args, **kwargs)
         return ret
 

<sniping>

> class Trace(Observer):
> 
>     def __init__(self, filename=None):
>         self.filename = filename
> 
>     def write(self, prefix, class_name, method_name, *args, **kwargs):
>         cls = class_name
>         s = "%s: %s.%s(%s, %s)" % (prefix, cls, method_name, args,
>         kwargs) if not self.filename:
>             f = sys.stdout
>         else:
>             f = file(self.filename)
>         f.write(s)
>         f.write("\n")
>         if self.filename:
>             f.close()
> 
>     def before(self, class_name, method_name, *args, **kwargs):
>         self.write("before", class_name, method_name, *args, **kwargs)
> 
>     def after(self, class_name, method_name, *args, **kwargs):
>         self.write("after", class_name, method_name, *args, **kwargs)
> 

This Trace class is an aspect.


> class Aspect(type):
> 
>     def __init__(cls, name, bases, dict):
>         super(Aspect, cls).__init__(name, bases, dict) for k,v in
>         dict.items():
>             if type(v) == type(lambda x:x):
>                 setattr(cls, k, wrapped_method(cls, v))
> 

This is not an Aspect. This is your way to implement method call interception.


<sniping>

> ************
> Observations
> ************
> 
> Obviously, Trace is designed so that I can specify a filename when
> creating an instance of it and that would use that file rather than
> sys.stdout.
> 
> Observers should probably be able to register for particular events
> (i.e., just before).  They should be able to unregister.  It should also
> be possible to register for some classes, some methods, and not others.

What about module functions ? Bounded methods ?
Using metaclass for call interception is too restrictive I believe.


> What is a good way to write the rules for enrolling in notifications?
> (And what's the word for this that AspectJ uses?)

Joinpoint... but you already found it ;)


> 
> I explicitly pass self from the wrapped_method instance to the actual
> wrapped method.  However, this is not really the instance of the class,
> it's the instance of the wrapped_method object.  How do I pass the
> actual instance to the wrapped method--whether explicitly or implicitly?
> 

This is the trickier part. When you do :
       setattr(cls, k, wrapped_method(cls, v))
you substitute a function (actually an unbounded method) by a callable
object. Unfortunately when you'll invoke this object with a classical
method call, Python will not pass the instance as the first argument.

I only know two ways to substitute an unbounded method by an object
that will allow to retrieve the instance argument :
- providing a function, and take benifit of nested_scopes to retrieve
  all the information from your context
- create an object through the 'new' package (new.instancemethod)
  that was broken for new type objects in 2.2, and fixed in 2.2.1

So your wrapped_method class is not sufficient per se to achieve this
goal.


> I should probably explicitly set the metaclass for wrapped_method,
> Trace, Observer, and even VirtualClassError to "type" (the default
> metaclass) to avoid weird infinite loops where the members of the aspect
> framework are themselves being aspected.
> 

I don't think that metaclass'ing is the good thing to do, but that's
just me.


> I need to think more about how this would be used.  My vague impression
> of AspectJ is that you use it from an IDE, specifying different
> conditions for filtering classes and methods (what's the word for that
> again?).  When you compile with aspect support turned on, the generated
> code has the aspects weaved in.  What are the equivalent developer use
> cases for weaving aspects into Python code? The whole point it seems to
> me is that when I author the classes to be aspected, I shouldn't have to
> do a thing--no subclassing, no method definition.  I just write my
> classes, omitting entirely any consideration of the future aspects to be
> weaved in.  I mean, that's the whole point, right?  To simplify the
> code.
> 

Yep. And this is why I consider the metaclass thing for being too intrusive.


> In the example, foo doesn't know diddly about the fact that it's being
> traced.  And we could turn that tracing on or off simply by changing the
> module-level __metaclass__ variable.
> 

Too intrusive. I don't believe that you can do it dynamically, at least
not for classes defined at module level. They will be created with the
metaclass defined at compilation time.


<sniping>

> It would be easy to add aspects for error handling.  This is a very
> quick and dirty example for error notification:
> 

<snip>

Yes. Interception of raised exception is a good (and easy ;) feature.
Just try to go a step further with 'around' methods.


Thanks, for this posting Mark. Reminds me that aop is my task list for
the moment I will have some spare time (not so far I hope ;)

Just to give some hints on aop, this is what I wrote as a reminder
6 months ago along with a more complete implementation of aop :


    Vocabulary

    pointcut
    an object identifying a set of joinpoints, and aggregating a set of
    advices

    joinpoint
    a callable function in Python (unbound method, bound method, function)

    advice
    a function that will be called when a joinpoint is reached. It may be
    triggered before the joinpoint is called, after, or when an exception
    occurs.
    An advice may also be set around a joinpoint, meaning that it will be
    able to do some job before and after the joinpoint runs, but it will
    have to call the joinpoint himself.

    aspect
    a class that contains poincuts and their related advices.


And a UML'ish model :
+-------------+       n  +-------------+       1 +-------------+
| PointCut    | - - - -> |  JoinPoint  |-------->|   Function  +
+-------------+          +-------------+         +-------------+
      |                    |
      |                    |
      V  1                 |
+-------------+  n         |
| AdviceStack |<-----------+
+-------------+
      |
      |
      V  n
+-------------+
|   Advice    |
+-------------+

- A pointcut uses several joinpoints
- Since a joinpoint may participate in several pointcuts, it will have
  to trigger several set of advices (AdviceStack) related to each of the
  pointcut it belongs to.
- A joinpoint is related to a Python function. It will intercept calls
  to it.



+-------------+         1 +-------------+
|  JoinPoint  |---------->|   Function  |
+-------------+           +-------------+
                                 ^ n
                                 |
                                 |
                          +-------------+
                          | CallableSet |
                          +-------------+
                                 A
                                 |
       +-----------------+-------+---------+----------------+
       |                 |                 |                |
+--------------+  +-------------+   +-------------+  +-------------+
|ClassicalClass|  |  TypeClass  |   |    Module   |  |  Instance   |
+--------------+  +-------------+   +-------------+  +-------------+

- A Python function related to a joinpoint belongs to a CallableSet.
- A CallableSet can be a classical python class, in which case the
  function is an unbound method.
- A CallableSet can be a new python class, in which case the function
  is an unbound method.
- A CallableSet can be a python instance, in which case the function
  is a bound method.
- A CallableSet can be a python module, in which case the function
  is a standard method.


Best regards,
Pedro



From Chris.Barker at noaa.gov  Fri Jun 21 15:53:02 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Fri, 21 Jun 2002 12:53:02 -0700
Subject: do...while
References: <mailman.1024613916.27915.python-list@python.org>
Message-ID: <3D13841D.F3C9EB5B@noaa.gov>

Bjorn Pettersen wrote:
> Doing a quick search over our C++ libraries (~900KLoc), "do {} while()"
> seems to be used in right around 3% of the loops (including while, for;
> excluding recursion).

I was surprised that it was used so rarely, but them I realized that in
C/C++, assignment returns a value, so you can do the equivalent of:

while line = file.readline():
	....

which is used a lot, but can't be done in Python. That's still not a
reason to add the construct, but it is a commonly occuring structure.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From emile at fenx.com  Thu Jun 27 21:56:08 2002
From: emile at fenx.com (Emile van Sebille)
Date: Fri, 28 Jun 2002 01:56:08 GMT
Subject: Most elegant python program to date.
References: <mailman.1025223937.24431.python-list@python.org>
Message-ID: <YmPS8.341189$cQ3.19250@sccrnsc01>

"Delaney, Timothy" <tdelaney at avaya.com> wrote in message
news:mailman.1025223937.24431.python-list at python.org...
> > From: PoulsenL at capecon.com [mailto:PoulsenL at capecon.com]
> >
> > examples of its prose.  Are there any nominations for the most
elegant
> > example of a python program to date.  Not sure under what
>
> print 'Hello, world!'
>

Cool... except the the single/double quotes, that's pure BASIC!  ;-)

OK-so-maybe-the-case-of-print-is-wrong-too ly y'rs,

--

Emile van Sebille
emile at fenx.com

---------



From bokr at oz.net  Thu Jun 13 16:44:35 2002
From: bokr at oz.net (Bengt Richter)
Date: 13 Jun 2002 20:44:35 GMT
Subject: Type subclassing: bug or feature
References: <aeau48$111$1@panix1.panix.com>
Message-ID: <aeb07j$kiv$0@216.39.172.122>

On 13 Jun 2002 16:08:40 -0400, aahz at pythoncraft.com (Aahz) wrote:

>Consider the following code:
>
>class MyStr(str):
>    def contains(self, value):
>        return self.find(value) >= 0
>
>s = MyStr("hello, world!")
>s = s.capitalize()
>if s.contains('Hello'):
>    print "Found it!"
>
>It fails with an AttributeError when it calls s.contains(), because
>s.capitalize() returned a str instead of a MyStr.  Anyone want to take a
>whack at defending this as the correct behavior?
I don't think you could generally coerce base class method results
to subclass instances, so I think it has to be explicit. Maybe it
could be done with a method list for a more compact way (using a metaclass?)
Anyway:

 >>> class MyStr(str):
 ...     def contains(self, value):
 ...         return self.find(value) >= 0
 ...     def capitalize(self):
 ...         return MyStr(str.capitalize(self))
 ...
 >>> s = MyStr("hello, world!")
 >>> s = s.capitalize()
 >>> if s.contains('Hello'):
 ...     print "Found it!"
 ...
 Found it!


Regards,
Bengt Richter


From marklists at mceahern.com  Thu Jun 13 08:56:22 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 13 Jun 2002 07:56:22 -0500
Subject: get a list of classes in this module
In-Reply-To: <E17INzC-000057-00@tswings.swing.cluster>
Message-ID: <NCBBLFCOHHDIKCAFGCFBGEGKLEAA.marklists@mceahern.com>

[Christian Tanzer]
> I'd recommend to use a custom metaclass to put the classes in question
> into the dictionary -- this way some other module can add credit card
> classes and it should still work.

Hmm, putting the other credit card classes in other modules creates a new
problem:  I have to import those modules.  Another way of looking at what
I'm trying to do--that I failed to state explicitly--is that I'm trying to
get a list of all the available classes at runtime, via introspection,
without having to specify them explicitly.  Having to import the other
modules amounts to having to specify them explicitly.

Here's another solution to the problem, fwiw, using classmethod to stick the
factory method inside the base class:

class CreditCardValidation(object):

  def factory(cls):
    classes = cls.__subclasses__()
    d = {}
    for c in classes:
      d[c.__name__] = c
    return d
  factory = classmethod(factory)

  def validate(self, card_number):
    raise NotImplementedError("Subclasses must override.")

class Visa(CreditCardValidation):

  def validate(self, card_number):
    print "Visa: %s" % card_number

class MasterCard(CreditCardValidation):

  def validate(self, card_number):
    print "MasterCard: %s" % card_number

f = CreditCardValidation.factory()
card_type = "Visa"
card_number = "5105 1051 0510 5100"

validator = f[card_type]()
validator.validate(card_number)

Thanks for the suggestion.  I find it much easier to just utilize
__subclasses__ than to peer into the mysteries of metaclass programming.

Cheers,

// mark

-





From jepler at unpythonic.net  Thu Jun 27 12:00:15 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Thu, 27 Jun 2002 11:00:15 -0500
Subject: Var substitution
In-Reply-To: <c07d0edc.0206270557.28180893@posting.google.com>
References: <c07d0edc.0206270557.28180893@posting.google.com>
Message-ID: <20020627160014.GE3710@unpythonic.net>

Guy,
"A good FORTRAN programmer can program FORTRAN in any language."
    -- Allen Brown.

You're asking questions about how to do things in Python that, while
possible, are not idiomatic.  It sounds like you're probably coming from
background in another language or trying to directly translate a program
written in another language.

While people are giving you good answers about how to do the things
you're asking about, I suspect that you would be better off in the long
run learning to write Python programs than to write programs for
language XXX in Python.

Jeff




From claird at starbase.neosoft.com  Tue Jun 25 08:33:14 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 25 Jun 2002 07:33:14 -0500
Subject: Python and Eclipse
References: <3D17FAF4.35CC9F1B@earthlink.net>
Message-ID: <81CC01E58BF2A99B.DFCD5EC6E4654907.32D731380951B9CA@lp.airnews.net>

In article <3D17FAF4.35CC9F1B at earthlink.net>,
Joseph A Knapka  <jknapka at earthlink.net> wrote:
>Hi folks,
>
>I am interested in having Python support in
>Eclipse <URL: http://www.eclipse.org> - interested
>enough to contribute code, or to tackle the
>task myself, if need be. Of course, I'd like
>the Python support to be at least as good as the
>Java support, which is a pretty tall order,
>'cause the Java support in Eclipse is awfully
>damn good --- it nearly allows me to *enjoy*
>writing Java code :-)
>
>Anyway, I know some other folks have expressed
>interest in this as well, and I'd like to hook
>up with them. If you're working on Python+Eclipse
>and you've got code in any state of completion,
>I'm willing to alpha- or beta-test, fix stuff,
>add new stuff, etc.
			.
			.
			.
Can you say a few more words about Eclipse?  It ... well, I'm
wary.  I understand its marketing more than its technology,
and on hostile days, I think it's *only* marketing.  Maybe you
can make it more appealing.  "[It] nearly allows me to *enjoy*
writing Java ..." is certainly a good start.
-- 

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


From sholden at holdenweb.com  Mon Jun 17 20:11:09 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Mon, 17 Jun 2002 20:11:09 -0400
Subject: why not "'in' in 'in'"?
References: <mailman.1023989567.18847.python-list@python.org> <aeb20i01psk@drn.newsguy.com>
Message-ID: <UaMP8.165344$Fp1.92799@atlpnn01.usenetserver.com>

"Grant Griffin" <not.this at seebelow.org> wrote ...
[ ... ]
>
> just-because-guido-allows-it-doesn't-mean-you-should-use
>    -it-injudicious-ly y'rs,


But you don't mind complaining about the rules he has painstakingly
established for sequence membership.

Characters have always been strings of length one in Python, and that *does*
make strings slightly anomalous. Your arguments that sequence membership
should differ between string and other sequences are unconvincing.

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From rajarshi at presidency.com  Tue Jun 11 12:35:29 2002
From: rajarshi at presidency.com (Rajarshi Guha)
Date: Tue, 11 Jun 2002 12:35:29 -0400
Subject: sftp interface for python
Message-ID: <pan.2002.06.11.12.35.29.475361.16259@presidency.com>

Hi,
  I need to Python'ize some scripts which involve sftp. Is there any sftp
interface for Python? Or is the best way system (or popen et al)?

TIA,
Rajarshi Guha


From sholden at holdenweb.com  Tue Jun  4 12:47:54 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 4 Jun 2002 12:47:54 -0400
Subject: Behavior of += (was Re: [Python-Dev] Customization docs)
References: <mailman.1023144725.29483.python-list@python.org> <slrnafo7qv.37c.huaiyu@gauss.almadan.ibm.com> <ufpb3sn6332ub4@news.supernews.com> <9A5L8.232032$%u2.84870@atlpnn01.usenetserver.com> <20020604112824.B24006@unpythonic.net>
Message-ID: <019301c20be7$933acb50$7201a8c0@holdenweb.com>

----- Original Message -----
From: "Jeff Epler" <jepler at unpythonic.net>
To: "Steve Holden" <sholden at holdenweb.com>
Cc: <python-list at python.org>
Sent: Tuesday, June 04, 2002 12:28 PM
Subject: Re: Behavior of += (was Re: [Python-Dev] Customization docs)


> On Tue, Jun 04, 2002 at 12:01:56PM -0400, Steve Holden wrote:
> > >>> a = b = [1,2,3]
> > >>> a += [4]
> > >>> a, id(a), b, id(b)
> > ([1, 2, 3, 4], 269472088, [1, 2, 3, 4], 269472088)
> > >>>
> >
> > Where did the rebinding take place? ISTM that "a" is still bound to the
same
> > list, which has been modified in place.
>
> Sure it rebinds a (just to the same object formerly bound to it) ..
> >>> class X:
> ...     def __setattr__(self, a, v):
> ...         print "rebinding %s, sucker" % a
> ...         self.__dict__[a] = v
> ...
> >>> x = X()
> >>> x.a = b = [1,2,3]
> rebinding a, sucker
> >>> x.a += [4]
> rebinding a, sucker
> >>> x.a is b
> True
>
Interesting. But there was no need to call me a sucker.

> Or are you saying that in
>     a = b = [1,2,3]
>     a = b
> the second statement isn't a rebinding operation?  That seems like an
> odd position to take....
>
It wouldn't be the oddest I've ever taken, but now you have clarified your
assertion I understand what you were saying.

> > Clearly an explicit assignment to a tuple element is always going to
fail,
> > so I don't really see what this would tell you.
>
> But you're claiming that there *is* no binding, so it would be
> surprising that this augmented assignment wouldn't work
>     >>> t = ([1])
>     >>> t[0] += [2]
> t[0] "isn't" being rebound (according to your above logic, since if the
> operation could complete, the "is" relation would hold between old t[0]
> and new t[0]) .. but, of course, this actually causes a traceback.
>
Indeed, I now understand exactly why the error occurs. I have to say I still
think that the existing behavior is reasonable.

> I hate augmented assignment more and more each day, but at least I think
> I understand it.  Your claim that there is no rebinding operation makes
> me think that you may not.
>
Possibly. In fact it seems to me that it's only *because* a rebinding does
in fact take place that the tuple element assignment fails, so it's just as
well it does.

There's no point hating augmented assignment. Just don't use it if you find
it gives confusing results.

happy-when-i-didn't-understand-and-happy-now-i-do'ly y'rs  - steve
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From martin at v.loewis.de  Sun Jun  9 15:50:02 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 09 Jun 2002 21:50:02 +0200
Subject: Embed python in multi-thread environment
References: <mailman.1023636372.9719.python-list@python.org>
Message-ID: <m3fzzwb5at.fsf@mira.informatik.hu-berlin.de>

Arlo <arlo.maillist at ozmaker.com> writes:

> it seems i can't call PyObject_CallObject twice at the same process?

You can, but you must hold the GIL while doing so. In each thread,
allocate a thread state (perhaps just for the lifetime of the Python
call), and allocate the GIL.

Regards,
Martin



From romany at actimize.com  Sun Jun  9 10:29:50 2002
From: romany at actimize.com (Roman Yakovenko)
Date: Sun, 9 Jun 2002 16:29:50 +0200
Subject: Problem with wxPython
Message-ID: <7647A9F4B298E74DB6793865DA67285004AD78@exchange.adrembi.com>

Could you post small piece of code. I have some computer configuration
and it works fine to me. Also try Boa IDE. It is very helpful.

Roman

-----Original Message-----
From: Stephane Larouche [mailto:steph_lar at videotron.ca]
Sent: Saturday, June 08, 2002 3:57 AM
To: python-list at python.org
Subject: Problem with wxPython

I use wxPython 2.3.2.1 with Python 2.2.1 on multiple computers running
WinNT 4.0 SP6. On one of those computers, which has a setup that is
identical to the setup of the other computers (I think), wxPython
software crashes when a wxFrame is initialized with any of the window
styles wxMAXIMIZE_BOX, wxRESIZE_BOX or wxSYSTEM_MENU (or
wxDEFAULT_FRAME_STYLE that includes those styles). Even the demo
provided with wxPython crashes. According to DrWatson, the crash is
caused by a stack overflow during the execution of the function
wxTransformMatrix::operator-.

Have your ever seen the same problem? Is there a workaround?

Thank you for your help!

Stephane Larouche
steph_lar at videotron.ca
--
http://mail.python.org/mailman/listinfo/python-list




From woodsplitter at rocketmail.com  Fri Jun 21 01:00:35 2002
From: woodsplitter at rocketmail.com (David Rushby)
Date: 20 Jun 2002 22:00:35 -0700
Subject: C extension/multithreading question
Message-ID: <7876a8ea.0206202100.b9d6383@posting.google.com>

Is it safe to call PyThread_acquire_lock and PyThread_release_lock on
a PyThread_type_lock from a native thread that:
a) was not created by Python
b) has no bootsrapped Python thread state
c) does not have the GIL locked
?

It appears that this is safe, since the GIL itself is of type
PyThread_type_lock, but I want to make sure that I'm not missing some
fatal subtlety.

Thanks.


From peter at engcorp.com  Sun Jun  9 03:06:45 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 09 Jun 2002 03:06:45 -0400
Subject: Program mysteriously STOPS!
References: <yZuM8.146064$ux5.186314@rwcrnsc51.ops.asp.att.net>
Message-ID: <3D02FE85.EA48FBEA@engcorp.com>

LJD wrote:
> 
> Hi all:
> 
> The following code is designed (at least thats my intent) to run forever.
> The program should sit in a loop until it connects with the server.  It
> should just loop and loop and loop.
> 
> It does, for a while, maybe five or ten minutes, but then the process just
> seems to vanish.  It's in the c1==0 loop when it quits. When I do a PS AUX,
> the process is just gone.  Obviously it quit, but more importantly WHY???,
> and what can I do so it doesn't?

You may need to investigate socket.shutdown() and socket.close() ...

Especially on Windows, you have a finite number of sockets you can have
open simultaneously.  Since socket resources are not necessarily freed by 
the underlying OS immediately upon Python releasing the reference as
it goes out of scope, you may need to be explicit.  Haven't done this 
enough under Linux to help beyond that...

-Peter


From menucool at yahoo.com.cn  Tue Jun 18 06:46:30 2002
From: menucool at yahoo.com.cn (coolmenu)
Date: 18 Jun 2002 03:46:30 -0700
Subject: Can python have lib to display PDF ?
Message-ID: <a6a2957e.0206180246.511fb932@posting.google.com>

i want a python application to display end edit pdf.
where are the lib pdf for python?
thank


From elfinman at yahoo.com  Wed Jun 26 17:50:25 2002
From: elfinman at yahoo.com (Timothy H)
Date: Wed, 26 Jun 2002 17:50:25 -0400
Subject: Memory Monitor
References: <pan.2002.06.25.12.14.06.994352.17223@pobox.com> <3D1894B5.720EB7F7@bioreason.com> <200206252139.QAA88038@starbase.neosoft.com> <3D18E752.7A5A4EAC@bioreason.com>
Message-ID: <20020626.175009.711845894.6390@yahoo.com>

In article <3D18E752.7A5A4EAC at bioreason.com>, "Mitch Chapman"
<Mitch.Chapman at bioreason.com> wrote:

> Cameron Laird wrote:
>> Mitch Chapman  <Mitch.Chapman at bioreason.com> wrote:
>> >        os.system("cat /proc/self/status | grep VmSize")
>> Rather than
>>   os.system("grep VmSize /proc/self/status")
>> ?
> 
> No, that's also concise =8)
> 

Concise, but it doesn't work. Then /proc/self/status is reporting the
memory usage for grep, not for the python interpreter+script.


From cliechti at gmx.net  Fri Jun 28 16:45:43 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 28 Jun 2002 22:45:43 +0200
Subject: Dialing out on MODEM
References: <mailman.1025270142.10532.python-list@python.org>
Message-ID: <Xns923BE7FFF29D3cliechtigmxnet@62.2.16.82>

"David McInnis" <david at dataovation.com> wrote in 
news:mailman.1025270142.10532.python-list at python.org:

> This is my object
> 
>>>> import serial
>>>> modem = serial.Serial
>>>> modem = serial.Serial(2)
>>>> modem.write("atdt3120992\n")

somtimes it it's a '\r\n' and maybe you need to send the escape sequence 
first, to enter the command mode. if i remember correctly thats 3 times ESC 
and then wait 100ms or so. either google or your modem manual should help. 
i had to do somthing like that to dial with pyserial, but unfortunately the 
code is on an other machine...

chris

-- 
Chris <cliechti at gmx.net>



From donn at drizzle.com  Mon Jun 10 22:34:58 2002
From: donn at drizzle.com (Donn Cave)
Date: Tue, 11 Jun 2002 02:34:58 -0000
Subject: M2Crypto: select() behaves weird on SSL socket
References: <3D04ECED.2030202@NOSPAMREMOVETHISxs4all.nl> <slrnag9rie.enn.akuchlin@ute.mems-exchange.org> <3D04FBF5.8090407@NOSPAMREMOVETHISxs4all.nl> <3D04FE2F.6070302@NOSPAMREMOVETHISxs4all.nl>
Message-ID: <1023762896.869994@yasure>

Quoth Irmen de Jong <irmen at NOSPAMREMOVETHISxs4all.nl>:
...
| The manual page of select(2) says:
| "select  waits  for  a number of file descriptors to change status.
|   Three independent sets of descriptors are watched.   Those
|   listed  in  readfds  will  be watched to see if characters
|   become available for reading (more precisely, to see if  a
|   read  will not block - in particular, a file descriptor is
|   also ready on end-of-file),  those  in  writefds  will  be
|   watched  to  see  if  a write will not block, and those in
|   exceptfds will be watched for exceptions."
|
| The way I interpret this is that -buffering or not- select must
| return a FD if there is more data available. In my case, the SSL
| socket indeed has more data available, and a recv() on the socket
| will not block. I therefore think that either M2Crypto or OpenSSL
| is doing the wrong thing because select() doesn't behave as it should?
| The recv() may read from a buffer, I don't care, but it doesn't
| block on my next call and thus I expect select() to return the socket
| as being ready for reading... am I wrong here?

Wrong, I'd say.  SSL's buffer, like the file object's stdio buffer,
is user process storage.  From select's point of view (the low level
device) you have already read that data, and it doesn't know or care
what you did with it.

I see my reply earlier today followed up a separate thread, after
you posted your initial question twice.  If you haven't seen it,
a hint:  SSL_pending(self->ssl.)  (I don't know much more than that,
anyway.)

	Donn Cave, donn at drizzle.com


From i.linkweiler at gmx.de  Sun Jun  9 05:10:20 2002
From: i.linkweiler at gmx.de (Ingo Linkweiler)
Date: Sun, 09 Jun 2002 11:10:20 +0200
Subject: Comparison: HTTP with Java, Perl or Python
References: <mailman.1023589393.995.python-list@python.org>
Message-ID: <3D031B7C.8010207@gmx.de>

>
>
>WHat you want to do reminds me of this (open source) project:
>
>http://www.bagley.org/~doug/shootout/
>
>- it already has a framework/examples for many things you 
>are measuring. (and for MUCH MUCH more languages).
>The project seem to be frozen but everybody can
>advance it further.
>  
>
thanks, that?s it!

Ingo



From sjmachin at lexicon.net  Fri Jun  7 18:40:41 2002
From: sjmachin at lexicon.net (John Machin)
Date: 7 Jun 2002 15:40:41 -0700
Subject: question on pattern
References: <3D007E96.3010505@rmb.co.za>
Message-ID: <c76ff6fc.0206071440.70ce82f6@posting.google.com>

Ulrich Pfisterer <ulrich.pfisterer at rmb.co.za> wrote in message news:<3D007E96.3010505 at rmb.co.za>...
> I want to do a pattern match on a string that spans multiple lines (c++ 
> style comments, re.match('(/\*.*?\*/)', filestring)
> If I use the re.findall(pattern, string) method it only finds those 
> matches that are on a single line. If I use re.match(pattern, string, 
> DOTALL) it spans multiple lines but only finds the first match.
> Any help on this one would be greatly appreciated.

You have got an answer to your problem of how to get DOTALL behaviour
with .findall(). However you should be aware that your pattern can
match strings that are not syntactically C or C++ comments. Here is an
example from page 172 of "Mastering Regular Expressions" by Jeffrey E.
F. Friedl:

   const char *cstart = "/*", *cend = "*/";

I'd agree that it's a pathological case. I'm just pointing out that in
most work on program source files, to be 100% correct you need to have
a lexical analyser for the language in question. Regexes won't take
you the whole distance.


From timo at alum.mit.edu  Fri Jun 28 00:09:52 2002
From: timo at alum.mit.edu (Timothy O'Malley)
Date: Fri, 28 Jun 2002 00:09:52 -0400
Subject: explain this pickle issue to me...
Message-ID: <timo-1418C7.00095228062002@corp.supernews.com>

hola.

Can someone explain why I can't seem to use Pickle (actually
cPickle) in combination with execfile?

Sample code:
-----------------------------------
Python 2.2.1 (#1, 06/20/02, 11:40:06) 
[GCC Apple devkit-based CPP 6.0] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> d = {}
>>> execfile("testme.py", d)
>>> q = d["writeSimple"]()
>>> print `q`
"(i__builtin__\nSimple\np1\n(dp2\nS'a'\nS'a'\nsS'b'\nI2\nsb."
>>> d["readSimple"](q)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
  File "testme.py", line 14, in readSimple
    s = cPickle.loads(ss)
AttributeError: 'module' object has no attribute 'Simple'
>>> d["readSimple"].func_globals.keys()
['__builtins__', 'Simple', 'cPickle', 'readSimple', 'writeSimple']


And the testme.py file used in execfile
---------------------------------------
import cPickle
class Simple:
    def __init__(self):
        self.a = "a"
        self.b = 2
# end class A

def writeSimple():
    s = Simple()
    return cPickle.dumps(s)
# end writeSimple

def readSimple(ss):
    s = cPickle.loads(ss)
    assert ininstance(s, A), "Not an instance of Simple"
# end readSimple


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


As I understand it from the docs, there is a restriction in
Pickle that requires class definitions to be in the global
namespace in order to be found.  And, it appears, I am running
afoul of that restriction.

But -- the last two lines of the posted sample code shows
that the class definition *is* in the global namespace of
the readSimple function.

What's the catch?

TimO


From claird at starbase.neosoft.com  Tue Jun 25 16:02:53 2002
From: claird at starbase.neosoft.com (Cameron Laird)
Date: 25 Jun 2002 15:02:53 -0500
Subject: Python daemon
References: <mailman.1024999953.15765.python-list@python.org> <F733C874CD014E16.148E57261FFD5E7A.AFFA4BD17B2F8D8B@lp.airnews.net> <v50S8.307703$352.31615@sccrnsc02>
Message-ID: <680C5E8D0C5D5E02.689CD98EE6CE4D7E.6018C29C8F1E8BF9@lp.airnews.net>

In article <v50S8.307703$352.31615 at sccrnsc02>,
Emile van Sebille <emile at fenx.com> wrote:
>Cameron Laird
>> This is one of the great FMMs under Unix--the under-utilization
>
>FMMs ?
			.
			.
			.
OK, I was being provocative.  It's the Federation of
Malaysian Manufacturers.

No, wait; while it *is* that, the more pertinent reading
appears at <URL: http://
starbase.neosoft.com/~claird/comp.lang.tcl/fmm.html >.
-- 

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


From peter at engcorp.com  Sat Jun  1 08:12:30 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 01 Jun 2002 08:12:30 -0400
Subject: Interactive socket connection
References: <3CF88F9B.1080308@saudi.net.sa>
Message-ID: <3CF8BA2E.A8C42539@engcorp.com>

Rayed Al-Rashed wrote:
> 
> Hello,
> 
> I am trying to build POP3 proxy, I want it to run from "inetd" so it should
> communicate with the user using regular stdin, stdout.
> I built a small code that should run as general proxy for any protocol, but
> I faced the following problem, it doesn't show any output from the server
> until I pass it a command [...]

Are you maybe in need of a sys.stdout.flush() to get the output to
be sent after the sys.stdout.write()?

-Peter


From glyph at twistedmatrix.com  Wed Jun 26 07:03:23 2002
From: glyph at twistedmatrix.com (Glyph Lefkowitz)
Date: Wed, 26 Jun 2002 06:03:23 -0500 (CDT)
Subject: Python 2.2 __slots__ Bug?
Message-ID: <20020626.060323.78699939.glyph@twistedmatrix.com>


I'm sorry if this is a known issue.  I couldn't get through to the bugtracker
at sourceforge.net; it's abysmally slow and I stopped waiting for the search to
run after about 15 minutes.  Searches on google yielded nothing similar for
'python __slots__ bug'.

I think I have discovered a weird bug in the Python interpreter.  Under
Python2.2, when I run this script:

    class foo(object):
        def __init__(self, n):
            self.n = n
        __slots__ = ['n']

    o = None
    for i in xrange(43550):
        o = foo(o)
    del o
    o = None
    print "I'm fine."
    for i in xrange(43551):
        o = foo(o)
    del o
    print "I'm dead."

I get "I'm fine" and then a segmentation fault.  This only happens when
__slots__ is defined on a new-style class.  Seriously, i'm not making those
numbers up -- reliably, 43551 objects will crash the interpreter; fewer will
not.

I don't know the code here, but I'm willing to bet Python is smashing the stack
because PyType_GenericAlloc() is recursive.  I get a stacktrace from gdb that
looks like this:

    #0  0x100a3fc4 in PyDict_New ()
    #1  0x100244e8 in _PyType_Lookup ()
    #2  0x100249d4 in _PyObject_SlotCompare ()
    #3  0x1001a3d4 in PyType_GenericAlloc ()
    #4  0x1001a4fc in PyType_GenericAlloc ()
    #5  0x1001a5b8 in PyType_GenericAlloc ()
    #6  0x1001a5b8 in PyType_GenericAlloc ()
    #7  0x1001a5b8 in PyType_GenericAlloc ()
    #8  0x1001a5b8 in PyType_GenericAlloc ()
    #9  0x1001a5b8 in PyType_GenericAlloc ()
    #10 0x1001a5b8 in PyType_GenericAlloc ()
    #11 0x1001a5b8 in PyType_GenericAlloc ()
    #12 0x1001a5b8 in PyType_GenericAlloc ()
    #13 0x1001a5b8 in PyType_GenericAlloc ()
    #14 0x1001a5b8 in PyType_GenericAlloc ()

... and so on, and so on, for longer than I had the patience to page through.
(I did get to about #400, though.)

Version Information:

    % python2.2
    Python 2.2.1 (#1, May  4 2002, 11:33:48) 
    [GCC 2.95.4 20011002 (Debian prerelease)] on linux2

And since those arguments to xrange() above look like some weird magic to me,
this may also be useful to know:

    % free
                 total       used       free     shared    buffers     cached
    Mem:        253968     146192     107776          0       8008      59184
    -/+ buffers/cache:      79000     174968
    Swap:       262136       3168     258968

    % uname -a
    Linux zelda 2.4.18 #1 SMP Fri Apr 26 19:18:50 CDT 2002 ppc unknown

Is there any in-development code I should try running?

-- 
 |    <`'>    |  Glyph Lefkowitz: Traveling Sorcerer   |
 |   < _/ >   |  Lead Developer,  the Twisted project  |
 |  < ___/ >  |      http://www.twistedmatrix.com      |
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 196 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020626/1fc4389b/attachment.sig>

From prema at prema.co.nz  Sun Jun 16 22:42:00 2002
From: prema at prema.co.nz (Prema)
Date: 16 Jun 2002 19:42:00 -0700
Subject: MS Word -- finding text
References: <mailman.1024247060.12127.python-list@python.org>
Message-ID: <ae58a20e.0206161841.415a100d@posting.google.com>

Hi Mike !
Excellent -- thanks very much for your time and sharing.
Would have taken a while to work through that -- I can't wait to try it out!
I'll post to this thread as to how I got on
Kind regards 
Thanks again
Mike




Mike Brenner <mikeb at mitre.org> wrote in message news:<mailman.1024247060.12127.python-list at python.org>...
> The COM objects (like Project, Word, Excel, etc.) sometimes return stuff in Unicode format. When they do, the python str() function dies when converting non-ASCII unicode characters. 
> 
> To avoid this problem, I use the following conversion routine. After making the necessary check for None, it attempts a quick conversion str() first. When necessary, it slowly goes through each character, handling the exceptions that are raised. 
> 
> The default is a prime because that is the most common character that hits me in Word and Excel documents. Instead of coding it as an ASCII single-quote characters, these applications code it as a more "beautiful" character, so it kills the python str() function.
> 
> You may or may not wish to change the return to eliminate the string.strip there, depending on your needs.
> 
> You could make a separate function that has just the TRY and the EXCEPT in it, in order to use the MAP function instead of the for loop.
> 
> Mike Brenner
> 
> 
> def phrase_unicode2string(message):
>     """
>     phrase_unicode2string works around the built-in function str(message)
>     which aborts when non-ASCII unicode characters are given to it.
>     """
>     if type(message)==types.NoneType:
>        return ""
>     try: st=str(message)
>     except: # untranslatable unicode character
>        list=[]
>        for uc in message:
>           try:
>              c=str(uc)
>           except:
>              c="`"
>           list.append(c)
>        # Note: because it raises exception instead of returning
>        # a default characters, we cannot use map() here.
>        st=string.join(list,"")
>     return string.strip(st)
> 
> ------------------------
> 
> Mike Prema wrote: 
> 
> #######
> from win32com.client import Dispatch
> W=Dispatch('Word.Application')
> D=W.Documents.Open('c:\\windows\\Desktop\\TOR.doc') ## Test Doc
> FindRange=D.Content
> F=FindRange.Find.Execute('Conman','True')
> print FindRange.Text
> #######
> str() doesn't seem to work in this case
> I tried using the codecs library but I think I am missing something


From neal at metaslash.com  Sat Jun  8 20:07:54 2002
From: neal at metaslash.com (Neal Norwitz)
Date: Sun, 09 Jun 2002 00:07:54 GMT
Subject: Simple pychecker question
References: <mailman.1023482712.8801.python-list@python.org>
Message-ID: <pan.2002.06.09.00.07.47.445563.6814@metaslash.com>

On Fri, 07 Jun 2002 16:44:50 -0400, Michael Gilfix wrote:

>   Pychecker doesn't seem to like named arguments (such as def
> f(f='blah'), or sometimes, it's handy in __init__).  Why is that? Are
> they considered bad style in some respects?  How about use in __init__
> functions?

Are you saying that there is a warning message?
Or that the code breaks pychecker?

In older versions, the default for this warning was turned on.
It is off in more recent versions.

I personally think that superfluous use of named args are bad,
but that the feature is good when used properly.  For example,
with your code above, calling f(f=2) would be better written
f(2).  However for Tkinter, named args can be useful when there
are a bunch of parameters and you only want to use the last.

Neal


From martin at v.loewis.de  Sun Jun  9 16:00:18 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 09 Jun 2002 22:00:18 +0200
Subject: (no subject)
References: <mailman.1023640273.18210.python-list@python.org>
Message-ID: <m37kl8b4tp.fsf@mira.informatik.hu-berlin.de>

Attila Horvath <mutsuura at mutsuura.com> writes:

> 	- provide [generic] CGI functionality
> 	and if possible...

Please have a look at the cgi module.

> 	- generate dynamic (DHTML) responses to CGI input parameters

There are several strategies, but I recommend to use plain print
statements. You can use HTML templates by putting multiline strings
into your scripts which offer %s or %(key)s interpolation.

Regards,
Martin



From jepler at unpythonic.net  Thu Jun 13 13:51:52 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Thu, 13 Jun 2002 12:51:52 -0500
Subject: Translating Tcl/Tk code into Python/Tkinter help please...
In-Reply-To: <B92E4241.6F6B%sarmstrong13@mac.com>
References: <B92E4241.6F6B%sarmstrong13@mac.com>
Message-ID: <20020613175152.GE25416@unpythonic.net>

If you're rewriting your tk into Python, don't use a sub-python to
create this data.

Otherwise,
    catch {exec echo hi} output
    puts "output was $output"
and
    import os
    output = os.popen("echo hi").read()
    print "output was %s" % output
seem to behave about the same.

The reason for the use of of 'catch { ... } output' instead of 'set output
[...]'  is because a nonzero exit status from an 'exec' is an error in Tk.
(Tk errors are similar to Python exceptions, with "if {[catch]} {...}"
being similar to 'try/bare except' and "error" being similar to "raise".

Jeff




From tjreedy at udel.edu  Mon Jun  3 09:54:45 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 03 Jun 2002 13:54:45 GMT
Subject: Graph distances, how can I speed it up?
References: <33803989.0206030235.4134d33e@posting.google.com>
Message-ID: <FyKK8.39522$4i.4685714@bin2.nnrp.aus1.giganews.com>

"Miki Tebeka" <tebeka at cs.bgu.ac.il> wrote in message
news:33803989.0206030235.4134d33e at posting.google.com...
> Hello All,
>
> I'm trying to calculate distances in a graph. (The ultimate goal is
> to calculate the distances between verbs in WordNet)
>
> The graph is built edge after edge and each time I recalculate the
> distances matrix. However this takes ages to complete. (will finish
in ~50H).

There is a reasonably efficient algorithm for computing the distance
between vertices in a graph.  I am sure that this is *much* more
efficient than doing so incrementally.  I am pretty sure that I would
be available in the graphical algorithms section of the C-coded Boost
package, which has a Python wrapper.  (Don't know details.)

>     Distances are held in hash table of pairs (x,y) -> distance.

A 2-d array would take about the same space and probably be quicker.
A 2-d Numerical Python array, which you would want for a large graph,
would be better both ways.

>     For each vertex a hash of all vertexes conncted to it is saved
>     to improve distance update.

As near as I can tell, you only access the lists of neighbors as
lists, so store them that way.

>         return self.forward.get(x, {}).keys()
>         return self.backwards.get(x, {}).keys()

But biggest boost will come from using C-coded function.

Terry J. Reedy





From sadams123 at optushome.com.au  Wed Jun 19 16:05:36 2002
From: sadams123 at optushome.com.au (Steven Adams)
Date: Thu, 20 Jun 2002 06:05:36 +1000
Subject: How to Store Binary data in a *.py module
References: <3D10E0E9.906DF517@noaa.gov>
Message-ID: <3d10e411$0$31829$afc38c87@news.optusnet.com.au>

> I know a Python string can contain an arbitrary set of bytes, but I
> don't know how to express binary data as a literal. If I were to try to
> use r" ..." and I happened to have a byte that matched the ", it would
> be mis-parsed. I'm imaging that UUencoding or something might solve my
> problem, but I don't know anything about that.
>
> Anyone have an idea??

I've stored GIFs using base64 encoding via the base64 module and it seemed
to work fine.

in the interpreter

read in the file in binary mode
base64 encode it
copy and paste the encoded string to your module
decode it when you want use it as binary data again

Steven




From max at alcyone.com  Wed Jun 12 22:33:38 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 12 Jun 2002 19:33:38 -0700
Subject: why not "'in' in 'in'"?
References: <ae8dg70217g@drn.newsguy.com>
Message-ID: <3D080482.7A7FD514@alcyone.com>

Grant Griffin wrote:

> So why is the "in" operator limited to single characters when used as
> a
> membership operator?  This peculiar lack of generality seems a bit
> (dare I say
> it?) "non-Pythonic".

On the contrary, it is exquisitely Pythonic.  You used the magic word
yourself:  _membership_.  A string is a sequence of characters, so when
testing membership with the `in' operator, the left-hand side of the
expression should be a character (really a string of length one).  `in'
tests membership, not for the existence of subsequences.  Would you have
expected

	[1, 2] in [1, 2, 3, 4]

to evaluate to true?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Who'd ever think it / Such a squalid little ending
\__/ The American and Florence, _Chess_
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From peter at engcorp.com  Mon Jun 17 21:32:33 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 21:32:33 -0400
Subject: pychecker + XEmacs
References: <3d0de483$0$3881$e4fe514c@dreader4.news.xs4all.nl>
Message-ID: <3D0E8DB1.E67264A6@engcorp.com>

Boudewijn Rempt wrote:
> 
> Has anyone succeeded in combining pychecker and XEmacs (or emacs)? I'm
> checking a large project I began in 1999, and which is full of the most
> atrocious errors, and I'd like to click on the line that mention the error
> and be transported there, just like when you run a python program from
> XEmacs and get a stacktrace.

Phlip's PyUnitBrowser (hope I got that right) seemed to be
sporting or about to sport an interface with PyChecker.... might
be a start.

-Peter


From emile at fenx.com  Thu Jun  6 15:25:01 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 06 Jun 2002 19:25:01 GMT
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de>
Message-ID: <hGOL8.119008$352.5927@sccrnsc02>

Shagshag13
>  I'm looking for any state of art, optimizing techniques, tips, that a
> python beginner couldn't know by itself, but should learn and
> would be aware to write "efficient and good looking"python's code...
>
>  (not the kind of thing which made you an obfuscated python nerd,
> but that "map" should be preferred to "for in loop" if possible, that
> dictionary are very well optimized, etc.)
>

I think the advise you'll generally get is to write first, then profile,
then optimize where needed.  Those optimizations may well involve using
map, or numpy, or c-extensions, etc...  But emphasis on optimization
first is probably time mis-spent.

--

Emile van Sebille
emile at fenx.com

---------



From pyth at devel.trillke.net  Thu Jun 13 18:38:53 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Fri, 14 Jun 2002 00:38:53 +0200
Subject: returning traceback strings
In-Reply-To: <20020613181109.A1019@gurthang.twcny.rr.com>; from mboedick@mboedick.org on Thu, Jun 13, 2002 at 06:11:09PM -0400
References: <20020613181109.A1019@gurthang.twcny.rr.com>
Message-ID: <20020614003853.M6609@prim.han.de>

Matthew Boedicker wrote:
> Hello,
> 
> I am looking for something exactly like traceback.print_exc(), but instead
> of writing the traceback text to a filehandle, I want it to return it to
> me as a string.  I was unable to find any way to do this using the traceback
> module, since everything there seems to be in terms of file handles being
> passed around.
> 
> The problem is I need to pass the entire multi-line traceback string to a
> file-like object's write() method in one call, not line by line.
> 
> This is the best workaround I was able to come up with:
> 
> import cStringIO
> import traceback
> 
> def print_exc_str():
>     buf = cStringIO.StringIO()
>     traceback.print_exc(file=buf)
>     return buf.getvalue()

this is a straightforward solution. if you don't want
cStringIO you could do:

import traceback

class fakefile(list):
    write=list.append

def get_exception():
    ff=fakefile()
    traceback.print_exc(file=ff)
    return "\n".join(ff)
    
> Any suggestions for a better way?  What about the traceback module using
> strings internally, and wrapping another layer around that to write to files
> and file-like objects?

The traceback module seems to route every print through '_print' (look
into traceback.py) but you can't just replace 'file' with a string,
because strings are immutable. You have to use a mutable type
like a list <wink>. 

    holger




From whisper at oz.net  Thu Jun 27 19:55:03 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 27 Jun 2002 16:55:03 -0700
Subject: Pythoniac: Thoughts on a hardware Python processor
In-Reply-To: <7xeletjxtm.fsf@ruckus.brouhaha.com>
Message-ID: <GCEDKONBLEFPPADDJCOEMEJODMAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Paul Rubin
> Sent: Wednesday, June 26, 2002 18:49
> To: python-list at python.org
> Subject: Re: Pythoniac: Thoughts on a hardware Python processor
>
>
> I don't think it makes sense to stick with existing Python bytecodes
> for hardware execution.  Take a look at the work done on Lisp machines
> in the 70's and 80's instead.

Ah yes, thanks for reminding me. The Self project claimed to have gotten
Self to run in the mid 90's percentile of "C speed", largely by making as
many primatives as possible actual machine ops of the host processor - not
too good for portability. OTOH, Self has many of the complexities of Python,
and if this is true, this is blazing speed. For anyone that's interested,
there is still a Self website (and by darned, a new release this year!) at
http://research.sun.com/self/index.html and there are some good papers there
on their compiler techniques.
http://research.sun.com/self/release/optionalFiles.html offers the sources
to the Self VM (in C++ and "Sparc-asm")

Of course, some of the Self guys went on to another project at Sun: Java.
I've read somewhere that Java HotSpot JIT compiler tech came out of the Self
work (oops, there it is on one of the Self pages).

Actually, Java and this idea of hosting Python on Forth brings up an
interesting question: If Sun can't make a VM that will run everywhere, why
would a Forth based Python implementation do any better?

Dave LeBlanc
Seattle, WA USA





From jdhunter at nitace.bsd.uchicago.edu  Sat Jun 29 16:14:31 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Sat, 29 Jun 2002 15:14:31 -0500
Subject: can't import generators
Message-ID: <m2hejlc060.fsf@mother.paradise.lost>

I am having trouble working with generators.  Anyone know what might
cause this behavior:

> /usr/local/bin/python2.2
Python 2.2.1 (#1, Apr 22 2002, 21:20:54) 
[GCC 3.0.4] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> from __future__ import generators
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
ImportError: cannot import name generators
>>> 


Thanks,
John Hunter


From fperez528 at yahoo.com  Thu Jun 20 15:43:48 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 20 Jun 2002 13:43:48 -0600
Subject: logical statements (in Python and Numeric)
References: <3D11F1B2.21587F31@lanl.gov> <aet2j4$h8c$1@peabody.colorado.edu> <3D122C5D.C6985686@noaa.gov>
Message-ID: <aetb9n$ksn$1@peabody.colorado.edu>

Chris Barker wrote:

> One solution is to use the bitwise operators instead:
> y = y + (x >= -2. & x < 0.) * (x + 6.)
> 

Are you sure? This is what I get:

In [11]: x = arrayrange(-4.,2.+.001,.001, Float)

In [12]: x >= -2. & x < 0.
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)

?

TypeError: unsupported operand type(s) for &: 'float' and 'float'


That's why I suggested logical_and() as a workaround. But maybe I'm missing 
something.

Cheers,

f.


From mfranklin1 at gatwick.westerngeco.slb.com  Wed Jun 26 15:51:46 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Wed, 26 Jun 2002 19:51:46 +0000
Subject: printing (again!!)
In-Reply-To: <afcusp$66r$1@newsg2.svr.pol.co.uk>
References: <afcusp$66r$1@newsg2.svr.pol.co.uk>
Message-ID: <200206261855.g5QItRM21915@helios.gatwick.geco-prakla.slb.com>

On Wednesday 26 Jun 2002 5:57 pm, Geoff Tarrant wrote:
> I want to use Python as the language in a Computing course that I teach.
> However, I can't find a way that is relatively straightforward to send data
> to a printer in a Windows environment.  If I can't find a way around this,
> I will have to revert to Delphi or VB.  Please help.  I like the idea of
> sending 20 or so Python programmers to University each year.
>
> Geoff Tarrant
Geoff, 


Get hold of the excelent win32 extension for python (url????)

Then somthing like this: (untested since I wrote it about a year or two ago!)


import win32print
printer_list=[]
for p in win32print.EnumPrinters(win32print.PRINTER_ENUM_CONNECTIONS):
    printer_list.append(p[2])
for p in win32print.EnumPrinters(win32print.PRINTER_ENUM_LOCAL):
    printer_list.append(p[2])
default_printer=win32print.GetDefaultPrinter()


printer = win32print.OpenPrinter(default_printer)

jid = win32print.StartDocPrinter(printer, 1, ('Page %i' %copy, None, 'RAW'))
bytes=win32print.WritePrinter(printer, open(file, 'rb').read())
win32print.EndDocPrinter(printer)
win32print.ClosePrinter(printer)


Please not this is sending a file's contents to the printer (as it happens a 
postscript file)  


Cheers
Martin




From steve.menard at vodeotron.ca  Sat Jun 29 20:21:54 2002
From: steve.menard at vodeotron.ca (Steve Menard)
Date: Sat, 29 Jun 2002 20:21:54 -0400
Subject: Is this a __slot__ bug?
Message-ID: <pan.2002.06.30.00.21.54.133520.5016@vodeotron.ca>

OK, this is probably me not understanding something, but then again, you
never know.

First things first, I love __slots__. For the paranoid programmer I am,
knowing I will never again assing to a variable with a typo and then
wonder ehat the hell is going wring is a godsend.

However, I also love(need) private variables (those __X). It seems they
dont work with slots. try the following simple example under ptyon 2.2 on
linux :

class O(object) :
	__slots__ = [
		'__a'
		]

	def __init__(self) :
		object.__init__(self);
		self.__a = 3
o = O()

I get thoe following error :
AttributeError: 'O' object has no attribute '_O__a'

Now, I know I can easily work around this by manually obfuscating the
private variables names in the __slots__, but I would rather not. Feels
wrong somehow to rely on what should be a hidden way of generating
private variables.

So if this is by design, I will not agree but I will do the workaround.
If not, well then seems I founda bug :)

			Steve


From chouster at uclink4.berkeley.edu  Wed Jun 19 15:51:16 2002
From: chouster at uclink4.berkeley.edu (Chris)
Date: 19 Jun 2002 12:51:16 -0700
Subject: Writing to Excel
Message-ID: <79b54d28.0206191151.52ee78a3@posting.google.com>

Hi I'm using a Python cgi script to write the contents of a table to
an Excel sheet. I retrieve all the information fine but it seems like
Excel doesnt want to print more than one line for me. It prints
'headerLine' but nothing more. Why???

... 
headerLine = "some crap"

sys.stdout.write( Html.header( 'application/x-msexcel' ) )
print string.strip(headerLine)
allText = ""

...
# set allText = "some other crap"
...

print allText

Im dismayed that this only will print the first print line I specified
to the Excel page. The Html module is a proven-to-work module I've
obtained so I'm 99.99% sure that's not the problem. I'm using the
latest version of python, winNT, and using IE Explorer 5.5.

Chris

Chris


From skip at pobox.com  Tue Jun 25 09:11:08 2002
From: skip at pobox.com (Skip Montanaro)
Date: Tue, 25 Jun 2002 08:11:08 -0500
Subject: Thread Memory Leak
In-Reply-To: <pan.2002.06.25.12.37.12.535288.17223@pobox.com>
References: <pan.2002.06.25.12.37.12.535288.17223@pobox.com>
Message-ID: <15640.27628.444843.841517@12-248-8-148.client.attbi.com>

    Marco> while 1:
    Marco>     listThread = []

    Marco>     for i in range(50):
    Marco>         listThread.append( threading.Thread( None, runthread ) )
    Marco>     for i in listThread:
    Marco>         i.start()
    Marco>     for i in listThread:
    Marco>         i.join()

    Marco>     pid = os.getpid()
    Marco>     print "==================================="
    Marco>     os.system( "ps -o vsize,rss %d" % pid )

You never remove any thread objects from your listThread list.  It fills up
with completed threads.

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html




From ethanw at rediffmail.com  Fri Jun 14 15:58:22 2002
From: ethanw at rediffmail.com (Ethan)
Date: 14 Jun 2002 12:58:22 -0700
Subject: PythonWin Problem!
References: <1a52c7f1.0206131455.6df50cf8@posting.google.com> <3D0973C5.60407@skippinet.com.au>
Message-ID: <1a52c7f1.0206141158.1c03b3f5@posting.google.com>

Thanks Mark. It worked like a charm. One thing though, the key I had
to delete was in HKCU rather than in HKLM. The second method would not
work with ME since like I said in my earlier post, I don't have a
right click menu for items in the taskbar.

Thanks again.



> Pythonwin has probably somehow remembered it's screen location as off 
> the screen.
> 
> Try deleting the "HKLM\Software\Python 2.1\Python for Win32" key from 
> the registry, and it should come back.
> 
> Alternatively, from the task-bar select "Move", then use the arrow keys 
> to try and bring it back on screen (presumably with the "Left" and "Up" 
> keys)
> 
> Mark.


From amuys at shortech.com.au  Tue Jun  4 21:25:51 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 4 Jun 2002 18:25:51 -0700
Subject: prototyping good OOdesign in Python?
References: <mailman.1022957762.16687.python-list@python.org>
Message-ID: <7934d084.0206041725.220fa6d3@posting.google.com>

Roman Suzi <rnd at onego.ru> wrote in message news:<mailman.1022957762.16687.python-list at python.org>...
> Maybe this was already discussed, but I'd liked to raise this topic again.
> 
> Python has superb OOProgramming support. Lets not dispute this for a
> moment. Now let's consider that we need to make well-designed OOP solution
> first prototyped in Python.
> 
> What troubles are there? Python is too dynamic and all that. And all
> those nifty features need to be translated properly into, say, C++.
> 

One thing I'm very interested in is knowing if anyone has any
experience with Python/Jython -> Java?  Many managers are now firm
believers in the gospel of Scott, but selling them Python as
prototyping support for their new favourite seems feasible.

Andrae Muys


From spahievi at vega.bg  Thu Jun  6 15:43:32 2002
From: spahievi at vega.bg (Niki Spahiev)
Date: Thu, 6 Jun 2002 22:43:32 +0300
Subject: [ANN] constraint-0.1
In-Reply-To: <20020606135835.GF26159@orion.logilab.fr>
References: <20020606135835.GF26159@orion.logilab.fr>
Message-ID: <1272291294.20020606224332@vega.bg>

6/6/2002, 16:58:35, Alexandre wrote:

A> Hello,

A> Logilab has released constraint-0.1.

Is it usable as Geometry Solver?

-- 
Best regards,
 Niki Spahiev


______________________________________
VEGA(TM) Internet Service Provider
      <http://www.vega.bg>
Scanned and protected by Inflex
Tova pismo e provereno i zashtiteno
ot programite "Inflex/Antivir".






From marcus1 at marcus1.de  Fri Jun 21 12:21:59 2002
From: marcus1 at marcus1.de (Marcus Klein)
Date: Fri, 21 Jun 2002 18:21:59 +0200
Subject: Newbie-Question: Oberserving an directory
Message-ID: <3d1352a7@news.piro.net>

Hi there,

I want to observe an directory to check if a new file has been put in, if 
that happens I want to start an event.

Does anyone have a good idea how to start, I've tried it with signal, but 
either I am too stupid or it is the complete wrong way :-/

Thx in advance, marcus

-- 
sick nature ?


From emile at fenx.com  Thu Jun  6 09:12:01 2002
From: emile at fenx.com (Emile van Sebille)
Date: Thu, 06 Jun 2002 13:12:01 GMT
Subject: HttpServer help!
References: <3CFF4C14.450850D8@millfilm.co.uk>
Message-ID: <BcJL8.4859$pw3.343@sccrnsc03>

Eric Texier
[snip]
> req.putrequest("POST",
>
"file:/usr/people/erict/python/srcToolskit/pythonXmlBook/c8/favquote.cgi
")
>
> .
> .
> .
>
> ##### That the Error after the ruprequest->"#####
> Traceback (most recent call last):
>   File "post.py", line 25, in ?
>     req.putrequest("POST",
>
"file:/usr/people/erict/python/srcToolskit/pythonXmlBook/c8/favquote.cgi
")
>
>   File "/usr/people/erict/bin//lib/python2.2/httplib.py", line 453, in
> putrequest
>     self.send(str)
>   File "/usr/people/erict/bin//lib/python2.2/httplib.py", line 395, in
> send
>     self.connect()
>   File "/usr/people/erict/bin//lib/python2.2/httplib.py", line 379, in
> connect
>     raise socket.error, msg
> socket.error: (111, 'Connection refused')
>


So, you've got 'Connection Refused' on req.putrequest() with a file...
Permissions isssues on the server or file?


--

Emile van Sebille
emile at fenx.com

---------



From rx.andrew at ieee.org  Sat Jun  1 23:47:23 2002
From: rx.andrew at ieee.org (Rex K. Andrew)
Date: Sun, 02 Jun 2002 03:47:23 GMT
Subject: Newbie question: is there a python wrapper for a distributed task?
Message-ID: <3cf99371.1554503@news.mindspring.com>

Hi folks,
 So I'm totally new here. I'm considering several scientific/data
crunching projects that need a lot of CPU hours. We have at hand
several linux clusters, but we want to expand (and take over the 
world.) 

I was looking at FIDA from the Chemistry Dept at UW and the Cosm
library from Mithral. These are ways to "glue" a server process to
the client processes on the nodes. (i.e., all the distributed
computers.) I got to wondering: has anybody used python as
the "glue"? python works on lots of computers now, and an
app can be compiled into it if necessary using SWIG. And the
python community has lots of netware contributions.

I dunno, just a troll, I guess. It seems like it might be a good 
marriage.

Any pointers, URLs, flames, etc, appreciated. Thanks,
Rex
************************************************
Rex K. Andrew
rx dot andrew at ieee dot org (you know what to do)


From occeanlinux at linuxmail.org  Thu Jun 27 20:14:54 2002
From: occeanlinux at linuxmail.org (SiverFish)
Date: Fri, 28 Jun 2002 00:14:54 +0000
Subject: How to get rid the new line
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org> <3D1AD534.A74EC9F4@ipm.fhg.de> <3D1B1A76.563E5394@engcorp.com>
Message-ID: <pan.2002.06.28.00.14.48.789371.16627@linuxmail.org>

On Thu, 27 Jun 2002 14:00:22 +0000, Peter Hansen wrote:

> Markus von Ehr wrote:
>> 
>>         f = open(filename, 'r')
>>         lines = f.readlines()
>>         line1 = lines[0] # exemplarily for first line line1 =
>>         line1[0:len(line1)-1]
>> 
>> or:
>>         f = open(filename, 'r')
>>         lines = f.readlines()
>>         line1 = lines[0][0:len(line1)-1] # exemplarily for first line
> 
> I believe this is unsafe.  The final line may not be terminated with \n.
> 
> -Peter
Any method like chomp in perl
I just one to cut read only the line that not empty into the list


From peter at engcorp.com  Thu Jun 27 19:52:21 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 27 Jun 2002 19:52:21 -0400
Subject: How to get rid the new line
References: <pan.2002.06.27.18.34.26.356672.9945@linuxmail.org> <3D1AD534.A74EC9F4@ipm.fhg.de> <3D1B1A76.563E5394@engcorp.com> <pan.2002.06.28.00.14.48.789371.16627@linuxmail.org> <afg513$2426$1@news.idiom.com>
Message-ID: <3D1BA535.1EE16646@engcorp.com>

"James T. Dennis" wrote:
> 
>  Could try this:
> 
>         def chomp(line):
>                 if line[-1]=='\n':
>                         line=line[:-1]
>                 return line

Just make sure you pass chomp() only lines that are not equal
to the empty string ''.

>>> line = ''
>>> line[-1]
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
IndexError: string index out of range


-Peter


From xsebbi at gmx.de  Sun Jun 23 05:18:29 2002
From: xsebbi at gmx.de (Sebastian Roth)
Date: Sun, 23 Jun 2002 11:18:29 +0200
Subject: Parsing
In-Reply-To: <3D14CEBC.C6655F86@vip.fi>
References: <3D14CEBC.C6655F86@vip.fi>
Message-ID: <200206231115.37654@xsebbi.de>

On Samstag, 22. Juni 2002 21:23, Pekka Niiranen wrote:
> Has anybody used python's parser -module to parse user defined
> text structure? It is possible ?

do you mean regular expressions?
If yes, 're' is your module ;)

> -pekka-

		Sebastian






From kragen at pobox.com  Mon Jun 10 16:53:40 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 16:53:40 -0400
Subject: Where is Berkeley DB on your system? Can you live without DB 1.85?
References: <mailman.1023698119.5905.python-list@python.org>
Message-ID: <83wut6q2i3.fsf@panacea.canonical.org>

"David LeBlanc" <whisper at oz.net> writes:
> > I don't think anybody has proposed including a modern version of BSD
> > DB in Python, only dropping support for an old version, which is also
> > not included in Python.
> 
> Hmmm... I have Python 2.2.1 binary installed from Sourceforge - what is that
> bsddb.pyd file dated 4/9/2002 in the dlls directory?

That itself is probably the Python module that interfaces to whatever
version of Berkeley DB you have, not Berkeley DB itself (although it
could be statically linked in there), but Tim Peters has asserted that
a binary version of db 1.85 is included, regardless.

So I was wrong.  Certainly Berkeley DB is not included in the Python
source tarball.

> > "commercial for-sale" is misleading.  The current license of BSD DB
> > allows distribution in commercial for-sale apps with no further hassle
> > unless said commercial for-sale apps are, approximately, free
> > software, and it forbids inclusion in noncommercial not-for-sale apps
> > unless said not-for-sale apps are, approximately, free software.
> 
> This makes no sense, so I'll just pass on it. I think you got your facts
> backwards.

It makes no sense because I got the sense of my first clause
backwards; thanks for the correction.  It should read as follows:

    It allows distribution in commercial for-sale apps with no further
    hassle unless they are *not* free software.

> > > prospects. It didn't work for Perl in it's early years and
> > that's why the
> > > Artistic License was change to allow "for fee" distributions
> > that included
> > > Perl.
> >
> > I thought the AL allowed this from the beginning.  Certainly
> > Sleepycat's license doesn't forbid it.
> 
> IIRC, in the early days of Pearl, the "AL" had the same tainting as GPL.
> Again, IIRC, some of the Pearl inner circle where the ones who agitated for
> the change so they could make some money from all the time they where
> pouring into Pearl. Today, the AL is more like the GLL.

My experience with Perl doesn't go back to the days before its public
release when it was spelled with an "a", so I'll take your word for it.

> Sleepycat's license doesn't prohibit commercial for-fee distribution, it
> simply requires that you negotiate a payment scheme with them for the use of
> versions after 1.85. They don't charge for redistribution with free open
> source apps like Python.

This is self-contradictory.  The second statement is accurate; the
first statement is not.  They don't require that you negotiate a
payment scheme with them *unless your commercial for-fee distribution
is a distribution of proprietary software*.  For example, Red Hat
doesn't have to negotiate a license to Berkeley DB to include it in their
commercial for-free Linux distribution.

> However, as I said in my last post (along with exerpts from
> Sleepycat's FAQ on the subject), if a Python developer was to
> include a Python distribution that included a version of bsddb after
> 1.85 in a for-fee commercial program, then they would have to
> negotiate a payment scheme with Sleepycat.

Again, *only* if it was a *proprietary* for-fee commercial program.
The same issue would come up if they were distributing a proprietary
not-for-fee noncommercial program.

> More to the point, they would have to know that they would have to
> do that or risk copyright infringement.

Yes, they would have to understand the licensing terms of the software
they were redistributing.

> > In any case, your irresponsible, alarmist accusation is a red herring,
> > as you will no doubt agree once you understand the situation properly.
> > Dropping support for db 1.85 will not require all Python apps to be
> > free software or open source; it won't even require all distributions
> > of Python to be free software or open source.  It will only require
> > distributions of Python with bsddb support to either link to an
> > external BSD DB library (like the one included with glibc) or be free
> > software.
> 
> I don't think I'm the one not understanding the situation properly, nor
> could anything I posted be reasonably construed as irresponsible, alarmest
> or an accusation!

You said:
> If you want to change Python's market dynamics and require _all_ Python apps 
> to be free and open source, then you can include any version of Sleepycat 
> BSDDB you like.

That was what I meant by "your irresponsible, alarmist accusation". It
isn't true, and it sounds scary and accusatory to me.  Only Python
apps that use Berkeley DB would be affected, and then only if they're
distributed in binary form, and then only if they're running on
operating systems that don't ship with Berkeley DB.

> > If it's not clear, I'm in favor of this change.  It helps everyone
> > everywhere.
> 
> Whatever gets done, it would be nice to have some useful db shipped in the
> standard Python distro. I have heard a rumor that gadfly is such a
> candidate - and also that gadfly has bugs of it's own.

I think dumbdbm and gdbm are more comparable to Berkeley DB than Gadfly is;
either or both of them could be included in a standard Python distro
with no licensing problems, even for proprietary software.

> BTW, since you're so keen on the failings of the BSDDB that seems to ship
> with standard Python, perhaps you'd like to contribute information on how to
> use it and what it's pitfalls are and ways to avoid them to the pythondoc?

I don't know what they are because I use sane versions of Berkeley DB.
I seem to recall that hash-style db 1.85 files will fail after a lot
of inserts, and Sleepycat has patches on their page that fixes a
couple of core dumps and one case of lost updates in the BTree code,
but ISTR that there are other problems, too.

> Hell, you could even help maintain BSD DB 1.x!

I think it's more cost-effective to just ship 4.0 in binary
distributions of Python and require users to switch.  If someone finds
4.0 (or 2.x or 3.x) unacceptable because they want to incorporate
Berkeley DB functionality into proprietary software, I would be
willing to work on 1.x for them at my regular contract rate, but
increasing the profits of proprietary software companies who are too
cheap to buy a license from Sleepycat isn't really my idea of a fun
way to spend my spare time.  Given that, it would probably be more
cost-effective for such a company to just pay Sleepycat for a 4.0
license.



From cpbotha at i_triple_e.org  Tue Jun 11 05:43:21 2002
From: cpbotha at i_triple_e.org (Charl P. Botha)
Date: Tue, 11 Jun 2002 09:43:21 +0000 (UTC)
Subject: Sorting list (maybe stupid)
References: <slrnagbh7m.h03.jarvin24@lehtori.cc.tut.fi>
Message-ID: <ae4gnp$l18$1@news.tudelft.nl>

In article <slrnagbh7m.h03.jarvin24 at lehtori.cc.tut.fi>, J?rvinen Petri wrote:
> I have a problem and can't solve it:
> 
> There is lines of text like:
> 
> 2. Header2
> 1. Header1
> 102. Header102
> 
> and so on...
> 
> And I need to sort them according to the number (1,2,3,4...)

Put each line in a tuple, first element the number, second element the
header.  Make a list of these tuples, and then use list.sort() with a comp
function that compares using the first element in the tuple.

Here's some help:
>>> print [].sort.__doc__  
L.sort([cmpfunc]) -- sort *IN PLACE*; if given, cmpfunc(x, y) -> -1, 0, 1

-- 
charl p. botha http://cpbotha.net/ http://visualisation.tudelft.nl/


From peter at engcorp.com  Sat Jun 29 20:23:28 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 29 Jun 2002 20:23:28 -0400
Subject: Newbie that don't understand
References: <AIlT8.13914$p56.4478653@newsb.telia.net> <3d1e262f.3755054@news.clara.net>
Message-ID: <3D1E4F80.D8EFBB3E@engcorp.com>

Just a dude wrote:
> 
> >how can 23 and -23 be equal?
> >
> To some this can be a hard concept to grasp. 

I can't grasp it either, because it's not possible.  Only their absolute
values are equal, not the numbers themselves.

If the answers were unclear, it might be because we aren't sitting
face to face reading body language and seeing confusion on your face.
The Usenet medium can require a couple of abortive attempts before
an answer works, as we experiment and try to find the proper level
and manner of replying.

> My son is an electronics
> design engineer, and I still can't get my head around negative
> voltages, how the hell do you get a -12 V supply... to him it's basic
> but to me..............

With a 12V supply, the power supply is trying to push electrons
out the negative terminal and into the positive terminal.  If you
have a -12V supply, it tries to push electrons out the positive
terminal and into the negative terminal.

-Peter


From max at alcyone.com  Thu Jun 13 01:50:16 2002
From: max at alcyone.com (Erik Max Francis)
Date: Wed, 12 Jun 2002 22:50:16 -0700
Subject: why not "'in' in 'in'"?
References: <ae8dg70217g@drn.newsguy.com> <3D080482.7A7FD514@alcyone.com> <3D082A62.CF800C75@seebelow.org>
Message-ID: <3D083298.470850D0@alcyone.com>

Grant Griffin wrote:

> In that case, perhaps I should have thought of a better one <wink>. 
> In
> any event, in the case of strings, I think one might stretch the
> concept
> of "membership" to include substrings.  In fact, note that since
> Python
> has no actual character type (because a character is adequately
> represented as a string of length one) the "in" operator actually
> _does_
> test membership of strings in strings.  But for some reason, it's
> limited to testing strings of length one.  Go figure!

A string is a sequence of one-character strings.  It's not a sequence of
all the subsequences of a string.

> Honest answer: although I can't see much use for that construct, I
> can't
> see much harm in it either.  So although it's probably not worth
> adding
> to Python, I'm sure _somebody_ would find a use for it.

But that's exactly the problem; if you want [1, 2] in [1, 2, 3, 4] to
return true (i.e., for the `in' operator to test substrings, not just
membership), then you have an ambiguity problem.  A list can contain
sublists as elements, after all.

> But anyway, let's "just say no" to the tuple case rather than confuse
> ourselves; let's just make the string version of "in" a special case.

Why make a special case for it when string.find and S.find already exist
and are included expressly for that purpose?

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Who'd ever think it / Such a squalid little ending
\__/ The American and Florence, _Chess_
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From inkedmn at earthlink.net  Fri Jun  7 22:59:38 2002
From: inkedmn at earthlink.net (inkedmn at earthlink.net)
Date: Sat, 08 Jun 2002 02:59:38 GMT
Subject: syntax error
References: <mailman.1023443290.23645.python-list@python.org>
Message-ID: <3d0171be.530166419@news.earthlink.net>

On Fri, 7 Jun 2002 11:48:45 +0200, "Marc Beyerlin"
<beyerlin at magic-garden.de> wrote:

>hello pythons,
>im new to python and have a strange error:
>if have installed python 2.4 with treads on my openbsd3.0 box
>with also installed zope.
>zope works fine, but when i try to start an script like
>
>python ipcheck.py
>
>i get an error msg like syntax error at .....
>
>but i m sure there is no error.
>
>what could be the problem?
>
>greetings,
>marc
>
>
>
>-------------------------------------------------------
>magic garden GmbH - Agentur f=FCr Informationsarchitektur
>
>Hermannstr. 15 - 70178 Stuttgart (Am Feuersee)
>
>www.magic-garden.de =A6 beyerlin at magic-garden.de
>
>Tel. (07 11) 619 57-42 =A6 Fax (07 11) 615 01 38
>
>

What's the exact error that's returned?  Please also paste the culprit
code :)

ink


From kragen at pobox.com  Mon Jun 10 21:27:10 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 20:27:10 -0500
Subject: Dr. Dobb's Python-URL! - weekly Python news and links (Jun 10)
Message-ID: <EE3CDB256E0A231C.577F9FCCC77B96EC.B451C214671092A5@lp.airnews.net>

QOTW:  "I favour Kent Beck's aphorism: 'Make it work, make it right, make
it fast.'" -- James Kew


Programming Libraries:

    Logilab's 'constraint', a constraint satisfaction problem
    programming environment for Python, which lets you program in
    Prolog style in Python, has released version 0.2.
        http://groups.google.com/groups?selm=mailman.1023468735.788.python-list%40python.org

    Twisted, an event-based framework for Python applications,
    including network servers, has released version 0.18.0.
        http://groups.google.com/groups?selm=mailman.1023121326.9358.clpa-moderators%40python.org

    Rdflib, which includes a store for RDF triples and XML
    serialization facilities for them, has released version 0.9.0.
        http://redfoot.sf.net/

    The proposed standard Python logging module has been released
    version 0.4.5.
        http://groups.google.com/groups?selm=mailman.1023398221.4689.clpa-moderators%40python.org
        http://www.red-dove.com/python_logging.html

    scgi, yet another FastCGI clone, with support in Apache and Python,
    has released verion 0.3.
        http://groups.google.com/groups?selm=mailman.1023213423.29409.clpa-moderators%40python.org

    Skunkweb, a web application framework for Python, has released
    version 3.2.1, which supports Apache 2, among other things.
        http://groups.google.com/groups?selm=mailman.1023380951.25710.python-list%40python.org

Python Development News

    Oleg BroytMann posted glue that automatically moves vim users to
    the locations of syntax errors in their Python source code, much
    like C-c - in Emacs.
        mailman.1023368949.9130.python-list at python.org
    Google may not save the attachment, but pipermail does.
        http://mail.python.org/pipermail/python-list/2002-June/107625.html

    Skip Montanaro wants help making Python better at detecting where
    various pieces of BSD DB are installed, and also wants to know if
    it's OK to stop supporting obsolete versions of BSD DB.
        http://groups.google.com/groups?selm=mailman.1023479590.27892.python-list%40python.org

    IDE Studio, an enhanced version of IDLE, including features like a
    graphical class browser, pydoc integration, and method syntax
    tips, released version 1.6.
        http://starship.python.net/crew/mike/Tide/idledev/IDEStudio.html

Discussion on Features of Python:

    Andrew Kuchling has written "What's new in Python 2.3?"
        http://www.python.org/dev/doc/devel/whatsnew/

    eval and exec still need real dicts for variable lookup, which is
    a wart.  This is not a new topic.
        http://groups.google.com/groups?selm=mailman.1023324129.19618.python-list%40python.org
        http://groups.google.com/groups?selm=mailman.1023322808.24525.python-list%40python.org
        http://groups.google.com/groups?th=2338421ec8397bec&seekm=mailman.1004338951.27778.python-list%40python.org

    The timbot described the implementation details of the built-in
    list type, how it struck a balance between efficient reallocation
    behavior and efficient space usage, and why simple benchmarks may
    not always reveal the whole truth.
        http://groups.google.com/groups?selm=mailman.1023657975.22641.python-list%40python.org

Problems and Solutions:

    There's a new Python Cookbook recipe describing how to produce
    lazily-computed attributes in Python 2.2.
        http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/131495

    David Mertz wrote an article about how to use generators to get
    cooperative multithreading.  It turns out that "yield" really does
    mean "yield" after all.
        http://www-106.ibm.com/developerworks/library/l-pythrd.html

    Vojin Jovanovic is building a system for describing systems of
    interdependent lazily computed nonscalar values in Python, and
    he's annoyed that his variable names all have to begin with
    "self.".  This is a little different from the standard "Python is
    different from C++ and therefore bad" thread.  Holger Krekel,
    among others, suggested a solution, but it was not sufficient.
        http://groups.google.com/groups?selm=mailman.1023224648.434.python-list%40python.org

    Thomas Heller posted a sort of emulation of classmethod that works
    in Python 2.1.
        http://groups.google.com/groups?selm=ado5v8$uih9$1%40ID-59885.news.dfncis.de

    Uwe Schmitt wanted help talking to his barcode printer directly on
    Microsoft Windows.
        http://groups.google.com/groups?selm=Xns9226F1A931224cliechtigmxnet%4062.2.16.82

    Cimarron Taylor posted his procedure for building a Python
    interpreter with Tkinter and Pmw.
        http://groups.google.com/groups?selm=29e28c51.0206090021.72466d4b%40posting.google.com

Miscellaneous:

    ONLamp.com interviewed Guido van Rossum about Python.
        http://www.onlamp.com/pub/a/python/2002/06/04/guido.html

    EuroPython has posted some more interviews, including one with
    Steve Alexander and Stephan Richter, in which they talk about
    Zope's place in the world and Zope 3.
        http://europython.zope.nl/interviews/entries/steve_stephan_zope

    ActiveState continues its annual sponsorship of Active Awards
    for toilers in the vineyards of its supported languages.  Voting
    closes on the 26th of this month.
	http://www.activestate.com/Corporate/Communications/Releases/Press1023137709.html

    Domenic Merenda, who used to work at BeOpen, is running a
    home-grown ERP system written entirely in Python on an AS/400 to
    manage the operations of a hundred-million-dollar manufacturing
    company.  He hopes to release it as open source. He's looking for
    other people who have done this.
        http://groups.google.com/groups?selm=KHkL8.21383$eD2.2399923271%40newssvr10.news.prodigy.com

    The discussion about ERP systems generalized to large Python
    projects --- more than a few tens of thousands of lines.  Michael
    Chermside posted a summary of a few such systems.
        http://groups.google.com/groups?selm=mailman.1023460091.14067.python-list%40python.org

    Domenic Merenda posted a list of several dozen large Python
    projects that BeOpen had compiled in 2000.
        http://groups.google.com/groups?selm=QX8M8.21882$EP.2548532334%40newssvr10.news.prodigy.com

    Tim Churches answered a question about how to compute confidence
    intervals in Python with an overview of available statistical
    software for Python, including SciPy, RPy, Gary Strangman's stats
    module, and NumPy's MLab.
        http://groups.google.com/groups?selm=mailman.1023574153.24062.python-list%40python.org

    Chris Armstrong is developing a Python-source serialization module
    for arbitrary object instances that works along the same lines as
    pickle, but is human-readable and human-editable.
        http://groups.google.com/groups?selm=mailman.1023668414.25769.python-list%40python.org

    Dean Goodmanson described what features of Python he avoided when
    trying to write Python that would be easy to translate into C++.
        http://groups.google.com/groups?selm=e81be8b2.0206022315.3a0bf9a9%40posting.google.com

    Peter Hansen eloquently praised test-first development.
        http://groups.google.com/groups?selm=3D02FB3B.CF75E925%40engcorp.com

    Dave Cole wrote about why he wrote Albatross, a web application
    framework for Python, when there were already so many web
    application frameworks for Python.
        http://groups.google.com/groups?selm=mailman.1023539113.15573.python-list%40python.org


========================================================================

Everything you want is probably one or two clicks away in these pages:

    Python.org's Python Language Website is the traditional
    center of Pythonia
        http://www.python.org
    Notice especially the master FAQ
        http://www.python.org/doc/FAQ.html

    PythonWare complements the digest you're reading with the
    daily python url
         http://www.pythonware.com/daily  
    Mygale is a news-gathering webcrawler that specializes in (new)
    World-Wide Web articles related to Python.
	 http://www.awaretek.com/nowak/mygale.html 
    While cosmetically similar, Mygale and the Daily Python-URL
    are utterly different in their technologies and generally in
    their results.

    comp.lang.python.announce announces new Python software.  Be
    sure to scan this newly-revitalized newsgroup at least weekly.
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python.announce

    Michael Hudson continued Andrew Kuchling's marvelous tradition
    of summarizing action on the python-dev mailing list once every
    other week, into July 2001.  Any volunteers to re-start this
    valuable series?
        http://starship.python.net/crew/mwh/summaries/
        http://www.amk.ca/python/dev

    The Vaults of Parnassus ambitiously collect Python resources
        http://www.vex.net/~x/parnassus/   

    Much of Python's real work takes place on Special-Interest Group
    mailing lists
        http://www.python.org/sigs/

    The Python Software Foundation has replaced the Python Consortium
    as an independent nexus of activity
        http://www.python.org/psf/

    Cetus does much of the same
	http://www.cetus-links.org/oo_python.html

    Python FAQTS
        http://python.faqts.com/

    The old Python "To-Do List" now lives principally in a
    SourceForge reincarnation.
	http://sourceforge.net/tracker/?atid=355470&group_id=5470&func=browse
	http://python.sourceforge.net/peps/pep-0042.html
     
    The online Python Journal is posted at pythonjournal.cognizor.com.
    editor at pythonjournal.com and editor at pythonjournal.cognizor.com
    welcome submission of material that helps people's understanding
    of Python use, and offer Web presentation of your work.

    *Py: the Journal of the Python Language*
        http://www.pyzine.com

    Links2Go is a new semi-automated link collection; it's impressive
    what AI can generate
        http://www.links2go.com/search?search=python

    Tenth International Python Conference 
        http://www.python10.org            

    Archive probing tricks of the trade:
        http://groups.google.com/groups?oi=djq&as_ugroup=comp.lang.python&num=100
        http://groups.google.com/groups?meta=site%3Dgroups%26group%3Dcomp.lang.python.*

Previous - (U)se the (R)esource, (L)uke! - messages are listed here:
  http://www.ddj.com/topics/pythonurl/         
  http://purl.org/thecliff/python/url.html (dormant)
or
  http://groups.google.com/groups?oi=djq&as_q=+Python-URL!&as_ugroup=comp.lang.python


Suggestions/corrections for next week's posting are always welcome.
[http://www.egroups.com/list/python-url-leads/ is hibernating.  Just
e-mail us ideas directly.]

To receive a new issue of this posting in e-mail each Monday morning
(approximately), ask <claird at neosoft.com> to subscribe.  Mention
"Python-URL!".


-- The Python-URL! Team--

Dr. Dobb's Journal (http://www.ddj.com) is pleased to participate in and
sponsor the "Python-URL!" project.
-- 

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


From yidaki2 at excite.com  Tue Jun  4 19:47:55 2002
From: yidaki2 at excite.com (Brent Miller)
Date: Tue, 04 Jun 2002 23:47:55 GMT
Subject: Fastest way to read / procsess / write text?
Message-ID: <3CFD5229.2060803@excite.com>

I'm somewhat new to python and programing in general and I trying to
write a simple script that I can use to format large text files so that
they are suitable for importing into MySQL. So far, I have came up with
this:

#!/usr/bin/python2

import sys, re

infile = sys.stdin
data = infile.read()
infile.close()

data = re.sub('[.](?=\d\d\d\d\d\d)|[.](?=\w+[ ][<|>])|[.](?=\w+[:])|[
](?!0x)', '\t', data)

outfile = open("/mnt/storage/output.txt", "w")
outfile.write(data)
outfile.close()


This works beautifully most of the time (it's super fast), except when I
pipe large files (>50megs) to it and then it usually dies half way
through complaining of memory errors because it ran out of ram.

Short of buying more ram, is there a way I can make this more effecient
without taking a big peformance hit? I tried the above where I used a
"for line in data:" loop so it would only process one line at a time but
this seemed to take forever! I'm imagining there's a way to process the
data in something like 1 meg chunks, but I'm not too sure how to do that.

I'm using python 2.2.1 on a RH linux 7.2 system with 256 ram and a
Athlon XP1700 if it makes any diffrence.

Any suggestions?

TIA,
Brent



From phd at phd.pp.ru  Thu Jun 27 17:43:41 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Fri, 28 Jun 2002 01:43:41 +0400
Subject: Python crashed on me
In-Reply-To: <afftgs$5p7$0@216.39.172.122>; from bokr@oz.net on Thu, Jun 27, 2002 at 08:45:16PM +0000
References: <mailman.1025196737.29740.python-list@python.org> <afftgs$5p7$0@216.39.172.122>
Message-ID: <20020628014341.B22132@phd.pp.ru>

On Thu, Jun 27, 2002 at 08:45:16PM +0000, Bengt Richter wrote:
> To eliminate pickle, you could substitute a dump pickle.
> If it's happening in pickle, you could substite a file object of your own,
> to get a clue what it died on

   Aha, that's good point, thank you. I'll try your code.

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From rganesan at myrealbox.com  Tue Jun 18 07:19:50 2002
From: rganesan at myrealbox.com (Ganesan R)
Date: 18 Jun 2002 16:49:50 +0530
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <KiEP8.44424$GY.13974606@twister.nyroc.rr.com>
Message-ID: <ou8z5cg7ft.fsf@andlx-anamika.cisco.com>

>>>>> "George" == George Hester <hesterloli at hotmail.com> writes:

> "If you can't get an answer, please don't take it personally that we don't
> feel we can help you. Sometimes the members of the asked group may simply
> not know the answer."

Yeah, right. Nobody in the group knows the answer to your question. We would
really appreciate if you do some research and figure it out for yourself. We
would be even more happy when you come back and enlighten us with your
knowledge. Thank you.

Ganesan

-- 
Ganesan R 


From gerhard at bigfoot.de  Wed Jun 12 12:23:07 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 12 Jun 2002 16:23:07 GMT
Subject: Extending distutils?
References: <mailman.1023894093.18616.python-list@python.org>
Message-ID: <slrnagetc2.19o.gerhard@lilith.my-fqdn.de>

Martin Sjoegren wrote:
> There is so far no chapter in the distutils manual about extending distutils.
> [...]

I can't answer your question, but if you want to look into a setup.py that
overrides a lot of the defaults, you can look into the one from the
mxExtensions.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id 86AB43C0
public key fingerprint: DEC1 1D02 5743 1159 CD20  A4B6 7B22 6575 86AB 43C0
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From fperez528 at yahoo.com  Thu Jun 13 15:37:25 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 13 Jun 2002 13:37:25 -0600
Subject: filter2
References: <mailman.1023975618.6474.python-list@python.org>
Message-ID: <aeas9k$e8l$1@peabody.colorado.edu>

Oleg Broytmann wrote:

>    I want to have new python builtin - filter2. It is exactly like filter,
> but it returns also a list of items that didn't pass the test.
> 

Why not instead propose that a new keyword parameter be added to filter? 
Something like filter(fn,seq,rejected=1) which would then behave as you want 
it? By default, rejected would be false and the current behavior would be 
maintained. This sounds like somehting you could put in a pep quickly.

f.


From res04ft9 at gte.net  Fri Jun 28 19:33:31 2002
From: res04ft9 at gte.net (Erv Young)
Date: Fri, 28 Jun 2002 19:33:31 -0400
Subject: Why the colon?
In-Reply-To: <XFMail.20020628153929.pedrobados@eresmas.net>
Message-ID: <4.3.2.7.2.20020628190712.02450328@mail.gte.net>

I've come a long way, baby!  Four days ago, I posted an inquiry on a MS 
Developer mailing list, to the following effect:  "How do I get MS Visual 
C++ 6.0 to act like good old non-visual, nonplussed (forgive me) plain old 
C?" and got back a response to the effect that I should forget C and use 
Perl.  And now here I am, having read the entire thread "Python vs. Perl: 
which one to learn?" and having identified the pieces of Python that I need 
for the job at hand.  I will omit several intermediate steps on the journey.

But I have a question.  Let me set it in the context of the first code 
snippet in section 4.7.1 
(http://www.python.org/doc/current/tut/node6.html#SECTION006710000000000000000) 
of the on-line tutorial

1. def ask_ok(prompt, retries=4, complaint='Yes or no, please!'):
2.     while 1:
3.         ok = raw_input(prompt)
4.         if ok in ('y', 'ye', 'yes'): return 1
5.         if ok in ('n', 'no', 'nop', 'nope'): return 0
6.         retries = retries - 1
7.         if retries < 0: raise IOError, 'refusenik user'
8.         print complaint

where I have added line numbers for the sake of the discussion.  I 
understand the colons in the middle of lines 4, 5, and 7 as statement 
separators.  Makes perfect sense.

Then what am I to make of the colons at the end of lines 1 and 
2?  "Yoo-hoo, interpreter!  I'm about to start indenting now!"  Hmmm....

Maybe one of you can help me understand this syntactic convention a little 
better.  An invitation to continue RTFM would be quite in order, if it were 
accompanied by a specific URL or section number.  (An invitation to read 
the whole thing like a novel would be a bit un-Python-like, as near as I 
can tell.)

Thanks.

--erv





From op73418 at mail.telepac.pt  Wed Jun 26 13:45:25 2002
From: op73418 at mail.telepac.pt (Gonçalo Rodrigues)
Date: Wed, 26 Jun 2002 18:45:25 +0100
Subject: question od default args
Message-ID: <o3vjhuguen6gfm8vnq7rbk5volmi80shl5@4ax.com>

Hi,

When I want default args I usually do

def (arg = None):
    ...

But what if I want to make None a reasonable argument too? That is, I
want to know if the user has in fact passed an argument, and if not to
do something special, with None (or whatever object) being a reasonable
arg to be passed.

All the best,
Gon?alo Rodrigues


From rapskat at hotmail.com  Fri Jun  7 00:16:38 2002
From: rapskat at hotmail.com (rapskat)
Date: Fri, 07 Jun 2002 00:16:38 -0400
Subject: Help with Black Adder...
Message-ID: <adpc36$179jq$1@ID-139866.news.dfncis.de>

Greetings All.  I'm just getting started with python and I chose Black
Adder as a RAD IDE.  Since my background is (ugh!) VB, I'm used to the
whole Top Down IDE thing.

My questions are:

1) I am having problems with BA not finding the PyQT libs.  I installed
both from the rpm's (I'm using RH 7.3).  But when I start BA it gives me
an error that it couldn't find the PyQT modules and that they won't be
available for the session.  I've tried removing them and reinstalling to
no avail.  I've run ldconfig as well.  Has anyone ele seen anything like
this and any ideas on how to resolve it?

2) Are there any good printed books out there for development swith BA and
PyQT?  I have Deitel's "Python - How To Program" which is really excellent
and comprehensive, however I was looking for something more specific to
development within BA itself.

I have briefly parsed the
documentation, but it seems to be more of a generic compilation of the
different elemnts than product specific instructions.

Any assistance that anyone could give I would really appreciate it.  I
already have like 3 projects that I want to tackle with this puppy, so I'm
really eager to get started.

Thanks in advance! :-)

-- 
rapskat  - 12:05am  up 3 days,  3:04,  5 users,  load average: 1.19, 1.09, 1.11
drop the hot to mail me

lovin' livin' the linux life


From tim.one at comcast.net  Wed Jun 12 18:46:01 2002
From: tim.one at comcast.net (Tim Peters)
Date: Wed, 12 Jun 2002 18:46:01 -0400
Subject: map -> class instance
In-Reply-To: <m2r8jcp3pv.fsf@video.paradise.lost>
Message-ID: <BIEJKCLHCIOIHAGOKOLHKEDPDEAA.tim.one@comcast.net>

[John Hunter]
> I want to convert dictionary instances to instances of a class that
> has the keys of the map as attributes. ...

Hmm.  Nobody gave the obvious <wink> way yet:

class Map2Class:
    def __init__(self, amap):
        self.__dict__.update(amap)

That's it.  Then

>>> m = {'first': 'John',
...      'last': 'Hunter',
...      'age': 34}
>>> c = Map2Class(m)
>>> print c.first, c.last, c.age
John Hunter 34
>>>

at-34-life-can-be-unclear-ly y'rs  - tim




From donn at u.washington.edu  Thu Jun 20 13:27:46 2002
From: donn at u.washington.edu (Donn Cave)
Date: 20 Jun 2002 17:27:46 GMT
Subject: AF_UNIX + SOCK_DGRAM
References: <aes1he$gv2$1@news.idiom.com>
Message-ID: <aet3ai$1l0o$2@nntp6.u.washington.edu>

Quoth "James T. Dennis" <jadestar at idiom.com>:

|  I was playing with the socket module and Unix domain sockets,
|  trying to make it work with SOCK_DGRAM.  I could get messages
|  recvfrom() but any attempt to send responses was failing, 
|  complaining about the return address.
|
|  Now I think I understand the problem.  The recvfrom() tuple
|  does contain two parts (which WOULD BE address and data in UDP)
|  but I'm guessing that the address is meaningless for Unix 
|  domain sockets.  It occurs to me that the "return" address 
|  must be the same socket (filesystem node of type 's') as the 
|  one that I did a .bind() to.  

Actually think you have run into a bug in Python's socket module.
If it's convenient for you to build your own socket module, change
it like this and see what happens (I'm looking at version 2.1 here.)

*** socketmodule.c.dist Sun Apr 15 17:21:33 2001
--- socketmodule.c      Thu Jun 20 10:10:12 2002
***************
*** 597,602 ****
--- 597,603 ----
        case AF_UNIX:
        {
                struct sockaddr_un *a = (struct sockaddr_un *) addr;
+               a->sun_path[addrlen - (sizeof(*a) - sizeof(a->sun_path))] = 0;
                return PyString_FromString(a->sun_path);
        }
  #endif


The address returned from recvfrom is (addr, len), but we've been
ignoring len on the theory that the string would be NUL-terminated.
It isn't, so there can be some garbage on the end.

|  This is academic curiosity, and I'll read through the C sources
|  to see what I can glean therefrom, eventually.  I'd like to know
|  how I can tell what process has opened a connection to my 
|  AF_UNIX socket and I'd love to know how to pass an open file
|  descriptor through an AF_UNIX socket (in Python).  I've read that
|  this is possible in C and that it is one of the few features of 
|  AF_UNIX sockets which is NOT available via other address/protocol 
|  families.

Yes, AF_UNIX supports that (SOCK_DGRAM or SOCK_STREAM.)  See man 2
sendmsg.  That would have to be added to the socket module.  Probably
easy enough to do, as long as you confine yourself to support for this
one particular thing (passing file descriptors.)

|  ... I've also heard that this can be used as a way for a
|  client to request read-only or append-only access to a file
|  --- since the UNIX APIs (in C) apparently offer ways for the server
|  to obtain credential information about the requesting process as
|  well as passing open file descriptors (which are then roughly akin
|  to "capabilities" since their access mode can't be changed)

Don't know, right off hand I can't think what API function you
might be thinking of there.

	Donn Cave, donn at u.washington.edu


From martin at v.loewis.de  Wed Jun 19 01:47:06 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 19 Jun 2002 07:47:06 +0200
Subject: how i use '=?iso-8859-15?q?=E3?=' in python???
References: <3D0F6623000017AA@www.zipmail.com.br> <mailman.1024463483.27572.python-list@python.org>
Message-ID: <m3ofe7g6qt.fsf@mira.informatik.hu-berlin.de>

Gerhard H?ring <gerhard at bigfoot.de> writes:

> You need to give more details than "doesn't work". Most probably there
> is no problem with the statement
> 
>     s='?'
>     
> but a problem occurs when you try to print a string that contains a
> non-ASCII character, right?

I guess the error occurs when entering that statement into IDLE. In
that case, I can only recommend to use a different editor, at the
moment.

Regards,
Martin



From fperez528 at yahoo.com  Fri Jun  7 15:24:47 2002
From: fperez528 at yahoo.com (Fernando Perez)
Date: Fri, 07 Jun 2002 13:24:47 -0600
Subject: map
References: <3D007666.2B6D3582@gol.ge>
Message-ID: <adr1a0$1u3$1@peabody.colorado.edu>

Giorgi Lekishvili wrote:

> Hi all!
> 
> Is there a way to map a function with several arguments to a list? E.g.,
> 
>>>>import math
>>>>import MLab
>>>>a=MLab.rand(3,5)
>>>>b=map(math.pow,a)
> 
> of course, this doesn't work.
> 
> What to do? My task is to optimize the power (i.e., y in pow(x,y)) to
> achieve maximal performance.
> 
> 
> thanx,
> Giorgi
> 
> PS: Perhaps I have missed the needed functionality, that's presented in
> NumPy?

It's hard to tell what you're trying to do, since a 3x5 array doesn't have an 
'obvious' way to be used as an argument to pow. But anyway, for 'maximal 
performance' you should use NumPy the way it was meant to be used. It already 
overloads ** so you can just write:

In [1]: import MLab

In [2]: a=MLab.rand(3,5)

In [3]: a
Out[3]:
array([[ 0.7906459 ,  0.9962979 ,  0.59841698,  0.13870062,  0.93503422],
       [ 0.03783634,  0.68916065,  0.79217309,  0.23443535,  0.87729597],
       [ 0.04226565,  0.4821128 ,  0.8859337 ,  0.45935223,  0.47994375]])

In [4]: a**3
Out[4]:
array([[  4.94249303e-01,   9.88934754e-01,   2.14294850e-01,   
2.66830335e-03,
               8.17490114e-01],
       [  5.41660550e-05,   3.27311607e-01,   4.97118874e-01,   
1.28845512e-02,
               6.75209283e-01],
       [  7.55027297e-05,   1.12058801e-01,   6.95350325e-01,   
9.69253713e-02,
               1.10553126e-01]])


As a rule of thumb, in Numeric NEVER roll your own loops (even using map). 
Right there you kill all of the speed advantage of NumPy. If you can't find a 
way to write things with Numeric's functions and absolutely need to loop by 
hand, it will be dog slow. At that point, your only option is to write the 
tight loops in C, which is actually extremely easy to do using weave 
(http://scipy.org)

Here's a quick comparison showing the disaster you're heading into if you try 
to use loops/map instead of properly using numeric:

In [1]: import MLab, Numeric

In [2]: def loop_pow3(arr):
   ...:   out = Numeric.zeros(len(arr),arr.typecode())
   ...:   for i in xrange(len(arr)):
   ...:     out[i] = arr[i]**3
   ...:   return out
   ...:

In [3]: numpow3 = lambda x:x**3

In [4]: pow3=lambda x: pow(x,3)

In [5]: a=MLab.rand(100000)

In [6]: time_test (1,numpow3,a)
Out[6]: 0.070000000000000062

In [7]: time_test (1,map,pow3,a)
Out[7]: 1.1199999999999999

In [8]: time_test (1,loop_pow3,a)
Out[8]: 0.54000000000000004

In [9]: _7/_6
Out[9]: 15.999999999999984

In [10]: _8/_6
Out[10]: 7.7142857142857082

So in summary, using map is a factor of 16 slower than using Numeric's 
builtin operators (or rather its natural overloading of **) and using loops 
is a factor of 8 slower.

Unless you have a lot of time to sit around and wait, I'd recommmend spending 
an hour or two cuddled up with the Numeric documentation. It's a complete (if 
'creatively' organized ;) reference.

Cheers,

f.


From news at lindbergs.org  Sat Jun  1 12:53:48 2002
From: news at lindbergs.org (VanL)
Date: Sat, 01 Jun 2002 10:53:48 -0600
Subject: Generating unique numbers?
References: <3CF7EBC8.4080802@verio.net>
Message-ID: <3CF8FC1C.7060903@lindbergs.org>

Hi,

I appreciate the various answers which have been proffered. 
  My application is thus:

I am writing an outliner.

TANGENT: Yes, I am aware of several other python outliner 
projects.  While each is good for various things, 1. none so 
far have the exact featureset that I want/need and 2. I 
thought it would be fun to try to write a GUI application. 
I have written several 1000+ line scripts, but those were in 
a completely different problem domain.
END TANGENT

I am using these IDs as keys to various nodes, all held 
within a B-Tree (I to O Btree) that is contained within a 
standalone ZODB.

The reason why I am wanting keys that are more likely to be 
globally unique is because I will be building in hooks to 
make this outliner network transparent (using ZEO), making 
it possible for multiple outliners to be used together as a 
real-time collaborative tool.  I want to be able to be able 
to have people merge stuctures that people have made in the 
standalone instance into a shared instance without tromping 
on another nodes data.

Based on what people have said, I was thinking of using some 
combination of DateTime.gmticks and the id of the obj, 
making it (I think) extremely unlikely that there would be a 
collision.

Any comments on this, or the project in general would be 
appreciated.

Thanks,

VanL



From BPettersen at NAREX.com  Wed Jun 26 13:23:38 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Wed, 26 Jun 2002 11:23:38 -0600
Subject: Problems using samples
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E201922151B2@admin56.narex.com>

> From: Raphael Ribeiro [mailto:gusraan at terra.com.br] 
> 
> 1 x = 8
> 2 y = 4
> 3 v = x - y
> 4
> 5 print v
> 6 if x = y:
> 7	print "Eles s?o o mesmo valor"
> 8	else:
> 		print "Eles n?o s?o iguais , manja?"
> 
> this is my code, it says there's an error in line 6, but i 
> can't find what **** error is that, can anyone help me?

Generally, it's easier to figure out what is going wrong if you give us the stacktrace. In this case:

>>> if x = y:
  File "<stdin>", line 1
    if x = y:
         ^
SyntaxError: invalid syntax

Note that the ^ is pointing to where in the source the error is. x = y is an assignment, assigning the value of y to x. Assignments are not allowed as a test. I'm guessing you wanted to compare the two values to see if they were equal, in that case you need to say

  if x == y:

note the equality operator is spelled ==. A better error message would probably have been helpful...

-- bjorn




From tim.one at comcast.net  Mon Jun  3 18:56:13 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 03 Jun 2002 18:56:13 -0400
Subject: New win32all versions for Python 2.1 and 2.2
In-Reply-To: <3CFBF2BE.6020508@skippinet.com.au>
Message-ID: <LNBBLJKPBEHFEDALKOLCIEDFPKAA.tim.one@comcast.net>

[Mark Hammond]
>    I have released new win32all builds.
> ...
> These binaries are new (and have been built with WISE 8 - thanks Tim!)

Tim who?  Thank Wise.  I have no idea what you're talking about.

> so are still considered experimental.  They seem to work wekk though :)

I'm sure they work as wekk as Python itself <wekk>.

congratulating!-ly y'rs  - tim




From dbasch at yahoo.com  Sat Jun 29 18:39:08 2002
From: dbasch at yahoo.com (Derek Basch)
Date: Sat, 29 Jun 2002 15:39:08 -0700 (PDT)
Subject: Win32/Python serial port event character problem
In-Reply-To: <200206292038.g5TKcHu10071@smtp.swissonline.ch>
Message-ID: <20020629223908.78455.qmail@web20808.mail.yahoo.com>

Chris,
It is funny that you wrote me. I just spent the last
hour looking at pySerial. I dug into it some more and
I think that using the event character is only usefull
for driving the serial port polling function
('ReadFile' in this case). It is not usefull for
breaking the serial data into logical units (ie..lines
that end with '\n').

http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnwbgen/html/msdn_serial.asp

The 'caveat' section of this paper almost answers my
question. 

> and read the characters in a separate receiver
> thread. that thread can then 
> trigger on any character or sequence (e.g. using a
> Queue to communicate 
> with other threads)

Yeah, I think that is what will work best as a
solution. Thanks for the suggestion.

Also, I have been using the 'cbInQue' member of the
COMSTAT structure returned by 'ClearCommError' to
figure out how many characters are waiting in the
serial buffer. I see that you are pre-allocating a
buffer, using 'win32file.AllocateReadBuffer', based
upon the size of 'cbInQue' for the subsequent
'ReadFile' operation. According to Mark Hammond I
should be doing the same too.:

http://starship.python.net/crew/mhammond/win32/RecentChangeHistory.html
(section 108)

If anyone reads this and can show me a better way to
work with serial event characters in win32 please feel
free to :) 

Thanks for the help,
Derek Basch

> > Sorry to post this again but it didn't get any
> > responses. Maybe better luck this time.
> 
> well 14hrs isn't that much time on a weekend... some
> people only look into 
> news once per week.
> 
> > I use the following code to read data from the
> serial
> > port when a certain event character ('\n') is
> > received:
> ...
> > Has anyone ever dealt with this? Is using the
> event
> > character not a reliable method in windows?
> 
> no sorry, i haven't used the event mode.
> 
> ....
> > This is driving me batty so any help is greatly
> > appreciated.
> 
> <shaemles plug>
> my tip is: use http://pyserial.sf.net
> 
> and read the characters in a separate receiver
> thread. that thread can then 
> trigger on any character or sequence (e.g. using a
> Queue to communicate 
> with other threads)
> </shaemles plug>


__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com




From skip at pobox.com  Mon Jun 24 18:39:13 2002
From: skip at pobox.com (Skip Montanaro)
Date: Mon, 24 Jun 2002 17:39:13 -0500
Subject: embedded Python - optimization for creating/destroying many small objects
In-Reply-To: <3D174EEA.EFAE79C@fxtech.com>
References: <3D174EEA.EFAE79C@fxtech.com>
Message-ID: <15639.40849.865213.135766@12-248-11-90.client.attbi.com>

    Paul> I would like to simply create ONE of these objects and reuse it
    Paul> for the entire loop, so I am not paying for the memory
    Paul> allocation/deallocation overhead.

Define your object to have something akin to a __setstate__ method, create a
single instance of it before the loop, then set its state during each pass
of the loop:

    holder = MyContainer()
    while 1:
        holder.setstate(data)
        dosomething(holder)

-- 
Skip Montanaro
skip at pobox.com
consulting: http://manatee.mojam.com/~skip/resume.html




From kragen at pobox.com  Sat Jun 15 15:02:28 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 15 Jun 2002 15:02:28 -0400
Subject: Numeric module won't load on Debian Woody
References: <pan.2002.04.21.20.58.43.196357.31994@uol.com.br>
Message-ID: <83ofectlff.fsf@panacea.canonical.org>

synthespian <synthespian at uol.com.br> writes:
> 	Numeric Python o Woody does not work, even though I did apt-get install
> python-numeric without any problem.Theoretically, it's installed.
> 
> 	Has anyone experienced problems with the Numeric module on Debian Woody?
> 	What should I check? 
> 	Is there a workaround?
> 
> 	I have Python2.2 installed woth no problem.

Is your Python 2.2 a Woody package?  Last I recall, Woody had Python
2.1, and there were two Numeric packages: one for 1.5.2 and one for
2.1.  Neither of these will install files in /usr/lib/python22.  HTH.

FWIW, I'm running potato.  You are allowed to call me lame now.


From jdhunter at nitace.bsd.uchicago.edu  Mon Jun 17 10:55:12 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Mon, 17 Jun 2002 09:55:12 -0500
Subject: MySQL database location
References: <aekasd$96t$1@news.onet.pl>
Message-ID: <m28z5evttb.fsf@video.paradise.lost>

>>>>> "Pawel" == Pawel Lewicki <lewy0lewy at poczta.onet.pl> writes:

    Pawel> Hi, Is is possible to get the physical path to the MySQL
    Pawel> database file using MySQLdb or any other pythonous way?

Depending on your platform, you could parse /etc/my.cnf.

for line in open('/etc/my.cnf').readlines():
    if line.find('datadir=')==0:
        dataDir = line[8:-1]

print dataDir







From fredrik at pythonware.com  Wed Jun 26 06:29:44 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 26 Jun 2002 10:29:44 GMT
Subject: Parsing strings (\n and \\)
References: <mailman.1025016731.901.python-list@python.org> <3D19753C.9070104@thomas-guettler.de>
Message-ID: <sIgS8.45144$n4.10659657@newsc.telia.net>

Thomas Guettler wrote:

> Look at the two functoins quote and unquote. I wrote them
> without regular expression because I think it faster.

faster to write, perhaps.

and faster to run, if you only use them on strings with no
more than 2-3 characters.

but if you use a different set of test strings with more ordinary
characters than escaped characters, e.g.

     strings = ['foo', '', '\\', ' ', '"', '\\"', '\\\\']
     strings = [(x+"spamspamspamspamspam")*10 for x in strings]

you'll find that a RE approach can be much faster.  the following
version is about four times faster than your code, under 2.2:

def re_quote(string, sub=re.compile(r"[\\\"]").sub):
    def fixup(m):
        return "\\" + m.group(0)
    return sub(fixup, string)

def re_unquote(string, sub=re.compile(r"(?s)\\(.)|\\").sub):
    def fixup(m):
        ch = m.group(1)
        if ch is None:
            raise 'Parse Error: Backslash at end of string'
        if ch not in r"\\\"":
            raise 'Parse Error: unsupported character after backslash'
        return ch
    return sub(fixup, string)

:::

note the use of callbacks instead of substitution templates.  it's
usually faster (and in my opinion, also more pythonic) to use e.g.

    def fixup(m):
        return "spam %s %s" % m.group(1, 2)
    re.sub(pattern, fixup, string)

or, if you prefer lambdas:

    re.sub(pattern, lambda m: "spam %s %s" % m.group(1, 2), string)

than the re.sub non-standard interpolation syntax:

    re.sub(pattern, "spam \\1 \\2", string)

(and where possible, it's also slightly faster to use m.groups() instead
of enumerating all the groups in m.group(...))

ymmv, as usual.

</F>

<!-- (the eff-bot guide to) the python standard library:
http://www.pythonware.com/people/fredrik/librarybook.htm
-->




From jknapka at earthlink.net  Tue Jun 25 16:24:39 2002
From: jknapka at earthlink.net (Joseph A Knapka)
Date: Tue, 25 Jun 2002 20:24:39 GMT
Subject: Python and Eclipse
References: <3D17FAF4.35CC9F1B@earthlink.net> <81CC01E58BF2A99B.DFCD5EC6E4654907.32D731380951B9CA@lp.airnews.net> <3D189A8A.577C8186@earthlink.net> <3D18A6D2.5080404@dmsware.com>
Message-ID: <3D18D159.2F66A9E8@earthlink.net>

Paolo Invernizzi wrote:
> Concerning elipse and python, I think there are 2 roads.
> The first one is using jpe (http://sourceforge.net/projects/jpe) the
> python-java framework. I've tried it some time ago... and I managed to
>   execute python code from a java VM (a jython prompt in reality! ;) but
> I stopped there. Perhaps Frederic Giacometti and jython folks can give
> some help!
> 
> The second one is to use socket (is possible?) to connect to a python
> server application... (with the plus that switching various python
> versions is very easy)... and move some logic in python (fast
> development times ;)
> 
> The problem, as usual It's that I have no time ;(

I think the latter is the best way. It's not clear, though,
that these are the only two roads. But in any case, we must
IMO be able to debug code under any Python version, preferably
selectable among a number of different installed versions
on the same machine (Eclipse allows this with the JVM, so
I don't see why it would be a problem for Python). That
way we could support CPython and Jython equally well.

Another poster suggests that the Ruby Eclipse plug-in
would be a good place to start, and I'm going to have
a look at that later this week.

-- Joe
   "Many ways to get what you want / I use the best /
    I use the rest!"
    -- Sex Pistols, "Anarchy in the UK"


From ftobin at neverending.org  Sun Jun 30 16:48:00 2002
From: ftobin at neverending.org (Frank Tobin)
Date: Sun, 30 Jun 2002 16:48:00 -0400
Subject: Python needs better error reporting
In-Reply-To: <mailman.1025340162.4405.python-list@python.org>
References: <mailman.1025340162.4405.python-list@python.org>
Message-ID: <Pine.LNX.4.44.0206301647360.7942-100000@palanthas.neverending.org>

Brian Quinlan, on 2002-06-29, wrote:

> There are a lot more tokens than that allowed. What about "+", "*", "\",
> etc?

Gadfly does this with its SQL parsing, and it helps a lot.

-- 
Frank Tobin			http://www.neverending.org/~ftobin/



From owen at nospam.invalid  Thu Jun 20 18:20:52 2002
From: owen at nospam.invalid (Russell E. Owen)
Date: Thu, 20 Jun 2002 15:20:52 -0700
Subject: tkinter font picker?
References: <3D122BB6.92593026@tds.net>
Message-ID: <aetkg6$1qbo$1@nntp6.u.washington.edu>

In article <3D122BB6.92593026 at tds.net>,
 "Edward K. Ream" <edream at tds.net> wrote:

>Does anyone know of tkinter code that will put up a font chooser
>dialog?  There are a number of such packages written in tk/tcl.  I'm
>wondering if anyone knows of a port to tkinter.

I have a simple one that is part of a Tkinter-based preferences package. 
You can find it at <http://www.astro.washington.edu/owen/> in the 
Util.zip package; unpack and look in the Prefs folder.

If you hear of a better one, please post. I'd also like a better color 
picker than the basic one included with Tkinter.

-- Russell


From sholden at holdenweb.com  Fri Jun 28 08:56:10 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 28 Jun 2002 08:56:10 -0400
Subject: M2Crypto: select() behaves weird on SSL sockets
References: <3D04ED17.5070606@NOSPAMREMOVETHISxs4all.nl> <ae2str$b0q$1@nntp6.u.washington.edu> <3D066E77.4070901@NOSPAMREMOVETHISxs4all.nl> <ae6bmb$frc$1@dahlia.singnet.com.sg> <3D0790E1.9080904@NOSPAMREMOVETHISxs4all.nl> <831ybavprx.fsf@panacea.canonical.org> <3D0919AF.7050903@NOSPAMREMOVETHISxs4all.nl> <83it45lxfb.fsf@panacea.canonical.org> <3D1B92A2.7030605@NOSPAMREMOVETHISxs4all.nl>
Message-ID: <q0ZS8.106386$mF.41854@atlpnn01.usenetserver.com>

"Irmen de Jong" <irmen at NOSPAMREMOVETHISxs4all.nl> wrote in message
news:3D1B92A2.7030605 at NOSPAMREMOVETHISxs4all.nl...
> Kragen Sitaker wrote:
>
> > There are many pieces of middle ground between complete threading (two
> > threads per connection) and complete event-loop-drivenness with just
> > one thread --- as you're obviously aware, since you're using select()
> > in a threaded program.
>
> Um, yeah. I have 1 loop using select() that waits for new incoming
> connections. Every thread that is started for a connection just
> recv()s data off the socket.
>
> > What I guess you're not aware of is that asyncore supports the middle
> > ground (middle colors?) as well as the event-loop-driven end of the
> > spectrum.  For example, Zope has an asyncore thread that handles I/O,
> > but actually processes requests and produces results in some child
> > threads.
>
> You're probably right. I don't know a lot about the workings of asyncore.
> Perhaps if I finally squash the bug that is still in Pyro 3
> -- *blushes* it's probably a threading race condition -- I'll spend
> some time looking at it.
>
I understand that Zope used the asyncore code as a basis for development
rather than adopting it outright, but I haven't actually checked the source
to ensure this is so.

> Doing stuff the Zope way, the worker threads must have some way to
> get to the input and output streams that are handled in asyncore, right?
>
Well, it's a relatively simple matter: when an asyncore channel's
handle-connect() method is called, one fo the arguments is the socket that
was created by the connection. handle_connect() simply uses this socket to
create another channel, which is added to the map used by the asynchronous
select() loop.

This new channel then starts to fire handle_read() method calls, which you
service by doing something with the data. If the something you do includes
putting data in an output buffer then handle_write() events also start to
occur.

asyncore/asynchat really repay the effort of study, so I would advise you to
take a fresh look when you get time.

To assist you, I plan to add documentation for asynchat for 2.3.

when-sourceforge-stops-playing-silly-buggers-ly y'rs  - steve
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From larooy at xtar.co.nz  Sun Jun 16 03:02:40 2002
From: larooy at xtar.co.nz (John La Rooy)
Date: Sun, 16 Jun 2002 19:02:40 +1200
Subject: a possibly foolish question about slices
References: <yu99sn3prbqm.fsf@europa.research.att.com>
Message-ID: <20020616190240.5c3cc72e.larooy@xtar.co.nz>

On Fri, 14 Jun 2002 17:50:09 GMT
Andrew Koenig <ark at research.att.com> wrote:

> Suppose x is a string or list, and I want to refer to
> a subsequence of x with length n starting at position p.
> under the assumption that this subsequence is entirely
> 
> To do so, I can use x[p:p+n] in all circumstances but one,
> namely when p < 0 and p == -n.  Is there a similarly concise
> expression that works even in this special case?
> 
> -- 
> Andrew Koenig, ark at research.att.com, http://www.research.att.com/info/ark

x[p:][:n]

John


From matt at clondiag.com  Tue Jun  4 11:01:39 2002
From: matt at clondiag.com (Matthias Kirst)
Date: Tue, 4 Jun 2002 17:01:39 +0200
Subject: Win32 Decimal point/comma
Message-ID: <3cfcd2e5$1@post.newsfeed.com>

*** post for FREE via your newsreader at post.newsfeed.com ***

Hi Kaci,

This little problem caused continious annoyances to me until I found the
locale module:

import locale

locale.setlocale( locale.LC_ALL, 'C' )
# sets portable locale -> point is decimal separator

Matt




 -----= Posted via Newsfeed.Com, Uncensored Usenet News =-----
http://www.newsfeed.com - The #1 Newsgroup Service in the World!
-----== 100,000 Groups! - 19 Servers! - Unlimited Download! =-----
                  


From tatebll at aol.com  Sun Jun 23 10:36:58 2002
From: tatebll at aol.com (Bill Tate)
Date: 23 Jun 2002 07:36:58 -0700
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at>
Message-ID: <cb4ba455.0206230636.418fc4da@posting.google.com>

Siegfried Gonzi <siegfried.gonzi at kfunigraz.ac.at> wrote in message news:<3D1482AE.8FA299D8 at kfunigraz.ac.at>...
> My conclusion: I will rewrite my simulation in Fortran 90 and will
> abadon Python. Python is maybe good for some small scripts but not
> appropiate for serious programming.

The hydrodynamic and sediment transport models I used to run had
simulation periods that required at least a week on a high-end
computer.  I wouldn't for
a moment conclude that python is an appropriate numerical engine for
computing the hydrodynamics in an estuary.  However, python worked
great for preprocessing large input data sets, post-processing very
large binary output files and visualizing results.  While that may not
rise to your level of what constitutes "serious" programming,
nevertheless, python performed an important purpose for my work and
made my life a hell of alot easier.

As far as python being appropriate for "small scripts but not
appropriate for serious programming" I don't see where your particular
application of python is sufficiently encompassing enough to make such
a blanket assertion to the language's application as a whole.  It
didn't work for you in the problem domain to which you used python, so
be it.  There are more than a few examples where python rises to the
level of performing mission-critical work outside of your particular
area of focus.


From roffe at aqualene.uio.no  Sat Jun 22 09:47:40 2002
From: roffe at aqualene.uio.no (Rolf Marvin Bøe Lindgren)
Date: Sat, 22 Jun 2002 15:47:40 +0200
Subject: [Q] numeric entry in boa constructor
Message-ID: <lbzk7orpgqr.fsf@aqualene.uio.no>

I've created a simple app in boa constructor.  I would like to add a
feature that does not seem to come with a pre-made component:

the user hits a number on the numeric keypad.  the number is displayed
in a field.  the user hits another number, the previous number is
erased and the new number appears.  

any clues?  

-- 
Rolf Lindgren                                            http://www.roffe.com/
roffe at tag.uio.no


From jubafre at zipmail.com.br  Fri Jun 21 01:24:06 2002
From: jubafre at zipmail.com.br (jubafre at zipmail.com.br)
Date: Fri, 21 Jun 2002 02:24:06 -0300
Subject: =?iso-8859-1?Q?text=20search=3F=3F?=
Message-ID: <3D12657900000468@www.zipmail.com.br>

a have one componnent Entry, i put a string in Entry and search in the text
your position to select it, but the searck doesn?t find a Upper letter W
if a put lower, how i can serch in the text a W puting w??????


string='winzip'
i want to find WinZip
How i can find WinZip putting the string in lower??????


Juliano Freitas
www.gebrasil.hpg.com.br




------------------------------------------
Use o melhor sistema de busca da Internet
Radar UOL - http://www.radaruol.com.br







From insyte-clp at emt-p.org  Thu Jun 20 00:50:06 2002
From: insyte-clp at emt-p.org (Ben Beuchler)
Date: 20 Jun 2002 04:50:06 GMT
Subject: M2Crypto / SimpleXMLRPCServer
Message-ID: <slrnah2nmh.p18.insyte-clp@petra.bitstream.net>

I'm trying to set up an SSL SimpleXMLRPCServer instance.  I've
successfully used the xmlrpc_srv.py demo script from the M2 distro, but
would like access to the convenience of .register_function() and
.register_instance().

It looks to me like SSL.SSLServer and its various derivatives (Threading,
Forking) *should* be a drop-in replacement for SocketServer.TCPServer when
instantiating SimpleXMLRPCServer.  I attempted to derive a
SimpleSSLXMLRPCServer class like this (after setting up an appropriate
ssl_context):

class SimpleSSLXMLRPCServer(SSL.SSLServer, SimpleXMLRPCServer):
    def __init__(self, ssl_context, address, 
                 RequestHandler=SimpleXMLRPCRequestHandler, logRequests=1):
      SSL.SSLServer.__init__(self, address, RequestHandler, ssl_context)
      self.funcs{}                     # These three lines are copied from
      self.logRequests = logRequests   # SimpleXMLRPCServer's __init__, as 
      self.instance = None             # calling it would create non-ssl
                                       # sockets. 

My understanding is that the methods in SSL.SSLServer would override the
methods inherited by SimpleXMLRPCServer from SocketServer.TCPServer.

It appears that the SSL part, at least, works great.  I'm able to make a
connection using xmlrpclib.ServerProxy("https://blah") but the actual
remote method call hangs indefinitely.

Is my basic premise flawed?  Or just my implementation?

Here's the output on the server side:

>>> ctx = init_context('server.pem', 'ca.pem', SSL.verify_none)
>>> bob = SimpleSSLXMLRPCServer(ctx, ('216.243.168.40', 9090))
>>> def q(a,b):
...     return a + b
...
>>> bob.register_function(q)
>>> bob.serve_forever()
LOOP: SSL accept: before/accept initialization
LOOP: SSL accept: SSLv3 read client hello A
LOOP: SSL accept: SSLv3 write server hello A
LOOP: SSL accept: SSLv3 write certificate A
LOOP: SSL accept: SSLv3 write server done A
LOOP: SSL accept: SSLv3 flush data
LOOP: SSL accept: SSLv3 read client key exchange A
LOOP: SSL accept: SSLv3 read finished A
LOOP: SSL accept: SSLv3 write change cipher spec A
LOOP: SSL accept: SSLv3 write finished A
LOOP: SSL accept: SSLv3 flush data
INFO: SSL accept: SSL negotiation finished successfully
petra.bitstream.net - - [19/Jun/2002 22:27:55] "POST /RPC2 HTTP/1.0" 200 -

The client side is unremarkable.  I instantiate a ServerProxy and attempt
to make a remote call.  It hangs.  Very boring.

Any thoughts?

-Ben

-- 
Ben Beuchler                                           There is no spoon.
insyte at emt-p.org                                            -- The Matrix


From dom at edgereport.put_a_c_o_m_here  Thu Jun  6 15:33:00 2002
From: dom at edgereport.put_a_c_o_m_here (Domenic R. Merenda)
Date: Thu, 06 Jun 2002 19:33:00 GMT
Subject: Medium to Large Scale Python Deployments
References: <KHkL8.21383$eD2.2399923271@newssvr10.news.prodigy.com> <82afc355.0206060043.d26564e@posting.google.com>
Message-ID: <MNOL8.21619$wd1.2487967893@newssvr10.news.prodigy.com>

   Sure, Pedro.  The database server is homegrown C and RPG II code running
on the AS/400 for the majority of our "archived" data.  The live database is
mySQL on a PC.  User interface is PC/400 by IBM, and communcation is handled
over IPX through twisted pair cabling.  Kludgy, but it works so far. :-)
OS/400 is a wacky operating system.

--
Domenic R. Merenda
Editor, The Edge Report
http://www.edgereport.com


"Pedro Vale Lima" <pedrovl at yahoo.com> wrote in message
news:82afc355.0206060043.d26564e at posting.google.com...
> "Domenic R. Merenda" <dom at edgereport.put_a_c_o_m_here> wrote in message
> >
> >    I am running a home-grown Enterprise Resource Planning system,
written
> > entirely in Python
>
> Hello,
>
> Could you tell us more information about your system (Database Server,
> User Interface, Communication protocol, etc)? I am running an ABAP
> based ERP but I'm pretty sure your is better :-)
>
> regards,
>
> Pedro Lima




From peter at engcorp.com  Thu Jun  6 21:08:00 2002
From: peter at engcorp.com (Peter Hansen)
Date: Thu, 06 Jun 2002 21:08:00 -0400
Subject: Interfacing to 1-Wire LAN, especially Dallas temperature sensors?
References: <a7iefuk650noamrqcmpvroqj60mtgbpf9v@4ax.com> <3CF82E8E.23C6FA10@engcorp.com> <3cfb83e2$1_4@goliath.newsgroups.com> <3CFC226B.B6CB8995@engcorp.com>
Message-ID: <3D000770.84B25F29@engcorp.com>

Peter Hansen wrote:
> 
> Brad Clements wrote:
> >
> > There is a technique where you can use a UART at some whacky baud rate and
> > bits/byte setting to communicate. There's an APPNOTE somewhere, I don't have
> > it handy.
> 
> If someone can dig that out (I searched before and couldn't find it
> in all the noise) I'd appreciate it.  I would also be happy to try
> to implement it under Python, since we'd have some uses for such a
> beast.
> 
> (But I still don't believe it can be done reliably.  Maybe I'm wrong.
> Would be cool.)

I stand corrected, by private email.  And now by Dave Moor's posting
elsewhere in the thread.  Sometimes it's _good_ to be wrong. :)

-Peter


From phil at river-bank.demon.co.uk  Thu Jun 20 15:50:55 2002
From: phil at river-bank.demon.co.uk (Phil Thompson)
Date: Thu, 20 Jun 2002 20:50:55 +0100
Subject: PyQT Complie problems
References: <a5384db6.0206201130.1e08b42a@posting.google.com>
Message-ID: <3D12321F.9090100@river-bank.demon.co.uk>

ANdrew wrote:

> First off,
> 
> using:
> 
> Windows 2000
> 
> MSVC 6.0 service pack 5
> QT eval 3.0.4
> Active State Python 2.2
> PyQt-3.2.4 (source)
> 
> 
> I am trying to compile for Qt 3.0.4 eval.  Following the win32
> instructions, i am able to compile the first makefile
> (sip_helper.exe).  However, when i change dir to the "qt" dir and
> modify the makefile for the correct version of the .lib file and fix
> the one type-o in the sipqtDeclqt.h file ( changed #include <sipQt.h>
> to #include <sipQtQT.h> )


Why do you think there is a typo? There is no file sipQtQT.h.

> i get a synatx error followed by an
> unexpected end of file error. This is stated as being in the file
> "sipqtQPaintDevice.h".  The error is "sipqtQPaintDevice.h(44) : error
> C2146: syntax error : missing ';' before identifier 'sipLazyAttrDef'"
> 
> I have downloaded the latest snapshot (PyQt-snapshot-20020619.zip)and
> there is nothing in there about the qt section.  I am at a loss and
> any help would be greatly appreciated.


What do you mean by "the qt section"? The build system in the snapshots 
is completely different to v3.2.4 - the C++ code is now generated as 
part of the configuration process.


Phil






From dyoo at hkn.eecs.berkeley.edu  Sat Jun 15 02:02:51 2002
From: dyoo at hkn.eecs.berkeley.edu (Daniel Yoo)
Date: Sat, 15 Jun 2002 06:02:51 +0000 (UTC)
Subject: PyArg_Parse() & review for memory leaks
References: <ugk9amdu63ec36@corp.supernews.com>
Message-ID: <aeelab$1gpv$1@agate.berkeley.edu>

Anton Graph <""aglyportat\"@n-o.spa__mmm.yahoo dott com> wrote:

[some code cut]

:        char *pstr;
:        if(!PyArg_Parse(result, "s", &pstr)) { // the plug-in just
:                // builds a string containing tag
:                return 0;
:        }
:        int st=strlen(pstr);
:        if(st>=len) {
:                delete [] *res;
:                len=st+1;
:                *res = new char[len];
:        }
:        astrncpy(*res, pstr, len);
:        Py_DECREF(result);
:        return st;
: }

: do I leak pstr here? Sample code I've googled up seems similar, but I 
: don't understand how does PyArg_Parse allocate enough space without 
: going to the heap?

I don't believe so: PyArg_Parse()ing a string shouldn't be doing any
allocation because pstr is simply directed to the string pointer that
Python is using internally to store that String object --- no string
copying is involved.

Section 5.5 of the Python/C API document explains more about
PyArg_Parse():

    http://python.org/doc/api/arg-parsing.html

Good luck to you!


From phd at phd.pp.ru  Wed Jun 19 10:19:53 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 19 Jun 2002 18:19:53 +0400
Subject: procmail replacement in Python
In-Reply-To: <NCBBLFCOHHDIKCAFGCFBAEGILFAA.marklists@mceahern.com>; from marklists@mceahern.com on Wed, Jun 19, 2002 at 09:08:53AM -0500
References: <oqelf3s7n4.fsf@carouge.sram.qc.ca> <NCBBLFCOHHDIKCAFGCFBAEGILFAA.marklists@mceahern.com>
Message-ID: <20020619181953.E4127@phd.pp.ru>

On Wed, Jun 19, 2002 at 09:08:53AM -0500, Mark McEahern wrote:
>   http://www.qcc.sk.ca/~charlesc/software/getmail-2.0/

   "A POP3 mail retriever..." Hence, it is not fetchmail replacement, as
fetchamil also does IMAP, and does it very well. Fetchmail also has good
GUI configuration program, written in Python and Tkinter.

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From fredrik at pythonware.com  Tue Jun 11 17:01:39 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Tue, 11 Jun 2002 21:01:39 GMT
Subject: frame.bind ?
References: <3d05fff1$0$8149$ba620e4c@news.skynet.be>
Message-ID: <TytN8.42841$n4.9774273@newsc.telia.net>

"kemu" wrote:

> frame.bind("<Button-1>", callback)
>
> I don't understand what .bind does ?
> I know button-1 is the first mouse button and he calls the function callback
> but what does bind do ??

tell Tkinter that it should call the function callback when
the first mouse button is pressed, if the mouse pointer is
inside the frame widget.  more here:

    http://www.pythonware.com/library/tkinter/introduction/events-and-bindings.htm

</F>




From peter at engcorp.com  Sun Jun 23 21:42:11 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 23 Jun 2002 21:42:11 -0400
Subject: Redirecting StdOut
References: <1024856420.145085@seux119> <Xns9236D1FF016AAcliechtigmxnet@62.2.16.82> <1024860416.580309@seux119> <3D163550.E525D840@engcorp.com> <1024873013.652528@seux119>
Message-ID: <3D1678F3.AD9B059B@engcorp.com>

"Ugo Garc?a" wrote:
> 
> "Peter Hansen" <peter at engcorp.com> escribi? en el mensaje
> > Why would you bother with C for a game that relies on simple print
> > statements for its interface?
> 
> What I'm making is a engine to make Graphical Adventures.
> 
>  Can't you implement the game entirely in Python?
> Probably it would be implemented entirely in Python. Why not? SPEED. I'm
> familiarized with C and prefer to make the core of the engine in C, which is
> faster and more reliable for this kind of things, and leave Python be ONLY
> the script language of the engine.

You'll doubtless get other replies on this, including pointers to PyGame.
I just want to ask a few more questions (yes, all just to make my point...
please forgive me :-).

What makes you think a game implemented in Python will not have
enough speed?  You certainly would not be writing the low level
graphics routines in Python, but then you aren't about to do that
in C, either (I sincerely hope).  Python can call on libraries 
which have already been written *in C* (usually) to do the grunt
work, so it is quite fast enough.

As for reliable, I have to most vigorously object to anyone 
characterizing C as "reliable".  C code is notoriously difficult
to make reliable, generally being strewn with pointer bugs, 
memory allocation problems, and so forth.  C was really written
only as a somewhat more human-readable Assembly language substitute
and has been supplanted by many more suitable alternatives for
something as sophisticated as the engine of a game.

Or to put it another way: don't knock it till you've tried it!
Python is quite possibly a far more suitable solution for
the entire application, not just for the scripting language.

-Peter


From david.abrahams at rcn.com  Sat Jun 29 18:52:29 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Sat, 29 Jun 2002 18:52:29 -0400
Subject: SSE With Swig
References: <uhs5g8hos5og51@corp.supernews.com>
Message-ID: <afle7n$esi$1@bob.news.rcn.net>

"Andrew Wilkinson" <ajw126NO at SPAMyork.ac.uk> wrote in message
news:uhs5g8hos5og51 at corp.supernews.com...

> I'm trying to develop a Vector class that uses Intel's SSE extension. It
> works very well when called using c or c++, but when I wrap the class with
> Swig I get a privileged instruction exception error (this is under Windows
> 2000 using MSVC++ 6sp5 with the processor pack installed).
>
> I've included the code below, any suggestions would be greatly appreciated
> Andrew Wilkinson

1. Get rid of the virtual destructor
2. Get a copy of Scott Meyers' Effective C++ CD and read about operator
overloads. Then write operator+ as a free function:
    vector operator+(vector const&, vector const&);
3. None of these are the reason for your crash, though. I don't know
anything about SSE, but don't you have to tell it how many doubles it's
going to operate on?


> swig.i
> %module Vector
> %{
> #include "Vector.h"
> %}
> class Vector
> {
> public:
>  Vector();
>  virtual ~Vector();
>  Vector* operator+(const Vector *other);
>  float x;
>  float y;
>  float z;
>  float w;
> };
> __
> vector.h
> class Vector
> {
> public:
>  Vector();
>  virtual ~Vector();
>
>  Vector* operator+(const Vector *other);
>
>  __declspec(align(16)) union {
>   __m128 v;
>   struct {
>    float x;
>    float y;
>    float z;
>    float w;
>   };
>  };
> };
> ---
> vector.cpp
> #include "Vector.h"
> #include <xmmintrin.h>
>
> Vector::Vector()
> {
>  x = 0.0f; y = 0.0f; z = 0.0f; w = 0.0f;
> }
>
> Vector::~Vector()
> {
>
> }
>
> Vector* Vector::operator +(const Vector *other)
> {
>  Vector* r = new Vector;
>
>  r->v = _mm_add_ps(v, (other->v)); // The exception occurs here.
>
>  return r;
> }
>
> --
> ---
> Ditch The Decimal System!
> Lets Use Hex - http://www.intuitor.com/hex/switch.html
>
>




From martim.oberhuber at no-spam.org  Thu Jun 20 15:09:46 2002
From: martim.oberhuber at no-spam.org (Martin Oberhuber)
Date: Thu, 20 Jun 2002 21:09:46 +0200
Subject: Java to Python converter?
References: <YrjQ8.34$g32.995@nreader1.kpnqwest.net> <mailman.1024575575.2971.python-list@python.org>
Message-ID: <%LpQ8.55$g32.1762@nreader1.kpnqwest.net>

Hi Gerhard,

thanks a lot for your answer.

The reason why I want to convert Java into Python is this:

We are selling a programmer's IDE (SNiFF+) that is written in C++ but has an
embedded python interpreter for scripting it. I'd like to write an import
filter for importing project info from an external tool.

We have a reader for the project files of that external tool, but the reader
is written in Java. If I could easily do a "no-brain" conversion from Java
syntax to python syntax, this would make the "brain-1.0" porting work a lot
easier for me.

I don't expect the Java -> Python converter to produce 100% correct code; it
should just convert the syntax (i.e. classes + methods into "def" lines,
braces {} into indentation) -- that's it.

Thanks anyway,

Martin

--

martin <dot> oberhuber <at> windriver.com

"Gerhard H?ring" <gerhard.haering at gmx.de> wrote in message
news:mailman.1024575575.2971.python-list at python.org...
> * Martin Oberhuber <martin.oberhuber at windriver.com> [2002-06-20 13:58
+0200]:
> > Hi all,
> >
> > I've got some Java source code that I would like to convert into Python.
>
> Why?
>
> > Does anybody know of a converter that can do this?
>
> Brain 1.0.
>
> Btw. the Jython incarnation of Python can use Java libraries.
>
> Gerhard
> --
> mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
> web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
> public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
> reduce(lambda x,y:x+y,map(lambda
x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))
>
>






From tony at sentienc.co.yuk  Wed Jun 12 08:32:34 2002
From: tony at sentienc.co.yuk (Tony B)
Date: Wed, 12 Jun 2002 13:32:34 +0100
Subject: Zope Newbie query
Message-ID: <ugefre2980kj09@corp.supernews.com>

Apologies if this is an innaproprate newsgroup - it's the closest I could
find.

I have been playing with zope and would like to use it properly now the
problem I'm encountering is: How do I remotely install zope on a raq4 on
which I am not the main raq4 administrator (i.e I have a subdirectory on
which my site is hosted).? Do I have to have root priveliges to install and
run zope?

Has anyone done this who can point me in the right direction?

Cheers

T




From campbells4 at ozemail.com.au  Thu Jun 27 07:00:49 2002
From: campbells4 at ozemail.com.au (Alistair Campbell)
Date: Thu, 27 Jun 2002 21:00:49 +1000
Subject: speech synthesis
Message-ID: <idCS8.378$yY2.9108@ozemail.com.au>

Hiya,

I am a newbie with Python so any response to this may be way over my head.

Basically, I am wanting to know whether there are any python modules that
allow you to write programs which include voice synthesis. The sort of thing
I am wanting to look at is an augmentive communication program for disabled
people with speech and language disorders. So, I want to be able to convert
typed text and stored blocks of text into audible speech using standard
sound card interfaces. The application would be event driven and, at this
stage, I would probably look at building the GUI using Tkinter.

I am using Python 2.2 on a Windows OS. I have a Linux OS as well but I am
not at all proficient with the file management on this so I don't really
want to get into learning that interface just yet.

If there is not a simple solution to this would there be a group of people
around who would be interested in porting the Festival speech synthesis
engine to the Windows OS and then writing the wrappers(is that the right
term?) to be able to call it from Python.

Alistair Campbell
Qld, Australia

t = ("Our 2 main weapons are...", "fear", "surpise", "and a fiendish
devotion to the Pope!", "Oh Damn.")




From ken at hotmail.com  Sat Jun  1 08:28:34 2002
From: ken at hotmail.com (Ken)
Date: Sat, 1 Jun 2002 22:28:34 +1000
Subject: How to open a HTML file when the python cgi program is executing?
References: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de> <ad95l1.i0.1@kserver.org> <ad9vsm$vhjnu$1@ID-49758.news.dfncis.de> <9m6ada.31b.ln@10.0.0.1>
Message-ID: <adaeh8$uum7t$1@ID-49758.news.dfncis.de>

> If you realize this before you've output anything, you can use one of the
> redirection / refresh headers, such as "Location".
>
> print """Content-type: text/html
> Location: http://replacement-url
>
> <p><a href="http://replacement-url">Click here</a></p>

Doesn't seem to work... :(


> """
>
> Theoretically it should be sufficient to just give the Location, but put
in
> a human-readable redirection just in case.
>
> Otherwise, if you've already started, you must continue. Just open the
file
> and print it.
>
>
> As for frames, there's the main document, which decides which frames get
> shown; and then the server gets separate requests for each of the frames.
> You might want to browse a few pages with telnet to get a feel for how
> things work.

Where can I find this document?

Thanks




From christophe.chappet at silogic.fr  Wed Jun  5 10:21:01 2002
From: christophe.chappet at silogic.fr (Christophe Chappet)
Date: Wed, 5 Jun 2002 16:21:01 +0200
Subject: Extending inline help utility of IDLE
Message-ID: <02b801c20c9c$38b36e70$01000001@TAKAROA>

Hi to all,
Is it possible to extend the inline help utility of the IDLE program to document my owns modules and functions ?
Many thanks,

    Agence Scientifique & Industrielle  
            Christophe CHAPPET

            Charg? d'Affaires

           
            T?l. : +33 534 619 246  -   Fax : +33 534 619 222
            e-mail : christophe.chappet at silogic.fr
           
            6, rue Roger Camboulives - BP 1133 - 31036 Toulouse Cedex 01 - FRANCE  
     

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020605/f84a8868/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: p3uilogo.gif
Type: application/octet-stream
Size: 2923 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020605/f84a8868/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Christophe Chappet.vcf
Type: text/x-vcard
Size: 563 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/python-list/attachments/20020605/f84a8868/attachment.vcf>

From gleki at gol.ge  Fri Jun 28 04:34:10 2002
From: gleki at gol.ge (Giorgi Lekishvili)
Date: Fri, 28 Jun 2002 10:34:10 +0200
Subject: Packages +reload()
References: <slrnaho66b.3of.jarvin24@tutor.cc.tut.fi>
Message-ID: <3D1C1F82.443F22CC@gol.ge>

Hi!

It would be nice, if you shell (trace) send the output as well.

Few questions:
Do you import eNact prior to import what's inside?
E.g.,
>From eNAct import *

ir import eNAct



J?rvinen Petri wrote:

> Hey,
>
> I feel like complite lost. I have created nice package structure,
>  but I encountered lot's of troubles when using it with IIS and ASP.
>
> Structure is:
>
>         eNAct\
>                 documents\
>                         listDocuments\
>                                 showPage()
>                 enactXML()
>
> -> main package eNAct and sub-packages eg. documents and it's
>  sub-packages eg. listDocuments
>
> The problem is that I can't reload the modules.
>
> I have tried various ways of importing listDocuments but IT doesn't
> get reloaded.
>
> eg.
> from eNAct.documents import listDocuments
> reload(listDocuments)
>
> listDocuments.showPage()
>
> or
>
> import eNAct
> reload (enact)

You surely remember that enact!=eNAct, don't you?

>
>
> eNAct.documents.listDocuments.showPage()
>
> Questions:
>
> 1. How should the structure be?
> 2. Is it possible that I only reload the eNAct and all sub-modules
>    get reloaded.
> 3. What does reload(eNAct.documents) reload?
>
> - Petri J?rvinen

Grtz,
Giorgi




From egbert at bork.demon.nl  Thu Jun  6 06:12:00 2002
From: egbert at bork.demon.nl (Egbert Bouwman)
Date: Thu, 6 Jun 2002 12:12:00 +0200
Subject: Program to an interface, not an implementation
Message-ID: <20020606121159.A975@bork.demon.nl>

Hello,
Can someone shed some light on the GoF adagium:
'program to an interface, not an implementation'
and especially what it means for programming in Python ?
GoF is not easy.
egbert
-- 
Egbert Bouwman - Keizersgracht 197 II - 1016 DS  Amsterdam - 020 6257991
========================================================================




From ngiven at lanl.gov  Thu Jun 20 12:20:06 2002
From: ngiven at lanl.gov (Nathan Given)
Date: Thu, 20 Jun 2002 10:20:06 -0600
Subject: changing the viewing area with gnuplot in python
Message-ID: <3D1200B6.639CEBE5@lanl.gov>

right now gnuplot auto scales everything (I don't have control over the
viewing area)...


How do I change it?  (I want to view from -4 to 2 on the x axis, and 0
to 10 on the y axis....  right now gnuplot automatically chooses -4 as
xmin, 2 as xmax, 4 as ymin, and 10 as ymax)

thanks
--
Nathan
--


From occeanlinux at linuxmail.org  Sun Jun  2 09:49:37 2002
From: occeanlinux at linuxmail.org (Gold Fish)
Date: Sun, 02 Jun 2002 13:49:37 GMT
Subject: Question on time module
Message-ID: <3cfa2287_1@news.iprimus.com.au>

Can anyone tell me how to sort the file in the directory according to their 
modification time. Then if i want to look up a find say python.txt Mar 04 
2002 in the directory can i use glob.glob to do it. I using 
time.localtime(time.time()) to get the current time in, how can i display 
only the file in the directory 5 days agos from the current time. Please 
help me to solve this problem. I tried to use os.stat to seperate the file 
time but it couldn't work.  


From candiazoo at attbi.com  Thu Jun 27 10:09:09 2002
From: candiazoo at attbi.com (candiazoo at attbi.com)
Date: Thu, 27 Jun 2002 14:09:09 GMT
Subject: Most efficient way to write data out to a text file?
References: <3d1a6993.248534625@netnews.attbi.com> <3D1A8CD9.C8B8EC28@engcorp.com>
Message-ID: <3d1b1bd7.294171125@netnews.attbi.com>

I am a newbie... this is my first Python "project" so I am probably writing
horribly inefficient code...  I have never used the profiler but I'll try it!

I am not opening/closing the file each time.  I am extracting 700000 rows from a
mysql database, extracting additional data from our primary, Oracle database per
row, then stuffing each piece of data into a class which preformats the data (I
need to output the data into a fixed format string/record for another
application which reads them) and returns a single string... which I write out
to the file.

Mike


On Wed, 26 Jun 2002 23:56:09 -0400, Peter Hansen <peter at engcorp.com> wrote:

>candiazoo at attbi.com wrote:
>> 
>> I assume some sort of block i/o or preallocating a large block of file space
>> would be the best way to do it?  I am outputting about 700,000 records to a text
>> file and my program only spits out about 20 records per second.  This is using
>> standard file open and write...
>
>You're not opening and closing the file for each record that is written, are you?
>
>Can you post more detail on the amount of data in a record, and the format?
>
>Why not profile the code using the profile.py module and actually see
>where the problem is?
>
>-Peter



From Dick.Zantow at lexisnexis.com  Tue Jun 25 12:09:31 2002
From: Dick.Zantow at lexisnexis.com (rzed)
Date: Tue, 25 Jun 2002 12:09:31 -0400
Subject: Parsing strings (\n and \\)
References: <3D188006.7050202@thomas-guettler.de> <mailman.1025017712.20580.python-list@python.org> <lt0S8.45025$n4.10623651@newsc.telia.net>
Message-ID: <afa4m5$a92$1@mailgate2.lexis-nexis.com>

"Fredrik Lundh" <fredrik at pythonware.com> wrote in message
news:lt0S8.45025$n4.10623651 at newsc.telia.net...
> Fran?ois Pinard wrote:
>
> > A simple avenue is to get Python itself to evaluate the string as
a constant
> > (you ensure this by removing evaluation context).  Something like
this:
> >
> >     PARSED = eval(UNPARSED, {}, {})
>
> >>> UNPARSED = "__import__('os').system('echo dream on!')"
> >>> PARSED = eval(UNPARSED, {}, {})
> dream on!
>

Even if you use
from __far_future__ import * ?





From shalehperry at attbi.com  Mon Jun 24 18:55:06 2002
From: shalehperry at attbi.com (Sean 'Shaleh' Perry)
Date: Mon, 24 Jun 2002 15:55:06 -0700 (PDT)
Subject: Suggestions for good programming practices?
In-Reply-To: <WEuR8.2074$5l4.75657@ozemail.com.au>
Message-ID: <XFMail.20020624155506.shalehperry@attbi.com>

On 24-Jun-2002 Dianne van Dulken wrote:
> Hi,
> 
> I'm fairly new to python, coming from a perl background, and was wondering
> if anyone has a list of things that they consider as good programming
> practice in Python (I thought this might be an interesting topic for
> discussion, anyway)
> 
> For example, in perl you obviously are always encouraged to use strict, and
> we always use eval around our main code to pick up any unexpected errors.  I
> presume the equivalent in python is the try statement.  Would you suggest
> that this should be used, or would that be redundant here?  Any other tips
> that you would suggest a new python user should be always doing as a good
> programming practice?
> 

one of the mantras you will hear on this list is:

the simpler the code the better

you hear this on style issues, on optimization issues, and well practically
everywhere.  Usually the simple, clean approach yields the best possible python
-- both in your time and in runtime.  So if you find yourself saying "man, why
is this so complex" you should stop and rethink the problem.

As a side item, tutor at python.org is a GREAT place to hang out and read about
other people learning python.  You can really pick up a lot of the niceties and
sublties of the language there.




From nospam at bigfoot.com  Mon Jun 24 11:10:07 2002
From: nospam at bigfoot.com (Gillou)
Date: Mon, 24 Jun 2002 17:10:07 +0200
Subject: Subclass of array
References: <3d1714c3_7@news.newsgroups.com>
Message-ID: <af7c7o$ejc$1@norfair.nerim.net>

The UserList package is your friend...

--Gilles

"JB" <jblazi at hotmail.com> a ?crit dans le message news:
3d1714c3_7 at news.newsgroups.com...
> Can I derive an own class from array? (I use 2.2.)
> --
> Janos Blazi
>
>
> -----------== Posted via Newsfeed.Com - Uncensored Usenet News
==----------
>    http://www.newsfeed.com       The #1 Newsgroup Service in the World!
> -----= Over 100,000 Newsgroups - Unlimited Fast Downloads - 19 Servers
=-----



From xsebbi at gmx.de  Sun Jun 23 05:18:22 2002
From: xsebbi at gmx.de (Sebastian Roth)
Date: Sun, 23 Jun 2002 11:18:22 +0200
Subject: bloody freshman question regarding strings
In-Reply-To: <1024821351.3424.12.camel@dorian.blaue-elise.net>
References: <1024821351.3424.12.camel@dorian.blaue-elise.net>
Message-ID: <200206231118.22619@xsebbi.de>

On Sonntag, 23. Juni 2002 10:35, Christian Guenther wrote:
> My biggest question at the moment is why can't I do something like
> this:
>
> def mailmsg(severity message opts target recipient):
> 	# first create the message body
> 	msgbody="computer: " + COMPUTER + "\n
> 	         system:   " + SYSTEMTYPE + "\n
> 		 cpu:      " + CPUTYPE + "\n
> 		 kernel:   " + KERNELVER + "\n
> 		 date:     " + DATE + "\n
> 		 \n
> 		 severity: " + severity + "\n
> 		 message:  " + message + "\n\n"
>

a single " means that your string goes over one line in your 
interpreter. If you want to write over more than one line, try """ 
instead of ".

> 	chris

		Sebastian






From amuys at shortech.com.au  Mon Jun  3 20:36:46 2002
From: amuys at shortech.com.au (Andrae Muys)
Date: 3 Jun 2002 17:36:46 -0700
Subject: 'for every' and 'for any'
References: <20020526135944.A32690@hishome.net> <20020526091742.A987@unpythonic.net> <mailman.1022567685.7803.python-list@python.org> <slrnaf7af3.8kn.quinn@schilling.ugcs.caltech.edu> <mailman.1022658469.27441.python-list@python.org>
Message-ID: <7934d084.0206031636.5c29b31f@posting.google.com>

Oren Tirosh <oren-py-l at hishome.net> wrote in message news:<mailman.1022658469.27441.python-list at python.org>...
> I don't have a personal utility library. It's on purpose. I don't have
> personalized key bindings. I try to avoid customization.  Customization is
> a big part of what makes one programmer's code difficult to maintain by
> another programmer, what makes on programmers's workstation unusable by 
> another and makes combining code from several sources difficult.
> 
> Just one little example: I hate it when different parts of the code use 
> their own customized constants for ints from specific sizes.  Is it 
> u_int32_t, uint32_t or guint32? I hate it when a big project starts and the 
> programmers all start to waste their time on writing a yet another library 
> with all this stuff.
> 

Now I just don't understand how this position can possibly make sense?
 C dosn't have any 32-bit int types standardised, so what do _you_ use
when you need one?  Please don't tell me you don't use int, or I'll
have to hunt you down and put you out of your misery ;).  I personally
have absolutely no problem with any of the 3 examples you gave above:

sometimes-the-word-length-is-important-ly yours 
Andrae Muys


From phawkins at connact.com  Mon Jun 24 18:58:24 2002
From: phawkins at connact.com (Patricia J. Hawkins)
Date: 24 Jun 2002 18:58:24 -0400
Subject: Tkinter and threads
References: <3D0DCC8F.9A6D3B81@ipm.fhg.de> <wk8z58w81y.fsf@mail.connact.com> <af0gjp$g67$1@panix1.panix.com>
Message-ID: <wk7kko2sj3.fsf@mail.connact.com>

>>>>> "A" == Aahz  <aahz at pythoncraft.com> writes:

A> In article <wk8z58w81y.fsf at mail.connact.com>,
A> Patricia J. Hawkins <phawkins at connact.com> wrote:
>> 
>> Also be aware that the early code examples aren't intended as models,
>> just as teaching tools -- you don't hit the really useful stuff till
>> page 48. And the format makes it tough to flip back and forth (not to
>> mention, tiny amounts of information per page drives me nuts, but
>> perhaps that's just me) -- and at 84 pages, I'm not going to print it
>> out to take it downstairs to an armchair and mark it up, which is what
>> I want to do.  And skimming for what you don't know already is
>> impossible.

A> <sigh>  Yeah, I know.  Last summer, I was negotiating with O'Reilly to
A> write a whole book on the subject (their idea).  Last November, they
A> canned the project.  I haven't had the time/energy to rewrite the slides
A> into a real tutorial, and I'm working on a completely different book now.

I hate those old projects.  But in this case, just having an
all-on-one page version in smaller fonts would help a lot.  

-- 
Patricia J. Hawkins
Hawkins Internet Applications, LLC





From penguin at bredband.no  Wed Jun 12 20:08:59 2002
From: penguin at bredband.no (penguin)
Date: Thu, 13 Jun 2002 02:08:59 +0200
Subject: fileupload error!
Message-ID: <002c01c2126e$84925400$6e00a8c0@arcticpenguin02>

Hi I try to upload a picture with a python cgi script. Tried many versions of the code, this is the current. 







Here's the code and error!

The error:

Traceback (most recent call last): File "C:\cgi-bin\upload.py", line 179,
in ? xmldokument = skjema['dok'].value File "C:\Python22\lib\cgi.py", line
550, in __getitem__ raise KeyError, key KeyError: dok 



The code:

#!c:\python22\python.exe
# FileTransfer.py - CGI file transfer, may require Python 1.5 or higher
# Author: JeffBauer at bigfoot.com
# Prior art:  Aaron Watters, Jim Fulton

import xml.dom
import xml.dom.minidom
import xml.parsers

import os, sys, traceback
import httplib
import tempfile
from whrandom import randint
from string import joinfields
from time import time
import cgi

test = 0

FileUploadRequestException="Dette gikk ikke :"

def parseDok(xmlstreng): 
s=[]

pars = xml.dom.minidom.parseString(xmlstreng)
elementene = pars.getElementsByTagName('element') 

for element in elementene:
ss={}
ss['dataid'] =
element.getAttribute('dataid').encode('iso_8859_1')
ss['type'] =
element.getAttribute('type').encode('iso_8859_1')
ss['format'] =
element.getAttribute('format').encode('iso_8859_1')

s += [ss]

return s

def lagFilblokk(elementliste,xmldokument):
#path = "http://192.168.0.32/upload/"
path = "http://192.168.0.32/upload/"
#path = "O:\\"
#Ting skal vist hete userfile og attach1 o.s.v.
s = [{'fieldname':'userfile','filename':xmldokument}]
att = 0
for element in elementliste:
att += 1
ss={}
ss['fieldname'] = "attach%s" % att
ss['filename'] = path + element['dataid'] +
filending(element['type'],element['format'])

s += [ss]

return s

def filending(type,format):
#Dette skal gj?res bedre
if format == 'jpeg':
return '.jpg'
if format == 'jpg':
return '.jpg'
return '.txt'


class FileUploadRequest:
    """
    File upload modul, og sender en liste filer, i med gitte feltnavn.
    """
    def __init__(self, uri, host, port):
        self.uri = uri
        self.host = host
        self.port = port
        self.queryString = ''
self.teller = 1
        #self.boundary= '%s%s_%s' % ('-tr--', int(time()), randint(1,10000))
self.boundary= '%s%s_%s' % ('rrrr', time(), randint(1,10000))

#print self.boundary


    def load(self, filenames, headerDict=None):
for filename in filenames:
if not '?' in filename['filename']:
#Dette er ikke et xml dokument (som jo begynner med
<?xml...
        f = open(filename['filename'],'rb')
        data = f.read(1000000)
        #Problem her kanskje
        f.close()
else:
#Det er ikke noe filnavn dataene
data = filename['filename']
filename['filename'] = 'test.xml'
        hdr = []; part = []; total = []
        hdr.append('Content-Disposition: form-data; name="%s";
filename="%s"' % (filename['fieldname'],filename['filename']))
        hdr.append('Content-Type: application/octet-stream')
        hdr.append('Content-Length: %d' % len(data))
        # Add user-defined header values here, if provided.
        # Reminder: FieldStorage will convert key to lowercase.
        if type(headerDict) == type({}):
            for k in headerDict.keys():
                hdr.append("%s: %s", (k, headerDict[k]))
        part.append("%s\n\n%s" % (joinfields(hdr,'\n'), data))
        total.append('--%s\n' % self.boundary)
        total.append(joinfields(part, "\n--%s\n" % self.boundary))
        #total.append('\n--%s--\n' % self.boundary)
#print joinfields(total, '')
        self.queryString += joinfields(total, '')
self.teller +=1
self.queryString +=('\n--%s--\n' % self.boundary) #Dette er
sluttmarkeringen

    def request(self):
        query = self.queryString
        contentType = 'multipart/form-data; boundary=%s' % self.boundary
        contentLength = str(len(query))
        h = httplib.HTTP()
        h.connect(self.host, self.port)
        h.putrequest('POST', self.uri)
        h.putheader('Accept', '*/*')
        h.putheader('Proxy-Connection', 'Keep-Alive')
        h.putheader('User-Agent', 'Glue0.1 [no] (Mac OSX; U)')
        h.putheader('Content-Type', contentType)
        h.putheader('Content-Length', contentLength)
        h.endheaders()
        h.send(query)
        rcode, rmsg, headers = h.getreply()
        response = h.getfile().read(10000)
"""
        if rcode != 200:
            msg = "error: %s, %s\n%s %s" % (rcode, self.uri, rmsg, response)
            raise FileUploadRequestException(msg)
"""
f=file('c:\\RESPONSE.html','w')
f.write(response)
f.close
##
f=file('c:\\qvery.txt','w')
f.write(query)
f.close
##
        return response
        return response

def lastoppListe(host,uri,filenames):
    
    
    port = 9091
    port = 8080
    #port = 80
    
    fileUploadRequest = FileUploadRequest(uri, host, port)
    
    fileUploadRequest.load(filenames)
    response = fileUploadRequest.request()
    return response

def lesFil(filNavn):
r = open(filNavn)
dok = r.read(19000)
r.close()
return dok

def main(self,xmldokument):
#F?rst finne hvilke elementer vi har ? jobbe med
elementliste = parseDok(xmldokument) 
filliste = lagFilblokk(elementliste,xmldokument)
#Denne returnerer en liste over dataposter
#print filliste

svar = lastoppListe('192.168.0.1','/files/upload',filliste)
print svar


if __name__ == '__main__':
print "Content-type: text/html"
print
if test:
xmldokument = lesFil('test.xml')
else:

skjema = cgi.FieldStorage()
xmldokument = skjema['dok'].value

alfa= main(None, xmldokument)


Regards stan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/python-list/attachments/20020613/386bf1e4/attachment.html>

From huaiyu at gauss.almadan.ibm.com  Wed Jun 19 14:57:22 2002
From: huaiyu at gauss.almadan.ibm.com (Huaiyu Zhu)
Date: Wed, 19 Jun 2002 18:57:22 +0000 (UTC)
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <mailman.1024182500.12788.python-list@python.org> <f_gP8.42357$GY.12362923@twister.nyroc.rr.com> <3D0E8AC8.6C5C96EE@engcorp.com> <37TP8.47886$GY.15021139@twister.nyroc.rr.com> <3D1010C3.C321F8B0@engcorp.com> <raXP8.47926$GY.15260255@twister.nyroc.rr.com> <3D107948.B3FEE4BB@engcorp.com> <CT%P8.33113$uk2.13710586@twister.nyroc.rr.com>
Message-ID: <slrnah1l0i.6o5.huaiyu@gauss.almadan.ibm.com>

George Hester <hesterloli at hotmail.com> wrote:
> But the reason I
>mentioned is that I hoped that my students would take an interest in the subject
>and learn the concept.  They were not interested but they needed the credit to
>graduate.  In my case I was not necessarily interested in learning Python.  But
>I did want to determine my version of Python in ASP.  I suppose I was like my
>students and ya'll were like the teacher.  Yes basically I wanted it handed to
>me on a platter.  What I did with the platter may have surprised you.  So just
>becasue somebody wants something handed to them on a platter does not mean the
>contribution has less value.

Being able to communicate effectively in Usenet may be an important skill
for your students to master.  You may want to teach them some lessons you
have learned yourself.  Here's a mini multiple choice questionair to get you
started:

1. Suppose you want to get an answer to a question handed down on a platter,
   which would be the more effective question?
   a) What is the simplest script to get Python version?
   b) I know nothing about Python. I just want to know its version from ASP, 
   what's the easiest way to get it?

2. Suppose you are given an answer that is all Greek to you, which would be
   the more effective response?
   a) I don't think you guys know anything about this.
   b) I think I need a script to do this.
   c) This is well over my head.  I'm newbie.  Please explain the details,
   keeping in mind that I only want to get the version from ASP.

3. If someone tells you that you are going about it the wrong way and gives
   you some advice which sounds like a detour to you, do you judge it based
   on
   a) whether he is your daddy.
   b) whether the advice has merits.

4. Once you realize that all the answers people gave you were real, some
   much simpler than you imagined before, and that your accusations earlier
   were largely based on arogance, do you 
   a) Say: Hey, you guys didn't gave me the answers the way I wanted, and
   your tones were derogotory.
   b) Go away quietly.
   c) Say: Thanks for all the answers.  Even though I didn't ask the
   question clearly, one of answer was exacted what I wanted.  Sorry for the
   tone of my earlier posts, which was largely because I didn't recognize
   the answers.


No doubt others can add more lessons learned from this episode.

Huaiyu


From fredrik at pythonware.com  Wed Jun 19 17:58:32 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Wed, 19 Jun 2002 21:58:32 GMT
Subject: file .seek() under MS-Windows
References: <4169349c.0206191318.7f0d99d9@posting.google.com>
Message-ID: <c87Q8.44208$n4.10290986@newsc.telia.net>

Frank Schaefer wrote:

> The following code:
>
> fh = open("myfile.pot", "r")
> ...
> fh.seek(-1,1)
> ...
>
> produces an error:
> [IOError 22] Invalid Argument
>
> under MS Windows (XP). It works fine under Unix, though
> (for years).

really?

$ uname -sr
Linux 2.4.7-10
$ python
>>> fh = open("somefile")
>>> fh.seek(-1, 1)
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
IOError: [Errno 22] Invalid argument

> uname -sr
SunOS 5.8
>>> fh = open("somefile")
>>> fh.seek(-1, 1)
Traceback (innermost last):
  File "<stdin>", line 1, in ?
IOError: [Errno 22] Invalid argument

> uname -sr
OSF1 V4.0
>>> fh = open("somefile")
>>> fh.seek(-1, 1)
Traceback (innermost last):
  File "<stdin>", line 1, in ?
IOError: [Errno 22] Invalid argument

(etc)

according to the Unix specification, the underlying fseek
operation is supposed to generate this error if you attempt
to set the file pointer to a negative value.

what did you expect it to do?

</F>




From jadestar at idiom.com  Mon Jun 10 19:19:13 2002
From: jadestar at idiom.com (James T. Dennis)
Date: Mon, 10 Jun 2002 16:19:13 -0700
Subject: Why does Python mix OO concepts and non OO concepts for operation  s on basic types?
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E20192158EBD@admin56.narex.com>; from Bjorn Pettersen on Mon, Jun 10, 2002 at 05:00:30PM -0600
References: <60FB8BB7F0EFC7409B75EEEC13E20192158EBD@admin56.narex.com>
Message-ID: <20020610161913.B34334@idiom.com>

On Mon, Jun 10, 2002 at 05:00:30PM -0600, Bjorn Pettersen wrote:
>> From: James T. Dennis [mailto:jadestar at idiom.com] 
> > 
> Your wish is granted:
 
> >>> x = []
> >>> list.append(x, 5)
> >>> x
> [5]
> >>> str.join(' ', ['a', 'b', 'c'])
> 'a b c'
> >>>
> 
> not-that-I-would-ever-do-something-like-that'ly y'rs
> -- bjorn

 Doh!
 Thanks.

 should-have-tried-that-before-I-posted'ly y'rs
 Jim.





From brueckd at tbye.com  Wed Jun 26 11:03:54 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 26 Jun 2002 08:03:54 -0700 (PDT)
Subject: Suggestions for good programming practices?
In-Reply-To: <lupafa.691.ln@grey.aerojockey.localdomain>
Message-ID: <Pine.LNX.4.44.0206260722070.29286-100000@ns0.tbye.com>

On Tue, 25 Jun 2002, Carl Banks wrote:

> So why did you ask what kind of teacher calls a newbie's attention to
> eval?  You seemed (not to be confused with "I have concluded") to have
> been advocating ignorance.

Nope. Just like on the first day of a C programming class, a lousy teacher
would say, "Avoid pointers to pointers!" - it doesn't help at all to bring
that up so early in the learning process, it's confusing, and those types
of problems are furthest from the students' minds (they're interesting in
*doing* something with this new language). There's a host of other things
that are more useful to a newbie, whether experienced in other languages
or not. But *if* you do bring it up, then you should take the time to
explain why it's potentially bad.

Likewise, it doesn't really help a newbie much to bring up eval/exec right
away and say, "don't use this!" because there's tons of other things that 
will help them so much more, and in the early learning stages they're not 
going to be hurt by using those functions anyway. But *if* you do bring 
those functions up and advise against using them, you should explain why.

And if your code will be running in an environment where it might be used
maliciously, you need more than "don't use exec/eval" anyway - you really
need to take some time to learn about basic program security as there are
plenty of other potential security holes (at the very least you need a
security mini-FAQ or checkoff list in which exec/eval would be listed
along with many other things).  With that knowledge you'd discover the
risk of eval/exec on your own - they'd jump right out at you with even
just a quick security audit.

-Dave





From financing at cfswebmail.com  Tue Jun 18 06:19:01 2002
From: financing at cfswebmail.com (Canadian Subsidy Directory)
Date: Tue, 18 Jun 2002 06:19:01 -0400
Subject: Available; Subsidies, Grants, Loans, Financing and General help.
Message-ID: <200206180527.g5I5RO726262@plugngo.travelnet.ca>

MG PUBLISHING
4865 HWY 138,R.R 1
ST-ANDREWS WEST
ONTARIO, KOC 2A0


PRESS RELEASE

CANADIAN SUBSIDY DIRECTORY YEAR 2002 EDITION
Legal Deposit-National Library of Canada

ISBN 2-922870-02-02 (2002)
ISBN 2-922870-01-4 (2001)



M.G. Publishing is offering to the public a revised edition of the Canadian
Subsidy Directory, a guide containing more than 2800 direct and indirect
financial subsidies, grants and loans offered by government departments and
agencies, foundations, associations and organizations.  In this new 2002
edition
all programs are well described.

The Canadian Subsidy Directory is the most comprehensive tool to start up a
business, improve existent activities, set up a business plan, or obtain
assistance from experts in fields such as: Industry, transport, agriculture,
communications, municipal infrastructure, education, import-export, labor,
construction and renovation, the service sector, hi-tech industries,
research
and development, joint ventures, arts, cinema, theatre, music and recording
industry, the self employed, contests, and new talents.
Assistance from and for foundations and associations, guidance to prepare a
business plan, market surveys, computers, and much more!

The Canadian Subsidy Directory is sold $ 49.95, to obtain a copy please
call one of the following distributors:

Canadian Business Ressource Center: (250)381-4822, 8am-4pm pacific time.
Fureteur bookstore: (450)465-5597 Fax (450)465-8144 (credit card orders
only).





From ken at hotmail.com  Sun Jun  2 01:41:33 2002
From: ken at hotmail.com (Ken)
Date: Sun, 2 Jun 2002 15:41:33 +1000
Subject: What does sys.exit(1) means?
Message-ID: <adcb22$100q9d$1@ID-49758.news.dfncis.de>

What does sys.exit(1) means?

Thanks





From kragen at pobox.com  Mon Jun 10 08:05:42 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 08:05:42 -0400
Subject: Looking for a list subclassing example...
References: <adi4dv$118tko$1@ID-146704.news.dfncis.de> <83bsaj4kgn.fsf@panacea.canonical.org> <3D04684C.BA67E64F@alcyone.com>
Message-ID: <83r8jf1gq1.fsf@panacea.canonical.org>

Erik Max Francis <max at alcyone.com> writes:
> Kragen Sitaker wrote:
> > Why do you think it's a good idea to subclass 'list'?
> > 
> > I don't have one to show, but I think it's generally not a good idea.
> 
> Why do you think it was made possible in 2.2, then?

As an experiment.  In October, Guido said:

    Subclassing list and dictionary etc. should be seen as an experimental
    feature; I don't want to fully fix the semantics yet in all cases.
     . . .
    Subclassing a built-in type is appropriate when either (a) you want to
    use it in a context where a genuine list/dictionary/file/etc.; or (b)
    you want the speed advantage of the built-in type.  In both cases you
    have to live with some restrictions.  Remapping the fundamental
    accessors (like __getitem__) is probably not a good idea in either
    case.  Adding new state and behavior is fine.

[from
<http://groups.google.com/groups?selm=mailman.1004342434.4625.python-list%40python.org>]

I'm tempted to be flippant and say "to confuse new users" or "as a
sick joke", but as Guido points out, there are at least two valid
reasons to do it: to optimize your code and to work around breakage
--- well, let's say inconsistent, inflexible behavior --- caused by
other people's optimizations.



From duncan at NOSPAMrcp.co.uk  Thu Jun 13 04:29:53 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Thu, 13 Jun 2002 08:29:53 +0000 (UTC)
Subject: Python Virtual File System: A spike test
References: <mailman.1023872902.2540.python-list@python.org> <Xns922B70C12302duncanrcpcouk@127.0.0.1> <mailman.1023900618.26500.python-list@python.org>
Message-ID: <Xns922C60466D172duncanrcpcouk@127.0.0.1>

"Mike C. Fletcher" <mcfletch at rogers.com> wrote in
news:mailman.1023900618.26500.python-list at python.org: 
> Equality-testing is useful for:
<various good answers snipped>

I think the important thing is that all these scenarios use equality to 
optimise something, so it isn't the end of the world if they occasionally 
fail to detect equality, although it could be nasty if they returned two 
paths as equal when they were actually different.

> it definitely is tricky, but it's one of the major things I want in
> the package (and, after all, if there's no tricky code in the system,
> why is someone going to use it instead of rolling their own).  I wind
> up with code to
> do this stuff all over my packages otherwise, I'd like to localise it 
> somewhere.
> 
Oh good, you want to do some tricky things. :-)


> Will need to add special-casing for "local machine" unc, though
> I don't see how to do it with cases other than \\.\ given that the
> local machine name isn't immediately available as far as I know.
I'm not sure I have even seen any good documentation that completely 
specifies UNC filenames on windows. You may have to handle \\? as well as 
\\., if you want completeness, but I suspect the law of diminishing returns 
kicks in somewhere.

IF you have the win32 extensions around you can always call 
win32net.NetWkstaGetInfo to get the machine name, although I must admit 
that the last time I wanted to get the machine name I used os.popen('NET 
NAME') and parsed the output. Another option would be to read the 
appropriate registry keys, which doesn't require the win32 stuff.


-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From zaka07 at hotmail.com  Wed Jun 19 06:33:32 2002
From: zaka07 at hotmail.com (Rad)
Date: 19 Jun 2002 03:33:32 -0700
Subject: CPU Usage Windows
Message-ID: <ad381f5b.0206190233.4fa1bf58@posting.google.com>

Hi,

I've noticed that when I run Python script on 
Windows 2k Pro
Dual Intel Xeon (2.2GHz)
CPU Usage for pythonw.exe is only 50%?!

Nothing else is running on the machine (apart form the OS). Some of
the python programs are running for hours (large data sets) and even
though python shell freeze's (doesn't refresh) CPU usage is still only
50%.

System Idle Process is about 40%-50%

Any ideas why is this and can I do anything about it?

Is it something to do with python not being able to use the power of
both processors?

thanks,


From brueckd at tbye.com  Wed Jun 26 11:17:41 2002
From: brueckd at tbye.com (brueckd at tbye.com)
Date: Wed, 26 Jun 2002 08:17:41 -0700 (PDT)
Subject: Installation problems DCOracle2 & W2000
In-Reply-To: <talon34-C09E81.19332125062002@netnews.attbi.com>
Message-ID: <Pine.LNX.4.44.0206260810420.29286-100000@ns0.tbye.com>

On Wed, 26 Jun 2002, Talon wrote:

> I am having a heck of a time getting DCOracle2 to funcion properly.  I 
> am running on a W2K machine, python2.1.3 and Zope 2.5.1, both of which 
> run just fine.  I installed DCOracle2 and the ZOracleDA under Zope and 
> it works fine.  But standalone python scripts don't see the dco2.py 
> module.  So I tried to install DCOracle2 in the python21/lib folder.  I 
> got that latest version from Zope.org and it said it installed properly.  
> But it doesn't run.  If you try to import it into a script, you get 
> complaints that it can't find this or that module, all of which are part 
> of the Zope path.

Hi Talon,

Some of our development machines use the same configuration you list 
above. You're probably _very_ close, so I'd stick with DCOracle2 if you 
can. Would you mind posting the specific error messages you're getting? 
Also, you mentioned that standalone scripts don't work, but just to be 
clear, using DCOracle2 in Zope _does_ work without problems? (your 
connection object is open, you can browse the database, you can run test 
queries, ZSQL methods, etc.)

If it's working for Zope, then standalone scripts should be able to use
DCO2 if you include the right directory in your PYTHONPATH environment
variable, so please tell us what your PYTHONPATH is too.

Good luck,
Dave





From idbaxter at semdesigns.com  Sat Jun  1 09:49:19 2002
From: idbaxter at semdesigns.com (Ira Baxter)
Date: Sat, 1 Jun 2002 08:49:19 -0500
Subject: symbolic python
References: <3CEFAE4C.38A8022B@doc.ic.ac.uk> <3CEFD881.D7992095@doc.ic.ac.uk>
Message-ID: <3cf8ce86$1@giga.realtime.net>

If you want program transformation capability,
you can get it in the form the of DMS Software Reengineering
Toolkit.
See http://www.semdesigns.com/Products/DMS/DMSToolkit.html.


--
Ira Baxter, Ph.D. CTO Semantic Designs
www.semdesigns.com  512-250-1018


"Benjamin Tai" <bt98 at doc.ic.ac.uk> wrote in message
news:3CEFD881.D7992095 at doc.ic.ac.uk...
> > Hi, Benjamin.  Not as you say it.
> >
> > However, I often met needs that could have been resolved as you say in
> > some other languages, and that were easily addressed using more Pythonic
> > paradigms.  My belief if that Python is very flexible for such problems,
> > if you accept to "negotiate" with the language.
> >
> > If you explain a bit more precisely what is the problem you are trying
to
> > solve, it is likely that many members of the Python list will have ideas
> > or solutions to offer, that you might find workable and even elegant.
> > This is my guess, do not take it as a promise :-).
> >
> I am afriad it is only simple (stupid + abstract) question out of
> cuirosity. It does sound like some simple form of program transformation
> (although I am not try to achieve any optimisation or static checking
> here).
>
>
>
> Example 1):
>
> Could it be possible to have a parser to rewrite my derived class:
>
> class base:
>     def b_fun(self):
>         pass
>
>     def overloaded(self):
>         print "base"
>
>
> class derived(base):
>     def overloaded(self):
>         print "derived"
>
>
>
> into something like:
>
> class derived(base):
>     def b_fun(self):
>         pass
>
>     def overloaded(self):
>         print "derived"
>
>
> such that methods inherited from the base class is also written inside
> the derived class.
>
>
>
>
> Example 2):
>
> Could it be possible to have a parser to rewrite my derived class:
>
> class base:
>     def b_fun(self):
>         pass
>
> class derived(base):
>     def d_fun(self):
>         self.b_fun()
>
>
>
> into something like:
>
> class derived(base):
>     def d_fun(self):
>         self.b_fun()
>
>     def b_fun(self):
>         pass
>
>
> such that methods called from the base class is also written inside the
> derived class.
>
>
>
>
>
> Simply speaking, I can look up a module to find out name of the
> functions located inside. Could I also find out their function body?
> I can get Python to evaluate the function, but is it possible to return
> the source code of the funciton body at the same time?
>
> In a symbolic language, I would imagine there should be direct access
> towards the body of the function. Could it be possible for Python to
> have a similar behaviour?
>
>
>
> Thanks
>
> Ben
>




From pyth at devel.trillke.net  Wed Jun  5 20:28:15 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Thu, 6 Jun 2002 02:28:15 +0200
Subject: self
In-Reply-To: <sWsL8.82$v_7.4260962@newnews.cc.stevens-tech.edu>; from vjovanov@stevens-tech.edu on Wed, Jun 05, 2002 at 02:40:45PM -0400
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org> <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu> <vsqL8.279692$%u2.233541@atlpnn01.usenetserver.com> <mailman.1023294728.2322.python-list@python.org> <sWsL8.82$v_7.4260962@newnews.cc.stevens-tech.edu>
Message-ID: <20020606022815.F30696@prim.han.de>

Vojin Jovanovic wrote:
> [me]
> > The next best solution IMO is to preprend a namescope like 'self.'
> > to every name in an expression.
> > But you may have to parse the expression to
> > do this.
> 
> I have been thinking about this for some time, and realized that there some
> tricky cases that one has to deal with, and I am not sure that the whole
> business of appending self. will be error free in every case. 
> For example in 'x + Geometry.x' the correct result should
> be 'self.x + self.Geometry.x' where Geometry is an instance of Geometry
> class that is bound to self.Geometry.

with a small modification to my previously posted class you *can* do this:

class Evaluator:
    def __init__(self, eqs={}):
        self.__dict__['equations']=eqs
    def __getattr__(self, name):
        value = self.equations[name]
        if type(value)!=str:
            return value   # if it's not a string, just give it back
        scope={}
        while 1:
            try:
                return eval(value, scope)
            except Namerror, n:
                var=str(n).split("'")[1]
                scope[var]=getattr(self, var)
    def __setattr__(self, name, value):
        self.equations[name]=value

class Berta:
    a=5

>>> e=Evaluator({'a':'3'])
>>> e.berta=Berta()
>>> e.g='berta.a*a'
>>> e.g
15

This is hopefully what you want. This version doesn't detect 
cycles in equation definitions, though. Shouldn't be too difficult 
to add but requires some more thought. ask again if you
are having problems with it.

have fun,

    holger




From mdtorre at freemail.it  Thu Jun 13 05:30:38 2002
From: mdtorre at freemail.it (Matteo)
Date: 13 Jun 2002 02:30:38 -0700
Subject: What language - platform use to data acquisiton/numerical work?
Message-ID: <1c849ea9.0206130130.6cdaa36f@posting.google.com>

I'm a programmer, I've worked with C, Java, Perl, Matlab, PHP, and a
bit of almost every widespread language. I love OO programming. I have
recently doscovered Python thanks to the great enthusuasm of its
supporters. I love Java for its elegance and OO-ness, and Perl for its
raw power. I don't like C++ for its long developng time and complexity
and hate Visual C++ for being the worst (most complex, most slow, most
... ugly) develop environment on the earth.

I have to do (from scratch and alone) a work without any language
restriction. The system will have to work under MSWin2000 and maybe
that, for stability, we will port it to unix-likes.

The work involve:
- a system of data acquisition (and control) from (of) a PLC via
serial port. The data acquisition is in real time with low frequency
(about 1 - 2 hz)
- a big system of data analisys, pattern recognition, data
visualization and heavy user interaction mostly for experiment with
algorithms and for diagnostics
- a end user system, mostly a wrapping of siplified GUI to the
previous two subsystems.

I think Python would be suitable for my work. What do you think?

I think I will handle serial port comm vith small C libraries, I don't
know if Python is suited to the problem.
Having the base drivers should not be a problem ro realize the data
acquisition system in Python Or not?

The data analisys numerical procedures have to be done in matlab due
to their complexity and specialization, and I fear even the plotting
will have to be done in Matlab. I have seen there are some bridges
from Phyton to Matlab, how they are?
My problem here is where to start developing in Pyyhon and where to
stop in ML. ML has a very simple and good sistem of building GUIs, how
is Pithon?

Finallly I hope to develop the wrapping GUI layer in Phyton. Again,
has Python some simple and good GUI support? I've seen that the most
widely used GUI module is something in tcl/tk (which I don't know at
all). Considering that I don't know python nor tcl/tk, how is the
learnig and developing curve?

Thank you all.

Matteo Della Torre


From b_mcerlean at yahoo.com  Fri Jun  7 10:10:59 2002
From: b_mcerlean at yahoo.com (Brian McErlean)
Date: 7 Jun 2002 07:10:59 -0700
Subject: Efficient python programming...
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> 	<20020606215349.GA13474@node1.opengeometry.net> <mailman.1023401650.18434.python-list@python.org>
Message-ID: <f040c4f9.0206070610.7de92dcb@posting.google.com>

gabor <gabor at realtime.sk> wrote in message news:<mailman.1023401650.18434.python-list at python.org>...
> 
> for file in directory:
> 	file.printsize()
> 
> 
> ran = range(len(directory))
> for r in ran:
> 	directory[r].printsize()
> 
> [file.printsize() for file in directory]
> 
> they're the same, and everyone would recommend you the last one..

I wouldn't.  The first one is better in this case.  Using the list
comprehension implys to me that you're interested in the result -
while in fact you're discarding it.  If there are a lot of files in
the directory (And after all, thats the only reason to optimise it),
then you'll end up with a large list of [None,None,None,...] that is
completely unneeded.

testing gives:

[foo(i) for i in range(100000)]
- took 0.240000009537 secs

for i in range(100000): foo(i)
- took 0.149999976158 secs

(foo(i) is just an empty function)

Using the list comprehension here is both misleading (because you are
relying on a side-effect, not using the returned list), and less
efficient.  Using the for loop is better, because it makes clear that
you're using an imperative technique, relying on the side-effect
(printing the size) and are uninterested in building a list at all.

On the other hand, if the code is:

l=[]
for file in directory:
 	l.append(file.getsize())

It is both clearer, and more efficient if implemented as:
l= [file.getsize() for file in directory]

Brian.


From peter at engcorp.com  Mon Jun 17 20:56:57 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 20:56:57 -0400
Subject: Zip makes sense! was: Re: why not "'in' in 'in'"?
References: <mailman.1024086417.5966.python-list@python.org> <ugkpmkj4f0sid4@news.supernews.com> <aej16d$t97$1@news.idiom.com>
Message-ID: <3D0E8559.118C82CE@engcorp.com>

"James T. Dennis" wrote:
> 
>  More over the term "yield" in English has a few distinct senses.

I'd argue all the senses you've presented are extremely similar,
perhaps identical in semantic meaning, but with distinct contexts 
which make them look a little different.

>  One (slightly archaic) is the imperative to "yield" to a conqueror
>  or authority as in:
>         "Yield, Sir knight!"
>  Another is a similar and common traffic imperative meaning to defer
>  to the rights-of-way for others (probably the inspiration for the
>  most common programming semantics associated with 'yield').
>  Another is to refer to the product of a farm or manufacturing process
>  as in: the yield of the corn field last year was X bushells.
> 
>  It is clearly this last sense which inspires Python's use of the
>  term.  The return operator "yields" a result *and* terminates an
>  execution contect.  The yield operator also "yields" a result but
>  it leaves the production machinery or the "field" available for
>  future results.  

In exactly the same manner, if the knight yields he is still
available for future results, such as another fight or, more
usefully, joining the conqueror's own crusade.  In all cases
here, it looks like yielding involves producing an output
of value while preserving the producer intact for future
results.

Were 't not so, the conqueror would but cry "Die, Sir Knight!"
and despatch the hapless knave forthwith.  ;-)

-Peter


From gustav at morpheus.demon.co.uk  Fri Jun  7 09:41:10 2002
From: gustav at morpheus.demon.co.uk (Paul Moore)
Date: Fri, 7 Jun 2002 15:41:10 +0200 
Subject: Gordon McMillan's Installer - Data files in the archive
References: <y9e0xy2x.fsf@morpheus.demon.co.uk> <uofewjgv4.fsf@ctwd0143.fitlinxx.com>
 <lm9sl8pw.fsf@morpheus.demon.co.uk>
 <Xns92265D7BF8D4Dgmcmhypernetcom@199.171.54.216>
Message-ID: <elfj8ave.fsf@morpheus.demon.co.uk>

Gordon McMillan <gmcm at hypernet.com> writes:

> [posted and mailed]
>
> Paul Moore wrote:
>
> [snip]
>
>> That worked remarkably easily. A trivial example:
>> 
>> from carchive import CArchive
>> me = CArchive(sys.executable)
>> readme_file = me.extract('README')[1]
>> 
>> One thought occurs - how "supported" is the CArchive class interface?
>
> I expect advanced users to do things like this.
> I also expect them to manipulate TOCs and put
> extra processing in their spec files. Installer's
> only a black box if you don't open it :-).

Fair enough. That sounds like a reasonable approach.

Of course, whether *I* should do this boils down to whether I consider
myself an "advanced user", or a "dumb blunderer who shouldn't be
allowed near a computer" :-)

pandora-opened-a-black-box-and-look-what-happened'ly y'rs
Paul.



From bt98 at doc.ic.ac.uk  Sat Jun  8 11:47:51 2002
From: bt98 at doc.ic.ac.uk (Benjamin Tai)
Date: Sat, 08 Jun 2002 16:47:51 +0100
Subject: PyCObject subtype PyObject
Message-ID: <3D022727.2147C50E@doc.ic.ac.uk>

Hi,


Any comments to clarify my confusion would be appreciated.

In Python/C API Reference Manual, section 7.5.9 CObjects,  the document
mentions that:

"PyCObject This subtype of PyObject represents an opaque value, ..."



1) What does it mean by "subtype"? As a C interface, is the document
talking about two structs which share some common fields?


2) For function "PyCObject_FromVoidPtr()", it takes in a "void*" as the
first argument.
For function "PyCObject_AsVoidPtr()", it returns a "void*".
When implementing Python extension, what is the potential danger of
casting pointers, of different size, to and from "void*"?


3) What if I can always determine the original type of the pointer?


4) Apart from exporting function between modules, are there any examples
of using PyCObject?



Thanks

Ben



From phawkins at connact.com  Fri Jun 21 19:01:13 2002
From: phawkins at connact.com (Patricia J. Hawkins)
Date: 21 Jun 2002 19:01:13 -0400
Subject: Tkinter and threads
References: <3D0DCC8F.9A6D3B81@ipm.fhg.de>
Message-ID: <wk8z58w81y.fsf@mail.connact.com>

See Aahz's FibThreaded.py example that accompanies his slides for his
thread talk.

http://starship.python.net/crew/aahz/OSCON2001/index.html

The whole thing is worth reading, but beware, it's really set up for a
speech, not reading -- it's 84 pages with +/- 10 lines of text per
page, and a lot of places where Aahz probably says a whole lot to the
audience, but all you get is "Share external objects: Files, GUI, DB
connections DON'T partial exception: print"

so, um, why is print a _partial_ exception?

Also be aware that the early code examples aren't intended as models,
just as teaching tools -- you don't hit the really useful stuff till
page 48. And the format makes it tough to flip back and forth (not to
mention, tiny amounts of information per page drives me nuts, but
perhaps that's just me) -- and at 84 pages, I'm not going to print it
out to take it downstairs to an armchair and mark it up, which is what
I want to do.  And skimming for what you don't know already is
impossible.

Whinging aside, it's got a lot of meat, and useful solutions for a
number of problems.

-- 
Patricia J. Hawkins
Hawkins Internet Applications, LLC





From kragen at pobox.com  Sun Jun  2 22:51:31 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 02 Jun 2002 22:51:31 -0400
Subject: HTML links in a GUI tkinter
References: <mailman.1022986323.16196.python-list@python.org> <ade2tq$3q4$1@newsg1.svr.pol.co.uk>
Message-ID: <83k7phrs5o.fsf@panacea.canonical.org>

"G. Willoughby" <never at mind.info> writes:
> import os
> os.startfile(page.html)
> 
> to execute the html file.

os.startfile is nonportable.



From martin at v.loewis.de  Wed Jun  5 16:51:32 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 05 Jun 2002 22:51:32 +0200
Subject: Freeze questions
References: <mailman.1023297729.24131.python-list@python.org>
Message-ID: <m3y9dtmotn.fsf@mira.informatik.hu-berlin.de>

"Sung H. Kim" <skim at adobe.com> writes:

> Can the Freeze tool be used with C libraries? i.e. extensions?

Sure. I recommend that you don't build the extensions as shared
libraries, but as static object files - that allows you to incorporate
them into the binary. The frozen binary cannot incorporate shared
libraries - but it can still use them.

> Will it create a truly standalone app (on Windows)?

Not in the default configuration. If you build python as a static
library (i.e. pythonxy.lib instead of pythonxy.dll), then you can
create truly standalone binaries. It requires some manual work to
create such a build process, though.

> Does it support tk/tkinter apps yet?

Depends on what you mean by "support". You can certainly freeze
Tkinter applications, but that won't incorporate the Tcl libraries
into the executable. To achieve that, you need
- static versions of the Tcl libraries,
- Tix
- Tix Stand-Alone Modules (SAM)
If you build Tix with SAM support, you get truly standalone Tkinter
binaries. I did that five years ago, but haven't tried since.

> Finally, is there a good how-to on the Windows platform?

Sure. The Microsoft online documentation (press Start, then Help)
includes quite reasonable Howtos for using Windows :-)

Regards,
Martin



From martin at v.loewis.de  Thu Jun  6 03:14:37 2002
From: martin at v.loewis.de (Martin v. Loewis)
Date: 06 Jun 2002 09:14:37 +0200
Subject: minidom and encoding problem
References: <17aafe08.0206051514.4af054d2@posting.google.com>
Message-ID: <m37klcubdu.fsf@mira.informatik.hu-berlin.de>

ehab_teima at hotmail.com (Ehab Teima) writes:

> I'm using Python 2.1. I wrote classes to create xml document from
> scratch. The code worked fine until I hit an encoding problem. The
> classes can read text and insert it as is to xml document using
> creatTextNode. This text had characters > 127, and I got this error.

This is a bug in your code. You must not insert (byte) string in a DOM
tree; always use Unicode objects.

> I know it's not possible to add an enconding attribute using writexml,
> so the generated document only has <?xml version="1.0"?>. Is there any
> way to get around this problem. 

Yes. Use Unicode strings when creating text nodes. When producing the
serialized document through .toxml, you will find that it produces a
Unicode string. Since (as you notice) the document has no encoding
declaration, you need to .encode("UTF-8") that string before saving it
into a file.

> Does any body know how to get the rootnode of a document? If I know
> the root node, I can add the proper header and then write the root
> node using writexml.

The document element is available through .documentElement on the
Document.

Regards,
Martin



From max at alcyone.com  Sat Jun 15 05:03:43 2002
From: max at alcyone.com (Erik Max Francis)
Date: Sat, 15 Jun 2002 02:03:43 -0700
Subject: Newbie deluxe
References: <hdDO8.29984$Hn4.969427@news1.east.cox.net>
Message-ID: <3D0B02EF.AA1AB471@alcyone.com>

Zach wrote:

> I've just started taking a look at Python (experienced in Delphi) and
> so far
> it looks very clean and comprehensible.  My question is, what are the
> fundamental differences and similarities between it and Perl?  I hear
> so
> much about these two languages, especially in Linux circles and just
> wondered what sets them apart.

	http://www.python.org/doc/Comparisons.html#perl

> Can full fledged graphical applications be
> built with Python or does it lean more towards scripting?

Of course.

-- 
 Erik Max Francis / max at alcyone.com / http://www.alcyone.com/max/
 __ San Jose, CA, US / 37 20 N 121 53 W / ICQ16063900 / &tSftDotIotE
/  \ Who'd ever think it / Such a squalid little ending
\__/ The American and Florence, _Chess_
    Church / http://www.alcyone.com/pyos/church/
 A lambda calculus explorer in Python.


From tim at vegeta.ath.cx  Thu Jun 13 05:13:13 2002
From: tim at vegeta.ath.cx (Tim Hammerquist)
Date: Thu, 13 Jun 2002 09:13:13 GMT
Subject: How to check version of Python under CGI
References: <rbmggus7ln1uiutktf5libf4bbngu29ji7@4ax.com>
Message-ID: <slrnaggonf.lfo.tim@vegeta.ath.cx>

Johann graced us by uttering:
> I would like to write a simple CGI script to get as much info about
> installed Python as possible. But I even don't know how to check
> version of Python. :( From shell command  it is easy, but how to do it
> from CGI script? Eg.
> 
> #!/usr/bin/env python
> print "Content-type: text/html\n\n"
> print "Python version: %s" %s (IDoNotKnowWhatToChoose(),)

import sys

print sys.version       # version info in string
print sys.version_info  # version info in tuple

HTH
Tim Hammerquist
-- 
Tools, of course, can be the subtlest of traps.
    -- Daniel, The Sandman


From nomailhere at lookitup.yourself.com  Sat Jun 22 12:01:01 2002
From: nomailhere at lookitup.yourself.com (Frihtiof Andreas Jensen)
Date: Sat, 22 Jun 2002 18:01:01 +0200
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at>
Message-ID: <3d149f19$0$53133$edfadb0f@dspool01.news.tele.dk>

"Siegfried Gonzi" <siegfried.gonzi at kfunigraz.ac.at> skrev i en meddelelse
news:3D1482AE.8FA299D8 at kfunigraz.ac.at...
> A few weeks ago I complaint here that Python on Windows XP sucks. I had
> a problem which would took theoretically about 1.5 hours; but in reality
> it takes 6 hours on Windows XP and Python 2.2. After that 6 hours I must
> reboot my system in order to work on.

So, memory is leaking away somewhere...

> In the following I got the advice to install a real user system on my
> laptop. Okay, I bought Linux SuSE 8.0 and installed it on my laptop (256
> MB RAM, 1000 MHz Celeron). I also installed Python 2.2 (the ActivePython
> version, otherwise I couldn't install idle due to the annoying
> "free_software_garbage_2_weeks_installation_nightmare").

If you think the software others have spent time to develop for you to use
as you please is "garbage" then why do you still use it and even whine about
it to boot?

I have no problem whatever installing the standard Python 2.2.x installs on
ANY flavour of windows - even my win2k corporate install with restricted
access is dealt with nicely by the install! I doubt many others had
problems...it is about as easy as it possible gets.


> My conclusion: I will rewrite my simulation in Fortran 90 and will
> abadon Python. Python is maybe good for some small scripts but not
> appropiate for serious programming.
>

*My conclusion: Lets look for Common Factors here: Python builds are
*different* on Linux and WinXP, GUI's are *different*, in fact the common
factors are YOU and YOUR Program....which leads me to believe that you have
made a mistake, in your code.

> Thank you Billy Boy for a forgiving user system Windows XP which in turn
> relieves Python's "wreak havoc" behavior.

Then write your software for MSC++ or C# - it is fine by me.

> I cannot believe that Linux is responsible for Python's ill behavior
> under Linux. Before you start ranting about note the following: In a
> first step it has nothing to do with my programming style (admittedly
> not very good and famous) it has to do that the same calculation cannot
> be performed on Linux and Python 2.2!

I think it has something to do with your coding style and everything to do
with your attitude - you are using tools, which are working well for
thousands of developers, yet when they do not work for you, it is the fault
of the tool and not because you are doing something wrong! It is the
hallmark of the neophyte programmer to consider oneself capable of *knowing*
that ones code is flawless.

Do you also blame the hammer, when you hit your thumb rather than the nail?

Then you come out and slag off the very people who might be able to help,
had you bothered to post a proper description of the problem, perhaps with
some sample code, thus ensuring that nobody will care at all whether it
really is a bug within Python, or not, just as long as it only affects your
application.

My bet is that once you cool off and sits down with some debugging tools
like GDB, Boundschecker or Electric Fence you will indeed find the
programming error and fix it.

My bet is also that you will not tell anyone either.




From jonathan at onegoodidea.com  Fri Jun 28 16:20:26 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Fri, 28 Jun 2002 21:20:26 +0100
Subject: testing type of an object
References: <pan.2002.06.28.14.49.33.858327.1787@presidency.com> <afibp0$ekn25$1@ID-76829.news.dfncis.de>
Message-ID: <B942839A.D235%jonathan@onegoodidea.com>

On 28/6/2002 20:00, in article afibp0$ekn25$1 at ID-76829.news.dfncis.de,
"Russell Blau" <russblau at hotmail.com> wrote:

> You can use
> 
> if type(s) == type('abc'):
>   do something
> 
> or, even better, import the types module (which you can look up in the
> docs).

Or, in Python 2.2, it's preferable to compare against the new type
constructor builtins:

>>> type('foo') is str
1
>>> type(5) is int
1
>>> 

or:

>>> isinstance( 'foo', str )
1
>>> isinstance( 5, int )
1
>>> 

The other ones are: 'dict', 'long', 'list', and 'unicode'.

Jonathan



From gmcm at hypernet.com  Wed Jun 26 08:20:51 2002
From: gmcm at hypernet.com (Gordon McMillan)
Date: 26 Jun 2002 12:20:51 GMT
Subject: import precedence
References: <mailman.1025011774.28523.python-list@python.org>
Message-ID: <Xns92395501ED461gmcmhypernetcom@199.171.54.215>

Mark McEahern wrote:

[snip]

> Can I just assume that:
> 
>   from spam import foo
> 
> will only import the foo.py module if the foo.dll and foo.so are not
> found...?

No. On any given installation, you could set things up
so it would work, but it would be fragile and when it
broke, it would be very mysterious.

--Gordon
http://www.mcmillan-inc.com/


From syver-en+usenet at online.no  Wed Jun 26 21:33:08 2002
From: syver-en+usenet at online.no (Syver Enstad)
Date: Thu, 27 Jun 2002 01:33:08 GMT
Subject: python mode for emacs
References: <u660cfwo3.fsf@online.no> <af008d$ft1$1@newsreader.mailgate.org>
Message-ID: <uvg85ecb7.fsf@online.no>

Alberto Griggio <albgrig at tiscalinet.it> writes:

> On Fri, 21 Jun 2002 18:02:14 +0200, Syver Enstad wrote: 
>> Is there a straight forward way of deciding how the 
>> py-pseudo-keyword-face
>> should look?



> (defun my-py-mode-hook ()
>   (set-face-foreground 'py-pseudo-keyword-face "CadetBlue")
>   ;; ... and anything else you want
>   )
> (add-hook 'python-mode-hook 'my-py-mode-hook)
> 
> Bye,
> Alberto

Thanks, I changed it to "Black" and (make-face-italic
'py-pseudo-keyword-face)

Very nice, just gives self and the other pseudo
keywords a little emphasis without being too "loud". 

-- 

Vennlig hilsen 

Syver Enstad


From eddie at holyrood.ed.ac.uk  Thu Jun 13 09:53:22 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Thu, 13 Jun 2002 13:53:22 +0000 (UTC)
Subject: parse tar file with python
References: <ae7498$4jven$1@ID-146704.news.dfncis.de> <ae7o6s$4oh4v$1@ID-50400.news.dfncis.de> <aea0hs$54qq4$1@ID-146704.news.dfncis.de>
Message-ID: <aea84i$36b$1@scotsman.ed.ac.uk>

"Shagshag13" <shagshag13 at yahoo.fr> writes:

>In fact, i don't want to tar/untar files, and especially not in main memory !

>I wish i could read the tar-ed file line by line (f.readline) and be able to check when i find the beginning of an "inside file" and
>get some info about it like name, how and so on... (that's because my original files are plain text file, and i think that tar will
>let them unchanged)

>In my tar file there is, for example, a kind of separator like this (but with everything in one long line) with :

>shag.py_0100744_0002033_0001750_00000004414_07500237361_0015314_0_ustar_00_shagshag_user_0000040_0000417_beginofmyfilehere

>where _ stand for a variable amount of another ascii code that i can't cut/paste...

>Do you kwow what each means (for example the first one is undoubtly the file name, but then...) ?
>And what are the fixed position of each of theses ?
>Have a clue ?

A google search on 'tar file format directory header size' came up with this:

  http://www.mkssoftware.com/docs/man4/tar.4.asp

I did write a tar program in Bliss10 many years (nay decades) ago and recall
it was quite easy to deal with.

Using tar is only useful if you know for certain that you're going to process
each file in exact sequence.  I think other solutions like making smaller
directories would be easier.

Surely if all you're doing is reading all files in sequence you can just
concatenate them into one?, if you need to know where each ends add a unique
seperator.

Eddie


From sholden at holdenweb.com  Fri Jun  7 16:48:37 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Fri, 7 Jun 2002 16:48:37 -0400
Subject: Creating Dynamic Web Pages (ZOPE)
References: <46ca81a0.0206062016.789d3af6@posting.google.com>
Message-ID: <J29M8.358696$%u2.156726@atlpnn01.usenetserver.com>

"Hugh Cowan" <hwcowan at hotmail.com> wrote ...
> Hello,
>
> I am not exactly sure where to post this question -- the only
> reference that I could find for ZOPE in the Newsgroups is here.  I
> appologize if this is not the correct place.
>
> I have the situation where I need to create a simple Intranet for my
> company to display dynamic / data driven web-pages.  We are running a
> Windows NT network with MS-SQL Server and IIS.
>
> I had originally thought of using ASP (Server Side Scripting) to
> generate the dynamic web-pages.  While I know how to create the
> Intranet using ASP, I have yet to find any RAD type Web-Tools that
> will allow me to quickly and easily generate the ASP pages required,
> and I don't fancy using the Notepad approach either.
>
> My scenario is typical where I need to put together a company Intranet
> with a few front end Web-Pages where users can select information,
> various links, and enter search criteria.  The information is then
> retrieved from the MS-SQL Server and the results are sent back to the
> user.  It's nothing fancy, no shopping carts, or remembering sessions
> states, etc.. just really a simple front-end interface for
> non-technical people to easily retrieve information.
>
> I don't know if ZOPE would fit my situation or not.  I know that it is
> / can be used to create complex E-commerce sites, but I am wondering
> if it is overkill for what I need done.  Would I be better off trying
> to find some sort of RAD Web Tool instead (anyone have any good
> suggestions) ?
>

At the risk of sounding like a heretic...

I recently started to use DreamWeaver UltraDev (one of my clients mandated
its use) and I've been relatively impressed. For a really small
database-driven web created in DreamWeaver take a look at

    http://www.holdenweb.com/TinySite/

This web has precisely four pages, and I built it to explain to beginners
how database contents can affect a site's navigation as well as its readable
content.

DreamWeaver MX is now out, and targets several more backends. It would be
nice if it had one that drove some suitable Python framework -- possibly
WebWare.

If you are already familiar with ASP, this will get you up to speed on your
tasks much more quickly than learning Zope. Won't be as much fun, though.

regards
 Steve
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From eddie at holyrood.ed.ac.uk  Fri Jun 14 13:17:21 2002
From: eddie at holyrood.ed.ac.uk (Eddie Corns)
Date: Fri, 14 Jun 2002 17:17:21 +0000 (UTC)
Subject: I feel stoopid ... this code is to verbose
References: <3D0A1791.3060307@mxm.dk>
Message-ID: <aed8f1$biu$1@scotsman.ed.ac.uk>

Max M <maxm at mxm.dk> writes:

>Hmm ... I am working on a problem. In Danish we have a number format 
>that looks like:

>42.000.000,00 which is 42 millions

>def stringSlicer(string, chunkSize=3):
>     chunkList = []
>     reverseString = list(string)
>     reverseString.reverse()
>     for i in range(0, len(string), chunkSize):
>         chunk = reverseString[i:i+chunkSize]
>         chunk.reverse()
>         chunk = ''.join(chunk)
>         chunkList.append(chunk)
>     chunkList.reverse()
>     return '.'.join(chunkList)


>I just find that it's a lot of code for such a little function an it 
>annoys my sense of aestetics. I have tried a lot of different approaches 
>including using zip on a list like ['','','.'], and other weird stuff :-)

Recursion and remembering that you can avoid reverse by chopping off the tail
end.

My contribution:

def commas (n,span=3,sep='.'):
    def chop (x, acc=[]):
        if not x: return acc
        return chop (x[:-span], [x[-span:]]+acc)
    return sep.join(chop(str(n)))

I called it that because that's what the Icon command is called (and I misread
the post first time through - eyesight must be going).

Eddie


From martin.stahl at telia.com  Sat Jun 29 13:00:16 2002
From: martin.stahl at telia.com (Martin Ståhl)
Date: Sat, 29 Jun 2002 17:00:16 GMT
Subject: Newbie that don't understand
Message-ID: <AIlT8.13914$p56.4478653@newsb.telia.net>

I have never been programming before and now I am reading the
"Non-Programmers Tutorial For Python"
http://www.honors.montana.edu/~jjc/easytut/easytut/node9.html#SECTION0091000
0000000000000
And don't understand the following
I quote

"To start off this chapter I am going to give you a example of what you
could do but shouldn't (so don't type it in):

a = 23
b = -23

if a < 0:
    a = -a

if b < 0:
    b = -b

if a == b:
    print "The absolute values of", a,"and",b,"are equal"
else:
    print "The absolute values of a and b are different"

with the output being:
The absolute values of 23 and 23 are equal"

how can 23 and -23 be equal?




From anton at vredegoor.doge.nl  Sat Jun 29 13:43:40 2002
From: anton at vredegoor.doge.nl (Anton Vredegoor)
Date: Sat, 29 Jun 2002 19:43:40 +0200
Subject: [MODULE] rubiks cube simulation for visual python
Message-ID: <afkrn2$obb$1@news.hccnet.nl>

A very basic rubiks cube simulation written for visual python.
Displays a 3D simulation of a rubiks cube. It is possible to view to cube from
all sides, the cube can be zoomed, the layers can be rotated
by typing single keystrokes and there is an undo function.

http://home.hccnet.nl/a.vredegoor/rubik/rubik.py

or (html-version)

http://home.hccnet.nl/a.vredegoor/rubik/rubik.html



From roelsch at comtrak.com  Tue Jun  4 16:44:58 2002
From: roelsch at comtrak.com (Robert L. Oelschlaeger)
Date: 4 Jun 2002 13:44:58 -0700
Subject: Version strings in executable created by Py2exe
References: <mailman.1023145033.32074.python-list@python.org>
Message-ID: <4323edf0.0206041244.7c9b9b7a@posting.google.com>

"Jimmy Retzlaff" <jimmy at retzlaff.com> wrote in message news:<mailman.1023145033.32074.python-list at python.org>...
> Probing a little deeper, it appears as if either - or   should work.
> From the part of distutils which parses setup.cfg:
> 
> opt = string.replace(opt, '-', ' ')
> 
> Here is a part of my setup.cfg (of course the word-wrapping is not in
> the file):
> 
<SNIP>
> 
> Jimmy

The problem was with the static name of the configuration
file: setup.cfg.

I am building several applications in the same directory
(e.g., demo4, demo5) and have created separate
demo4_setup.cfg and demo5_setup.cfg files (and corresponding
demo4_setup.py and demo5_setup.py files). These .cfg files
were not being used.

The problem has been sidestepped by having the makefile copy
the proper demo<x>_setup.cfg to setup.cfg, generating the
demo<x>.exe file, and then deleting setup.cfg.

Thanks for the help.

Bob


From carlca at dircon.co.uk  Thu Jun 13 18:16:51 2002
From: carlca at dircon.co.uk (Carl Caulkett)
Date: Thu, 13 Jun 2002 23:16:51 +0100
Subject: Python 2.2.1
References: <II7O8.622$e27.275217555@twister1.starband.net>
Message-ID: <2v4igu4640ru57rv43kjrf43vvbn4h6btr@4ax.com>

On Thu, 13 Jun 2002 20:58:58 GMT, "Dave" <talbotdr at starband.net>
wrote:

>    Hi, I am 12 years old and would like to begin learning Python over the summer. I have Windows 98 SE. I have python 2.2.1 and all of its utilities on my computer. I am taking a tutorial from  http://softwaredev.earthweb.com/sdopen/article/0,,12382_626311,00.html  If you go to it and scroll down to the paragraph entitled Getting Started there is a paragraph that says, "You will need to cause the directory containing the file named python.exe to be listed in your system environment variable named path. If you already know how to set the path, go ahead and do it.  If you don't already know how, you may need to get some help." 
>
>
>    How do I set the path and what should the name of the path be? All help and information would be appreciated. Thanks.:)

Hi Dave,

Step 1 ... First of all you need to make a note of where Python is
installed. If you used the default settings, I would think that it
will be in C:\Python22 or something similar. The file python.exe
should be in a sub-folder called "bin" underneath the C:\Python22
folder. So here you would note down C:\Python22\bin.

Step 2 ... (*** VERY IMPORTANT ***) The next step is to make a backup
of your autoexec.bat file (this will be in the root of your C: drive.
Just use Windows Explorer to make a copy of it.

Step 3 ... Then you will need to edit autoexec.bat by opening it in
Notepad. Find the line which says something like PATH
C:\WINDOWS;C:\WINDOWS\COMMAND. You then need to add the folder name
that you noted in step 1 to the end of this line. If the line already
ends with a ; character, then just add the folder name straight after.
If there is no ; then you need to add one to the end of the existing
path entry before adding the new folder name.

Step 4 ... Save the edited autoexec.bat and close Notepad. 

Step 5 ... Reboot the machine. 

Step 6 ... Once the PC has rebooted, you can check that the path
change is correct by doing Strat -> Run -> "command". This will bring
up a DOS prompt. Just type in the word PATH and <enter>. You should
see a line that shows your newly added Python folder at the end. Type
EXIT and <enter> to leave the DOS prompt. That's it - you should be
ready for Python now.

By the way, well done for choosing Python. It's a remarkable language
that will help you learn some good programming skills. Just as
importantly, it's great fun because you can start to see useful
results very quickly, which is a great way of motivating you to learn
even more. Good luck <g>.

P.S. You might want to check your newsgroup settings and see if you
can switch on the wordwrap option. Your message came through as one
continuous line.

-- 
Cheers,
Carl


From pyth at devel.trillke.net  Sat Jun 15 07:44:10 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sat, 15 Jun 2002 13:44:10 +0200
Subject: Pronouncing '__init__'
In-Reply-To: <20020615132519.A14452@hishome.net>; from oren-py-l@hishome.net on Sat, Jun 15, 2002 at 01:25:19PM +0300
References: <20020615132519.A14452@hishome.net>
Message-ID: <20020615134410.A15079@prim.han.de>

Oren Tirosh wrote:
> There has been a thread about this issue in the past. Below are some of the
> proposals people made.  I'd like to have an informal vote - how do *you*
> say Python's magic names when talking with other Pythoneers?
> 
>  (all my Python-related communication is written)
>  double underscore init double underscore
>  under under init under under
>  double bar init double bar
>  bar bar init bar bar
>  unders init unders
>  under init under
>  bar init bar
>  magic init
>  pyinit
> 
> Most of these pronounciations also have a shorter version with the trailing 
> part omitted.

for oral language:

    init constructor

and if the context allows

    init

for written languange __init__

Isn't it obvious? :-)

    holger




From dguo at ux1.cso.uiuc.edu  Thu Jun  6 15:26:26 2002
From: dguo at ux1.cso.uiuc.edu (Carolyn Guo)
Date: Thu, 6 Jun 2002 14:26:26 -0500
Subject: alignment excel written in Python
In-Reply-To: <admips$7a7$1@bob.news.rcn.net>
References: <6bd9f01b.0206050736.f469082@posting.google.com> <3CFE9861.5090200@skippinet.com.au> <Pine.GSO.4.31.0206051810190.22089-100000@staff1.cso.uiuc.edu> <admips$7a7$1@bob.news.rcn.net>
Message-ID: <Pine.GSO.4.31.0206061423570.2488-100000@staff3.cso.uiuc.edu>

Thank you very much. This time, it works well.
Could you please tell me where I can get the detailed reference about
writing Excel command from Python? I have worked out something like put
color, borders in cells.But still I need more detailed information on this
issue so that I can get a more clear picture of it.
Thanks again.
Carolyn

On Wed, 5 Jun 2002, Raymond Hettinger wrote:

> > I have a questions now: How can I deal with the alignment of Excel from
> > Python? for example, put the value in a cell with alignment right, or
> > center.
>
> xlLeft, xlRight, xlCenter = -4131, -4152, -4108
> from win32com.client.dynamic import Dispatch
> xl = Dispatch('Excel.Application')
> xl.Range('A1').HorizontalAlignment = xlRight
>
>
> Raymond Hettinger
>
>
>



From kragen at pobox.com  Mon Jun 10 17:49:21 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 17:49:21 -0400
Subject: Interrupting a continuous script
References: <_6JK8.1535$376.74360@newsfep1-win.server.ntli.net>
Message-ID: <83k7p6pzxa.fsf@panacea.canonical.org>

"Julian Gollop" <julian.gollop at ntlworld.com> writes:
> I have a few programs continually processing text files on a Linux machine.
> Occasionally I need to stop them, but I don't want to use a keyboard
> interrupt because it may cause the script to stop in the middle of doing
> something.
> Does anyone know a simple way to allow me to stop the process by pressing a
> key on the keyboard - without stopping inside something crucial?

You could use select() to see if stdin is readable.

My suggestion, though, would be to try to design the program so that
it can be stopped at any point without anything bad happening.  If the
machine gets unplugged, after all...

There are two general strategies for this: restartability, or
idempotence, and atomicity.

A restartable operation is one that, if interrupted in the middle and
restarted from the beginning, will produce the same results as if it
had been done once straight through.

An atomic operation is one that is either completed or not and cannot
be interrupted in the middle.  On Linux, renaming a file and creating
a directory are atomic operations.



From mgerrans at mindspring.com  Sun Jun  9 23:58:33 2002
From: mgerrans at mindspring.com (Matt Gerrans)
Date: Sun, 9 Jun 2002 20:58:33 -0700
Subject: How to capture stdout into a string?
References: <2YLM8.905$Yw3.119357@news20.bellglobal.com> <3D039285.72DFC71D@engcorp.com> <moMM8.91$Vr2.130316@news20.bellglobal.com> <3D039A73.210CDF7F@engcorp.com>
Message-ID: <ae185e$77j$1@slb3.atl.mindspring.net>

> :-)  Actually, I assumed you already had RTFM....
> Still helpful sometimes to read the source.

Maybe you guys can coin (or have coined) a new acronym: RTFS




From mfranklin1 at gatwick.westerngeco.slb.com  Wed Jun  5 09:31:41 2002
From: mfranklin1 at gatwick.westerngeco.slb.com (Martin Franklin)
Date: Wed, 5 Jun 2002 13:31:41 +0000
Subject: Fastest way to read / procsess / write text?
In-Reply-To: <3CFD5229.2060803@excite.com>
References: <3CFD5229.2060803@excite.com>
Message-ID: <200206051234.g55CYsB06651@helios.gatwick.geco-prakla.slb.com>

On Tuesday 04 Jun 2002 11:47 pm, you wrote:
> I'm somewhat new to python and programing in general and I trying to
> write a simple script that I can use to format large text files so that
> they are suitable for importing into MySQL. So far, I have came up with
> this:
>
> #!/usr/bin/python2
>
> import sys, re
>
> infile = sys.stdin
> data = infile.read()
> infile.close()

This was the standard idiom (untill xreadlines came along I guess)


while 1:
    data=infile.readlines(100000) # hint to readlines for 100000 bytes....
    if not data:
        break ## EOF reached
    for line in data:
        ## process data




>
> data = re.sub('[.](?=\d\d\d\d\d\d)|[.](?=\w+[ ][<|>])|[.](?=\w+[:])|[
> ](?!0x)', '\t', data)
>
> outfile = open("/mnt/storage/output.txt", "w")
> outfile.write(data)
> outfile.close()
>
>
> This works beautifully most of the time (it's super fast), except when I
> pipe large files (>50megs) to it and then it usually dies half way
> through complaining of memory errors because it ran out of ram.
>
> Short of buying more ram, is there a way I can make this more effecient
> without taking a big peformance hit? I tried the above where I used a
> "for line in data:" loop so it would only process one line at a time but
> this seemed to take forever! I'm imagining there's a way to process the
> data in something like 1 meg chunks, but I'm not too sure how to do that.
>
> I'm using python 2.2.1 on a RH linux 7.2 system with 256 ram and a
> Athlon XP1700 if it makes any diffrence.
>
> Any suggestions?
>
> TIA,
> Brent




From jadestar at idiom.com  Mon Jun 10 17:10:53 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 10 Jun 2002 21:10:53 GMT
Subject: 'for every' and 'for any'
References: <20020526135944.A32690@hishome.net> <20020526091742.A987@unpythonic.net> <mailman.1022567685.7803.python-list@python.org> <slrnaf7af3.8kn.quinn@schilling.ugcs.caltech.edu> <mailman.1022658469.27441.python-list@python.org>
Message-ID: <ae34kt$6r3$1@news.idiom.com>

Oren Tirosh <oren-py-l at hishome.net> wrote:

>> Then drop those functions into your personal utility library and use 'em
>> as often as you want :)

> I don't have a personal utility library. It's on purpose. I don't have
> personalized key bindings. I try to avoid customization.  Customization is
> a big part of what makes one programmer's code difficult to maintain by
> another programmer, what makes on programmers's workstation unusable by 
> another and makes combining code from several sources difficult.

 ...

> One of the things I like about Python is that it comes with everything in
> the box. 

> 	Oren

 Yet, this very discussion proves that "everything" is not already in
 the box.  I'll grant that Python includes a very large set of batteries
 --- but I'd still say that there are many useful functions and classes
 that can still be added.

 Those should start as personal "utility" libraries, be discussed here
 and in PEPs, and (when their general utility is established) added to
 appropriate (standard or optional) libraries.

 Obviously you don't re-invent the wheel for each project that you join;
 you are re-using some code that you've used before, even if it's from 
 "finger macros."   I can see the argument in favor of adapting your
 functions to conform to local conventions and every programming team has
 to collaborate to find compatible conventions and standards for their
 projects.  Any non-trivial project (non-trivial, in this case meaning
 "requires a *team* of programmers (> 3)) is going to find some elements
 of their project that go beyond established coding conventions.  The
 alternative to "building a (domain/application specific) library" is
 to eschew code re-use.

 In this particular case the all() and every() functions that were
 presented are elegant and seem useful.  As others have pointed out the
 implementations are trivial; so the questions become: "does it 'pollute
 the namespace' to add them to the core library?" and (if so) "where would 
 we put them?"

 Personally I think that a language that support containers as first order
 data types should also include a reasonably comprehensive set of 
 functions to operate on and/or support them.  The in operator is good,
 the iteration protocol is great, the map(),zip(), and filter() functions
 are good.  Notably zip() and filter() can be done with map() (trivially
 in most cases).  I'd say that the inclusion of zip() and filter()
 argue *for* the (eventual/future) inclusion of (every() or all()) and 
 (any() or some()).


From tdelaney at avaya.com  Wed Jun 26 00:21:44 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Wed, 26 Jun 2002 14:21:44 +1000
Subject: autocoder redux (was Re: Reminder of a python Project.)
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A3C7@natasha.auslabs.avaya.com>

> From: Peter Hansen [mailto:peter at engcorp.com]
> 
> Timothy Rue wrote:
> > 
> > On 25-Jun-02 22:38:16 Peter Hansen <peter at engcorp.com> wrote:
> > >I'm afraid there is no longer any need for an autocoder.
> > 
> > Geee, maybe you should tell that to NIST
> 
> No need.  They called me the other day and I let them know, but
> it was too late to get it into the Computerworld article.
> 
> Seriously, there's really no issue with software development any
> more.  The problems are solved.  I guess if you've invented the
> world's first autocoder, it kinda sux to be you then, eh?

OK Peter - time to stop trolling. It's no fun when the bait is taken so
easily ... ;)

Tim (no - not *that* one) Delaney




From peter at engcorp.com  Sun Jun 23 14:55:06 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 23 Jun 2002 14:55:06 -0400
Subject: printing on the browser
References: <d9531d76.0206230935.6e7a7cab@posting.google.com>
Message-ID: <3D16198A.B298D0CD@engcorp.com>

"Ivan J." wrote:
> 
> This is actually cgi quesion, but ...
> Is there a way to print both HTML and IMAGE parts using:
> >>>print "Conent-type: ...",
> statement, if not is there any other way to create web page with text
> and images without savingimages onto disk?
> Thank you!

The question is very unclear.  Are you really talking about 
"printing" as in sending data to a printer, with real paper
and all?  It doesn't look that way...

I guess you're talking about create a web page which has
both text and images embedded in it.  Basically, the answer
is No, you cannot do that.  (There are some sort of exceptions
if you talk about leading edge stuff with embedded SVG graphics,
but that's not relevant to your situation I think.)

Can you clarify a little (actually, a lot :-)?

-Peter


From kosh at aesaeion.com  Fri Jun 21 04:20:40 2002
From: kosh at aesaeion.com (kosh at aesaeion.com)
Date: Fri, 21 Jun 2002 02:20:40 -0600 (MDT)
Subject: compile python witout tkinter
In-Reply-To: <20020621082846.GA933@lilith.my-fqdn.de>
Message-ID: <Pine.LNX.4.44.0206210219470.585-100000@aesaeion.com>

On Fri, 21 Jun 2002, Gerhard [iso-8859-15] H?ring wrote:

> It looks like in Python 2.1.3, Tkinter gets built with distutils in
> Python's setup.py. I fear I don't have a solution to your problem, as
> detect_tkinter in Python 2.1.3's setup.py looks fine to me. Maybe you
> have a very strange or broken Tcl/Tk setup on your machine. As a
> temporary workaround, you could try to insert a return right at the
> beginning of detect_tkinter in Python's setup.py.
>
> I sincerly hope some Tkinter expert here can come up with a better
> solution and help finding the real reason behind this.

The problem is that a lot of tk stuff is installed on the system but the
xlib devel packages are not so it passes that test but there is no Xlib.h
so it fails to compile.





From dw133 at cs.york.ac.uk.go.away  Thu Jun  6 19:34:30 2002
From: dw133 at cs.york.ac.uk.go.away (Darren Winsper)
Date: Fri, 07 Jun 2002 00:34:30 +0100
Subject: Detecting OS and Window Manager
References: <mailman.1023304388.9659.python-list@python.org>
Message-ID: <ufvsco8mlhf213@corp.supernews.com>

In <mailman.1023304388.9659.python-list at python.org>, Sean 'Shaleh' Perry
wrote:

> the only fool proof way to detect KDE, GNOME, etc is to check for X atoms.  The
> bigger question is why do you care?  As long as they can display a GUI you
> should not be bothered about which one it is.  Especially under the X Window
> System it was designed so that random applications DIDN'T know what they were
> running under.

As the person writing the KDE front-end to the mentioned application, let
me answer your question.  The idea is that if the user is running KDE,
then they get a KDE front-end by default.  I also hope that should
somebody write a GTK/GNOME front-end, if the user was running GNOME they
would get a GTK-based front-end by default.  It's good UI design.

> Also remember KDE is not a window manager, it is a desktop environment.  It is
> perfectly reasonable to run KDE and use fvwm (or even twm) as the window
> manager.  Same applies to GNOME.

OK, Andrew meant DE instead of window manager, but it's still a valid
question.

> Also, what about someone who only runs say
> Window Maker and never uses GNOME or KDE?

Well, in that case we just choose a default.  The thing is, we want to
detect if PyKDE is installed.  If it is, then we can do "import kdecore"
without risk of it throwing an exception.  If it isn't, I'd rather not
have to try something along the lines of attempting "import kdecore" and
then having to catch an exception if it's not found.

-- 
ICQ: 8899775 Jabber ID: darren at rooter.dyndns.org
Stop the EUCD before it's too late:
http://uk.eurorights.org/issues/eucd/
http://www.tomchance.uklinux.net/articles/darkages.shtm



From christophe.delord at free.fr  Fri Jun 21 12:34:59 2002
From: christophe.delord at free.fr (Christophe Delord)
Date: Fri, 21 Jun 2002 18:34:59 +0200
Subject: How to represent the infinite ?
References: <mailman.1024636889.6135.python-list@python.org>
Message-ID: <20020621183459.63fe2e4d.christophe.delord@free.fr>

On Fri, 21 Jun 2002 01:19:50 -0400
Tim Peters <tim.one at comcast.net> wrote:

> It can, but it's a x-platform crapshoot as to exactly how.  For example,
> under current Windows CVS, this still "works":
> 
> >>> 1e300 * 1e300
> 1.#INF
> >>>
> 
> Here's a clumsier way <wink>:
> 
> >>> 1.6e308 + 1.6e308
> 1.#INF
> >>>
> 
> Both of those assume the user hasn't installed and enabled the fpectl
> module, whose purpose in life is to cause even this to complain.
> 


>>> 1e300
1.0000000000000001e+300
>>> 1e300*1e300
inf
>>> 1e300**2
Traceback (most recent call last):
  File "<pyshell#12>", line 1, in ?
    1e300**2
OverflowError: (34, 'Numerical result out of range')
>>> 

(under Linux with yesterday CVS)

So I was wrong. But addition, multiplication and exponentiation seem to be implemented differently (1e300+1e300 is inf and 1e300**2 is an overflow).
Is there a reason for these different behaviours?


Christophe.


-- 

(o_   Christophe Delord                   _o)
//\   http://christophe.delord.free.fr/   /\\
V_/_  mailto:christophe.delord at free.fr   _\_V


From jadestar at idiom.com  Sun Jun 16 17:50:05 2002
From: jadestar at idiom.com (James T. Dennis)
Date: 16 Jun 2002 21:50:05 GMT
Subject: Zip makes sense! was: Re: why not "'in' in 'in'"?
References: <mailman.1024086417.5966.python-list@python.org> <ugkpmkj4f0sid4@news.supernews.com>
Message-ID: <aej16d$t97$1@news.idiom.com>

John Roth <johnroth at ameritech.net> wrote:


> "David LeBlanc" <whisper at oz.net> wrote in message
> news:mailman.1024086417.5966.python-list at python.org...
>> Aaaah, a logger. Obviously the first on his block to be a logger. The
> joys
>> of bravely loggin where no one has logged before! ;->
>>
>> Perhaps if he was an english major, or english was his first language,
> he
>> could tell that the common meaning of yield isn't produce. I have no
> idea
>> what zip means beyond being a name for a compression utility or a verb
>> meaning to exercise a zipper.

> But that's what it is: exercising a zipper...
> It puts two (or more) sequences together into a single
> sequence, just like a zipper puts the two sides
> together into one.

> John Roth

 More over the term "yield" in English has a few distinct senses.
 One (slightly archaic) is the imperative to "yield" to a conqueror 
 or authority as in:
 	"Yield, Sir knight!"
 Another is a similar and common traffic imperative meaning to defer
 to the rights-of-way for others (probably the inspiration for the
 most common programming semantics associated with 'yield').
 Another is to refer to the product of a farm or manufacturing process
 as in: the yield of the corn field last year was X bushells.

 It is clearly this last sense which inspires Python's use of the
 term.  The return operator "yields" a result *and* terminates an
 execution contect.  The yield operator also "yields" a result but
 it leaves the production machinery or the "field" available for 
 future results.  That is in keeping with the English semantics
 since saying that a field "yielded" a crop doesn't imply that I've
 taken out the field (building a shopping mall over it, perhaps).
 To say that my field "returned" a profit *could* mean that it
 was a "yield" or it could mean that I've sold it off (ROI, 
 return on investment).

 Of course these nuances are not strictly part of the English usage.
 The are tendencies in usage since the terms can be used almost 
 interchangeable in many cases.
 



From catunda at pobox.com  Tue Jun 25 08:14:07 2002
From: catunda at pobox.com (Marco Catunda)
Date: Tue, 25 Jun 2002 09:14:07 -0300
Subject: Memory Monitor
Message-ID: <pan.2002.06.25.12.14.06.994352.17223@pobox.com>

Hello,

	What is the best way to monitor usage memory?

	I have used "os.system( 'ps %d' % os.getpid() )" but I
	don't know if it is the best way?

	The function getusage of resource module returns only
	zero value.

-- Marco Catunda


From pyth at devel.trillke.net  Sun Jun 16 09:27:31 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 16 Jun 2002 15:27:31 +0200
Subject: Snort alert tail...
In-Reply-To: <AB0P8.107$W91.886@news1.global-one.fi>; from jan.eric-enlund@multi.fi on Sun, Jun 16, 2002 at 03:51:45PM +0300
References: <AB0P8.107$W91.886@news1.global-one.fi>
Message-ID: <20020616152731.I15079@prim.han.de>

Jan-Eric wrote:
> HEllo !
> I'm trying to write a dynamic firewall script in python that scans the Snort 
> alert file like 'tail -f' and takes action based on the infomation it gets 
> from that file. But I can't get the 'tail' function to work.It reads the 
> file, but any new information that Snort is writing to the file doesn't 
> show up to the script.
> 
> ex.
> file = open('/var/log/snort', 'r')
> while 1:
>         file = file.read()
>         print file
> ....
> 
> Any suggestions???

yes, here's a snippet for unix/posix:

    import os

    tailoutputfile = os.popen('tail -f syslog')
    while 1:
        line = tailoutputfile.readline()
        if len(line)==0:
            break
        process_line(line)


the regular popen call returns a file from which you can
read. It's a blocking read while 'tail -f' has nothing to offer.

the alternative is to implement the 'tail -f' algorithm yourself.
basically you have to do (untested)

    lastsize=os.path.getsize('filename')
    while 1:
        size=os.path.getsize('filename')
        if size>lastsize:
            f=open('filename')
            f.seek(lastsize)
            part = f.read(size-lastsize)
            f.close()
            process_next_part(part)
        elif size==lastsize:
            time.sleep(0.2) # seconds to wait
        elif size<lastsize:
            print >>sys.stderr, "file filename got truncated"
            lastsize=size
       
i guess you get the idea.

regards,

    holger




From phr-n2002b at NOSPAMnightsong.com  Sun Jun 30 02:52:55 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 29 Jun 2002 23:52:55 -0700
Subject: yield equivalent in C/JavaScript?
References: <60FB8BB7F0EFC7409B75EEEC13E201922151B8@admin56.narex.com> <mailman.1025390799.23641.python-list@python.org>
Message-ID: <7xelepi7g8.fsf@ruckus.brouhaha.com>

Robin Becker <robin at jessikat.fsnet.co.uk> writes:
> thanks I know how to do this with threads, but I suspect I will end up
> doing this in JavaScript and threading is a complete mystery there (if
> it exists).

If it's a Netscape browser or server you can reach the java.threads
package from JS.  I don't know what happens if you actually try
using them though.


From miracle at paradise.net.nz  Thu Jun 27 16:50:16 2002
From: miracle at paradise.net.nz (Matthew Sherborne)
Date: Fri, 28 Jun 2002 08:50:16 +1200
Subject: How to find out DNS ?
References: <3D1B6186.3142.295959@localhost>
Message-ID: <3D1B7A88.7030705@paradise.net.nz>

A wrote:

>Hi,
>Is there a way how to find out, from Python , what primary or 
>secondary DNS I use when connecting to internet?
>Thanks for help
>Ladislav
>
>_______________________________________________
>ActivePython mailing list
>ActivePython at listserv.ActiveState.com
>To unsubscribe: http://listserv.ActiveState.com/mailman/mysubs
>Other options: http://listserv.ActiveState.com/mailman/listinfo/ActivePython
>
>
>
>  
>
    def getServersFromOS(self):
        if sys.platform in ('win32', 'nt'): return 
self.getServersFromWin32()
        elif sys.platform == 'posix': return self.getServersFromPosix()
        else: return []
       
    def getServersFromWin32(self):
        import _winreg
        servers = []
        x = _winreg.ConnectRegistry(None,_winreg.HKEY_LOCAL_MACHINE)
        try:
            y = _winreg.OpenKey(x, 
r'SYSTEM\CurrentControlSet\Services\Tcpip\Parameters')
        except EnvironmentError: # so it isn't NT/2000/XP
            # windows ME, perhaps?
            try: # for Windows ME
                y = _winreg.OpenKey(x, 
r'SYSTEM\CurrentControlSet\Services\VxD\MSTCP')
            nameserver = _winreg.QueryValueEx(y, 'NameServer')[0]
            if nameserver and not (nameserver in servers):
                servers += nameserver.split(',')
        except EnvironmentError: pass
        return servers

    def getServersFromPosix(self):
        """"Parses the /etc/resolv.conf file and sets defaults for name 
servers""""
        lines = open('/etc/resolv.conf').readlines()
        servers = []
        for line in lines:
            line = line.strip()
            if not line or line[0] == ';' or line[0] == '#': continue
            fields=string.split(line)
            if fields[0] == 'domain': defaults['domain'] = fields[1]
            elif fields[0] == 'search': pass
            elif fields[0] == 'options': pass
            elif fields[0] == 'sortlist': pass
            elif fields[0] == 'nameserver': servers.append(fields[1])
        return servers

GBU
Matthew Sherborne







From whisper at oz.net  Fri Jun 28 22:10:11 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 28 Jun 2002 19:10:11 -0700
Subject: Python needs better error reporting
In-Reply-To: <Pine.LNX.4.44.0206281948130.17292-100000@aesaeion.com>
Message-ID: <GCEDKONBLEFPPADDJCOEEEOMDMAA.whisper@oz.net>

That's not the point - I know that.

David LeBlanc
Seattle, WA USA 

> -----Original Message-----
> From: kosh at aesaeion.com [mailto:kosh at aesaeion.com]
> Sent: Friday, June 28, 2002 18:49
> To: David LeBlanc
> Cc: Python-List at Python. Org
> Subject: Re: Python needs better error reporting
> 
> 
> On Fri, 28 Jun 2002, David LeBlanc wrote:
> 
> > "Syntax Error: invalid syntax" isn't very informative, yet it's 
> tossed out
> > frequently. HOW is the syntax invalid? For example:
> > if s == ' '
> >
> 
> That is supposed to start a new block so it should be
> if s == ' ':
> 
> 




From peter at engcorp.com  Mon Jun 24 13:33:19 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 24 Jun 2002 13:33:19 -0400
Subject: How to stop a SocketServer?
References: <7xhejv9af2.fsf_-_@ruckus.brouhaha.com> <3d1744d4$1_1@hpb10302.boi.hp.com>
Message-ID: <3D1757DF.776A428F@engcorp.com>

Daniel Fackrell wrote:
> 
> I faced this problem with a simple chat server I wrote.  I expected that
> termination of the server with <ctrl-C> should have stopped it immediately,
> but it instead waits until some communication is received from each thread
> that was stopped at a read().
> 
> It appears that the current threading implementation does not provide for
> termination of threads inside a single python statement (such as a
> filelike.read()), so I am considering a move to character I/O, checking to
> verify that a character is available before each read.

The proper solution to this kind of problem is to use timeouts
and select.  There's generally no need to check for the data
being available ahead of time by _polling_, which is inefficient.

-Peter


From jepler at unpythonic.net  Thu Jun 13 13:42:59 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Thu, 13 Jun 2002 12:42:59 -0500
Subject: system command
In-Reply-To: <aeajuh$5b5eo$1@ID-108351.news.dfncis.de>
References: <aeajuh$5b5eo$1@ID-108351.news.dfncis.de>
Message-ID: <20020613174258.GD25416@unpythonic.net>

On Thu, Jun 13, 2002 at 07:19:48PM +0200, Thor wrote:
> Having this code
>  
>   print "source "+root+".in"
>   os.system("source "+root+".in")
> 
> the script does not run the root+".in" file, but copying and pasting the 
> output form teh 1st line work.. what i'm doing wrong? Thanks in advance

It is not a Python problem.  It is the difference between running
something in a shell or a subshell.  The former is like typing
    source $ROOT.in
vs
    sh -c "source $ROOT.in"
the latter will not make any changes to the environment of the parent
shell.  So setting environment variables, creating aliases, etc., won't
work like you'd hoped.

If you want to have a Python program that changes something in the
parent shell's environment, you have to arrange to execute it in the
shell.  For instance,
    source `python print_name_of_file_to_source.py`
or
    eval `python print_shell_code_to_execute.py`

Jeff




From sholden at holdenweb.com  Tue Jun  4 16:41:44 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 4 Jun 2002 16:41:44 -0400
Subject: DCOracle2 - Can't get cursor.execute to return data when SQL query involves datetimes.
References: <be0d725a.0206041139.1f8f6647@posting.google.com>
Message-ID: <IG9L8.239957$%u2.56293@atlpnn01.usenetserver.com>

"Jim" <jtt at hnc.com> wrote in message
news:be0d725a.0206041139.1f8f6647 at posting.google.com...
> Hello,
>
> I am trying to write a simple program to retrieve 2 colums from a
> table.
> I am running on Solaris 2.8, Python 2.1.1, DCOracle2 and Oracle
> 8.1.7.3.
>
>
> The columns are:
>
> acct_number   ( one char long)
> transaction_datestamp (oracle datetimestamp)
>
> Here is my program:
>
> import DCOracle2
> trandt = '19970110121212'
>
> db = DCOracle2.connect("uid/pwd")
> c = db.cursor()
>
> dt = DCOracle2.Timestamp(int(tran_date[0:4]), int(tran_date[4:6]),
> int(tran_date[6:8]), int(tran_date[8:10]), int(tran_date[10:12]),
> int(tran_date[12:14]))
>
> results = c.execute,('select acct_nbr, trn_dt from testtable where
> ACCT_NBR = :1 and TRN_DT = :2',"1",dt)
>
> aaa = c.fetchone()
>
> print aaa
>
>
> The execute never return any thing from query when the date timestamp
> is part of the query.  If I only query on the acct_nbr it returns the
> expected row.
>
> What am I missing ???
>

Possibly the cxOracle module from www.computronix.com :-)

Otherwise, please go away and *copy and paste* the real code: I'll overlook
the line wrapping, which your mailer probably inserted gratuitously, but you
appear to be using tran_date where you should be using trandt, and trn_dt
where you should be using transacttion_datestamp.

Finally, doesn't the .execute() method call want two arguments [i.e.,
shouldn't your data be a tuple ("1", dt)], not three?

regards
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From nhodgson at bigpond.net.au  Fri Jun 28 06:50:48 2002
From: nhodgson at bigpond.net.au (Neil Hodgson)
Date: Fri, 28 Jun 2002 10:50:48 GMT
Subject: Module Help
References: <mailman.1025164967.15063.python-list@python.org>
Message-ID: <ccXS8.368204$o66.976008@news-server.bigpond.net.au>

Juan M. Casillas:

> I want to create an empty module from C, and then add
> functions and definitions stored in another file, and
> some code that I will create on-the-fly, from C.

   Here is some Python code by Michael Hudson to create a new module and add
some code to it:

newm = new.module(name)
sys.modules[name] = newm
newm.__file__ = filepath
exec compile(open(filepath).read(), filepath, "exec") in \
        newm.__dict__

   You can change the open(filepath).read() to provide the code text in ay
way you want. I think you can use empty strings for the filepath variable.
Now you just have to work out how to call this from C which IIRC is well
covered in the embedding part of the "Extending and Embedding the Python
Interpreter" manual.

   Neil




From parker at gol.com  Thu Jun 13 21:00:56 2002
From: parker at gol.com (Ian Parker)
Date: Fri, 14 Jun 2002 01:00:56 GMT
Subject: Medium to Large Scale Python Deployments
References: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com>
 <mailman.1023409393.3361.python-list@python.org>
 <3D003EC7.C72D27BE@engcorp.com> <3D006320.5040805@web.de>
 <3D00A209.EC24F671@engcorp.com> <adr2ao$bei$0@216.39.172.122>
 <3D0140BF.A239E39E@engcorp.com>
 <7hUN8.529345$%u2.28370@atlpnn01.usenetserver.com>
Message-ID: <CEe53JEBBUC9Ew8D@hiredata.gol.com>

In article <7hUN8.529345$%u2.28370 at atlpnn01.usenetserver.com>, Steve 
Holden <sholden at holdenweb.com> writes
>"Peter Hansen" <peter at engcorp.com> wrote in message
>news:3D0140BF.A239E39E at engcorp.com...
>> Bengt Richter wrote:
>> >
>> > Is the "comments" number a count of _pure_ comment lines,
>> > # (like this one) ?
>>
>> I believe so.  I haven't examined the pycount code yet.
>>
>> > I think it would be interesting to analyze comments further,
>> > to see what the relative comment byte count is overall,
>> > and for lines with both code and comments,
>> > """like this line:""" # (this has code and comment).
>>
>> I consider it a mild stylistic no-no to use line-ending
>> comments, personally.  I believe it's almost always better
>> to put a block-comment of one or more lines above a block
>> of code with related purpose.  A good comment there tends to
>> be more maintainable, and useful, than lots of little comments
>> on each line.
>>
>I wouldn't necessarily disagree, but I used (when programming (gulp) Basic
>Plus) to make the line-end comments into a free-form narrative that rough;y
>parallelled the code.
>
>[...]
>regards
>-----------------------------------------------------------------------
>Steve Holden                                 http://www.holdenweb.com/
>Python Web Programming                http://pydish.holdenweb.com/pwp/
>-----------------------------------------------------------------------
>
>
>
>
>

I did the same on assemblers (double gulp) and possibly FORTRAN. I think 
it's an effective technique on languages where the statements are 
relatively short compared to the window width.

Regards

Ian
-- 
Ian Parker


From opengeometry at yahoo.ca  Wed Jun 26 17:06:00 2002
From: opengeometry at yahoo.ca (William Park)
Date: Wed, 26 Jun 2002 17:06:00 -0400
Subject: Help with an algorithm wanted
In-Reply-To: <rowen-CE3C0F.13020226062002@nntp1.u.washington.edu>
References: <rowen-CE3C0F.13020226062002@nntp1.u.washington.edu>
Message-ID: <20020626210600.GA2890@node1.opengeometry.net>

On Wed, Jun 26, 2002 at 01:02:02PM -0700, Russell E. Owen wrote:
> For instance, an object with A as the given would return D by solving 
> A->C->D.

> The real issue is creating such objects without hand coding each one 
> (since in the real world I have more representations and the number of 
> casese proliferates badly).

If pairing (ie. mapping between 2 objects) occurs randomly, then it's
virtually impossible, because you don't know which direction to go from any
one object.  It would be travelling from city to city, without knowing
which state or county. 

-- 
William Park, Open Geometry Consulting, <opengeometry at yahoo.ca>
8-CPU Cluster, Hosting, NAS, Linux, LaTeX, python, vim, mutt, tin




From Chris.Barker at noaa.gov  Thu Jun 20 15:42:09 2002
From: Chris.Barker at noaa.gov (Chris Barker)
Date: Thu, 20 Jun 2002 12:42:09 -0700
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <aer2fn$frc$1@peabody.colorado.edu> <3D122E79.A189C72E@astro.cornell.edu>
Message-ID: <3D122FDF.D64181DA@noaa.gov>


Tom Loredo wrote:
> Redhat
> uses Python, but does their reliance on an old version still cause
> problems?.

They use 1.5, and all their scripts use "python" on the #! line, so if
you use a more recent version, you need to put python2 (or python2.2 or
whatever) on your #! lines. Nat a big deal, really, but a little
annoying. I'd love to see RedHat put python1.5 on THEIR #! lines, which
you wouldn't think would be difficult. Maybe RedHat 8.*, they have been
getting complaints.

> Actually, I need to develop stuff that other Win32 users can use,
> some probably distributed via the McMillan installer (which I presume
> works as well with Active State as with the python.org version).
> So I'm stuck here....

I'd do most of your coding under Linux, and just use Windows to test an
package it up, unless you're doing com or something Windows only.

> The fact that Dell shipped it with a defective
> install of Win98 did not help its cause.

Is there such a thing an a non defective Win98? Seriously, use Win2k if
you can, I still don't like it, but it provides a modicum of stability
and a almost usable command line.

-Chris




-- 
Christopher Barker, Ph.D.
Oceanographer
                                    		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov


From tim.one at comcast.net  Sat Jun 29 07:25:58 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sat, 29 Jun 2002 07:25:58 -0400
Subject: Python needs better error reporting
In-Reply-To: <GCEDKONBLEFPPADDJCOEOEPADMAA.whisper@oz.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCEEOKAAAB.tim.one@comcast.net>

[David LeBlanc]
> Cute anthropomorphism is lots of fun, but doesn't _quite_ obscure the fact
> that computers don't really "tell" anybody anything, programmers do.

When we ship Guido with every compiler, the world will be a better place for
sure.

> IIRC, Python throws this "Syntax Error: invalid syntax" in a number of
> places.

Yes.  Study the code a little and you'll soon discover why.

> In the place that I used as an example, looking at the language
> reference, an appropriate error would be more like:
> 	Syntax Error: expected "and", "or", "xor" or ":"

Well, "xor" isn't legit there, but other tokens that are OK after

    if s == ' '

include

    '
    "
    '''
    """
    ==
    !=
    <>
    <
    <=
    >
    >=
    +
    -
    /
    *
    **
and
    \

Call me prophetic, but you wouldn't be happy with a list of all
possibilities either.  Collapse oodles of them into "binary operator" and
you lose half the people you're trying to help.  Etc.

> This would at least narrow down what should be looked at to resolve the
> problem.

Not really.  Such lists tend to baffle more than clarify.  I once used a
Pascal compiler whose favorite trick was to point a caret at a trailing
semicolon, and then complain "semicolon expected"; the message was always
correct when it appeared, while appearing dead wrong, and was useless in
determining the *real* problem regardless.  It did appear friendlier than
"syntax error" the first 719 times I saw it, though <wink>.

> Vague error reporting doesn't help new user's self confidance or
> acceptance of the language.
>
> I really DO have a point!

Error-reporting in parsers has been the subject of intense research.  Do a
little digging and come back when you can make a case that your point can be
addressed effectively.  Keep in mind that most consumer-grade computers
still don't have telepathy chips <wink>.





From mcalla at insightbb.com  Sun Jun 16 16:13:24 2002
From: mcalla at insightbb.com (Mike Callahan)
Date: Sun, 16 Jun 2002 20:13:24 GMT
Subject: Question about asksaveasfilename
References: <ed0O8.23891$nZ3.2215@rwcrnsc53> <aea3ha$lv$1@wanadoo.fr>
Message-ID: <zj6P8.66783$6m5.52153@rwcrnsc51.ops.asp.att.net>

I did find defaultextension, but as you pointed out, it only works with a
single string and does not pay attention to the user's choice in the file
type listbox. I guess I could make my own dialog but I thought that built-in
one would do the trick. Since no one else has responded, I guess it cannot.

Mike C.

"Eric Brunel" <eric.brunel at pragmadev.com> wrote in message
news:aea3ha$lv$1 at wanadoo.fr...
> Mike Callahan wrote:
>
> > I am using asksaveasfilename from tkFileDialog. I want to restrict my
> > users to save filenames with these extensions .unl or .sql. This is my
> > call:
> >
> > from tkFileDialog import asksaveasfilename
> > fn = asksaveasfilename(filetypes=[('Unloaded', '*.unl'),
> > ('Queries','*.sql')])
> >
> > This gives me the correct dialog box, but how do I force the correct
> > extension on fn?  Thanks.
>
> There's a defaultextension option, but it does not depend on the selected
> file type, i.e. you'll have to do either:
>
> fn = asksaveasfilename(filetypes=[('Unloaded', '*.unl'),
>                                   ('Queries',  '*.sql')],
>                        defaultextension='.unl')
>
> or:
>
> fn = asksaveasfilename(filetypes=[('Unloaded', '*.unl'),
>                                   ('Queries',  '*.sql')],
>                        defaultextension='.sql')
>
> which isn't what you're looking for, I fear...
>
> I don't know any means to force an extension depending on the selecetd
file
> type with asksaveasfilename, so my advice would be to split the two save
> options: one "Save unloaded" and one "Save query"... But if there's
another
> solution, I'm interested too ;-).
>
> HTH
> --
> - Eric Brunel <eric.brunel at pragmadev.com> -
> PragmaDev : Real Time Software Development Tools -
http://www.pragmadev.com




From R.Barrett at ftel.co.uk  Tue Jun 11 13:09:53 2002
From: R.Barrett at ftel.co.uk (Richard Barrett)
Date: Tue, 11 Jun 2002 18:09:53 +0100
Subject: Defining a method final
In-Reply-To: <ae4tvv$rvd$1@wanadoo.fr>
References: <6glN8.35376$86.884025@twister1.libero.it>
 <mailman.1023796095.1427.python-list@python.org>
 <phmN8.36164$K4.893500@twister2.libero.it>
 <KwmN8.153269$352.7754@sccrnsc02>
Message-ID: <5.1.0.14.2.20020611161524.0341a008@pop.ftel.co.uk>

At 15:35 11/06/2002 +0000, Eric Brunel wrote:
>Emile van Sebille wrote:
> > Accepted practice is to use a single underscore for objects to be
> > considered private.  This does not prevent their use, but strongly
> > signals that re-use ought not to be done.
> >
> >>>> class Test:
> > ...     def _private(self): pass
> > ...
> >
>
>Python now have an "official" means to make attributes private by prefixing
>them with a double-underscore. And it *does* prevent their use or
>overloading:

I'm not criticising the double underscore prefix feature but it prevents 
neither, if you factor in the name mangling that supports the feature. See 
the following transcript from Python 2.2.1 command line:

 >>> class A:
...     def __private(self): print "class A private"
...     def printit(self): self.__private()
...
 >>> class B(A):
...     pass
...
 >>> class C(A):
...     def _A__private(self): print "class C private"
...
 >>> a1 = A()
 >>> a1.printit()
class A private
 >>> a1._A__private()
class A private
 >>> b1 = B()
 >>> b1.printit()
class A private
 >>> b1._A__private()
class A private
 >>> c1 = C()
 >>> c1.printit()
class C private
 >>> c1._A__private()
class C private





From occeanlinux at linuxmail.org  Thu Jun 27 10:36:03 2002
From: occeanlinux at linuxmail.org (SiverFish)
Date: Thu, 27 Jun 2002 14:36:03 +0000
Subject: Questiion on list
Message-ID: <pan.2002.06.27.14.36.03.151659.5728@linuxmail.org>

I got the list of words now i want to calculate how many word in that
list which is the same word (insensitive case),can anyone show me how to do
it for example ls = ['abc','cd','abc','adf',abc','dwc','cd']
it will show abc = 3 ,cd = 2, adf = 1, dwc=1


From quinn at retch.ugcs.caltech.edu  Tue Jun 25 17:33:26 2002
From: quinn at retch.ugcs.caltech.edu (Quinn Dunkan)
Date: 25 Jun 2002 21:33:26 GMT
Subject: 'parent object'?
References: <3d18228c$0$221$4d4ebb8e@news.nl.uu.net> <3D182662.CBC925DB@alcyone.com> <3D18AF67.2010207@nospam.free.fr>
Message-ID: <slrnahhod6.v18.quinn@retch.ugcs.caltech.edu>

On Tue, 25 Jun 2002 13:59:03 -0400, laotseu <bdesth at nospam.free.fr> wrote:
>That's ok for 'parent/child' in the inheritence meaning. But the OP 
>asked about something else : a container instance (the 'parent'), and 
>the instances it contains (the 'childrens'). This has nothing to do with 
>  subclassing.
>
><to the op>
>You should define a 'container/child' interface. it may be very simple 
>in Python, something like
>
>class container:
>     def __init__(self):
>         self.childrens = []
>
>     def add_child(self, child):
>         if child not in self.childrens:
>             child.parent = self     # create a 'parent' attribute
>             self.childrens.append(child)
>
>     def remove_child(self, child):
>          if child in self.childrens:
>	    del self.childrens[self.childrens.index(child)]
>             del child.parent
>
>
>Now whatever class your child objects is an instance of, it gets a new 
>'parent' attribute when you add it to the container instance, and looses 
>this attribute when you remove it...

This also introduces a cycle.  In older pythons, discarded containers would
never deallocate their children.  In newer pythons, the garbage will get
cleaned eventually.

As a random aside, I've found that almost always when I thought I needed to
have a child know about its parent, it's turned out to be a design mistake.
Cyclical dependency is messy to think about.  Things can wind up getting called
in weird ways, and getting in tricky infinite recursion.

As another random aside, using __getitem__ and __setitem__ is usually more
natural for this kind of thing.


From BPettersen at NAREX.com  Fri Jun 14 19:59:56 2002
From: BPettersen at NAREX.com (Bjorn Pettersen)
Date: Fri, 14 Jun 2002 17:59:56 -0600
Subject: why not "'in' in 'in'"?
Message-ID: <60FB8BB7F0EFC7409B75EEEC13E20192F1A575@admin56.narex.com>

> From: Tim Peters [mailto:tim.one at comcast.net] 

[snip]

> So if somebody wants this change enough to submit a patch 
> (code, doc, + test suite changes), it's likely to be accepted 
> for 2.3.  As code changes go, this one should be particularly 
> easy.  We have an internal bet going as to whether this 
> invitation is enough to kill the idea <wink>.
> 
> some-parts-of-this-were-actually-true-ly y'rs  - tim

Here are the code changes. (This patch is public domain, etc. etc.)

I'm getting a two warnings:

   1: "incompatible types - from 'struct _object*' to 'struct 
      PyStringObject'" on the line

        result = string_find_internal(a, args, +1);

   2: "incompatible types - from 'struct _object*' to 'struct 
      PyUnicodeObject'" on the line

        findResult = (PyObject*) unicode_find(container, args);

which I can't understand (anyone?)

If anyone feels like writing the doc/test suite changes feel free to use
my patch.

-- bjorn


*** stringobject.orig.c	Fri Jun 14 16:46:58 2002
--- stringobject.c	Fri Jun 14 17:51:16 2002
***************
*** 46,51 ****
--- 46,56 ----
     not counting the null terminating character, and is therefore equal
to the 
     `size' parameter.
  */
+ 
+ /* forward declarations */
+ long
+ string_find_internal(PyStringObject *self, PyObject *args, int dir);
+ 
  PyObject *
  PyString_FromStringAndSize(const char *str, int size)
  {
***************
*** 824,848 ****
  static int
  string_contains(PyObject *a, PyObject *el)
  {
! 	register char *s, *end;
! 	register char c;
  #ifdef Py_USING_UNICODE
! 	if (PyUnicode_Check(el))
! 		return PyUnicode_Contains(a, el);
  #endif
! 	if (!PyString_Check(el) || PyString_Size(el) != 1) {
! 		PyErr_SetString(PyExc_TypeError,
! 		    "'in <string>' requires character as left operand");
! 		return -1;
! 	}
! 	c = PyString_AsString(el)[0];
! 	s = PyString_AsString(a);
! 	end = s + PyString_Size(a);
! 	while (s < end) {
! 		if (c == *s++)
! 			return 1;
! 	}
! 	return 0;
  }
  
  static PyObject *
--- 829,855 ----
  static int
  string_contains(PyObject *a, PyObject *el)
  {
!     long result;
!     PyObject* args;
!     
  #ifdef Py_USING_UNICODE
!     if (PyUnicode_Check(el))
!         return PyUnicode_Contains(a, el);
  #endif
!     if (!PyString_Check(el)) {
!         PyErr_SetString(PyExc_TypeError,
!             "'in <string>' requires string as left operand");
!         return -1;
!     }
!     args = Py_BuildValue("(O)", el);
!     result = string_find_internal(a, args, +1);
!     Py_DECREF(args);
!     if (result >= 0) {
!         return 1;
!     }
!     else {
!         return 0;
!     }
  }
  
  static PyObject *
***************
*** 1329,1335 ****
  
  	return -1;
  }
- 
  
  static char find__doc__[] =
  "S.find(sub [,start [,end]]) -> int\n\
--- 1336,1341 ----


*** unicodeobject.orig.c	Fri Jun 14 17:12:09 2002
--- unicodeobject.c	Fri Jun 14 17:53:44 2002
***************
*** 76,81 ****
--- 76,85 ----
  # define BYTEORDER_IS_LITTLE_ENDIAN
  #endif
  
+ /* forward declaration */
+ PyObject *
+ unicode_find(PyUnicodeObject *self, PyObject *args);
+ 
  /* --- Globals
------------------------------------------------------------
  
     The globals are initialized by the _PyUnicode_Init() API and should
***************
*** 3789,3830 ****
      return -1;
  }
  
! int PyUnicode_Contains(PyObject *container,
! 		       PyObject *element)
  {
      PyUnicodeObject *u = NULL, *v = NULL;
      int result;
!     register const Py_UNICODE *p, *e;
!     register Py_UNICODE ch;
  
      /* Coerce the two arguments */
      v = (PyUnicodeObject *)PyUnicode_FromObject(element);
      if (v == NULL) {
! 	PyErr_SetString(PyExc_TypeError,
! 	    "'in <string>' requires character as left operand");
! 	goto onError;
      }
      u = (PyUnicodeObject *)PyUnicode_FromObject(container);
      if (u == NULL) {
! 	Py_DECREF(v);
! 	goto onError;
      }
  
      /* Check v in u */
!     if (PyUnicode_GET_SIZE(v) != 1) {
! 	PyErr_SetString(PyExc_TypeError,
! 	    "'in <string>' requires character as left operand");
! 	goto onError;
      }
!     ch = *PyUnicode_AS_UNICODE(v);
!     p = PyUnicode_AS_UNICODE(u);
!     e = p + PyUnicode_GET_SIZE(u);
!     result = 0;
!     while (p < e) {
! 	if (*p++ == ch) {
! 	    result = 1;
! 	    break;
! 	}
      }
  
      Py_DECREF(u);
--- 3793,3835 ----
      return -1;
  }
  
! int PyUnicode_Contains(PyObject *container, PyObject *element)
  {
      PyUnicodeObject *u = NULL, *v = NULL;
      int result;
!     long tmpresult;
!     PyObject *args;
!     PyObject *findResult = NULL;
  
      /* Coerce the two arguments */
      v = (PyUnicodeObject *)PyUnicode_FromObject(element);
      if (v == NULL) {
!         PyErr_SetString(PyExc_TypeError,
!             "'in <string>' requires string as left operand");
!         goto onError;
      }
      u = (PyUnicodeObject *)PyUnicode_FromObject(container);
      if (u == NULL) {
!         Py_DECREF(v);
!         goto onError;
      }
  
      /* Check v in u */
!     args = Py_BuildValue("(O)", element);
!     findResult = (PyObject*) unicode_find(container, args);
!     Py_DECREF(args);
!     if (findResult == NULL) {
!         goto onError;
      }
!     else {
!         tmpresult = PyInt_AsLong(findResult);
!         Py_DECREF(findResult);
!         if (tmpresult >= 0) {
!             result = 1;
!         }
!         else {
!             result = 0;
!         }
      }
  
      Py_DECREF(u);
***************
*** 3832,3837 ****
--- 3837,3843 ----
      return result;
  
  onError:
+     Py_XDECREF(findResult);
      Py_XDECREF(u);
      Py_XDECREF(v);
      return -1;




From james.kew at btinternet.com  Tue Jun 11 18:52:06 2002
From: james.kew at btinternet.com (James Kew)
Date: Tue, 11 Jun 2002 23:52:06 +0100
Subject: Sorting list (maybe stupid)
References: <slrnagbh7m.h03.jarvin24@lehtori.cc.tut.fi> <ae4gnp$l18$1@news.tudelft.nl> <ae4hn9$jm4$1@wanadoo.fr> <3D062B7C.FE6A4963@noaa.gov>
Message-ID: <ae5uuo$4atsq$1@ID-71831.news.dfncis.de>

"Chris Barker" <Chris.Barker at noaa.gov> wrote in message
news:3D062B7C.FE6A4963 at noaa.gov...
> Eric Brunel wrote:
>
> > l = [(102, 'foo'), (2, 'bar'), (1, 'baz')]
> > l.sort()
> > print l
> >
> > Apparently, the default sort function already sorts tuples according to
> > their first element...
>
> Note that this only works if the rest of the elements in the tuples are
> also sortable. In this case, they are strings, so they are.

Although that does mean that, in the OP's problem, lines with identical
numbers will be further sorted by the following string -- this might not be
the behaviour he requires.

Is sort() stable -- that is, is the relative order of elements which compare
equal preserved? A quick shufti through the ActivePython docs wasn't
illuminating.

--
James Kew
james.kew at btinternet.com





From radix at twistedmatrix.com  Wed Jun 12 11:36:02 2002
From: radix at twistedmatrix.com (Christopher Armstrong)
Date: 12 Jun 2002 11:36:02 -0400
Subject: Bits and bytes?
In-Reply-To: <3d0760ec$0$223$4d4ebb8e@news.nl.uu.net>
References: <3d074e6d$0$224$4d4ebb8e@news.nl.uu.net>
	<mailman.1023890477.2592.python-list@python.org>
	<3d0760ec$0$223$4d4ebb8e@news.nl.uu.net>
Message-ID: <87adq0v7a5.fsf@twistedmatrix.com>

>>>>> "Guyon" == Guyon Mor?e <gumuz at looze.net> writes:

    Guyon> ok, thanx a lot.  there's one thing i am missing though... how do i
    Guyon> write '011010110101' as 'binary bits' to a file?

You can only write strings to a file. So use the unbinarify function to convert
the bit-string to a "data" string, and write that.


    Guyon> "Christopher Armstrong" <radix at twistedmatrix.com> wrote in message
    Guyon> news:mailman.1023890477.2592.python-list at python.org...

    Guyon> http://twistedmatrix.com/users/moshez/binary.py

    Guyon> Moshe Zadka wrote this because he was annoyed at all the people in
    Guyon> #python asking how to do this ;-)


-- 
                                Chris Armstrong
                         << radix at twistedmatrix.com >>
                http://twistedmatrix.com/users/carmstro.twistd/





From gerhard at bigfoot.de  Tue Jun  4 12:25:14 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 4 Jun 2002 16:25:14 GMT
Subject: Python command line?
References: <B9224E06.628B%sarmstrong13@mac.com>
Message-ID: <slrnafpq0l.os.gerhard@developer2.opus-gmbh.net>

In article <B9224E06.628B%sarmstrong13 at mac.com>, SA wrote:
> Hi Everyone-
> 
> I would like to run python code line by line using:
> 
> python -c some line of code
> 
> Will this work?

You need to put your code into quotes, like in:

    python -c "print 5"

Code spanning multiple lines will also work, just be sure to use ' instead of "
in your code.

Gerhard


From peter at engcorp.com  Fri Jun 28 18:39:29 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 28 Jun 2002 18:39:29 -0400
Subject: generate all combinations of a list (with variable output length)
References: <3D1CBE6D.30002@huno.net>
Message-ID: <3D1CE5A1.1A0F5E58@engcorp.com>

thomas wrote:
> 
> hi,
> 
> i'm looking for a way to generate all possible combinations of all items
> in a list. list = ["A","D","$","5","R"]
> 
> but i also want it to use one item several times or not at all and in no
> particular order. the length of the string is 1-10. so a possible output
> would be:
> 
> DD$
> A5RD$D555
> A
> 5RR$$A
> DRRR55$5$A
> ...
> 
> i find it particualy difficult to tell python to use one item several
> times or not at all, and the variable length thing. i don't want to
> reinvent the wheel so someone probably has already a soultion for that
> :) and if not someone can probably push me in some direction.

Why not post your best effort so far and we can critique it?
That way you'll learn a lot more than if someone just hands
you the answer on a silver platter.  We'll also be happier
helping when we know it's not just another homework question
you're supposed to do on your own. ;-)

-Peter


From emile at fcfw.fenx.com  Tue Jun 25 10:06:52 2002
From: emile at fcfw.fenx.com (Emile van Sebille)
Date: Tue, 25 Jun 2002 07:06:52 -0700
Subject: Suggestions for good programming practices?
References: <31575A892FF6D1118F5800600846864DCBD432@intrepid>
Message-ID: <349201c21c51$eef186d0$0a06a8c0@FENX.COM>

Simon Brunning 
> It isn't in the 2.2 series until 2.2.1, IIRC.
> 

Am I missing something?

F:\Python22>python
Python 2.2 (#28, Dec 21 2001, 12:21:22) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import this
The ...

--

Emile van Sebille
emile at fenx.com

---------





From kragen at pobox.com  Sun Jun  2 22:49:34 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 02 Jun 2002 22:49:34 -0400
Subject: Which is fastest dict or list.index() ?
References: <adaih2$vtb8c$1@ID-146704.news.dfncis.de> <adbnkh$s57$1@bob.news.rcn.net>
Message-ID: <83n0udrs8x.fsf@panacea.canonical.org>

"Raymond Hettinger" <python at rcn.com> writes:
> There are a lot of ways to implement an OrderedDictionary.
> Take a look at
> http://home.arcor.de/wolfgang.grafen/Python/Modules/Modules.html
> and http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/107747
> If you want a skiplist version, let me know and I'll send it.

I'm curious how your skiplist version compares with the built-in
Python dict type on performance; I'd been thinking about building a
skiplist ordered dictionary extension type myself, but if you've
already done it, maybe I should just use yours.



From sanjay2kind at yahoo.com  Wed Jun 26 07:45:28 2002
From: sanjay2kind at yahoo.com (sanjay)
Date: 26 Jun 2002 04:45:28 -0700
Subject: how to remove HTML  attributes from web pages ( HTML parseing)
Message-ID: <63170f57.0206260345.5b9b3397@posting.google.com>

Hi,

 New to python and doing one application. i would like to  web page
content
after removing specific html tag, attribute etc. Like taking out
color,bgcolor and <font>,<p> tag from the web pages. Which one will be
helpful : regular expression  or HTMLParser  helpfully.


Thx,


From Norman_Shelley-RRDN60 at email.sps.mot.com  Mon Jun  3 18:44:03 2002
From: Norman_Shelley-RRDN60 at email.sps.mot.com (Norman Shelley)
Date: Mon, 03 Jun 2002 15:44:03 -0700
Subject: ?Tcl's embedded virtual filesystem and Scripted documents
Message-ID: <3CFBF133.7DF60BA4@email.sps.mot.com>

Tcl now has an embedded virtual filesystem that looks very useful.
http://mini.net/tcl/2138.html

Is there any motivation in the Python camp to have something similiar?


Also, a "scripted document" like methodology looks to be very useful also.
http://www.equi4.com/jcw/scripdoc.html

Any work going on in this arena?


Norman Shelley



From mhammond at skippinet.com.au  Sun Jun 30 03:53:50 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Sun, 30 Jun 2002 07:53:50 GMT
Subject: Python COM installation and NT Administrative rights.
References: <3281a460.0206290652.51b800c9@posting.google.com>
Message-ID: <3D1EB9C7.70901@skippinet.com.au>

P Adhia wrote:
> Hi,
> 
> I'd like to install Python COM on my work PC - on which I don't have
> administrative rights. Is there a Python COM installation available
> that does not require administrative rights on NT ? (I don't know if
> administrative rights is must for Python COM to install/work; but I
> have installed some other softwares that detect current privilege
> levels and install regitry entries reserved for per user instead of
> system. I am hoping that Python COM can be made to do something
> similar.)
> 

Installing Python 2.2 + the latest win32all *should* work here - but I 
am sure there will be some problems.

If you are willing, give it a try, then mail me with whatever problems 
you have.

Mark.



From Sarah_L at gmx.net  Fri Jun  7 05:57:47 2002
From: Sarah_L at gmx.net (Sarah Mayer)
Date: 7 Jun 2002 02:57:47 -0700
Subject: Memory leaks when  embedding python (debug-version)  ?
Message-ID: <1eed5b9.0206070157.6d281f65@posting.google.com>

Hi,
maybe someone could help me,

I just embedded python 2.2.1 into a dll , 

Program looks like:

MyClass::MyClass()
{
    Py_Initialize();
    Py_Finalize();
}


and after the program finishes I get a lot of memory-leaks.
What is missing ???


Greetings
Sarah


Memorydump:
Dumping objects ->
{36193} normal block at 0x00936DE0, 52 bytes long.
 Data: <                > B0 AC 15 1E 10 9A 92 00 CD CD CD CD CD CD CD CD 
{36182} normal block at 0x00936E40, 44 bytes long.
 Data: <     o          > 10 9A 92 00 00 6F 93 00 CD CD CD CD CD CD CD CD 
{36179} normal block at 0x00936EA0, 41 bytes long.
 Data: < o  Pn          > 10 6F 93 00 50 6E 93 00 04 00 00 00 88 F2 17 1E 
{36178} normal block at 0x00936F00, 44 bytes long.
 Data: <@n   o          > 40 6E 93 00 C0 6F 93 00 CD CD CD CD CD CD CD CD 
{36175} normal block at 0x00936F60, 45 bytes long.
 Data: < o   o          > D0 6F 93 00 10 6F 93 00 04 00 00 00 88 F2 17 1E 
{36174} normal block at 0x00936FC0, 44 bytes long.
 Data: < o   1          > 00 6F 93 00 00 31 93 00 CD CD CD CD CD CD CD CD 
{36170} normal block at 0x00933100, 44 bytes long.
 Data: < o  `1          > C0 6F 93 00 60 31 93 00 CD CD CD CD CD CD CD CD 
{36166} normal block at 0x00933160, 44 bytes long.
 Data: < 1   R          > 00 31 93 00 80 52 93 00 CD CD CD CD CD CD CD CD 
{36164} normal block at 0x00935070, 384 bytes long.
 Data: <   $p0   M   C) > 00 98 09 24 70 30 93 00 F0 4D 93 00 C0 43 29 1F 
{36162} normal block at 0x00935220, 41 bytes long.
 Data: < R  p1          > 90 52 93 00 70 31 93 00 04 00 00 00 88 F2 17 1E 
{36161} normal block at 0x00935280, 44 bytes long.
 Data: <`1   V          > 60 31 93 00 A0 56 93 00 CD CD CD CD CD CD CD CD 
{36158} normal block at 0x009352E0, 44 bytes long.
 Data: < V   R          > B0 56 93 00 90 52 93 00 04 00 00 00 88 F2 17 1E 
{36157} normal block at 0x009356A0, 44 bytes long.
 Data: < R   W          > 80 52 93 00 00 57 93 00 CD CD CD CD CD CD CD CD 
{36153} normal block at 0x00935700, 44 bytes long.
 Data: < V   W          > A0 56 93 00 C0 57 93 00 CD CD CD CD CD CD CD CD 
{36150} normal block at 0x00935760, 40 bytes long.
 Data: < W   W          > D0 57 93 00 10 57 93 00 04 00 00 00 88 F2 17 1E 
{36149} normal block at 0x009357C0, 44 bytes long.
 Data: < W   M          > 00 57 93 00 E0 4D 93 00 CD CD CD CD CD CD CD CD 
{36146} normal block at 0x00933070, 44 bytes long.
 Data: < M   W          > F0 4D 93 00 D0 57 93 00 04 00 00 00 88 F2 17 1E 
{36145} normal block at 0x00934DE0, 44 bytes long.
 Data: < W  @N          > C0 57 93 00 40 4E 93 00 CD CD CD CD CD CD CD CD 
{36141} normal block at 0x00934E40, 48 bytes long.
 Data: < M   N          > E0 4D 93 00 A0 4E 93 00 CD CD CD CD CD CD CD CD 
{36139} normal block at 0x00934EA0, 48 bytes long.
 Data: <@N   D          > 40 4E 93 00 C0 44 93 00 CD CD CD CD CD CD CD CD 
{36137} normal block at 0x009344C0, 48 bytes long.
 Data: < N   E          > A0 4E 93 00 20 45 93 00 CD CD CD CD CD CD CD CD 
{36135} normal block at 0x00934520, 48 bytes long.
 Data: < D   E          > C0 44 93 00 80 45 93 00 CD CD CD CD CD CD CD CD 
{36133} normal block at 0x00934580, 48 bytes long.
 Data: < E   6          > 20 45 93 00 A0 36 93 00 CD CD CD CD CD CD CD CD 
{36131} normal block at 0x009336A0, 48 bytes long.
 Data: < E   $          > 80 45 93 00 F0 24 93 00 CD CD CD CD CD CD CD CD 
{36130} normal block at 0x009324F0, 148 bytes long.
 Data: < 6   I          > A0 36 93 00 E0 49 93 00 CD CD CD CD CD CD CD CD 
{36124} normal block at 0x009349E0, 40 bytes long.
 Data: < $  `           > F0 24 93 00 60 CC 92 00 CD CD CD CD CD CD CD CD 
{35866} normal block at 0x00932F00, 41 bytes long.
 Data: <p    I          > 70 CC 92 00 F0 49 93 00 04 00 00 00 88 F2 17 1E 
{35297} normal block at 0x0092CC60, 52 bytes long.
 Data: < I              > E0 49 93 00 B0 CB 92 00 CD CD CD CD CD CD CD CD 
{35295} normal block at 0x0092CAD0, 171 bytes long.
 Data: <    p           > C0 CB 92 00 70 CC 92 00 01 00 00 00 88 F2 17 1E 
{35293} normal block at 0x0092CBB0, 44 bytes long.
 Data: <`               > 60 CC 92 00 20 CD 92 00 CD CD CD CD CD CD CD CD 
{35284} normal block at 0x0092CCC0, 42 bytes long.
 Data: <0               > 30 CD 92 00 C0 CB 92 00 04 00 00 00 88 F2 17 1E 
{35283} normal block at 0x0092CD20, 44 bytes long.
 Data: <                > B0 CB 92 00 E0 CD 92 00 CD CD CD CD CD CD CD CD 
{35280} normal block at 0x0092CD80, 42 bytes long.
 Data: <    0           > F0 CD 92 00 30 CD 92 00 04 00 00 00 88 F2 17 1E 
{35279} normal block at 0x0092CDE0, 44 bytes long.
 Data: <                > 20 CD 92 00 A0 CE 92 00 CD CD CD CD CD CD CD CD 
{35276} normal block at 0x0092CE40, 38 bytes long.
 Data: <                > B0 CE 92 00 F0 CD 92 00 04 00 00 00 88 F2 17 1E 
{35275} normal block at 0x0092CEA0, 44 bytes long.
 Data: <    `           > E0 CD 92 00 60 CF 92 00 CD CD CD CD CD CD CD CD 
{35272} normal block at 0x0092CF00, 38 bytes long.
 Data: <p               > 70 CF 92 00 B0 CE 92 00 04 00 00 00 88 F2 17 1E 
{35271} normal block at 0x0092CF60, 44 bytes long.
 Data: <    `           > A0 CE 92 00 60 BC 92 00 CD CD CD CD CD CD CD CD 
{35268} normal block at 0x0092BC00, 38 bytes long.
 Data: <p   p           > 70 BC 92 00 70 CF 92 00 04 00 00 00 88 F2 17 1E 
{35267} normal block at 0x0092BC60, 44 bytes long.
 Data: <`               > 60 CF 92 00 20 BD 92 00 CD CD CD CD CD CD CD CD 
{35264} normal block at 0x0092BCC0, 37 bytes long.
 Data: <0   p           > 30 BD 92 00 70 BC 92 00 04 00 00 00 88 F2 17 1E 
{35263} normal block at 0x0092BD20, 44 bytes long.
 Data: <`               > 60 BC 92 00 E0 BD 92 00 CD CD CD CD CD CD CD CD 
{35260} normal block at 0x0092BD80, 37 bytes long.
 Data: <    0           > F0 BD 92 00 30 BD 92 00 04 00 00 00 88 F2 17 1E 
{35259} normal block at 0x0092BDE0, 44 bytes long.
 Data: <                > 20 BD 92 00 A0 BE 92 00 CD CD CD CD CD CD CD CD 
{35256} normal block at 0x0092BE40, 37 bytes long.
 Data: <                > B0 BE 92 00 F0 BD 92 00 04 00 00 00 88 F2 17 1E 
{35255} normal block at 0x0092BEA0, 44 bytes long.
 Data: <    0           > E0 BD 92 00 30 B0 92 00 CD CD CD CD CD CD CD CD 
{35253} normal block at 0x01076FD0, 1536 bytes long.
 Data: <               /> 00 00 00 00 00 00 00 00 00 00 00 00 81 A8 AB 2F 
{35251} normal block at 0x0107A018, 24576 bytes long.
 Data: <   Op`          > 00 E8 1E 4F 70 60 8C 00 00 00 00 00 00 00 00 00 
{35250} normal block at 0x0092CFC0, 41 bytes long.
 Data: <@               > 40 B0 92 00 B0 BE 92 00 04 00 00 00 88 F2 17 1E 
{35249} normal block at 0x0092B030, 44 bytes long.
 Data: <                > A0 BE 92 00 F0 B0 92 00 CD CD CD CD CD CD CD CD 
{35246} normal block at 0x0092B090, 40 bytes long.
 Data: <    @           > 00 B1 92 00 40 B0 92 00 04 00 00 00 88 F2 17 1E 
{35245} normal block at 0x0092B0F0, 44 bytes long.
 Data: <0               > 30 B0 92 00 B0 B1 92 00 CD CD CD CD CD CD CD CD 
{35242} normal block at 0x0092B150, 37 bytes long.
 Data: <                > C0 B1 92 00 00 B1 92 00 04 00 00 00 88 F2 17 1E 
{35241} normal block at 0x0092B1B0, 44 bytes long.
 Data: <                > F0 B0 92 00 10 B2 92 00 CD CD CD CD CD CD CD CD 
{35237} normal block at 0x0092B210, 44 bytes long.
 Data: <                > B0 B1 92 00 D0 B2 92 00 CD CD CD CD CD CD CD CD 
{35234} normal block at 0x0092B270, 38 bytes long.
 Data: <                > E0 B2 92 00 20 B2 92 00 04 00 00 00 88 F2 17 1E 
{35233} normal block at 0x0092B2D0, 44 bytes long.
 Data: <                > 10 B2 92 00 90 B3 92 00 CD CD CD CD CD CD CD CD 
{35230} normal block at 0x0092B330, 38 bytes long.
 Data: <                > A0 B3 92 00 E0 B2 92 00 04 00 00 00 88 F2 17 1E 
{35229} normal block at 0x0092B390, 44 bytes long.
 Data: <    P           > D0 B2 92 00 50 B4 92 00 CD CD CD CD CD CD CD CD 
{35226} normal block at 0x0092B3F0, 37 bytes long.
 Data: <`               > 60 B4 92 00 A0 B3 92 00 04 00 00 00 88 F2 17 1E 
{35225} normal block at 0x0092B450, 44 bytes long.
 Data: <                > 90 B3 92 00 B0 B4 92 00 CD CD CD CD CD CD CD CD 
{35221} normal block at 0x0092B4B0, 44 bytes long.
 Data: <P   p           > 50 B4 92 00 70 B5 92 00 CD CD CD CD CD CD CD CD 
{35218} normal block at 0x0092B510, 38 bytes long.
 Data: <                > 80 B5 92 00 C0 B4 92 00 04 00 00 00 88 F2 17 1E 
{35217} normal block at 0x0092B570, 44 bytes long.
 Data: <                > B0 B4 92 00 D0 B5 92 00 CD CD CD CD CD CD CD CD 
{35213} normal block at 0x0092B5D0, 44 bytes long.
 Data: <p   0           > 70 B5 92 00 30 B6 92 00 CD CD CD CD CD CD CD CD 
{35209} normal block at 0x0092B630, 44 bytes long.
 Data: <                > D0 B5 92 00 90 B6 92 00 CD CD CD CD CD CD CD CD 
{35205} normal block at 0x0092B690, 44 bytes long.
 Data: <0               > 30 B6 92 00 F0 B6 92 00 CD CD CD CD CD CD CD CD 
{35201} normal block at 0x0092B6F0, 44 bytes long.
 Data: <                > 90 B6 92 00 B0 B7 92 00 CD CD CD CD CD CD CD CD 
{35198} normal block at 0x0092B750, 42 bytes long.
 Data: <                > C0 B7 92 00 00 B7 92 00 04 00 00 00 88 F2 17 1E 
{35197} normal block at 0x0092B7B0, 44 bytes long.
 Data: <    p           > F0 B6 92 00 70 B8 92 00 CD CD CD CD CD CD CD CD 
{35194} normal block at 0x0092B810, 39 bytes long.
 Data: <                > 80 B8 92 00 C0 B7 92 00 04 00 00 00 88 F2 17 1E 
{35193} normal block at 0x0092B870, 44 bytes long.
 Data: <    0           > B0 B7 92 00 30 B9 92 00 CD CD CD CD CD CD CD CD 
{35190} normal block at 0x0092B8D0, 39 bytes long.
 Data: <@               > 40 B9 92 00 80 B8 92 00 04 00 00 00 88 F2 17 1E 
{35189} normal block at 0x0092B930, 44 bytes long.
 Data: <p               > 70 B8 92 00 F0 B9 92 00 CD CD CD CD CD CD CD CD 
{35186} normal block at 0x0092B990, 39 bytes long.
 Data: <    @           > 00 BA 92 00 40 B9 92 00 04 00 00 00 88 F2 17 1E 
{35185} normal block at 0x0092B9F0, 44 bytes long.
 Data: <0               > 30 B9 92 00 B0 BA 92 00 CD CD CD CD CD CD CD CD 
{35182} normal block at 0x0092BA50, 39 bytes long.
 Data: <                > C0 BA 92 00 00 BA 92 00 04 00 00 00 88 F2 17 1E 
{35181} normal block at 0x0092BAB0, 44 bytes long.
 Data: <    p           > F0 B9 92 00 70 BB 92 00 CD CD CD CD CD CD CD CD 
{35178} normal block at 0x0092BB10, 39 bytes long.
 Data: <                > 80 BB 92 00 C0 BA 92 00 04 00 00 00 88 F2 17 1E 
{35177} normal block at 0x0092BB70, 44 bytes long.
 Data: <                > B0 BA 92 00 E0 A2 92 00 CD CD CD CD CD CD CD CD 
{35174} normal block at 0x0092A280, 39 bytes long.
 Data: <                > F0 A2 92 00 80 BB 92 00 04 00 00 00 88 F2 17 1E 
{35173} normal block at 0x0092A2E0, 44 bytes long.
 Data: <p               > 70 BB 92 00 A0 A3 92 00 CD CD CD CD CD CD CD CD 
{35170} normal block at 0x0092A340, 39 bytes long.
 Data: <                > B0 A3 92 00 F0 A2 92 00 04 00 00 00 88 F2 17 1E 
{35169} normal block at 0x0092A3A0, 44 bytes long.
 Data: <                > E0 A2 92 00 00 BF 92 00 CD CD CD CD CD CD CD CD 
{35164} normal block at 0x0092BF00, 44 bytes long.
 Data: <    `           > A0 A3 92 00 60 BF 92 00 CD CD CD CD CD CD CD CD 
{35160} normal block at 0x0092BF60, 44 bytes long.
 Data: <                > 00 BF 92 00 C0 BF 92 00 CD CD CD CD CD CD CD CD 
{35156} normal block at 0x0092BFC0, 44 bytes long.
 Data: <`   p           > 60 BF 92 00 70 A0 92 00 CD CD CD CD CD CD CD CD 
{35152} normal block at 0x0092A070, 44 bytes long.
 Data: <                > C0 BF 92 00 D0 A0 92 00 CD CD CD CD CD CD CD CD 
{35147} normal block at 0x0092A0D0, 48 bytes long.
 Data: <p   0           > 70 A0 92 00 30 A1 92 00 CD CD CD CD CD CD CD CD 
{35145} normal block at 0x0092A130, 48 bytes long.
 Data: <                > D0 A0 92 00 90 A1 92 00 CD CD CD CD CD CD CD CD 
{35143} normal block at 0x0092A190, 48 bytes long.
 Data: <0               > 30 A1 92 00 F0 A1 92 00 CD CD CD CD CD CD CD CD 
{35141} normal block at 0x0092A1F0, 48 bytes long.
 Data: <                > 90 A1 92 00 10 A6 92 00 CD CD CD CD CD CD CD CD 
{35139} normal block at 0x0092A610, 48 bytes long.
 Data: <    p           > F0 A1 92 00 70 A6 92 00 CD CD CD CD CD CD CD CD 
{35137} normal block at 0x0092A670, 48 bytes long.
 Data: <                > 10 A6 92 00 00 A4 92 00 CD CD CD CD CD CD CD CD 
{35134} normal block at 0x0092A400, 48 bytes long.
 Data: <p   `           > 70 A6 92 00 60 A4 92 00 CD CD CD CD CD CD CD CD 
{35132} normal block at 0x0092A460, 48 bytes long.
 Data: <                > 00 A4 92 00 C0 A4 92 00 CD CD CD CD CD CD CD CD 
{35130} normal block at 0x0092A4C0, 48 bytes long.
 Data: <`               > 60 A4 92 00 20 A5 92 00 CD CD CD CD CD CD CD CD 
{35128} normal block at 0x0092A520, 48 bytes long.
 Data: <                > C0 A4 92 00 80 A5 92 00 CD CD CD CD CD CD CD CD 
{35126} normal block at 0x0092A580, 48 bytes long.
 Data: <                > 20 A5 92 00 D0 A6 92 00 CD CD CD CD CD CD CD CD 
{35123} normal block at 0x0092A6D0, 48 bytes long.
 Data: <    0           > 80 A5 92 00 30 A7 92 00 CD CD CD CD CD CD CD CD 
{35121} normal block at 0x0092A730, 48 bytes long.
 Data: <                > D0 A6 92 00 90 A7 92 00 CD CD CD CD CD CD CD CD 
{35119} normal block at 0x0092A790, 48 bytes long.
 Data: <0               > 30 A7 92 00 F0 A7 92 00 CD CD CD CD CD CD CD CD 
{35117} normal block at 0x0092A7F0, 48 bytes long.
 Data: <    P           > 90 A7 92 00 50 A8 92 00 CD CD CD CD CD CD CD CD 
{35115} normal block at 0x0092A850, 48 bytes long.
 Data: <                > F0 A7 92 00 B0 A8 92 00 CD CD CD CD CD CD CD CD 
{35113} normal block at 0x0092A8B0, 48 bytes long.
 Data: <P               > 50 A8 92 00 10 A9 92 00 CD CD CD CD CD CD CD CD 
{35112} normal block at 0x0092A910, 148 bytes long.
 Data: <                > B0 A8 92 00 90 80 92 00 CD CD CD CD CD CD CD CD 
{35068} normal block at 0x00929A10, 44 bytes long.
 Data: < m  @n          > E0 6D 93 00 40 6E 93 00 CD CD CD CD CD CD CD CD 
{35056} normal block at 0x00928090, 40 bytes long.
 Data: <     L          > 10 A9 92 00 E0 4C 92 00 CD CD CD CD CD CD CD CD 
{34705} normal block at 0x00924CE0, 40 bytes long.
 Data: <                > 90 80 92 00 B0 AC 15 1E CD CD CD CD CD CD CD CD 
{34589} normal block at 0x00922BF0, 39 bytes long.
 Data: < J   L          > F0 4A 90 00 F0 4C 92 00 03 00 00 00 88 F2 17 1E 
{34101} normal block at 0x00904AF0, 40 bytes long.
 Data: <P    +          > 50 E9 8F 00 F0 2B 92 00 04 00 00 00 88 F2 17 1E 
{33763} normal block at 0x008FE950, 36 bytes long.
 Data: <     J          > 10 C3 8F 00 F0 4A 90 00 04 00 00 00 88 F2 17 1E 
{33696} normal block at 0x008FC310, 37 bytes long.
 Data: <p   P           > 70 C3 8F 00 50 E9 8F 00 04 00 00 00 88 F2 17 1E 
{33695} normal block at 0x008FC370, 39 bytes long.
 Data: <                > C0 20 8E 00 10 C3 8F 00 04 00 00 00 88 F2 17 1E 
{33048} normal block at 0x008E20C0, 55 bytes long.
 Data: <@%  p           > 40 25 8E 00 70 C3 8F 00 01 00 00 00 88 F2 17 1E 
{33043} normal block at 0x008E24E0, 24 bytes long.
 Data: <                > E0 A0 92 00 80 A0 92 00 02 00 00 00 D0 D2 16 1E 
{36194} normal block at 0x008E22E0, 64 bytes long.
 Data: < ?   ]  @Z   W  > D0 3F 8C 00 90 5D 8B 00 40 5A 8B 00 F0 57 8B 00 
{33041} normal block at 0x008E2530, 52 bytes long.
 Data: <p   p#          > 70 1B 8E 00 70 23 8C 00 85 FF FF FF CD CD CD CD 
{33039} normal block at 0x008E2350, 255 bytes long.
 Data: < $  @%          > 90 24 8E 00 40 25 8E 00 01 00 00 00 88 F2 17 1E 
{33037} normal block at 0x008E2480, 44 bytes long.
 Data: <P   p           > 50 08 8C 00 70 17 8E 00 85 FF FF FF CD CD CD CD 
{33028} normal block at 0x008E2590, 41 bytes long.
 Data: < &   $          > 00 26 8E 00 90 24 8E 00 04 00 00 00 88 F2 17 1E 
{33027} normal block at 0x008E25F0, 44 bytes long.
 Data: <p   p'          > 70 14 8E 00 70 27 8E 00 85 FF FF FF CD CD CD CD 
{33024} normal block at 0x008E2650, 42 bytes long.
 Data: < &   &          > C0 26 8E 00 00 26 8E 00 04 00 00 00 88 F2 17 1E 
{33023} normal block at 0x008E26B0, 44 bytes long.
 Data: <P               > 50 16 8E 00 B0 16 8E 00 85 FF FF FF CD CD CD CD 
{33020} normal block at 0x008E2710, 40 bytes long.
 Data: < '   &          > 80 27 8E 00 C0 26 8E 00 04 00 00 00 88 F2 17 1E 
{33019} normal block at 0x008E2770, 44 bytes long.
 Data: < %  P           > F0 25 8E 00 50 16 8E 00 85 FF FF FF CD CD CD CD 
{33016} normal block at 0x008E27D0, 36 bytes long.
 Data: <0(   '          > 30 28 8E 00 80 27 8E 00 04 00 00 00 88 F2 17 1E 
{33015} normal block at 0x008E2820, 44 bytes long.
 Data: <                > 20 13 8E 00 E0 13 8E 00 85 FF FF FF CD CD CD CD 
{33012} normal block at 0x008E2880, 37 bytes long.
 Data: < (  0(          > F0 28 8E 00 30 28 8E 00 04 00 00 00 88 F2 17 1E 
{33011} normal block at 0x008E28E0, 44 bytes long.
 Data: <`               > 60 12 8E 00 A0 11 8E 00 85 FF FF FF CD CD CD CD 
{33007} normal block at 0x008E10E0, 44 bytes long.
 Data: <`/              > 60 2F 8E 00 F0 15 8E 00 85 FF FF FF CD CD CD CD 
{33003} normal block at 0x008E1140, 44 bytes long.
 Data: <                > B0 16 8E 00 80 13 8E 00 85 FF FF FF CD CD CD CD 
{32999} normal block at 0x008E11A0, 44 bytes long.
 Data: < (              > E0 28 8E 00 D0 14 8E 00 85 FF FF FF CD CD CD CD 
{32995} normal block at 0x008E1200, 44 bytes long.
 Data: <     -          > 90 15 8E 00 F0 2D 8E 00 85 FF FF FF CD CD CD CD 
{32993} normal block at 0x008E2940, 768 bytes long.
 Data: <              g > 00 00 00 00 00 00 00 00 00 00 00 00 C1 06 67 CF 
{32991} normal block at 0x008E2C70, 39 bytes long.
 Data: < ,              > E0 2C 8E 00 10 12 8E 00 04 00 00 00 88 F2 17 1E 
{32990} normal block at 0x008E2CD0, 44 bytes long.
 Data: <    P.          > E0 13 8E 00 50 2E 8E 00 85 FF FF FF CD CD CD CD 
{32987} normal block at 0x008E2D30, 42 bytes long.
 Data: < -   ,          > A0 2D 8E 00 E0 2C 8E 00 04 00 00 00 88 F2 17 1E 
{32986} normal block at 0x008E2D90, 44 bytes long.
 Data: < /  P           > 00 2F 8E 00 50 10 8E 00 85 FF FF FF CD CD CD CD 
{32982} normal block at 0x008E2DF0, 44 bytes long.
 Data: <     /          > 00 12 8E 00 C0 2F 8E 00 85 FF FF FF CD CD CD CD 
{32978} normal block at 0x008E2E50, 44 bytes long.
 Data: < ,  `/          > D0 2C 8E 00 60 2F 8E 00 85 FF FF FF CD CD CD CD 
{32975} normal block at 0x008E2EB0, 24 bytes long.
 Data: < /  `.          > 10 2F 8E 00 60 2E 8E 00 02 00 00 00 D0 D2 16 1E 
{32972} normal block at 0x008E2F00, 48 bytes long.
 Data: <     -          > D0 14 8E 00 90 2D 8E 00 85 FF FF FF CD CD CD CD 
{32970} normal block at 0x008E2F60, 48 bytes long.
 Data: <P.              > 50 2E 8E 00 E0 10 8E 00 85 FF FF FF CD CD CD CD 
{32968} normal block at 0x008E2FC0, 48 bytes long.
 Data: < -  0           > F0 2D 8E 00 30 15 8E 00 85 FF FF FF CD CD CD CD 
{32966} normal block at 0x008E1470, 48 bytes long.
 Data: <     %          > C0 AC 15 1E F0 25 8E 00 85 FF FF FF CD CD CD CD 
{32964} normal block at 0x008E14D0, 48 bytes long.
 Data: <     /          > A0 11 8E 00 00 2F 8E 00 85 FF FF FF CD CD CD CD 
{32962} normal block at 0x008E1050, 48 bytes long.
 Data: < -              > 90 2D 8E 00 20 13 8E 00 85 FF FF FF CD CD CD CD 
{32959} normal block at 0x008E1260, 48 bytes long.
 Data: <     (          > 10 17 8E 00 E0 28 8E 00 85 FF FF FF CD CD CD CD 
{32957} normal block at 0x008E12C0, 48 bytes long.
 Data: <0   p           > 30 15 8E 00 70 08 8E 00 85 FF FF FF CD CD CD CD 
{32955} normal block at 0x008E1320, 48 bytes long.
 Data: <P    (          > 50 10 8E 00 20 28 8E 00 85 FF FF FF CD CD CD CD 
{32953} normal block at 0x008E1380, 48 bytes long.
 Data: <@               > 40 11 8E 00 10 17 8E 00 85 FF FF FF CD CD CD CD 
{32951} normal block at 0x008E13E0, 48 bytes long.
 Data: < (   ,          > 20 28 8E 00 D0 2C 8E 00 85 FF FF FF CD CD CD CD


From koancui at lucent.com  Wed Jun 26 01:29:48 2002
From: koancui at lucent.com (Cui, Zhi Gang (Zhigang))
Date: Wed, 26 Jun 2002 13:29:48 +0800
Subject: subscriber
Message-ID: <31C0F08B0D18D511ACC800508BAE7B470159B91F@CI0026EXCH001U>

THX




From msoulier at nortelnetworks.com_.nospam  Thu Jun 13 11:19:21 2002
From: msoulier at nortelnetworks.com_.nospam (Michael P. Soulier)
Date: 13 Jun 2002 15:19:21 GMT
Subject: import problem
References: <2se7guc3kkmm3bgooj776j55igpkmgpst4@4ax.com>
Message-ID: <slrnaghe2q.12v.msoulier@pmerd071.ca.nortel.com>

On Sun, 09 Jun 2002 13:37:05 -0700, Michael Eager <eager at eagercon.com> wrote:
> I'm trying to use Python 2.2 and PyQt on a RedHat-7.2 system.
> 
> I get an error when I execute "from qt import *" -- there is a failure
> in qt.py when it executes "import libsup".  The error is "No module
> named libsup".
> 
> libsup.so.9 is a link to libsup.so.9.0.3 and is in
> /usr/lib/python2.2/site-packages.  If I enter "import libsup", I get
> the same error message.
> 
> Any suggestions?

    Are you sure that you have the required dependencies for qt.py? Python
does not import .so files directly, to my knowledge.

    Mike

-- 
Michael P. Soulier, QX41, SKY  Tel: 613-765-4699 (ESN: 39-54699)
Optical Networks, Nortel Networks, SDE Pegasus
"...the word HACK is used as a verb to indicate a massive amount
of nerd-like effort."  -Harley Hahn, A Student's Guide to Unix


From walterm at parque.homelinux.net  Tue Jun 18 09:32:06 2002
From: walterm at parque.homelinux.net (Walter Moreira)
Date: Tue, 18 Jun 2002 10:32:06 -0300
Subject: __subclasses__
In-Reply-To: <lk4rg6matg.fsf@pc150.maths.bris.ac.uk>
References: <a9bcagho74.22t.jimrowe@optilink.com> <lk4rg6matg.fsf@pc150.maths.bris.ac.uk>
Message-ID: <20020618133206.GA19082@parque.homelinux.net>

On Fri, Jun 14, 2002 at 10:09:31AM +0000, Michael Hudson wrote:
> > class Foo(object):
> >     def __init__(self):
> >         object.__init__(self)
> > 
> > class Bar(Foo):
> > 	pass
> > 
> > dir(Foo) # returns:
> > 
> >  ['__class__', '__delattr__', '__dict__', '__doc__',
> >  '__getattribute__', '__hash__', '__init__', '__module__',
> >  '__new__', '__reduce__', '__repr__', '__setattr__',
> >  '__str__', '__weakref__']
> 
> Whether things show up in dir() is a bit of a lottery (think about
> __getattr__ methods!).  Notice __bases__ isn't in there either, for
> example.

The __bases__ and __subclasses__ (and others) are listed in the 'type'
object:

    >>> class Foo:
    ...     pass
    ...
    >>> dir(type(type(Foo)))
    ['__base__', '__bases__', '__basicsize__', '__call__', '__class__',
    '__cmp__', '__delattr__', '__dict__', '__dictoffset__', '__doc__',
    '__flags__', '__getattribute__', '__hash__', '__init__', '__itemsize__',
    '__module__', '__mro__', '__name__', '__new__', '__reduce__',
    '__repr__', '__setattr__', '__str__', '__subclasses__',
    '__weakrefoffset__', 'mro']

Walter




From jh at web.de  Mon Jun  3 15:40:13 2002
From: jh at web.de (Jürgen Hermann)
Date: Mon, 3 Jun 2002 21:40:13 +0200
Subject: python and windows and pydoc and favorite editor == True
References: <un0ugnxa3.fsf@online.no>
Message-ID: <adggk2$psv$00$1@news.t-online.com>

"Syver Enstad" <syver-en+usenet at online.no> schrieb im Newsbeitrag
news:un0ugnxa3.fsf at online.no...
> Post if there is any interest.

OK, post. :)




From sholden at holdenweb.com  Thu Jun 13 22:50:54 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Thu, 13 Jun 2002 22:50:54 -0400
Subject: libftp.size function gives wrong results!
References: <RNKM8.560$Yw3.96921@news20.bellglobal.com> <3D038A19.D991A885@engcorp.com> <%ULM8.903$Yw3.118334@news20.bellglobal.com>
Message-ID: <AQcO8.9665$Fp1.7818@atlpnn01.usenetserver.com>

"Michael Davis" <michael at damaru.com> wrote ...
> Peter Hansen wrote:
>
> > Michael Davis wrote:
> >>
> >> I'm writing a python program which will upload or download files via
ftp.
> >> It's intended to be used to deploy a web site from a local machine to a
> >> server. It tries to determine which files are modified on the local
> >> machine by comparing file sizes and modification times.
> >>
> >> I try to compare files using os.stat for the local file and FTP.size
for
> >> the remote. The problem is, FTP.size is giving me incorrect results. If
I
> >> use the ftp command line client, and look at one of my files on the
> >> server (the output looks like 'ls -l') then I can see that the file
size,
> >> is, say, 1200. But the FTP.size function returns, say, 1242.
> >
> > I think you'll find the number of lines in that file is 42.  The
> > difference
> > is probably between CRLF and LF line terminations in the file.  If you
> > transfer files with FTP and want to preserve the line endings, you have
> > to specify the binary type, not ascii.  Don't know how to do this with
> > ftplib, but from the ftp command line you would use 'type binary' or
> > 'type ascii' to switch.
>
> Thanks Peter,
>
> I had already thought of that. Both my local and the remote systems are
> unix, and so don't have CRs. But you're right - the difference in size is
> the same as the number of lines, as though the ftp size command assumed
> that I wanted to convert the end-of-lines to CRLF. I had specified binary
> mode transfers. And if I do 'ftp size' on true binary files, the same
> problem happens!
>
> Anyway, I just found out that one of the servers I need to talk to doesn't
> allow me to use 'size' anyway, so I'm going to ask a different python
> question in a different message.
>

You should take a look at "ftpmirror.py" in the Tools subdirectory of your
Python distribution.

Also, you might find it easier to use "make" to determoine what needs
uploading to you web system. This has worked very well for me.

regards
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------







From gerhard.haering at gmx.de  Tue Jun 25 20:56:18 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 26 Jun 2002 02:56:18 +0200
Subject: Reminder of a python Project.
In-Reply-To: <2218.941T781T12224051threeseas@earthlink.net>
References: <2218.941T781T12224051threeseas@earthlink.net>
Message-ID: <20020626005618.GA1145@lilith.my-fqdn.de>

* Timothy Rue <threeseas at earthlink.net> [2002-06-26 00:40 +0000]:
> First off, it's ok that I got such a high level of resistance to my
> initial posting here about this project.

ACK.

> Thanks to some research that /.

That's an oxymoron.

> had an article on, regarding the size of GPL project teams.
                                           ^^^^^^^^^^^^^^^^^

What's that?  Somebody mistaking GPL with Free Software with Open
Source? Again?

> But anyway.... Remember the autocoding project I mentioned?

Yes, unfortunately.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From kragen at pobox.com  Sat Jun  1 02:17:33 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 01 Jun 2002 02:17:33 -0400
Subject: The email package and KLEZ mails
References: <oqn0ukzfsu.fsf@titan.progiciels-bpi.ca> <mailman.1022585747.21423.python-list@python.org> <acvmf3.t8.1@kserver.org> <mailman.1022725133.16968.python-list@python.org>
Message-ID: <834rgnwmiq.fsf@panacea.canonical.org>

pinard at iro.umontreal.ca (Fran?ois Pinard) writes:
> For example, I expect compilers to raise diagnostics and help me at being
> strict, because being overly liberal for a compiler is just not helpful.
> Another example, a sad one, is the messy state of HTML all around us,
> it comes from browsers having been by far too liberal, and for too long.

If you had designed the early web browsers, the web never would have
caught on, as indeed many other networked hypertext systems predating
the Web did not; being liberal in what you accept is a crucial
principle in building decentralized systems, and its embodiment in the
Web made the Web possible.

Having "bad HTML" warnings is, of course, extremely helpful.  It makes
it possible to be conservative in what you send.



From jubafre at zipmail.com.br  Mon Jun 24 07:45:52 2002
From: jubafre at zipmail.com.br (jubafre at zipmail.com.br)
Date: Mon, 24 Jun 2002 08:45:52 -0300
Subject: =?iso-8859-1?Q?module=20py2exe?=
Message-ID: <3D163E400000171F@www.zipmail.com.br>

i don?t know how to use a py2exe module, i have made the download but i
don?t know to install. the direcroty of installation is lib\site-packages.

i have a program using tkinter module and i want to compile to .exe the
name of the file is lfa.py, i read in http://starship.python.net/crew/theller/py2exe/,
i have to make a setup.py, so doesn?t work, how i can make this rigth???????????????????

# setup.py
from distutils.core import setup
import py2exe

setup(name="lfa",
      scripts=["lfa.py"],
)

Juliano Freitas
www.gebrasil.hpg.com.br




------------------------------------------
Use o melhor sistema de busca da Internet
Radar UOL - http://www.radaruol.com.br







From d2002xx at myrealbox.com  Sun Jun 23 04:31:12 2002
From: d2002xx at myrealbox.com (d2002xx)
Date: 23 Jun 2002 01:31:12 -0700
Subject: Python hits the spot
References: <3D1482AE.8FA299D8@kfunigraz.ac.at> <m21yazxtgs.fsf@mother.paradise.lost> <3D148E71.2070201@kfunigraz.ac.at> <m2wusrw76y.fsf@mother.paradise.lost>
Message-ID: <8d3f4438.0206230031.7838d0d7@posting.google.com>

John Hunter <jdhunter at nitace.bsd.uchicago.edu> wrote in message > unix> at now +6min
> at> killall myprog
> at> CTRL-d                # Hit Control and D together
> unix> ./myprog

There is a faster method named SAK in linux, it can kill any processes
in current console (i.e. also kill X-window)

the document is in:

1.YOUR_LINUX_KERNEL_SOURCE_PATH/Docmentation/sysrq.txt
2.YOUR_LINUX_KERNEL_SOURCE_PATH/Docmentation/SAK.txt

Be careful:
The command is executed by kernel and send SIGKILL signal to kill
processes, the signal can't be caught, so the process killed may lost
some things such as working files which was not saved. And you can't
use it with enabling the virtual screen in X-windows (I think few
people use it, most people use virtual desktop instead).
Also, if you start X-window in the first console, some daemons may be
killed, it's something boring...... (although I have not tested)

Good luck!


From mhammond at skippinet.com.au  Fri Jun 14 00:38:13 2002
From: mhammond at skippinet.com.au (Mark Hammond)
Date: Fri, 14 Jun 2002 04:38:13 GMT
Subject: PythonWin Problem!
References: <1a52c7f1.0206131455.6df50cf8@posting.google.com>
Message-ID: <3D0973C5.60407@skippinet.com.au>

Ethan wrote:
> I have been using ActiveState's port of Python for a while. PythonWin
> has been very helpful. Since a couple of days PythonWin developed a
> problem. When I start it, it appears in the taskbar and tray but does
> not appear on the screen.
> 
> My Win98 did recently develop a problem which prevents me from getting
> the right click menu on the taskbar entries. This was secondary to
> some registry cleaning I did. I am not sure if this is related to
> Pythonwin problem but all others apps are working fine. I do get the
> menu for the tray icon and the "Close Program" window does not say
> "not responding"
> 
> I uninstalled Python - deleted the Python21 folder and reinstalled. It
> did not help. Any suggestions? I really prefer PythonWin over IDLE.

Pythonwin has probably somehow remembered it's screen location as off 
the screen.

Try deleting the "HKLM\Software\Python 2.1\Python for Win32" key from 
the registry, and it should come back.

Alternatively, from the task-bar select "Move", then use the arrow keys 
to try and bring it back on screen (presumably with the "Left" and "Up" 
keys)

Mark.



From rganesan at myrealbox.com  Wed Jun 19 23:20:56 2002
From: rganesan at myrealbox.com (Ganesan R)
Date: 20 Jun 2002 08:50:56 +0530
Subject: Variable Interpolation - status of PEP 215
Message-ID: <ouit4e63fr.fsf@andlx-anamika.cisco.com>

Hi,

I am a relative newbie to Python. I have a reasonably good perl background
(don't worry I am not about to start a language flame war :-), and one of
the things I really miss is variable interpolation in strings. I searched
the net and found PEP 215 (http://www.python.org/peps/pep-0215.html). This
seems to be exactly what I need. The proposal seems to be quite old, so I'd
like to know if it's still in consideration or has it been dropped
completely?

Ganesan

-- 
Ganesan R 



From cliechti at gmx.net  Thu Jun 13 14:36:28 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 13 Jun 2002 20:36:28 +0200
Subject: Embedded systems
References: <aeaajv$5db01$1@ID-17980.news.dfncis.de>
Message-ID: <Xns922CD20A73219cliechtigmxnet@62.2.16.82>

"Stuart" <stuart at stu.org.uk> wrote in
news:aeaajv$5db01$1 at ID-17980.news.dfncis.de: 

> Is Python practical for embedded systems?   I am working on a project
> at the moment that uses a GSM/GPRS modem and a GPS receiver.  I have
> prototyped the whole thing in Visual Basic, and now I need to port it
> over to a suitable processor (possibly on a 68000 device using Linux).

python need some resources as its interpreted but there is a posrt for the 
palm (http://sourceforge.net/projects/pippy). as you have linux there 
shouldn't be a real problem. download the sources and try to compile it fo  
your target system.

> I don't really want  to get into C for this project, and was wondering
> if Python was suitable?  Are memory requirements an issue?

i don't know any precise numbers, depending on project maybe starting at 
100k RAM and 200k ROM? (pippy uses 137k in ROM)

chris


-- 
Chris <cliechti at gmx.net>



From gerhard at bigfoot.de  Wed Jun 19 04:38:47 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 19 Jun 2002 10:38:47 +0200
Subject: procmail replacement in Python
In-Reply-To: <3u8pea.617.ln@127.0.0.1>
References: <mailman.1024444823.19953.python-list@python.org> <3u8pea.617.ln@127.0.0.1>
Message-ID: <20020619083847.GA9664@lilith.my-fqdn.de>

* Radovan Garabik <spam at melkor.dnp.fmph.uniba.sk> [2002-06-19 08:38 +0200]:
> Gerhard H?ring <gh_pythonlist at gmx.de> wrote:
>  : Does such a thing exist? I've recently looked into the procmail sources
> 
> <shameless advertising plug>
> http://melkor.dnp.fmph.uniba.sk/~garabik/pycmail.html
> </shameless advertising plug>

Whoa! _Exactly_ what I was searching. Would you mind registering it at
Freshmeat, so other people can find it more easily?

Looks like a good project for my next weekend to convert my
fetchmail/procmail setup to getmail/pycmail.

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 26.6 ?C      Wind: 1.3 m/s




From peter at engcorp.com  Sun Jun  2 07:50:14 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 02 Jun 2002 07:50:14 -0400
Subject: Pop a list from beginning ? and memory saving...
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de>
Message-ID: <3CFA0676.68F80DE6@engcorp.com>

Shagshag13 wrote:
> 
> I had a huge list containing consuming data, as i only need a one pass in a
> for statement, i'm wondering if by doing this with a pop i could save memory ? 

Maybe, maybe not.  The implementation _probably_ makes a copy of
the entire list at some point as you shrink it.  That would mean that
at least at some interval, you would temporarily consume almost twice
the memory of the original list as the original list is copied, then 
freed.  You might end up with less memory used, but with double the 
memory needed.  Of course, the same would have held true as you built
the list in the first place...

Why are you trying to save memory?  Are you really working with such
a large amount of data in one chunk that you will run out?  Maybe
there is some other approach you haven't thought of.  This effort
almost smells like premature optimization (the root of all evil).

-Peter


From oren-py-l at hishome.net  Mon Jun 24 16:01:40 2002
From: oren-py-l at hishome.net (Oren Tirosh)
Date: Mon, 24 Jun 2002 23:01:40 +0300
Subject: PEP 294: Type Names in the types Module
Message-ID: <20020624230140.B3555@hishome.net>

PEP: 294
Title: Type Names in the types Module
Version: $Revision: 1.1 $
Last-Modified: $Date: 2002/06/23 23:52:19 $
Author: oren at hishome.net (Oren Tirosh)
Status: Draft
Type: Standards track
Created: 19-Jun-2002
Python-Version: 2.3
Post-History:


Abstract

    This PEP proposes that symbols matching the type name should be
    added to the types module for all basic Python types in the types
    module:

        types.IntegerType -> types.int
        types.FunctionType -> types.function
        types.TracebackType -> types.traceback
         ...    

    The long capitalized names currently in the types module will be
    deprecated.

    With this change the types module can serve as a replacement for
    the new module.  The new module shall be deprecated and listed in
    PEP 4.


Rationale

    Using two sets of names for the same objects is redundant and
    confusing.

    In Python versions prior to 2.2 the symbols matching many type
    names were taken by the factory functions for those types.  Now
    all basic types have been unified with their factory functions and
    therefore the type names are available to be consistently used to
    refer to the type object.

    Most types are accessible as either builtins or in the new module
    but some types such as traceback and generator are only accssible
    through the types module under names which do not match the type
    name.  This PEP provides a uniform way to access all basic types
    under a single set of names.


Specification

    The types module shall pass the following test:

        import types
        for t in vars(types).values():
            if type(t) is type:
                assert getattr(types, t.__name__) is t

    The types 'class', 'instance method' and 'dict-proxy' have already
    been renamed to the valid Python identifiers 'classobj',
    'instancemethod' and 'dictproxy', making this possible.


Backward compatibility

    Because of their widespread use it is not planned to actually
    remove the long names from the types module in some future
    version.  However, the long names should be changed in
    documentation and library sources to discourage their use in new
    code.


Reference Implementation
 
    A reference implementation is available in SourceForge patch
    #569328: http://www.python.org/sf/569328
  

Copyright

    This document has been placed in the public domain.





From python at tykebsd.net  Fri Jun  7 03:55:16 2002
From: python at tykebsd.net (Dave Moor)
Date: 7 Jun 2002 00:55:16 -0700
Subject: Interfacing to 1-Wire LAN, especially Dallas temperature sensors?
References: <a7iefuk650noamrqcmpvroqj60mtgbpf9v@4ax.com> <1bf96e7f.0206060356.417a7394@posting.google.com> <fqkufuoljmafu6si319bhmt3f83pr587uc@4ax.com>
Message-ID: <1bf96e7f.0206062355.4718e1c1@posting.google.com>

Hi Everyone

Since one or two people are interested in this I have put a page on my
website (http://www.tykebsd.net) with some info and some files to
download.  This module is only in its initial stages of developement
so has limited functionality.  There is a Linux and Windows version of
the module.

The module is currently only compiled against Python version 2.1 if
you need a version for Python 2.2 you will have to wait until the
middle of next week.  I plan to upgrade but I don't think I'll have
time before then. If you have any questions just drop me a line and
I'll try to answer them.

Dave Moor


From irmen at NOSPAMREMOVETHISxs4all.nl  Sun Jun 16 12:21:54 2002
From: irmen at NOSPAMREMOVETHISxs4all.nl (Irmen de Jong)
Date: Sun, 16 Jun 2002 18:21:54 +0200
Subject: ANN: pyro 3.0 beta 1
Message-ID: <mailman.1024324999.4143.clpa-moderators@python.org>

I've released the first beta of Pyro 3.0.
It has some very exiting new features over the previous release (2.8);
major speed increase, SSL support, safe XML pickling, authentication,
timeouts, and much more.

If you're interested in helping me improve Pyro 3.0,
download it from http://www.sourceforge.net/projects/pyro

Please read the change log carefully (available on request,
but it's also in the manual in the distribution), and test
it to the max!

Please remember, it is still a beta release so there will be some
rough edges, and there are still a few issues that need to be resolved
for the final 3.0. Check the Pyro mailing list to stay up-to-date.

Irmen de Jong
  - irmen(at)users.sourceforge.net


What is Pyro?:
Pyro is an acronym for PYthon Remote Objects. It is a basic Distributed 
Object Technology system written entirely in Python. With this, it 
closely resembles Java's Remote Method Invocation (RMI). Pyro is small, 
simple, fun and free!

For more information, visit http://pyro.sourceforge.net





From zopestoller at thomas-guettler.de  Thu Jun 27 07:51:56 2002
From: zopestoller at thomas-guettler.de (Thomas Guettler)
Date: Thu, 27 Jun 2002 13:51:56 +0200
Subject: print() with unicode strings
Message-ID: <3D1AFC5C.2040708@thomas-guettler.de>

Hi!

 >>>print '?' #(German Umlaut)

works

but

 >>>print u'?'

does not.

Is there a reason why print does not accept unicode?
Why not default to a latin-1 encoding?

Version:

Python 2.2 (#1, Dec 31 2001, 15:21:18)
[GCC 2.95.3-5 (cygwin special)] on cygwin

thomas



From phlip_cpp at yahoo.com  Sun Jun  2 22:26:24 2002
From: phlip_cpp at yahoo.com (Phlip)
Date: 03 Jun 2002 02:26:24 GMT
Subject: cursors, etc
References: <20020602221459.05355.00002083@mb-mm.aol.com>
Message-ID: <adek4g$3cp@dispatch.concentric.net>

BCBOOKER wrote:

> Is there any way for a user to actively move a cursor or some other marker
> on a
> plane? 

Not until you start naming specifics. What's this "plane"? Is it a window? 
If so, is it a Tkinter Canvas?

Contrarily, is it a virtual terminal? such as the curses module addresses?

> That is, if there is a 4x4 square with the cursor at (1,1), can
> the
> user press up, n, 1 or whatever and make the x increase?  I tried making
> something along the lines of this, but i had to go type in every situation
> - that is, make the map for (2, 3) or (1, 2).

So if you had variable row and col, could you let the Up key trigger a row 
-= 1, for example?

-- 
  Phlip
               http://www.c2.com/cgi/wiki?PhlIp
  --  Creationism: The belief that God is too stupid to be able
      to invent a Universe where life can spontaneously arise  --



From jimmy at retzlaff.com  Fri Jun  7 20:06:47 2002
From: jimmy at retzlaff.com (Jimmy Retzlaff)
Date: Fri, 7 Jun 2002 17:06:47 -0700
Subject: Efficient python programming...
Message-ID: <A7B5C22F20EE3C4E95744E2F03F1887F56A6@MAYOR.retzlaff.com>

Peter Hansen [mailto:peter at engcorp.com] writes:
> I disagree, if the tests are written first (and run, proving that
> they will fail when the code is wrong or missing).

I think the other perspective here is that the writing of the unit tests
is design/coding, and hence subject to the same types of mistakes as the
designing/writing of the code they test. A contrived example:


def addEmUp(x, y):
    return x * y

def testAddEmUp():
    # try some limit cases and something in between
    for testCase in (0, 2, 1.79e+308):
        assert addEmUp(testCase, testCase) == testCase+testCase

if __name__ == '__main__':
    testAddEmUp()
    print 'Success!!!'


That test runs successfully on Python 2.2.1 on Windows despite the
flawed implementation of the function the test is exercising. Having
unit tests may make you more confident in correctness than not having
them, but they are still prone to mistakes themselves. In order for unit
tests to prove the code is correct, you'd need to prove the unit tests
are correct and exhaustive (despite its correctness, the above unit test
is not exhaustive which leads to a false validation).

> The tests are executable requirement specifications.  Running
> them "proves" (yes, not 100%... nothing could) that the code
> meets the specs.  The specs could be wrong of course, but that
> doesn't mean the code is broken...

You can say the tests are your spec and hence your program is correct
with regards to the spec, but that doesn't make the user any happier if
your code costs them time, money, safety, etc. just because the part of
the code you call the spec was also flawed or incomplete.

Unit testing is just one of many important tools. On any significant
project, it is important to have a well-rounded quality strategy that
also employs things like design/code reviews, pair programming, user
testing, etc.

But if you're really after correctness, there may be only one true
measure - let a sales person demo the product! Good old Murphy ensures
that bugs are very good at exposing themselves during important sales
presentations. :)

Jimmy




From merkosh at hadiko.de  Fri Jun 14 22:39:37 2002
From: merkosh at hadiko.de (Uwe Mayer)
Date: Sat, 15 Jun 2002 04:39:37 +0200
Subject: file.tell()  --  obscure IOError
References: <MPG.1774c0f11ff79749989686@news.rz.uni-karlsruhe.de>
Message-ID: <MPG.1774c753f4d1f6fe989687@news.rz.uni-karlsruhe.de>

In article <MPG.1774c0f11ff79749989686 at news.rz.uni-karlsruhe.de>, 
merkosh at hadiko.de says...
> hi,
> 
> i have an object which conforms to the file protocoll. while calling the 
> self.file.tell()   method, i get the following error message:
> 
> >>> s1 == s2
> Traceback (most recent call last):
>   File "<stdin>", line 1, in ?
>   File "riff.py", line 153, in __cmp__
>     data1 = self.read(self.value.step)
>   File "riff.py", line 205, in read
>     maxsize = len(self) -self.tell()
>   File "riff.py", line 234, in tell
>     return self.file.tell() -self.value.start
> IOError: (0, 'Error')
> 

more info:
the tell() does not fail every time. what i do:
i compare two files by reading count bytes and calling cmp() on both of 
them until end of file or cmp() != 0 

the tell() failes for different numbers of 'count' bytes. f.e. 
count = 1     fails after 7 bytes
count = 5	works
count = 6 	fails right at the beginning

...?

any ideas?
Ciao
Uwe


From janto_d at hotmail.com  Sun Jun 30 10:30:01 2002
From: janto_d at hotmail.com (Janto Dreijer)
Date: 30 Jun 2002 07:30:01 -0700
Subject: squared functions--most Pythonic way?
References: <mailman.1025423271.21617.python-list@python.org>
Message-ID: <f9068f8e.0206300630.64e248b9@posting.google.com>

I think that's a typo:
>>> print AddNumbers(5)
5

How about:
reduce(lambda a,b: a(b), range(10), addNumbers)
or
reduce(apply, [[i] for i in range(10)], addNumbers)

"Opus" <opus at value.net> wrote in message news:<mailman.1025423271.21617.python-list at python.org>...
> Shouldn't addNumbers(5) equate to 5?  In other words, it should evaluate that 
> as addNumbers(5)(0) or is that addNumbers(0)(5)?
> 
> How would you send a list of numbers (or objects that represent numbers) to 
> this?
> 
> On 30 Jun 2002 at 19:24, greg wrote:
> 
> > Janto Dreijer wrote:
> > > 
> > > def addNumbers(k):
> > >     def f(x):
> > >         a = addNumbers(x + k)
> > >         a.val = x+k
> > >         return a
> > >     return f
> > > 
> > > >>> addNumbers(9)(5)(2)(4)(6).val
> > > 26
> > > 
> > > Now if only I could figure out how to use __repr__() so I don't 
> > > need that ".val". It also fails when passed only one number. i.e
> > > addNumbers(5). Help?
> > 
> > class AddNumbers:
> > 
> >   def __init__(self, x):
> >     self.val = x
> > 
> >   def __call__(self, k):
> >     return AddNumbers(self.val + k)
> > 	
> >   def __repr__(self):
> >     return repr(self.val)
> > 
> > addNumbers = AddNumbers(0)
> > 
> > >>> print addNumbers(9)(5)(2)(4)(6)
>  26
> > >>> print addNumbers(5)
> > 0
> >


From whisper at oz.net  Sun Jun  9 05:19:39 2002
From: whisper at oz.net (David LeBlanc)
Date: Sun, 9 Jun 2002 02:19:39 -0700
Subject: Socket error problem
Message-ID: <GCEDKONBLEFPPADDJCOEAEKFDHAA.whisper@oz.net>

I have this problem on windows where socket.socket(...) keeps returning
10022, 'Invalid argument'. It's actually dieing in the
poplib.POP3.__init__(...) method. I have verified that host is reasonable
(mymail.myisp.net or something like that ;)), and port is 110.

If I use the test script in poplib.__main__(...) with the correct host, user
and password, all is well, but if I call it from this other program i'm
trying to get working I get the above error.

I can't figure out what's wrong - any pointers?

TIA,

David LeBlanc
Seattle, WA USA





From loredo at astro.cornell.edu  Thu Jun 20 15:35:21 2002
From: loredo at astro.cornell.edu (Tom Loredo)
Date: Thu, 20 Jun 2002 15:35:21 -0400
Subject: Windows versions of Python---pros and cons?
References: <3D110B55.4D8AD7FA@astro.cornell.edu> <aer2fn$frc$1@peabody.colorado.edu>
Message-ID: <3D122E79.A189C72E@astro.cornell.edu>

Fernando P?rez wrote:
> 
> If you're an old unix hand and want to preserve your sanity, you have two
> options:
> 
> 1- put linux on that laptop and be happy.

Actually, I've reserved a big chunk of the disk for linux.  Just gotta
decide which version (is there a "best" one for Python?  E.g., Redhat
uses Python, but does their reliance on an old version still cause
problems?).  I'll stay there when I'm developing for *me* on it!

> 2- if (1) is not an option for some reason, put cygwin on it. 

Actually, I need to develop stuff that other Win32 users can use,
some probably distributed via the McMillan installer (which I presume
works as well with Active State as with the python.org version).
So I'm stuck here....

> If you have to work in the normal,
> crippled windows environment, that laptop is going to suffer a painful death
> after being hurled through the nearest window very soon :)

I've only had it two weeks, and only the power cord to the battery
charger has saved it!  The fact that Dell shipped it with a defective
install of Win98 did not help its cause.

Thanks, to everyone!
-Tom


From fredrik at pythonware.com  Thu Jun  6 18:57:31 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 06 Jun 2002 22:57:31 GMT
Subject: minidom and encoding problem
References: <17aafe08.0206051514.4af054d2@posting.google.com> <m37klcubdu.fsf@mira.informatik.hu-berlin.de> <17aafe08.0206061419.6366b12b@posting.google.com>
Message-ID: <vNRL8.42073$n4.9541344@newsc.telia.net>

Ehab Teima wrote:

> > This is a bug in your code. You must not insert (byte) string in a DOM
> > tree; always use Unicode objects.
>
> I do not have control over the sent text. The issue started when some
> bullets were copied from a word document and pasted into a file and
> the whole file was passed to my classes.

if you don't know what encoding the file is using, what
makes you think Python can figure it out?

> I tried to encode the string using different encodings but I could
> not.

the string is already encoded.  you need to *decode* it.

> Here is what I got when I tried .encode("UTF-8"):
> UnicodeError: ASCII decoding error: ordinal not in range(128)

this means that you have non-ASCII characters in an
ASCII string.  to convert this to a unicode string, use

    u = s.decode(encoding)

where "encoding" is the source encoding (if you haven't
the slightest idea, try "iso-8859-1")

also see:

    http://effbot.org/guides/unicode-objects.htm

</F>




From peter at engcorp.com  Fri Jun 21 21:21:08 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 21 Jun 2002 21:21:08 -0400
Subject: [newbie] import fails first time, then works :-[
References: <3D1379D1.5020705@nospam.free.fr> <3D132C65.34B38F8C@engcorp.com> <3D138E31.4070409@nospam.free.fr>
Message-ID: <3D13D104.A49AF73C@engcorp.com>

laotseu wrote:
> 
> Well... I added '/usr/local/lib/lib-dynload' to my PYTHONPATH and
> 'import' does not complain anymore. Everything seems just fine now. But
> that is puzzling me ! [being a python newbie,] I thought the
> '/lib-dynload/' directorie, that is in the '/lib/python<version>/'
> directories, was automagically searched for external modules. Was I wrong ?

Searching the executable with "strings" and "grep", I find references
to /usr and /lib/python2.0 (in my case) and both /lib/lib-dynload 
and /lib-dynload.  When I run python and check sys.path I get

['', '/usr/bin', '/usr/lib/python2.0', '/usr/lib/python2.0/plat-linux2', 
'/usr/lib/python2.0/lib-tk', '/usr/lib/python2.0/lib-dynload', 
'/usr/lib/python2.0/site-packages'] 

All with default values, under RedHat 7.1.  Not sure why you've
got troubles... Linux is still Black Magic much of the time to me. :)

-Peter


From mlh at vier.idi.ntnu.no  Sat Jun  8 17:56:15 2002
From: mlh at vier.idi.ntnu.no (Magnus Lie Hetland)
Date: Sat, 8 Jun 2002 21:56:15 +0000 (UTC)
Subject: Removing ^M
References: <mailman.1023546912.19781.python-list@python.org>
Message-ID: <slrnag4vbv.gac.mlh@vier.idi.ntnu.no>

In article <mailman.1023546912.19781.python-list at python.org>, Michael
Hall wrote:
>
>I am trying remove ^M characters (some kind of newline character) from an
>HTML file. I've tried all sorts of string.replace and sed possibilities
>but the
>things just won't go away. Does anyone have a way of removing such
>characters?

If you happen to run on a Windows system (including Cygwin) you could
simply do this:

  open('outfile', 'wb').write(open('infile').read())

Python will then take care of the translation for you.

>TIA 
>
>Mick

--
Magnus Lie Hetland                                  The Anygui Project
http://hetland.org                                  http://anygui.org


From theller at python.net  Wed Jun 12 03:17:29 2002
From: theller at python.net (Thomas Heller)
Date: Wed, 12 Jun 2002 09:17:29 +0200
Subject: Can Distutils include .pyd files directly?
References: <mailman.1023837255.23228.python-list@python.org>
Message-ID: <ae6slo$4h0rf$1@ID-59885.news.dfncis.de>

"Bjorn Pettersen" <BPettersen at NAREX.com> wrote in message news:mailman.1023837255.23228.python-list at python.org...
> I've built my extension modules using the MS Devstudio IDE and now I
> would like to use Distutils to create a Windows installer for me that
> puts both the debug and release versions of the extension in the
> python22\DLLs directory. Is this possible? If not, is there a way to
> have Distutils build my extension in both debug and release mode and
> install both?
>
> -- bjorn

Sure. The 'build' command builds a 'pseudo installation tree' in the
directory build\lib.win32-2.2, all of this will later be installed by
the 'install' command.
So you can
  python setup.py build
to build the release version,
  python setup.py build -g
to build the debug version, and
  python setup.py install
will install all of this into Python22\lib\site-packages.
  python setup.py bdist_wininst
will create a windows installer containing the debug _and_
the release version (everything that's in the build directory).

Thomas






From whisper at oz.net  Mon Jun  3 18:30:11 2002
From: whisper at oz.net (David LeBlanc)
Date: Mon, 3 Jun 2002 15:30:11 -0700
Subject: getting windows processes list
In-Reply-To: <adgacv01l41@drn.newsguy.com>
Message-ID: <GCEDKONBLEFPPADDJCOEMEGHDGAA.whisper@oz.net>

I'm not sure that mere mortal programs are allowed access to that
information due to Window's (NT's) arcane idea of what constitutes security.
I think you're doing well to get your own PID.

David LeBlanc
Seattle, WA USA

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Grant Griffin
> Sent: Monday, June 03, 2002 10:53
> To: python-list at python.org
> Subject: getting windows processes list
>
>
> Hi Gang,
>
> Does anybody know offhand how to use Python to get the list of
> processes that
> are currently running on Windows?  (I'm essentially trying to get the
> information that appears in Window's Task Manager's "Processes" tab.)
>
> not-gifted-at-finding-needles-in-haystacks-ly y'rs,
>
> =g2
>
> _________________________________________________________________________
>
> Grant R. Griffin                                           g2 at dspguru.com
> Publisher of dspGuru                               http://www.dspguru.com
> Iowegian International Corporation                http://www.iowegian.com
>
> --
> http://mail.python.org/mailman/listinfo/python-list





From peter at engcorp.com  Mon Jun 17 21:25:58 2002
From: peter at engcorp.com (Peter Hansen)
Date: Mon, 17 Jun 2002 21:25:58 -0400
Subject: What If.....    Strong Types
References: <3D0C0F0D.4030900@bgb.cc>
Message-ID: <3D0E8C26.7BF0772D@engcorp.com>

Don Garrett wrote:
> 
>    I'm not suggesting any real changes to any, only proposing a thought
> experiment.
> 
>    What if Python had been built from the start with strong types? And I mean
> strong types, not strong typing. The distinction I am trying to draw is that
> variables would not be typed, but that all types would have rigid interfaces.
> 
>    Primarily, what if classes always had rigidly defined interfaces. I mean
> that public members had to be declared to exist, and that methods couldn't be
> modified on an instance. Private members could exist but would be really private.

It takes work to define those interfaces.  In my opinion, unnecessary,
inconvenient, and unreadable work...  And Python does have private
members, of course, for the right definition of private.

Preventing modifying methods on an instance would make some of my 
automated testing really difficult, if not impossible.

I don't understand the "must be declared to exist" part.  If they are declared
but not defined, they don't exist anyway.  If they are defined, then they
already exist and a declaration is redundant.  Or were you using some other
definitions of "define" and "declare"?

>    My belief is that almost all the convenience of Python would be maintained,
> but that compilation, optimization, and automated code analysis would be simpler.

Given that other languages have these things, but do not have nearly the
maintainability and convenience of Python, I doubt that.

>    I'm just wondering if the change would be good or bad. Would it really
> break anything important in Python today? Would it prevent any common errors?
> Would it help with automated code analysis (compilation and optimization
> included) as much as I think it would?

Fork a branch of Python and experiment!

I suppose, since Python is flexible and adding all these things to the
language would tend to make it more restrictive and, well, brittle, you 
could probably call the language "Stick". :-)

-Peter


From mertz at gnosis.cx  Sat Jun  8 13:14:04 2002
From: mertz at gnosis.cx (Lulu of the Lotus-Eaters)
Date: Sat, 08 Jun 2002 13:14:04 -0400
Subject: Counting lines
References: <mailman.1023405789.30972.python-list@python.org>
 <aJTL8.21666$ZF5.2505752733@newssvr10.news.prodigy.com>
 <3D000B36.5C7BCEBF@engcorp.com>
 <mailman.1023468709.32328.python-list@python.org>
Message-ID: <ctjA9kKkX4ZI092yn@gnosis.cx>         

Since this thread has drifted into line counts, I think it worth sharing
a resource that a fellow-developer shared with me:

    SLOCCount:
    http://www.dwheeler.com/sloccount/

Take a look.  It's not written in Python (I don't think), but it seems
to do a large superset of what pycount.py does.  SLOCCount counts the
lines of code in a large number of different programming languages,
using rather clever rules and heuristics both to identify what language
a file actually is, and to discern which lines are code versus comments
(including, e.g. Python docstrings).

SLOCCount also reports some stuff about cost and development efforts.  I
would take that with a grain of salt, for lots of reasons.  But it's fun
seeing just how many kilodollars you -should- have been paid to write
your open source application :-).

Yours, Lulu...

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From jonathan at onegoodidea.com  Thu Jun 27 09:18:40 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Thu, 27 Jun 2002 14:18:40 +0100
Subject: Suggestions for good programming practices?
References: <mailman.1024951532.5616.python-list@python.org> <af8n9i$3at$1@panix1.panix.com> <afafee$fc$0@216.39.172.122> <afaohi$682$1@panix1.panix.com> <afb3a2$skm$0@216.39.172.122> <B93F45F6.CEA6%jonathan@onegoodidea.com> <afd648$3b4$0@216.39.172.122>
Message-ID: <B940CF40.D106%jonathan@onegoodidea.com>

On 26/6/2002 20:53, in article afd648$3b4$0 at 216.39.172.122, "Bengt Richter"
<bokr at oz.net> wrote:

> LOL ;-)

It continually amazes me that Python is so backward as to require the
programmer to explicitly note that a function takes arguments ;-)

> ISTM caller-friendly is good, but the *args implementation of this overloading
> is not as nice
> as it could be. Is it better to allow foo(x,y,s) and foo(s) via overloaded
> foo() than, e.g.,
> to use separate names like fooxy(x,y,s) and foo(s), and have foo() call
> fooxy()?
> (That, at least, is an OT question ;-)

I'm of the opinion that when a function can take a sensible set of unrelated
optional arguments they should be keyword arguments, and one should call the
function with those keywords explicitly:

>>> def writeString( s, x=0, y=0 ):
...     pass
... 
>>> writeString( "Hello World" )
>>> writeString( "This is a test", x=10, y=35 )
>>> 

When the function can take a variable number of related arguments, I use
*args:

>>> def writeStrings( *strings ):
...     for s in strings:
...         writeString( s )
... 
>>> writeStrings( "Hello World" )
>>> writeStrings( "This is a test", "I love tests" )
>>> 

These styles, of course, can be mixed. I'm not a big fan of the idea of
distinguishing between not supplying an optional argument and supplying the
equivalent default. I think a default should mean a default, not an
indicator of no-argument. I don't much like:

>>> def writeString( s, x=None, y=None ):
...     if x is None:
...         # do something else
...         pass
...     pass
... 

I think that calling a function with a default argument explicitly given
should always make sense, i.e., read sensibly. So:

>>> writeString( "This is a test", x=None, y=None )
>>>

doesn't feel right to me since I'm not convinced that 'None' makes sense as
a position. If the default is being used to indicate, for example, the
current cursor position, then it should be explicitly named like:

>>> class _CurrentPosition: pass
... 
>>> CurrentPosition = _CurrentPosition()
>>> 
>>> def writeString( s, x=CurrentPosition, y=CurrentPosition ):
...     if x is CurrentPosition:
...         # obtain current x co-ordinate
...         pass
...     pass
... 
>>> writeString( "Hello World", x=CurrentPosition, y=CurrentPosition )
>>> 

The exception that proves the rule is where None really is a sensible
default.

Hoo hah. That's enough of that.

Jonathan



From mutsuura at mutsuura.com  Thu Jun 13 11:25:49 2002
From: mutsuura at mutsuura.com (Attila Horvath)
Date: Thu, 13 Jun 2002 11:25:49 -0400 (EDT)
Subject: Poll: Readable languages
In-Reply-To: <200206131409.JAA05450@scotland.k12.mo.us>
Message-ID: <Pine.LNX.4.44.0206131052500.20311-100000@adams.patriot.net>

On Thu, 13 Jun 2002, Chris Gonnerman wrote:

> On Thu, 13 Jun 2002 09:36:06 -0400 (EDT), Attila Horvath wrote:
> 
> > In my years of experience I have come across VERY readable assembler and
> > very unreadable COBOL, PASCAL, FORTRAN, ADA, C/C++, JAVA, etc. So I
> > suggest that a langauge does not inherently make resultant code readable.
> 
> True, and false.  It is possible to write bad code in any language; but in my
> opinion, any programmer who has mastered Python (takes what, a couple of
> weeks? :-) *tends* to write more readable code than an equivalently 
> accomplished programmer in any other language.

Sorry. What I didn't say but meant to is that 'readability' is determined
by several factors - 'structure' being only one. Clarity and simplicity in
logic, nomenclature, packaging, comments, documentation [to mention a
few] add greatly to readability. All these can be bastardized in ANY
language.

So my question/point to Ingo was, is 'readability' or 'fundamentals' of
good programming at issue?! Either way, a survey will not answer that 
question:

	'Good programming cannot be taught by preaching generalities.
	The way to learn to program well is by seeing, over and over,
	how real programs can be improved by the application of a few
	principles of good practice and a little common sense.
	Practice in critical reading leads to skills in rewriting,
	which in turn leads to better writing.'

The above excerpt 'punch card' bookmarked in my copy of:

	"The Elements of Programming Style"; 
		Brian W. Kernighan and P. J. Plauger
		Bell Telephone Laboratories
		1974

If I understand the excerpt, I think it's saying that experience through
practice from good examples makes for good programming skills.

attila






From kragen at pobox.com  Mon Jun  3 00:03:52 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 03 Jun 2002 00:03:52 -0400
Subject: Why does this work: 5<"five" ?
References: <3CF80435.689AE006@icc-computer.de> <nig9da.7s9.ln@10.0.0.1>
Message-ID: <831ybprot3.fsf@panacea.canonical.org>

Jiri Baum <jiri at baum.com.au> writes:
> It's documented that way. When comparing values of different types, you get
> an arbitrary (but fixed) ordering. I think the docs say it can change from
> version to version.
> 
> You'll note that 5>"five" gives you a 0.
> 
> This means you can do things like binary search trees without worrying
> about how to sort disparate values.

That's what I thought, too, but (in 2.1.1):

>>> 1<1j
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
TypeError: cannot compare complex numbers using <, <=, >, >=

So if you put complex numbers into your binary search tree, you will
have problems.

Oh, enjoy this:
>>> (1,) < (2,)
1
>>> (1j,) < (1j,)
0

I'm not sure what's going on there.  I thought maybe tuple comparison
did an 'is' before doing an actual comparison, but timing suggests
that it doesn't.



From James_Althoff at i2.com  Wed Jun  5 16:17:41 2002
From: James_Althoff at i2.com (James_Althoff at i2.com)
Date: Wed, 5 Jun 2002 13:17:41 -0700
Subject: Thought on PEP 204 and 276
Message-ID: <OF905873A5.7AEF9394-ON88256BCF.006DE528@i2.com>

[holger krekel ]
>in many cases (where table is iterable that is).
>I think that 'for i in 10:' is hardly ever used literally.
>By far the most common case is:
>
>       for i in xrange(len(somelist)):
>            item=somelist[i]
>            # use it
>            somelist[i]=newvalue
>
>which becomes:
>
>       for i,item in enumerate(somelist):
>            # use item
>            somelist[i]=newvalue
>
>So i don't see many applications for PEP276 anymore.

Except that tables -- GUI tables, Database tables, and other
multi-dimensional structures -- *aren't* likely to be interable in the
sense implied above (especially when they come from non-Python sources such
as Swing/JDBC/Jython).  As suggested in a previous thread, one might work
around this limitation by writing wrappers that allow iteration across
virtual rows and virtual columns of tables, for example.  "enumerate" only
helps (in the examples like the one above) if one is willing to write and
use such wrappers.  On the other hand, such wrappers often (like in the
example above) make "enumerate" unnecessary.

Jim





From kragen at pobox.com  Sat Jun  8 18:33:08 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 08 Jun 2002 18:33:08 -0400
Subject: How could i get execution times ?
References: <adkjro$12183k$1@ID-146704.news.dfncis.de>
Message-ID: <83r8jh5rkr.fsf@panacea.canonical.org>

"Shagshag13" <shagshag13 at yahoo.fr> writes:
> I'm looking for a way to get execution time at different points of a program, by now i use the
> code below, but i've seen in some post that time.clock() could be a better way to measure this,
> 
> So my question, is how you pythoners do it ??? and to understand _why_ ?

time.time() gives wallclock time; time.clock() gives CPU time.



From tim.one at comcast.net  Sun Jun 30 16:17:49 2002
From: tim.one at comcast.net (Tim Peters)
Date: Sun, 30 Jun 2002 16:17:49 -0400
Subject: Python needs better error reporting
In-Reply-To: <U2iT8.45587$n4.10825231@newsc.telia.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCMECDABAB.tim.one@comcast.net>

[Tim]
> Error-reporting in parsers has been the subject of intense
> research.  Do a little digging and come back when you can
> make a case that your point can be addressed effectively.

[/F]
> fwiw, I've seen Lisp advocates argue that this may be
> one of Lisp's greatest advantages...

If so, the quality of Lisp advocacy has suffered a precipitous decline since
my day <wink>.  I spend less time puzzling over syntax errors in Python than
I spent tracking down stuttered or missing parens in Lisp code, and in the
presence of read macros errors in Lisp input can be arbitirarily difficult
to spot.  I routinely spend gobs of time trying to figure out what's wrong
in C source, though -- e.g., forget a semicolon after a struct declaration
and it's likely to take the next identifier as declaring a variable of the
struct type, and the subsequent cascade of nonsensical error msgs is truly
amazing.  Lisp is mounds better than C in that respect -- but so is Python.

don't-even-think-about-c++-ly y'rs  - tim





From ffjhenon at club-internet.fr  Wed Jun 26 18:03:37 2002
From: ffjhenon at club-internet.fr (Fabien =?iso-8859-1?Q?H=E9non?=)
Date: Thu, 27 Jun 2002 00:03:37 +0200
Subject: Embed terminal into a Tkinter application
References: <3D1758B4.8FA63EBE@club-internet.fr> <Xns9237F055C1CABcliechtigmxnet@62.2.16.82> <653AB281E1346AA6.7E81367537EEC2BE.CF6BC8C39D83D26D@lp.airnews.net>
Message-ID: <3D1A3A39.A63B1038@club-internet.fr>

Thanks for the link.
Yes that is what I want : Insert a terminal as Konsole or xterm into a
Tkinter application instead of a Text (Tkinter widget) to which I never
managed to redirect stdout and stderr of an external program.
I have reposted a message about it today.
Continuous redirection of stdout and stderr to a Tkinter app

Fabien

Cameron Laird a ?crit :

> In article <Xns9237F055C1CABcliechtigmxnet at 62.2.16.82>,
> Chris Liechti  <cliechti at gmx.net> wrote:
> >Fabien H?non <ffjhenon at club-internet.fr> wrote in
> >news:3D1758B4.8FA63EBE at club-internet.fr:
> >
> >> The title says it all.
> >> Has someone already managed to include a Linux terminal into a Tkinter
> >> GUI ?
> >>
> >> Redirecting the output of a console into the application does not do
> >> what I want.
> >
> >well, what do you want? let us help you!
> >
> >there is also popen[2,3,4] - you can also redirect stdin.
> >
> >chris
> >
> >--
> >Chris <cliechti at gmx.net>
> >
>
> My guess is he wants an xterm(-like thing) in a Tkinter frame.
>
> Information on the subject appears in <URL: http://wiki.tcl.tk/xterm >.
> I haven't made the time to translate this into idiomatic Python.  Speak
> up if you need help doing so.
> --
>
> Cameron Laird <Cameron at Lairds.com>
> Business:  http://www.Phaseit.net
> Personal:  http://starbase.neosoft.com/~claird/home.html



From logiplexsoftware at earthlink.net  Fri Jun 21 03:39:20 2002
From: logiplexsoftware at earthlink.net (Cliff Wells)
Date: 21 Jun 2002 00:39:20 -0700
Subject: Inhibiting duplicate apps in Windows?
In-Reply-To: <3D12BCCA.3A0F7544@engcorp.com>
References: <3D11F9E8.DFFFC8E4@tds.net>
	<Xns9233F02DB217Ccliechtigmxnet@62.2.16.82> 
	<3D1283EC.A28A45CB@engcorp.com>
	<mailman.1024636228.21073.python-list@python.org> 
	<3D12BCCA.3A0F7544@engcorp.com>
Message-ID: <1024645162.10524.8.camel@localhost.localdomain>

On Thu, 2002-06-20 at 22:42, Peter Hansen wrote:
> Cliff Wells wrote:
> > 
> [on locks 'n such]
> > A fairly fool-proof/portable method is to open a socket. If the app is
> > run a second time it will fail on the attempt.  Even if the app crashes,
> > the socket will be closed by the OS (although it may take a couple of
> > seconds on occasion).
> 
> Interesting, although I take it you mean "bind a socket to a port".

Uh, yeah, sorry to be ambiguous.

> Are there any complications with this?  I created one and it's true
> another instance could not be opened ("address already in us") but
> the socket is listed with "netstat" as being LISTENING even though
> I have not called listen() or accept().  Any problems with doing 
> that, if something connects to the socket?  Could it cause the
> socket to be held open for more than a couple of seconds after
> closing?  Sometimes sockets can take minutes to be cleaned up.

You know, I haven't actually tested such a situation, but I don't think
another program can connect to that socket without accept() being called
(but I could be wrong).  Probably best just to try it an find out.


Cliff





From dmitri.gouliaev at telkel.net  Thu Jun 20 08:47:52 2002
From: dmitri.gouliaev at telkel.net (Dmitri I GOULIAEV)
Date: Thu, 20 Jun 2002 07:47:52 -0500
Subject: latex-tex-packages
In-Reply-To: <29A97D00F387D411AC7900902770E148058E5A11@lcoeexc01.coelce.net>; from "Alves, Carlos Alberto - Coelce" on Wed, Jun 19, 2002 at 04:24:36PM
References: <29A97D00F387D411AC7900902770E148058E5A11@lcoeexc01.coelce.net>
Message-ID: <20020620074752.A7257@lifebook>

Hi, Alves, Carlos Alberto - Coelce !

 On Wed, Jun 19, 2002 at 04:24:36PM -0300, Alves, Carlos Alberto - Coelce wrote:

> You forgive me, c'os I know this issue it's not to this list; But, anybody
> can tell me where can I get tex/latex packages to install. I just browse
> some url but can't find'em.

May I suggest to browse some more ?..

http://directory.google.com/Top/Computers/Software/Typesetting/TeX/


Best regards,

-- 
DIG (Dmitri I GOULIAEV)




From larooy at xtar.co.nz  Thu Jun 13 03:50:38 2002
From: larooy at xtar.co.nz (John La Rooy)
Date: Thu, 13 Jun 2002 19:50:38 +1200
Subject: Why does Python mix OO concepts and non OO concepts for operation s  on basic types?
References: <3D054A98.3090708@bgb.cc>
	<Pine.LNX.4.33.0206120008180.9668-100000@rnd.onego.ru>
	<mailman.1023827360.23248.python-list@python.org>
Message-ID: <20020613195038.281d513d.larooy@xtar.co.nz>

On Tue, 11 Jun 2002 22:28:44 +0200
holger krekel <pyth at devel.trillke.net> wrote:

> Roman Suzi wrote:
> > I only forgot how will I do this:
> > 
> > map(string.split, s)
> > 
> > if GvR will deprecate string module...
> 
> what about
> 
>     map(str.split, s) 
> 
> :-) holger
> 
> 
what about 

map(string.capwords, s)

I'm guessing it's an oversight that there is no str.capwords??

John ;oP


From sandeep182 at hotmail.com  Thu Jun 20 11:12:48 2002
From: sandeep182 at hotmail.com (Sandeep Gupta)
Date: Thu, 20 Jun 2002 15:12:48 GMT
Subject: FreeTDS vs Inline TDS: Problem with Inline TDS
Message-ID: <QhmQ8.215123$9F5.15887558@typhoon.austin.rr.com>

I'm attempting to access a SQL Server 2000 DB from Linux using ODBC.  My
setup is
Python --> mxODBC --> unixODBC --> <ODBC Driver> --> SQL Server

When I use the Inline TDS driver directly from unixODBC, all queries return
the correct results, but the final line is always "-1 rows affected".  The
FreeTDS driver returns the correct results as well as the correct number of
rows affected.  The incorrect "rows affected" from Inline TDS seems to be
causing problems with mxODBC.

Has anyone experienced this same problem with Inline TDS's driver?

Are there any other suggestions for querying against SQL Server 2000 from
Python running on Linux?

Thanks
Sandeep




From phd at phd.pp.ru  Wed Jun 12 14:57:29 2002
From: phd at phd.pp.ru (Oleg Broytmann)
Date: Wed, 12 Jun 2002 22:57:29 +0400
Subject: dict = urllib.urldecode(string) ?
In-Reply-To: <jjb5-743571.14235312062002@newsstand.cit.cornell.edu>; from jjb5@cornell.edu on Wed, Jun 12, 2002 at 02:23:53PM -0400
References: <jjb5-743571.14235312062002@newsstand.cit.cornell.edu>
Message-ID: <20020612225729.B642@phd.pp.ru>

On Wed, Jun 12, 2002 at 02:23:53PM -0400, Joel Bender wrote:
> I'm processing GET and POST data and I would like the functional 
> inverse of urllib.urlencode().  That is, split the string apart by
> the '&' chars, split each of those by '=' and build a dict of the 
> results.  The key and the value should be passed to unquote_plus().
> 
> Here is my solution, comments?
> 
> >>> def urldecode(s):
> ...     rslt = {}
> ...     for item in s.split('&'):
> ...         keyValue = item.split('=')
> ...         rslt[ urllib.unquote_plus(keyValue[0]) ] = 
> urllib.unquote_plus(keyValue[1])
> ...     return rslt
> ... 
> 
> >>> urldecode('a=b+c&d%26e=%20f')
> {'a': 'b c', 'd&e': ' f'}

   You cannot handle URLs like "city=London&city=Paris". BTW, modern
urlencode accept a list of 2-lists, not only a dict.

Oleg.
-- 
     Oleg Broytmann            http://phd.pp.ru/            phd at phd.pp.ru
           Programmers don't die, they just GOSUB without RETURN.




From pinard at iro.umontreal.ca  Thu Jun 20 19:53:46 2002
From: pinard at iro.umontreal.ca (=?iso-8859-1?q?Fran=E7ois?= Pinard)
Date: 20 Jun 2002 19:53:46 -0400
Subject: do...while
In-Reply-To: <60FB8BB7F0EFC7409B75EEEC13E20192158EEC@admin56.narex.com>
References: <60FB8BB7F0EFC7409B75EEEC13E20192158EEC@admin56.narex.com>
Message-ID: <oqu1nxjylx.fsf@titan.progiciels-bpi.ca>

[Bjorn Pettersen]

> Doing a quick search over our C++ libraries (~900KLoc), "do {} while()"
> seems to be used in right around 3% of the loops (including while, for;
> excluding recursion).  Seems like overkill to add special syntax for a
> feature used so rarely...

Some of you might remember C.A.R. Hoare's axiomatisation of Pascal, which
also gives good insight for many other languages.  The `do {} while()' was
called `repeat' in Pascal.  The axiom for `while' is one of the simplest
of the whole language, much simpler than the axiom for `repeat'.

This taught me that there is a theoretical foundation for the popularity of
the `while' construct.  After all, program axiomatisation is nothing more
than the formalisation of programmer's thought when s/he produces a program,
and our intuition naturally drives our thoughts towards simpler concepts.

-- 
Fran?ois Pinard   http://www.iro.umontreal.ca/~pinard




From bowman at montana.com  Tue Jun  4 00:11:22 2002
From: bowman at montana.com (bowman)
Date: Mon, 03 Jun 2002 22:11:22 -0600
Subject: getting windows processes list
References: <adgacv01l41@drn.newsguy.com>
Message-ID: <L2XK8.387$ZM.144840@newsfeed.slurp.net>

Grant Griffin wrote:

> Does anybody know offhand how to use Python to get the list of processes
> that are currently running on Windows???(I'm?essentially?trying?to?get?the
> information that appears in Window's Task Manager's "Processes" tab.)

Take a look at WSH/WMI. Probably the easiest way.
 



From emile at fenx.com  Sun Jun 30 11:42:57 2002
From: emile at fenx.com (Emile van Sebille)
Date: Sun, 30 Jun 2002 15:42:57 GMT
Subject: yield equivalent in C/JavaScript?
References: <wL9dFLAeifH9EwD$@jessikat.fsnet.co.uk> <7xy9cxhlvx.fsf@ruckus.brouhaha.com> <w3xdKUAuljH9EwgH@jessikat.fsnet.co.uk> <7xit41i7i9.fsf@ruckus.brouhaha.com> <14ms5VAygsH9Eww2@jessikat.fsnet.co.uk> <7xelepnk84.fsf@ruckus.brouhaha.com> <WrXUtJA3EvH9EwER@jessikat.fsnet.co.uk>
Message-ID: <5GFT8.369062$cQ3.24122@sccrnsc01>

Robin Becker
> The IE stuff seems to rely too much on the URL extension so provided
it
> thinks it's htm or html IE 5.x seems OK.
>

Cool... last time I looked into this it didn't work on IE, and I found
this on msdn

If the server-provided MIME type is either known or ambiguous, the
buffer is scanned in an attempt to verify or obtain a MIME type from the
actual content.

... that I read to mean 'if you tell it, it will make its own check
anyway and ignore you' so I gave it up.

Can you post an example url that causes IE to properly interpret
Content-Encoding: x-gzip?

Thanks,

--

Emile van Sebille
emile at fenx.com

---------



From marklists at mceahern.com  Thu Jun 13 12:11:52 2002
From: marklists at mceahern.com (Mark McEahern)
Date: Thu, 13 Jun 2002 11:11:52 -0500
Subject: For and Dates
In-Reply-To: <67F0CEAF5F0ED411B53900508BC5C8BC700195@admin.datacraft.com.br>
Message-ID: <NCBBLFCOHHDIKCAFGCFBIEHILEAA.marklists@mceahern.com>

> I need to do a loop that goes from 2002-6-1 to 2002-7-31, but I don't know
> how to do that...
>
> Is there a way to do such range using for !?
>
> Do you have any suggestions !?

Short answer:

	Use mx.DateTime + generators

Demo:

from __future__ import generators

import mx.DateTime

def date_span_gen(d1, d2):
    """Return an iterator for the dates between d1 and d2."""
    if d2 <= d1:
        raise RuntimeError("d2 (%s) must be later than d1 (%s)." % (d2, d1))
    diff = d2 - d1
    for d in range(diff.days):
        yield (d1 + mx.DateTime.RelativeDateTime(days=d))

num_days = 30
d1 = mx.DateTime.now()
d2 = d1 + mx.DateTime.RelativeDateTime(days=num_days)

for d in date_span_gen(d1, d2):
    print d

-





From Bill.Scherer at VerizonWireless.com  Wed Jun 12 11:27:09 2002
From: Bill.Scherer at VerizonWireless.com (Scherer, Bill)
Date: Wed, 12 Jun 2002 11:27:09 -0400 (EDT)
Subject: Bits and bytes?
In-Reply-To: <3d0760ec$0$223$4d4ebb8e@news.nl.uu.net>
Message-ID: <Pine.LNX.4.44.0206121117460.9455-100000@bagend>

On Wed, 12 Jun 2002, Guyon Mor?e wrote:

> ok, thanx a lot.
> 
> there's one thing i am missing though... how do i write '011010110101' as
> 'binary bits' to a file?

I've only had need for this once, but when I 'realized' that a 
byte was just eight bits, my problem was solved. For you that 
depends on if you're ok with writing bits in multiples of eight, 
and that you can create the proper 8 bit seqeunces for your 
needs.  chr(0) through chr(255) should cover your needs, I think.

chr(107) + chr(168) will give you

'011010110101000'

CAVEAT: I might be completely off my rocker on this. 8-)

Good luck.

> 
> 
> 
> 
> "Christopher Armstrong" <radix at twistedmatrix.com> wrote in message
> news:mailman.1023890477.2592.python-list at python.org...
> >>>>> "Guyon" == Guyon Moree <gumuz at looze.net> writes:
> 
>     Guyon> ok, this might be a strange question.... but i was wondering: how
>     Guyon> can i read or write 'bits'?
> 
>     Guyon> if i have a string, how can i convert to it's basic 0's and 1's?
> 
>     Guyon> and offcourse, the other way around.... let's say i want to store
>     Guyon> some bit-sequence(0011101110) to a file.
> 
> 
>     Guyon> how do i do this?
> 
> http://twistedmatrix.com/users/moshez/binary.py
> 
> Moshe Zadka wrote this because he was annoyed at all the people in #python
> asking how to do this ;-)
> 
> 
> --
>                                 Chris Armstrong
>                          << radix at twistedmatrix.com >>
>                 http://twistedmatrix.com/users/carmstro.twistd/
> 
> 
> 
> 
> 
> 

-- 
Bill.Scherer at Verizon Wireless
RHCE 807101044903581





From wuhsinyee at hotmail.com  Thu Jun 13 10:51:46 2002
From: wuhsinyee at hotmail.com (Mark Wu)
Date: 13 Jun 2002 07:51:46 -0700
Subject: How to install BITS in zope?
Message-ID: <c7902822.0206130651.1ff67d30@posting.google.com>

Hi All:

    Does any one successfully install Bug & Issues Tracking System on ZOPE?
I already followed the installation guide, and installed

1. MySQLdb for python 2.1.1
2. ZMySQLDA in Zope 2.5.1

But when I want to install the sql script to create the table in my SQL, it
not worked. It seems the script is only for Oracle. Can any one tell me how
to change the script and make it suitable for mysql??

Best Regards, Mark

** My environment:

OS: win2000
ZOPE: Zope 2.5.1
MySQL: 3.2.23
Python: 2.1.1


From mwh at python.net  Wed Jun 12 06:24:05 2002
From: mwh at python.net (Michael Hudson)
Date: Wed, 12 Jun 2002 10:24:05 GMT
Subject: unknown opcode Exception then AV
References: <ae546s$bg5$1@ctb-nnrp2.saix.net> <lkk7p5yh2w.fsf@pc150.maths.bris.ac.uk> <ae5eov$m8i$1@ctb-nnrp1.saix.net>
Message-ID: <lkfzzsyevl.fsf@pc150.maths.bris.ac.uk>

"Geoff Bien" <gbien at intekom.co.za> writes:

> Found the problem. There is a known bug
> in version 2.1 ?? when "continue" is used
> within a try...except block

Oh yeah, that.  It's fixed in 2.1.something though, isn't it?  Time
for an upgrade perhaps.

Cheers,
M.

-- 
  C is not clean -- the language has _many_ gotchas and traps, and
  although its semantics are _simple_ in some sense, it is not any
  cleaner than the assembly-language design it is based on.
                                        -- Erik Naggum, comp.lang.lisp


From TuxTrax at fortress.tuxnet.net  Wed Jun  5 17:23:13 2002
From: TuxTrax at fortress.tuxnet.net (TuxTrax)
Date: Wed, 05 Jun 2002 21:23:13 -0000
Subject: Compiling Python
References: <slrnafrqit.lbm.TuxTrax@fortress.tuxnet> <3CFE23DB.994F8A6C@engcorp.com>
Message-ID: <slrnaft0p7.3ac.TuxTrax@fortress.tuxnet>

On Wed, 05 Jun 2002 10:44:43 -0400, Peter Hansen <peter at engcorp.com> wrote:
> TuxTrax wrote:
>> 
>> Hello all
>> 
>> I am new to this forum, and to Python, please excuse the newbie question.
>> 
>> I have started reading the o'reilly book, "learning python" and find python
>> to be an amazing language so far. My first impression of it is a language that
>> is easier than basic to learn and use, and more powerful than many other
>> high level languages such as pascal.
>> 
>> My question is this; I realize that Python compiles to bytecode automatically
>> as part of the interpreters tokenization run, but is there a way to permanently
>> compile python code outputting to an executable binary?
> 
> Why do you want to do that?  For presumably increased speed, or because
> you don't want the hassle of multiple .py files all over the place (in
> other words, better packaging)?
> 
> -Peter

Hi Peter

I was more interested in if there were existing tools to do it, already
available. 

I really don't care to "hide" my source code, as I hail from the open source
Linux community (kind of like "I'm from star fleet, I never lie" <g>)
and I would be releasing any source for any python app that I wrote (that was 
any good <g>) anyway. Besides, if you know others will be seeing your source 
you tend to write cleaner code <g>.

As I noted, it was a newbie question, a matter of curiousity. I am just
starting to learn python and I really love it, and I am full of such
questions that probably others have already gotten past.

Packaging is a thought, but speed (so far) isn't an issue, as python is pretty 
fast IMHO.
 
Cheers,

Mathew

-- 
TuxTrax   (n.) An invincible, all terrain, Linux driven armored assault
vehicle that can take as much fire as it gives ;-)

<Carnivore activation string> 
ASSASINATION ANTHRAX PRESIDENT NUCLEAR TALIBAN AMMONIUM NITRATE
</Carnivore activation string>

Yes, I am a Penguin cult high priest. Fin readings upon request.
ROT13 this email address to mail me: uvtuqrfregzna at lnubb.pbz



From gerhard at bigfoot.de  Fri Jun 14 21:47:31 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Sat, 15 Jun 2002 03:47:31 +0200
Subject: Using Python to script a game?
In-Reply-To: <cdc2589b.0206141718.7dc2ae13@posting.google.com>
References: <cdc2589b.0206141718.7dc2ae13@posting.google.com>
Message-ID: <20020615014731.GA2856@lilith.my-fqdn.de>

* Ben Blonder <ben at sourcecod.com> [2002-06-14 18:18 -0700]:
> I'm writing a 3D naval tactics in C++ for the Mac. I'm planning on
> having all the object behaviors, movements, and AI controlled through
> a scripting layer. The scripts would be text files the game would
> parse and then carry out actions. All the scripts would be read at
> runtime, not compile time, so they'd all have to be interpreted.
> 
> For example, I might have the C++ code
> 
> class foo { 
> public:
> int x;
> void print() {
> cout << x << endl;   
> }
> };
> 
> In the scripting system, there would have to be some way to create a
> new foo object, define x, and then call print() while the user was
> playing.
> 
> Can I use Python to accomplish this,

Yes. And it would be a pretty good match, too. There's even been a
commercial game that uses Python for internal scripting, its name
escapes me, though.

> and if so, can you point me to some tutorials on integratring Python
> with my existing code?

Python's documentation has (now) good documentation about it's C API.
Python's API is _C_, though, not C++. There are several C++ wrappers for
it than can simplify the integration a lot. My personal favourite is
Boost::Python. http://www.boost.org/ As a C++ developer, you might be
interested in the other free libraries from Boost, too. What these
people accomplish is really amazing!

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From peter at engcorp.com  Fri Jun  7 08:07:37 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 07 Jun 2002 08:07:37 -0400
Subject: Medium to Large Scale Python Deployments
References: <38C31EBA-79A4-11D6-9A7D-003065B33450@intersight.com> <mailman.1023409393.3361.python-list@python.org> <3D003EC7.C72D27BE@engcorp.com> <3D006320.5040805@web.de>
Message-ID: <3D00A209.EC24F671@engcorp.com>

J?rgen Hermann wrote:
> 
> Peter Hansen wrote:
> > How big is that when run through pycount?
> >
> > (I.e. comments and blank lines don't count!)
> 
> Summary on "twisted"
> ====================
> files                  256
> lines                51654
> bytes              1728211
> comments              5993
> empty lines          10029
> non-commentary lines 29910

Thanks!  Similar results to what I've seen over about 50000 
lines in 235 files.  Roughly 50-60% code content, and 
almost twenty percent blank lines.

I suspect such numbers (perhaps excluding lines of comments)
are more consistent for Python code, whoever wrote it,
than for many other languages.

-Peter


From drew_csillag at yahoo.com  Wed Jun 26 03:14:06 2002
From: drew_csillag at yahoo.com (Drew Csillag)
Date: Wed, 26 Jun 2002 12:14:06 +0500
Subject: [ANN] SkunkWeb 3.2.3
Message-ID: <3D1969BE.2040901@yahoo.com>

SkunkWeb 3.2.3 has just been released.

The release is available from the SkunkWeb home page at
  http://skunkweb.sourceforge.net

or directly from
      http://prdownloads.sourceforge.net/skunkweb/skunkweb-3.2.3.tgz

This is mainly a bugfix release.

Changes in 3.2.3
========================================
* added support for the SCGI Protocol
  - http://www.mems-exchange.org/software/scgi/
* document updates/revamps (currently stmlrefer)
* now is more persistant about the timeout exception (it's reraised each
  second after a timeout occurs)
* now should build with non-GNU makes (specifically BSD make)
* many fixes -- thanks to pychecker!
* fixed error handling weirdnesses

Drew Csillag





From gh_pythonlist at gmx.de  Tue Jun  4 03:48:51 2002
From: gh_pythonlist at gmx.de (Gerhard =?unknown-8bit?Q?H=E4ring?=)
Date: Tue, 4 Jun 2002 09:48:51 +0200
Subject: 'for every' and 'for any'
In-Reply-To: <7934d084.0206031636.5c29b31f@posting.google.com>
References: <20020526135944.A32690@hishome.net> <20020526091742.A987@unpythonic.net> <mailman.1022567685.7803.python-list@python.org> <slrnaf7af3.8kn.quinn@schilling.ugcs.caltech.edu> <mailman.1022658469.27441.python-list@python.org> <7934d084.0206031636.5c29b31f@posting.google.com>
Message-ID: <20020604074850.GC23780@gargamel.hqd-internal>

* Andrae Muys <amuys at shortech.com.au> [2002-06-03 17:36 -0700]:
> C dosn't have any 32-bit int types standardised, so what do _you_ use
> when you need one?

Do you need 32 bit ints or >= 32 bit ints?

I only know that Ada 95 and C++ with Boost can guarantee a certain bit length
for ints. Dunno if that's what you wanted to hear :)

Does Python run on any machine where ints are 16 bits long at all?

Gerhard
-- 
This sig powered by Python!
Au?entemperatur in M?nchen: 18.7 ?C      Wind: 2.0 m/s




From idbaxter at semdesigns.com  Sun Jun 16 16:21:52 2002
From: idbaxter at semdesigns.com (Ira Baxter)
Date: 16 Jun 2002 13:21:52 -0700
Subject: symbolic python
References: <3CEFAE4C.38A8022B@doc.ic.ac.uk> <3CEFD881.D7992095@doc.ic.ac.uk> <3cf8ce86$1@giga.realtime.net> <83sn46vkgi.fsf@panacea.canonical.org> <3CF942C7.2E4AA4CF@engcorp.com> <838z5xtcla.fsf@panacea.canonical.org>
Message-ID: <6a625eea.0206161221.50d351ad@posting.google.com>

Kragen Sitaker <kragen at pobox.com> wrote in message news:<838z5xtcla.fsf at panacea.canonical.org>...
> Peter Hansen <peter at engcorp.com> writes:
> > Kragen Sitaker wrote:
> > > "Ira Baxter" <idbaxter at semdesigns.com> writes:
> > > > DMS Software Reengineering Toolkit.
> > > It sounds very interesting.  How can I get a copy?
> > 
> > No doubt by sending them a pile of money.  :-)  (Given the customer
> > list mentioned, it would probably be a large pile.)

It isn't cheap by most people's standards.
But then, we don't sell tens of thousand of copies,
and so the cost must reflect the engineering effort
and volume.  And we (at least) think it is well worth the price
in terms of time saved for the tasks which it does well.

> I looked around the web site for a while without finding any
> information about how to do that, so I posted the above message.  Ira
> Baxter responded in private, saying it "was available commercially",
> and suggesting that I look at the web site for more details.
> 
> This all leaves a bad taste in my mouth; the Web site makes a point of
> explaining that the information they would like to sell me was
> developed with my tax dollars.

And frankly, so was the basic VLSI technology that powers the computers
you seem happy to pay for.

Yes, we did get some tax dollars to help build an early,
*demo* version of the technology.
And, like most other beneficiaries of US Government
Research grants, we have put a huge amount of effort into
it above and beyond what the original grants provided.

> > no support for Python is (yet) mentioned in the list of supported
> > languages.

That's because our tools are designed to accept langauge definitions
easily.   A Python grammar would take only a few days to develop.

If you want to transform Python symbolically, DMS will do the
job very well, and I don't see many alternatives out there.

-- IDB


From aahz at pythoncraft.com  Mon Jun 10 12:15:34 2002
From: aahz at pythoncraft.com (Aahz)
Date: 10 Jun 2002 12:15:34 -0400
Subject: Correct code/unit testing (was Re: Efficient python programming...)
References: <adoa9u$vqed$1@ID-146704.news.dfncis.de> <3D02FB3B.CF75E925@engcorp.com> <Pine.SOL.4.30.0206092211170.15771-100000@mimosa.csv.warwick.ac.uk> <3D03CDD2.78C01DE1@engcorp.com>
Message-ID: <ae2jb6$933$1@panix1.panix.com>

In article <3D03CDD2.78C01DE1 at engcorp.com>,
Peter Hansen  <peter at engcorp.com> wrote:
>
>The nastiest bugs were found by lots and lots of work, which usually
>included repeated attempts to reproduce the problem, manual runs, 
>and sometimes using a debugger.  Staring helped little.  And then,
>having found the bug, we look at the code and say "oh yeah, we 
>should have seen that." 

Not necessarily.

>Better not to put the bugs in the code in the first place, however,
>so staring or those other things won't be necessary.  Writing the 
>tests and the code together, tests first, in small iterative steps, 
>tends not to lead to buggy code.  The payback is high compared to
>staring.

Maybe.  The most difficult bug I recall tracking down turned out to
involved a web page created by a program that put a null character at
the end of an URL, like this:

    <a href="http://www.example.com/foo/bar.html\x00">

Except, of course, that looking at the page's source in the obvious ways
didn't display the damn thing.  I finally used Python to download the
page and display chunks of the pagestring.  The problem was compounded
by a bug in MS SQL Server that considered the null-ending and
non-null-ending versions of the string to be either duplicate or unique,
depending on the precise operation.
-- 
Aahz (aahz at pythoncraft.com)           <*>         http://www.pythoncraft.com/

"I had lots of reasonable theories about children myself, until I
had some."  --Michael Rios


From tdelaney at avaya.com  Mon Jun 24 20:29:31 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 25 Jun 2002 10:29:31 +1000
Subject: Redirecting StdOut
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A3BF@natasha.auslabs.avaya.com>

> From: Ugo Garc?a [mailto:plugin at supercable.es]
>
> or, perhaps this is the right way to do it. Do u know many 
> games that are
> made with an interpreted languaged? I don't. (I'm talking 
> about GAMES, not
> little demos). Well, I known one: BLADE. It use Python to 
> define the world,
> the behavour of the characters, and other things like the 
> save games... If u

http://groups.google.com/groups?hl=en&lr=lang_en&ie=UTF-8&frame=right&th=54a
844b1ecd059e9&seekm=3D0D9317.6371F155%40del_cygnus-software_del.com

In particular, look at Bruce Dawson's response.

Tim Delaney




From Mitch.Chapman at bioreason.com  Tue Jun 25 12:05:09 2002
From: Mitch.Chapman at bioreason.com (Mitch Chapman)
Date: Tue, 25 Jun 2002 10:05:09 -0600
Subject: Memory Monitor
References: <pan.2002.06.25.12.14.06.994352.17223@pobox.com>
Message-ID: <3D1894B5.720EB7F7@bioreason.com>

Marco Catunda wrote:
> 
> Hello,
> 
>         What is the best way to monitor usage memory?
> 
>         I have used "os.system( 'ps %d' % os.getpid() )" but I
>         don't know if it is the best way?

Can't tell if you're using Linux or some other *nix variant.  If
you are using Linux, this might do what you want:

	inf = open("/proc/self/status", "r")
	print inf.read()
	inf.close()


'course, you probably want more concise output, e.g.:

        os.system("cat /proc/self/status | grep VmSize")

-- 
Mitch Chapman
Mitch.Chapman at bioreason.com


From mwh at python.net  Sat Jun  8 12:50:36 2002
From: mwh at python.net (Michael Hudson)
Date: Sat, 8 Jun 2002 16:50:36 GMT
Subject: __imul__ broken for 'objects' in 2.2.1
References: <yfsd6v3qa86.fsf@black132.ex.ac.uk>
Message-ID: <lk660tlnon.fsf@pc150.maths.bris.ac.uk>

Alexander Schmolck <a.schmolck at gmx.net> writes:

> OK, after unsuccessfully trying for about half an hour to submit this to the
> sf bug tracker, I gave up:
> 
> class Breaky(object):
>     def __imul__(self, other):
>         print "imuling"
>         return self
> sq = Breaky()
> sq *=1.
> 
> gives:
> 
> Traceback (most recent call last):[...] line 10, in ?
>     sq *=1.
> TypeError: can't multiply sequence to non-int
> 
> Unless I'm overlooking something, this is a fairly serious bug and I
> can't see a way to work around it (getattribute would presumably
> work, but slow everything down unacceptably, so the only 'solution'
> seems to be to have no inplace-multiplication).
> 
> I had the same behavior on two different machines (running Mandrake 8.2 and
> Suse 7.3).

Hmm.  You should file a bug on sf so this doesn't get forgotten
(unless it's already been reported, but I don't think so).

Cheers,
M.

-- 
  It's a measure of how much I love Python that I moved to VA, where
  if things don't work out Guido will buy a plantation and put us to
  work harvesting peanuts instead.     -- Tim Peters, comp.lang.python


From mertz at gnosis.cx  Fri Jun 28 14:29:09 2002
From: mertz at gnosis.cx (David Mertz, Ph.D.)
Date: Fri, 28 Jun 2002 14:29:09 -0400
Subject: metaclasses vs. inheritance
References: <ghTS8.58749$mh.1826422@news1.telusplanet.net>
Message-ID: <1rKH9kKkX0cE092yn@gnosis.cx>                

|Can someone give an example of a problem that is solved with metaclasses,
|which would be difficult or impossible to solve with inheritance?
|I (very barely) understand metaclasses, but to me it just seems like
|inheritance, where the classes who get the benefit of the meta-class get the
|metaclass' features.

Well...  I am at the "barely" level also.  But my Gnosis_Utils package
contains a moderately useful example of a metaclass.

By way of background, gnosis.xml.pickle pickles Python objects to XML.
There are several ways to use it, but one is:

    # By inheritence
    from gnosis.xml.pickle import XML_Pickler
    class MyClass(XML_Pickler):
    # create some behavior and attributes for MyClass...
    o1 = MyClass()
    # o1 knows how to dump itself
    xml_str = o1.dumps()

Good enough.  But every class that wants to know how to serialize itself
needs to inherit from XML_Pickler.  Here's an alternate approach:

    """Perform black magic of unearthly and ungodly sorts

    Quick Example 1:

        Python 2.2 (#0, Dec 24 2001, 18:42:48) [EMX GCC 2.8.1] on os2emx
        Type "help", "copyright", "credits" or "license" for more information.
        >>> import gnosis.magic
        >>> __metaclass__ = gnosis.magic.MetaPickler
        >>> class Boring:
        ...     def __init__(self):
        ...         self.this = 'that'
        ...         self.spam = 'eggs'
        ...     def print_spam(self):
        ...         print self.spam
        ...
        >>> boring = Boring()
        >>> boring.print_spam()
        eggs
        >>> print boring.dumps()
        <?xml version="1.0"?>
        <!DOCTYPE PyObject SYSTEM "PyObjects.dtd">
        <PyObject module="__main__" class="Boring" id="1276364">
        <attr name="this" type="string" value="that" />
        <attr name="spam" type="string" value="eggs" />
        </PyObject>

    """
    from gnosis.xml.pickle import dumps
    class MetaPickler(type):
        def __init__(cls, name, bases, dict):
            super(MetaPickler, cls).__init__(name, bases, dict)
            setattr(cls, 'dumps', dumps)

--
mertz@  | The specter of free information is haunting the `Net!  All the
gnosis  | powers of IP- and crypto-tyranny have entered into an unholy
.cx     | alliance...ideas have nothing to lose but their chains.  Unite
        | against "intellectual property" and anti-privacy regimes!
-------------------------------------------------------------------------





From jeremy at jdyallop.freeserve.co.uk  Sat Jun 15 04:44:15 2002
From: jeremy at jdyallop.freeserve.co.uk (Jeremy Yallop)
Date: 15 Jun 2002 08:44:15 GMT
Subject: html parser etc help
References: <7fe97cc4.0206111445.43b196de@posting.google.com> <ae63pm$47b95$1@ID-114079.news.dfncis.de> <7fe97cc4.0206150003.65fe8821@posting.google.com>
Message-ID: <aeeuov$6709t$1@ID-114079.news.dfncis.de>

* Xah Lee
| Thanks Jeremy Yallop for help. Some follow up questions...
| 
| in the following code:
| 
| import os 
| def visit(arg, dirname, fnames): 
|     for file in fnames: 
|         if file.endswith('.html'): 
|             print file 
|             print arg 
| os.path.walk('/export/home/xah/unixnotes/',visit,'---')
| 
| 
| * in the os.path.walk, what's the third argument for?

Here's what the interactive help says:

  >>> import os
  >>> help(os.path.walk)
  Help on function walk in module posixpath:
  
  walk(top, func, arg)
      Directory tree walk with callback function.
      
      [...] No semantics are defined for, or required of, arg, beyond
      that arg is always passed to func.  It can be used, e.g., to
      pass a filename pattern, or a mutable object designed to
      accumulate statistics.  Passing None for arg is common.

| * what if i want to print the full path?

Use os.path.join():

  def visit(arg, dirname, fnames): 
      for file in fnames: 
          if file.endswith('.html'): 
              print os.path.join(dirname, file)

| * what if i want to print a particular type of file, for example only
| directories.

Use os.path.isdir()

| * how to find out about a particular function or method, or what
| modules are available?

In python 2.2, type help() at a (python) prompt, and then 'modules' to
give a list of all modules available on your system.  You can use
help() on a type or function to read its documentation (try
'help(file)' for example).  The full current library reference is at:

   http://www.python.org/doc/current/lib/lib.html

Jeremy.


From jason at mastaler.com  Tue Jun 25 17:53:06 2002
From: jason at mastaler.com (Jason R. Mastaler)
Date: Tue, 25 Jun 2002 15:53:06 -0600
Subject: bug in os.getgroups?
References: <1025033152.17806.TMDA@nightshade.la.mastaler.com> <20020625202442.GG4435@unpythonic.net>
Message-ID: <hh3cvbja9p.fsf@nightshade.la.mastaler.com>

Jeff Epler <jepler at unpythonic.net> writes:

> This is liely to be because /tmp is "sticky":

/tmp was a bad choice for my example.  The problem is the same in a
non-sticky directory such as /home.

# python
Python 2.2.1 (#1, Apr 22 2002, 10:19:01) 
[GCC 2.95.3 20010315 (release) [FreeBSD]] on freebsd4
Type "help", "copyright", "credits" or "license" for more information.
>>> import os
>>> os.setgid(31)
>>> os.setuid(667)
>>> os.getgroups()
[31, 0, 2, 3, 4, 5, 20, 31]
>>> os.unlink('/home/junk')
Traceback (most recent call last):
  File "<stdin>", line 1, in ?
OSError: [Errno 13] Permission denied: '/home/junk'
>>> 

-- 
(http://tmda.sourceforge.net/)






From tdelaney at avaya.com  Mon Jun 10 21:26:15 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Tue, 11 Jun 2002 11:26:15 +1000
Subject: Efficient python programming...
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A36E@natasha.auslabs.avaya.com>

> From: John J. Lee [mailto:jjl at pobox.com]
> 
> On Sun, 9 Jun 2002, Peter Hansen wrote:
> [...]
> > What tests do, however, is vastly increase your confidence in the
> > correctness of the code, and when written properly (in advance
> > of each small morsel of code that is written, and having been
> > shown to fail without that code) they are sufficient to deliver
> [...]
> 
> This point -- that tests should fail before they pass -- is 
> easy to forget
> about.  Thanks for reminding me.

It's one of the things that you need to be veryt rigourous about. It's very
tempting, upon receiving a bug report, to go charging into the code, finding
where the problem is, "fixing" it, then realising you need to test it.

Of course, by this time you can't write a test that *fails* with the
previous behaviour unless you go back to a previous version.

Always write a (correct ;) test that fails before "fixing" anything.

Unit tests may include 100% code coverage, but they rarely cover all
possible situations. When a new situation is found, the test should be
written.

Tim Delaney




From tim.one at comcast.net  Mon Jun 10 08:19:09 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 10 Jun 2002 08:19:09 -0400
Subject: Negative long literals (was Re: Does Python need a '>>>' operator?)
In-Reply-To: <Pine.GSO.4.33.0206100850550.12342-100000@techst02.technion.ac.il>
Message-ID: <LNBBLJKPBEHFEDALKOLCCELGPLAA.tim.one@comcast.net>

[Beni Cherniavksy]
> I just got another idea: use 0x1234 for 0-filled numbers and 1xABCD for
> 1-filled ones.  That way you impose no restrictions on what follows the
> prefix and keep backward compatibility.  0xFFFFFFFF stays a 2^n-1
> _positive_ number, as it should be.  The look of 1x is weird at first but
> it is very logical...

Indeed, on both counts <wink>.  I like it a lot!

For the other points related to getting rid of machine long-size
assumptions, see PEP 237:  <http://www.python.org/peps/pep-0237.html>.





From peter at engcorp.com  Fri Jun  7 21:01:37 2002
From: peter at engcorp.com (Peter Hansen)
Date: Fri, 07 Jun 2002 21:01:37 -0400
Subject: Efficient python programming...
References: <mailman.1023494898.26737.python-list@python.org>
Message-ID: <3D015771.4BDCB22F@engcorp.com>

Sean 'Shaleh' Perry wrote:
> 
> > I disagree, if the tests are written first (and run, proving that
> > they will fail when the code is wrong or missing).
> >
> > The tests are executable requirement specifications.  Running
> > them "proves" (yes, not 100%... nothing could) that the code
> > meets the specs.  The specs could be wrong of course, but that
> > doesn't mean the code is broken...
> 
> this also assumes the test covers every possible failure.  this also assumes
> one can write tests for that section of code.

And all these sorts of things are the direct focus of test-driven
development.  For example, on the issue of testability: code developed
this way is inherently testable because the test was written first,
and the code had to be written to pass the test.  Code developed in
a more traditional way is much less likely to be testable and often
ends up tightly coupled and very hard to test or maintain as a result.
This, of course, tends to lead to having no tests at all, which is
a terrible thing.

I guess I was just taking a moment to point out there are some new
approaches which are more useful than the traditional (and generally
neglected) approach of writing a handful of tests after the fact.

I was also rebelling momentarily against the idea of shipping code
to a customer without having a suite of unit tests in place to give
confidence in the code.

I'll step back from any claim that unit tests really prove anything,
but in a very Pythonic and pragmatic fashion they are Good Enough to 
make them a /necessity/ for anyone trying to deliver reliable software.

-Peter


From tjreedy at udel.edu  Mon Jun  3 09:37:03 2002
From: tjreedy at udel.edu (Terry Reedy)
Date: Mon, 03 Jun 2002 13:37:03 GMT
Subject: Efficient posting-list
References: <adffdk$10mep9$1@ID-146704.news.dfncis.de>
Message-ID: <3iKK8.99777$%y.10468246@bin4.nnrp.aus1.giganews.com>

"Shagshag13" <shagshag13 at yahoo.fr> wrote in message
news:adffdk$10mep9$1 at ID-146704.news.dfncis.de...
> A posting list is something like that (hope i'm not too wrong) :
>
> [key_0] -> ...
> ...
> [key_i] -> [id_j, informations] -> [id_k, informations] -> ...
> ...
> [key_n] -> ...
>
> where in information retrieval :
> - key are often words,
> - id are document id (so we have key_i in document id_j and in
document id_k and ...)
> - informations are often in document frequency, but might be more...
>
> I'm looking for an efficient way of implementing this in full
python, by now i use a dict
> for keys and a python list containing node objects for [id_j,
informations].
> But i have two troubles this is very slow to populate, and too
memory consuming do you
> have an idea for optimizing this ?
>
> (keep in mind that : there are more than 500,000 text keys ; id are
numbered from 1 to
> 150,000 ; length of a posting list might be from 1 to 500,000...)

For English text, one would almost certainly reduce the number of text
keys by deleting some suffixes, so that 'statistic', 'statistics',
'statistical', 'statistically', and maybe 'statistician' would be one
key.  But I have no idea of word structure of non-Indo-European
languages and what one should do to reduce key number.

By 'length of a posting list' do you mean number of text keys?  Ie,
size of key dict?  The length of list attached to each key is at most
number of documents.

You do not specify 'node object'.  This is crucial since these take up
most of memory.  (This a good reason not to index the 2000 or so most
common keys.)  For pure Python, tuple is most space efficient.  If
each node really is (integer id, integer count), one could devise a
system using a pair of array (module) objects or 2-dimensional
numerical python arrays; which store actual integers rather than
pointers to int PyObjects.  When filled, they would have to be copied
to larger arrays in much the same manner as done automatically by
Python lists.

Terry J. Reedy





From koenig at v-i-t.de  Sat Jun 29 16:17:17 2002
From: koenig at v-i-t.de (Mirko Koenig)
Date: Sat, 29 Jun 2002 22:17:17 +0200
Subject: dynamic buttons with tkinter
Message-ID: <afl51r$emn$1@lnx107.hrz.tu-darmstadt.de>

Hi

In my i want to make a button for every file
in a special directory, so that i can load it.
I did:
 
 for file in os.listdir( dir ):
        Button( load_dlg_frame, text=file, relief=GROOVE, 
		command=load).pack()

But how can i find out which button is pressed?
Or another question: How can i pass a event funktion more than 1
parameter (the event and other things).
I can't give any button a new event function, because it's dynamic,
because of the number of files in the directory.

thanx 

Mirko


From seeger at sitewaerts.de  Wed Jun 12 05:34:59 2002
From: seeger at sitewaerts.de (Felix Seeger)
Date: Wed, 12 Jun 2002 11:34:59 +0200
Subject: SyntaxWarning: name 'mailcounter' is used prior to global declaration
Message-ID: <ae74k3$49l33$1@ID-79019.news.dfncis.de>

Hi

I get this warning.
I set mailcounter as a global var because I have to change it's value from a 
method in a class.
This was the only way. How can I remove this warning ?

Mailcounter is defined at the beginning of the script and "global 
mailcounter" is called in the method.


thanks
have fun
Felix


From vjovanov at stevens-tech.edu  Wed Jun  5 10:51:59 2002
From: vjovanov at stevens-tech.edu (Vojin Jovanovic)
Date: Wed, 5 Jun 2002 10:51:59 -0400
Subject: self
References: <mailman.1023225848.26315.python-list@python.org> <0%hL8.22575$tK.5490038@news02.optonline.net> <mailman.1023280507.13313.python-list@python.org>
Message-ID: <_zpL8.81$pq2.4342892@newnews.cc.stevens-tech.edu>

> If you are going to always store the equation, never the value, what
> happens when the user does
>     c.y = 37
>     c.y = 'y+1'

that is very simple, look at my response to Tim Peters
 and the method __setattr__

def __setattr__(self,arg,val):
  self.__dict__['Equations'][arg]=str(val)

str(val) takes care of c.y=37

I suggets that you play with my Foo class to see how all that works.

> Then you're left with detecting stuff like
>     c.x = "y"
>     c.y = "x+1"
cycle staff can be taken care with a counter, I don't see a problem with
that.

> Well, good luck with your project -- and don't spring any more
> requirements on us, it's getting harder each time to solve your problem.

Sorry, but it is not that my problem is hard.  It is that the concept of
having
to reference instance variables always through self. is making it hard and
inconvenient which was my point in my original post.
My opinion is that those points explained in FAQ are irrelevant.  The only
one relevant is .....

"When classes were added to Python, this was (again) the simplest way of
implementing methods without too many changes to the interpreter. "

This sentence says it all, everything after that is said in FAQ is of less
importance and
it looks like "well, we did it in the simplest way, but, you know, this way
is really
better for you, let us show you  how ..."
The consequence is that the simplest solution is not always the best one.

Vojin Jovanovic






From SteveC at nospam.innocent.com  Fri Jun 21 15:10:29 2002
From: SteveC at nospam.innocent.com (Steven M. Castellotti)
Date: Fri, 21 Jun 2002 19:10:29 GMT
Subject: Dispatching a Win32 COM in the background?
References: <mailman.1024642348.12905.python-list@python.org> <3D132A7C.94CE4670@engcorp.com>
Message-ID: <FSKQ8.1061$cE5.914@nwrddc02.gnilink.net>

On Fri, 21 Jun 2002 08:30:36 -0500, Peter Hansen wrote:

> David LeBlanc wrote:
>> 
>> Hmmm... if this is working for you, that's great - it just looks
>> strange to me.
>> 
>> If Background_TTS is the background speacher thread, shouldn't it have
>> the queue.get() in it and the foreground have the queue.put()?
>> 
>> It's also worth noting that self isn't used or needed at the global
>> level, so self.speech is kind of different ;-)
> 
> I'm pretty sure from previous postings he was going to call this from
> within another object, so the "self." used elsewhere might refer to it.
> 
> As near as I can tell, with that in mind, the .get() and .put() are in
> exactly the right places.


	That's right, this object is getting called from another class object.
All of this is part of a GPL'd program that allows children to create
their own video games using drawings and pictures, without them having to
learn how program.

	Since the program runs under Linux and Windows, there's separate
objects loaded for each enviroment, both having a reference set to
"self.speech" in order to abstract the exact implementation away from
other parts of the program.


Cheers,

Steve Castellotti
SteveC (at) innocent.com
http://cogengine.sourceforge.net/


From emile at fenx.com  Mon Jun 17 21:49:52 2002
From: emile at fenx.com (Emile van Sebille)
Date: Tue, 18 Jun 2002 01:49:52 GMT
Subject: PythonWorks Pro problem!
References: <q3nrgu019nlnc2bc19jaouhfo0833fknrj@4ax.com>
Message-ID: <3lwP8.112157$pw3.6716@sccrnsc03>

Johann
> I registered some examples from above link. The problem is, I cannot
> find  this "small, coloured icon" , so I cannot cycle through avalable
> settings. :-(
>
> I tried to clink on everything, but without results :( Anybody can
> help?


Hmmm... that icon appears to be missing from 1.3...

It's there in 1.2


--

Emile van Sebille
emile at fenx.com

---------



From duncan at NOSPAMrcp.co.uk  Tue Jun 18 07:34:28 2002
From: duncan at NOSPAMrcp.co.uk (Duncan Booth)
Date: Tue, 18 Jun 2002 11:34:28 +0000 (UTC)
Subject: python version?
References: <K7PO8.16057$uk2.8406537@twister.nyroc.rr.com> <aegfck$6tkre$1@ID-114079.news.dfncis.de> <wEPO8.16066$uk2.8447171@twister.nyroc.rr.com> <3tingu8psfmefoab93pl8gips8hg7sgpvp@4ax.com> <gYgP8.42356$GY.12361719@twister.nyroc.rr.com> <aekcug$7js8n$1@ID-114079.news.dfncis.de> <9YBP8.44418$GY.13865843@twister.nyroc.rr.com>
Message-ID: <Xns92317F85D48EDduncanrcpcouk@127.0.0.1>

"George Hester" <hesterloli at hotmail.com> wrote in 
news:9YBP8.44418$GY.13865843 at twister.nyroc.rr.com:

> I think you guys don't how to do it.

And I think you are one of the most impolite posters we have seen around 
here.

Try this:
<html>
<head>
  <title>Python Version</title>
</head
<body>
<SCRIPT Language="Python" RUNAT=Server>
import sys
Response.Write(sys.version)
</SCRIPT>
</body>
</html>


-- 
Duncan Booth                                             duncan at rcp.co.uk
int month(char *p){return(124864/((p[0]+p[1]-p[2]&0x1f)+1)%12)["\5\x8\3"
"\6\7\xb\1\x9\xa\2\0\4"];} // Who said my code was obscure?


From daniel.dittmar at sap.com  Fri Jun 21 09:36:49 2002
From: daniel.dittmar at sap.com (Daniel Dittmar)
Date: Fri, 21 Jun 2002 15:36:49 +0200
Subject: RE strings (was: Variable Interpolation - status of PEP 215)
References: <ouit4e63fr.fsf@andlx-anamika.cisco.com> <4PfQ8.44245$n4.10307683@newsc.telia.net> <KhiQ8.44269$n4.10312432@newsc.telia.net> <29e28c51.0206201156.7e61a7f5@posting.google.com> <3D127B89.EE1C17F4@engcorp.com> <aeuprv$ek$1@news1.wdf.sap-ag.de> <3D132ADC.D04B52EA@engcorp.com>
Message-ID: <aeva5h$bn4$1@news1.wdf.sap-ag.de>

>> The implementation could choose to cache the result of re.compile at
>> module level, thus
>> - speeding up the 'naive' re usage
>> - freeing the programmer from doing the caching in code
>
> Any reason the re.match() method could not do the same, internally?

The caching strategies would be different:

using re"...": the compiler would know that there are only x literals in the
module, thus he would store the re.compile results in an array for fast
lookup.

using module re: this requires a dictionary indexed by the patterns. And it
wouldn't be wise to keep all patterns around, so it depends on the locality
of pattern usage how well the cache performs.

Daniel





From mcfletch at rogers.com  Mon Jun  3 17:19:53 2002
From: mcfletch at rogers.com (Mike C. Fletcher)
Date: Mon, 03 Jun 2002 17:19:53 -0400
Subject: Guru advice needed for mxTextTools
References: <3CFBCC68.D35A03D8@vip.fi>
Message-ID: <3CFBDD79.5020809@rogers.com>

I'm curious, if you have an EBNF grammar, why not just go ahead and use 
it with SimpleParse to generate the tag-table.  As for the two-line 
parsing, just define the grammar such that a matched line is:

file := (match/fail)+

match := properHeaderLine, properContentLine
fail := -[\n]*,"\n",-[\n]*,"\n"? # two lines, second newline optional

<properHeaderLine> := whateveryouneedtotest
properContentLine := (group/-("\n"/STOPCHAR))*, "\n"?

group := STARTCHAR,(group/-("\n"/STOPCHAR))*,STOPCHAR

(That's untested, of course ;) .  Grouping code is always a pain to 
write without having corner cases fry you :) ).

2MB is a small file, so shouldn't take long to parse using mx.TextTools, 
and it's normally much faster to just slurp that size of file into 
memory and process the whole thing in the mx.TextTools C loop.  I often 
worked with 5-6MB VRML files (far more complex grammar generating large 
numbers of in-memory nodes) and never had a problem with speed there.

Note: the above grammar will silently ignore improperly formatted 
content lines after a properHeaderLine.  Marc-Andre has added suport 
that would allow creating an error in that case, but I've been busy 
elsewhere and don't yet support it in SimpleParse.

HTH,
Mike

Pekka Niiranen wrote:
> I am trying to optimize a function that searches nested strings from
> set of (allmost) flat files (about 2 MB each) . If I use regular
> expressions, I must
> fix the amount of nesting:
> 
<head-hurting code deleted>
...
> 
> In the code above support two nested strings. If prefix is "?" and
> suffix is "!" then
> it will evaluate into:
> 
> 
>>>>pattern = re.compile("(\?[^?!]+(\?[^?!]+\!)*[^?!]+\!)")
>>>>Line = "?AA?BB!CC!?DD!ee?EE!ff?FF?GG!HH!"
>>>>print re.findall(pattern, Line)
>>>
> [('?AA?BB!CC!', '?BB!'), ('?DD!', ''), ('?EE!', ''), ('?FF?GG!HH!',
> '?GG!')]
> 
...
> 2)    The file is not flat: I also need to check the contents of the
> previous line. If previous line
>        does not contain correct value, I do not have to run the regular
>        expression on the current line:
>             for i in range(1,len(lines),2):
>                         test = lines[i-1].strip()
>                         if (test == 'x' or test == 'y'):
>                             matches = re.findall(pattern,
> lines[i].strip())
>                             if matches:
>                                 # Remove empty results with filters
>                                 pars =
> filter(operator.truth,reduce(operator.add, matches))
> 
> 3)    Amount of nesting may vary in the future
...
> I have thought of EBNF -notation that should be supported with
> Simpleparse  + mxtexttools
> 
> Questions are:
> 
> 1)    What is the mxtexttool tagtable for the regular expression above
> with additions of unlimited nesting.
>        If suffix is the same as prefix, no nesting is assumed
> 2)    Is it possible to parse the file without keeping the record of the
> current line number since
>        values to be checked are allways on odd line numbers and regular
> expression is allways run
>        on even line numbers. If I could read two lines at a time and
> parsing them both simultaneously
>        (as a single line) with mxtexttools (with lookAhead or whatever
> ), I could gain some speed ?
> 3)    Should I seek examples from XML -tools instead OR write my own
> parser with C + SWIG ?
...
_______________________________________
   Mike C. Fletcher
   http://members.rogers.com/mcfletch/






From FooWeissBarMike at hotmail.com  Fri Jun  7 14:33:45 2002
From: FooWeissBarMike at hotmail.com (Michael Weiss)
Date: Fri, 7 Jun 2002 14:33:45 -0400
Subject: map
References: <3D007666.2B6D3582@gol.ge>
Message-ID: <3d00fe37@news.mhogaming.com>

Would this work?

##########
x = [1,2,3,4]
y = 2

#this?
print map(lambda z: pow(z,y), x)

# or this?
print map(pow, x, [y for z in xrange(len(x))])
##########

"Giorgi Lekishvili" <gleki at gol.ge> wrote in message
news:3D007666.2B6D3582 at gol.ge...
> Hi all!
>
> Is there a way to map a function with several arguments to a list? E.g.,
>
> >>>import math
> >>>import MLab
> >>>a=MLab.rand(3,5)
> >>>b=map(math.pow,a)
>
> of course, this doesn't work.
>
> What to do? My task is to optimize the power (i.e., y in pow(x,y)) to
> achieve maximal performance.
>
>
> thanx,
> Giorgi
>
> PS: Perhaps I have missed the needed functionality, that's presented in
> NumPy?
>
>
>




From peter at engcorp.com  Sun Jun 23 12:16:39 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sun, 23 Jun 2002 12:16:39 -0400
Subject: Newbie question (anyone programming with curses?)
References: <af4ovd$1bt3$1@pingwin.acn.pl>
Message-ID: <3D15F467.F66309EE@engcorp.com>

hellboy wrote:
> 
> I'm quite new to Python, but I'm positive that version 2.2 (for Windows)
> comes without curses module, though documentation states it is included.
> import curses gives me "Import Error: no module named curses" message.
> Anyone knows where to download curses module (for Windows)?

A quick google search (http://www.google.ca/search?q=python+curses+windows)
brought up a number of results, including a link to the curses How-To
by Kuchling and Raymond which has this to say:

  "No one has made a Windows port of the curses module. On a Windows 
   platform, try the Console module written by Fredrik Lundh."

-Peter


From jonathan at onegoodidea.com  Thu Jun 27 08:19:58 2002
From: jonathan at onegoodidea.com (Jonathan Hogg)
Date: Thu, 27 Jun 2002 13:19:58 +0100
Subject: Python 2.2 __slots__ Bug?
References: <mailman.1025089474.17780.python-list@python.org> <B93F7990.CECB%jonathan@onegoodidea.com> <mailman.1025105074.31122.python-list@python.org> <afdlv6$bu6$0@216.39.172.122>
Message-ID: <B940C17E.D101%jonathan@onegoodidea.com>

On 27/6/2002 1:24, in article afdlv6$bu6$0 at 216.39.172.122, "Bengt Richter"
<bokr at oz.net> wrote:

> Wondering if there's any 12*n byte chunks involved, fitting
> into some 2**n (-overhead) allocation space? Just noticing
> 
>>>> divmod(2**19, 43551)
> (12, 1676)
>>>> divmod(2**16, 5430)
> (12, 376)
> 
> Sometimes these things mean something, sometimes they don't ;-)
> Maybe it will jog someone's thought. E.g., a last-item index/limit
> problem?

It appears that GjR is on the case ;-) but I did some digging into the
source out of curiosity and it's definitely a stack overrun problem. If
anyone is interested, here's my analysis:

At the moment the deallocators for the builtin container types pull a neat
trick to stop the stack getting too deep. As they're recursively
deallocating objects they keep a count of how deep deallocation has gotten
so far. If this gets over a set limit then instead of deallocating the
object it jams it onto a list and returns immediately.

When the stack has unwound back to the top a new piece of magic is invoked
that checks to see if anything was put onto the "delete later" list. If
something was then it grabs the first thing off the list and starts
recursively deleting at that point again.

A hack is used to make this "delete later" list: The object's type is
checked and, since this trick is only used by a few builtin object types, a
numeric type code is stored in the reference count slot of the object (which
is no longer interesting since it must be 0). The type object pointer slot
is then re-used as a pointer to the next item in the "delete later" list.
When the item is pulled off the list and actually deallocated the reverse
trick is pulled to correct the type object pointer and reset the reference
count to 0.

This is all fine except that this trick is not pulled for __slots__
variables as these are not stored in a dict that can be added to the "delete
later" list (the point of slots variables). In the case of Glyph's example,
because the chain is made up entirely of slots pointers, there is never a
dict, tuple, or list - the only* types that can take part in this trick -
that can be added to the "delete later" list allowing the stack to be
unwound, so the deallocation simply recurses until it runs out of stack.

I can't see an easy way to fix this without changing the way that the
"delete later" list is constructed.

Chewy problem.

Jonathan


* OK, I lied slightly there. Actually frames and traceback objects take part
  in this crime as well, but they're not interesting here.



From phr-n2002b at NOSPAMnightsong.com  Sat Jun 22 00:58:41 2002
From: phr-n2002b at NOSPAMnightsong.com (Paul Rubin)
Date: 21 Jun 2002 21:58:41 -0700
Subject: How to stop a SocketServer?
Message-ID: <7xhejv9af2.fsf_-_@ruckus.brouhaha.com>

I want to run a simple network server based on
SocketServer.ThreadingTCPServer.  I followed the docs
and implemented a handler class which works fine.  
Then I instantiate the server and run the serve_forever() method.
The server listens for connections, starts threads for them
and handles them just like it should.

Question: how do I shut down the server once it's running, besides
doing something crude like killing the process?  I'd like to be able
to connect to the server and enter a "shutdown" command and have the
server gracefully exit and free the port.  I haven't been able to
figure out how to do that.

Suggestions?  Thanks.


From ljd at nospam.com  Sun Jun  9 12:11:18 2002
From: ljd at nospam.com (LJD)
Date: Sun, 09 Jun 2002 16:11:18 GMT
Subject: Program mysteriously STOPS (Solved)!
References: <yZuM8.146064$ux5.186314@rwcrnsc51.ops.asp.att.net>
Message-ID: <G6LM8.151484$ux5.188000@rwcrnsc51.ops.asp.att.net>

Thank you to all those who responded!

The problem seemed to be that I was writing output to the console.  When the 
program ran after I started it from a console session it worked fine 
because there was a place for the ouput to go.

When I started the program from a startup script, the program would run, but 
I noticed I never saw any console output.  Apparently, this output built up 
somewhere and eventually caused the program to stop.

Once I took the PRINT statement out of the code, the program started via 
startup script and ran without quitting.

I am very new to Linux and Python, so I am only surmising that this is what 
happened.  All I know is the program now runs "forever" just like I wanted!

Thanks again!!!

Larry


From tew at wiencko.com  Sat Jun  1 22:35:23 2002
From: tew at wiencko.com (Tom Wiencko)
Date: Sun, 02 Jun 2002 02:35:23 +0000
Subject: win32com under Cygwin
Message-ID: <3CF9846B.579606ED@wiencko.com>

There may be an easy answer for this, but I have not yet found it...

I noticed that the Python distribution with Cygwin does not include the
win32 extensions.  The only distribution I can find for them is a
Windows executible, which does not seem to know about Cygwin.

Can these libraries be loaded under Cygwin?  Do they even work under
Cygwin?  Can somebody point me toward how to load them if it is
possible?

Thanks in advance.

Tom
-- 
------------------------------------------------------------------------
Tom Wiencko                                              tew at wiencko.com
President - Wiencko & Associates, Inc.                    (404) 255-2330
Telecom Consulting & Project Development -- Wireline, Wireless, Internet


From chris.gonnerman at newcenturycomputers.net  Sun Jun 16 09:12:00 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Sun, 16 Jun 2002 08:12:00 -0500
Subject: What If.....    Strong Types
References: <3D0C0F0D.4030900@bgb.cc> <332c7c11.0206160305.304b7cb3@posting.google.com>
Message-ID: <001f01c21537$67b69a00$0101010a@local>

----- Original Message -----
From: "Benno" <benjl at cse.unsw.edu.au>


> Don Garrett <garrett at bgb.cc> wrote in message
news:<3D0C0F0D.4030900 at bgb.cc>...
> > I'm not suggesting any real changes to any, only proposing a thought
> > experiment.
> >
> >    What if Python had been built from the start with strong types? And I
mean
> > strong types, not strong typing. The distinction I am trying to draw is
that
> > variables would not be typed, but that all types would have rigid
interfaces.
>
> <snip>
>
> >    I'm just wondering if the change would be good or bad. Would it
really
> > break anything important in Python today?
>
> Well, it would break my stuff, which is pretty damned important to me :).
> Dynamically creating/modifying things on the fly can produce nicer,
cleaner
> code than would otherwise be possible.

Indeed.  Those who truly grasp the Zen of Python never miss "strong typing"
of any sort.  Python is dynamic.  Failure to follow interface rules usually
results in raising an exception; in the cases where it is more subtle, it
seems to me that Python takes less time to debug than strong-typed languages
due to the parallel feature of introspection.

> > Would it prevent any common errors?
>
> Possibly but not that many imho.

I have rarely experienced errors related to an object's interface; usually
if I have it has been a simple typo which raised an exception.  What exactly
would I gain therefor with strong typing?  More typing (of the keyboard
variety)
for little gain.

> > Would it help with automated code analysis (compilation and optimization
> > included) as much as I think it would?
>
> Possibly, it is probably rather difficult to follow through all the
> implications of such as change however.

Exactly.  Trust Guido... he knows far better than most of us.


Chris Gonnerman -- chris.gonnerman at newcenturycomputers.net
http://newcenturycomputers.net





From gerhard.haering at gmx.de  Thu Jun 27 12:23:51 2002
From: gerhard.haering at gmx.de (=?ISO-8859-1?Q?Gerhard=20H=E4ring?=)
Date: Thu, 27 Jun 2002 18:23:51 +0200 (CEST)
Subject: "multi" zip
In-Reply-To: <lrGS8.13892$bt.307@nwrddc04.gnilink.net>
References: <lrGS8.13892$bt.307@nwrddc04.gnilink.net>
Message-ID: <20020627162143.C13AC624@gargamel.hqd-internal>

Am Thu, 27 Jun 2002 15:46:25 GMT schrieben Sie (Mark <Aristotle_00 at yahoo.com>):
> So,
> 
> Let line be a list of lists where the sublists are the same length.
> 
> For example:
> 
> line = [ [0,1,2], [3,4,5], [6,7,8] ]
> 
> Now, I can do:
> 
> zip(line[0], line[1], line[2])
> 
> to "invert" the array.  Now, I'd like to do two different things:
> 
> 1) instead of producing tuples, I'd like to produce lists.
> 
> 2) I'd like to extend this to an arbitrary number of lists.  If it were
> possible, I'd do zip(line[0], line[1], ..., line[n]) but this doesn't seem to
> compile :).
> 
> Is there a way to make zip() do this, or should I stop trying to hammer with
> a wrench?

zip(*list)

The *sequence form of calls is described in chapter 5.3.4 of the language
reference: http://www.python.org/doc/current/ref/calls.html

Gerhard
--
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))





From dgallion1 at yahoo.com  Tue Jun  4 20:35:32 2002
From: dgallion1 at yahoo.com (darrell)
Date: Wed, 05 Jun 2002 00:35:32 GMT
Subject: Looking for a list subclassing example...
References: <adi4dv$118tko$1@ID-146704.news.dfncis.de> <b0c5da3a.0206040833.456d184e@posting.google.com> <mailman.1023209527.22088.python-list@python.org>
Message-ID: <o1dL8.3078$i17.1047407@twister.nyroc.rr.com>

Thanks Jeff.

I don't use this code now but I checked and it had a memory leak.
Which the "del self[:-self.limit]" fixed.
I didn't need support for slices. Guess I should have coded an assert
into those other methods.

--Darrell

Jeff Epler wrote:
> I don't think the line
>>             self = self[-self.limit:]
> can do what you want (it only changes something about the locals in
> the method).  You almost certainly need to use slice deletion
> del self[:-self.limit]
> or
> if diff > 0: del self[:diff]
> (these seem to work in all cases, even when diff == 1). You also haven't
> overridden the slice assignment method, extend(), __add__, or __mult__
> to make sure the length invariant is preserved in cases like
>     x[:0] = ['ha ha'] * 1000
>     x = x + ['ha ha'] * 1000
>     x = x * 1000
> 
> Jeff



From tdelaney at avaya.com  Mon Jun  3 02:53:55 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Mon, 3 Jun 2002 16:53:55 +1000 
Subject: Why no open(f, "w").write()?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A345@natasha.auslabs.avaya.com>

> From: John La Rooy [mailto:larooy at xtar.co.nz.avaya.com]
>
> It's unlikely to be explicitly documented. from the docs...
> 
> The file() constructor is new in Python 2.2. The previous 
> spelling, open(),
> is retained for compatibility, and is an alias for file().
> 
> So you shouldn't be using open() in new programs anyway

Yes - but *file objects* have been part of Python for a long time, and the
docs do not state anywhere I could find (I may be wrong) that *file objects*
have a destructor that closes the file.

The name used to create a file object (be it file, open or wibble) is
unimportant.

Tim Delaney




From rnd at onego.ru  Thu Jun 27 15:59:26 2002
From: rnd at onego.ru (Roman Suzi)
Date: Thu, 27 Jun 2002 23:59:26 +0400 (MSD)
Subject: Pythoniac: Thoughts on a hardware Python processor
In-Reply-To: <F7DB8D13DB61D511B6FF00B0D0F06233015DE661@mail.hebmex.com>
Message-ID: <Pine.LNX.4.44.0206272357530.12877-100000@rnd.onego.ru>

On Thu, 27 Jun 2002, Gustavo Cordova wrote:

>> The only route that has any chance of success is 
>>     Python -> Forth -> stack engine
       ---------------
       -------
       ---
Fython by analogy of Jython ;-)

>> which leverage works that has already gone into Forth area.  
>> But, my days of hand assembling 68000 or programming HP calculators
>> are long gone... :-)

Sincerely yours, Roman Suzi
-- 
rnd at onego.ru =\= My AI powered by Linux RedHat 7.2





From thenault at nerim.net  Mon Jun 17 07:12:36 2002
From: thenault at nerim.net (Sylvain Thenault)
Date: 17 Jun 2002 13:12:36 +0200
Subject: [ANN] PyReverse 0.3.1
Message-ID: <mailman.1024325000.4173.clpa-moderators@python.org>

this release fixes a packaging problem which was broken the installation
process of 0.3.

What's new ?
------------
See the ChangeLog file for more information.


What's PyReverse ?
------------------

Pyreverse is a set of utilities to reverse enginering Python code. So
far, it features dependency analysis tools, unittest generation,
and XMI generation for importation in a UML modeling tool. A special
module can be used to generate files readable by Argo UML. 
It uses a representation of a Python project in a class hierarchy
which can be used to extract any information (such as generating UML
diagrams and make a few statistics from the Python code, as "pyargo"
and "pystats") 


Home page
---------
http://www.logilab.org/pyreverse/

Download
--------
ftp://ftp.logilab.org/pub/pyreverse/

Mailing list
------------
mailto://xml-logilab at logilab.org

-- 
Sylvain Th?nault 




From gerhard at bigfoot.de  Thu Jun 13 10:08:34 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 13 Jun 2002 14:08:34 GMT
Subject: Threading tutorial wanted.
References: <mailman.1023954137.22915.python-list@python.org>
Message-ID: <slrnagh9r0.4f1.gerhard@lilith.my-fqdn.de>

Alex Polite wrote in comp.lang.python:
> Okidoki. My little webspider has reached the point where further
> optimization won't pay off. But my DSL pipe is not saturated by
> far. So I guess it's time for me to start looking into threads.

You could take a look at the linkchecker source code. It uses threads
for web spidering.

> I few minutes of googling on python + threads doesn't yield a
> lot. Maybe someone here can point me to a good primer. Maybe someone
> here can write a good primer.

Aahz has written a tutorial that you were already referred to. It's
useful info, but nowhere near complete. Maybe you can write a tutorial
yourself once you've mastered threading? ;-) Maybe there's a Python
book that covers multithreading in depth? I'd be interested which one
this is if so.

> When is threading good, when is it bad?

I've recently experimented myself with multitheading in Python. It can
open a whole new can of worms for your app, that's for sure.

> How much memory do they consume?

Depends on the threading implementation - just try it out. I don't
think it will be a problem unless start _lots_ (100+) of threads.

> Is it safe to let multiple threads use the same db connection?

Look into the DB-API specification for the threadsafety attribute of
your respective DB module. The ones I use have threadsafety=1, which
means you can't safely share the connection among threads, but it's ok
if every thread opens its own connection.

Alternatively, you could make the connection threadsafe by wrapping it
with a proxy class like the one I yesterday proposed here in:

Message-ID: <20020611205423.GA28113 at lilith.my-fqdn.de>
RfD: Automagically making a class instance 'synchronized' (draft impl.)

Or you could use a database connection pool. Webware includes one that
you can borrow.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From usenet at thinkspot.net  Sat Jun  1 12:20:06 2002
From: usenet at thinkspot.net (Sheila King)
Date: Sat, 01 Jun 2002 09:20:06 -0700
Subject: How to open a HTML file when the python cgi program is executing?
References: <ad9q91$vlrsb$1@ID-49758.news.dfncis.de> <ad95l1.i0.1@kserver.org> <ad9vsm$vhjnu$1@ID-49758.news.dfncis.de> <9m6ada.31b.ln@10.0.0.1> <adaeh8$uum7t$1@ID-49758.news.dfncis.de>
Message-ID: <ada3k7.lo.1@kserver.org>

On Sat, 1 Jun 2002 22:28:34 +1000, "Ken" <ken at hotmail.com> wrote in
comp.lang.python in article <adaeh8$uum7t$1 at ID-49758.news.dfncis.de>:

> > If you realize this before you've output anything, you can use one of the
> > redirection / refresh headers, such as "Location".
> >
> > print """Content-type: text/html
> > Location: http://replacement-url
> >
> > <p><a href="http://replacement-url">Click here</a></p>
> 
> Doesn't seem to work... :(
> 
> 
> > """

You must remove the line about Content-type.
The Location header will essentially redirect to a different HTML page.
When using the location header, your entire output that is printed should
simply be:

print "Location: http://replacement-url\n"

You should have nothing before this, and there's no point in having
anything after it, since the page is being redirected.

--
Sheila King
http://www.thinkspot.net/sheila/
http://www.k12groups.org/



From rivers at cars.uchicago.edu  Sun Jun 23 20:06:43 2002
From: rivers at cars.uchicago.edu (Mark Rivers)
Date: Sun, 23 Jun 2002 19:06:43 -0500
Subject: Problems with Pmw.Blt
Message-ID: <LttR8.19$U4.2252@news.uchicago.edu>

Folks,

I am having some problems with Pmw.Blt running Python 2.1.3 on Win32 and
Python 2.1.1 on Linux.  Pmw is version 0.8.5 and Blt is version 2.4.

I found descriptions of the first problem in the newsgroup archives, but no
solutions were posted.
Here is a short script that demonstrates the problem:
#################################
from Tkinter import *
import Pmw
t1 = Tk()
g1 = Pmw.Blt.Graph(t1); g1.pack()
t2 = Tk()
#reload(Pmw.Blt)
g2 = Pmw.Blt.Graph(t2); g2.pack()
#################################

When this script is run right after starting Python, the following happens.
The first Tk() window (t1) and the first Blt.Graph (g1) are created fine.
The second Tk() window (t2) is created and then the following error occurs:

Traceback (most recent call last):
  File "U:\Rivers\python\epics\testBlt1.py", line 7, in ?
    g2 = Pmw.Blt.Graph(t2); g2.pack()
  File "C:\PROGRA~1\Python21\Pmw\Pmw_0_8_5\lib\PmwBlt.py", line 260, in
__init__
    Tkinter.Widget.__init__(self, master, _graphCommand, cnf, kw)
  File "C:\PROGRA~1\Python21\lib\lib-tk\Tkinter.py", line 1756, in __init__
    self.tk.call(
TclError: invalid command name "::blt::graph"

I can eliminate this error if I uncomment the line "reload(Pmw.Blt).  I have
been using this workaround in my applications with 99% success.  Just before
every call to create a Pmw.Blt.Graph I execute the statement
"reload(Pmw.Blt)".  However, this workaround cannot be the right way to do
things, and it seems like it is occasionally creating problems, as evidenced
in the next example.

The second problem is illustrated in the following simple script:
#########################################
from Tkinter import *
import Pmw
t1 = Tk()
reload(Pmw.Blt)
g1 = Pmw.Blt.Graph(t1); g1.pack()
t2 = Tk()
l = Label(t2, text='A label'); l.pack()
b = Button(t2, text='Press me'); b.pack()
e = Pmw.EntryField(t2, labelpos=W, label_text='Enter something')
e.pack()
#########################################

When this script is run, the first Tk() window (t1) and the Pmw.Blt.Graph
(g1) appear OK.  The second Tk() window (t2) is created, and the Label and
Button widgets are created.  However, when the Pmw.EntryField is being
created the following error occurs:

Traceback (most recent call last):
  File "U:\Rivers\python\epics\testBlt2.py", line 9, in ?
    e = Pmw.EntryField(t2, labelpos=W, label_text='Enter something')
  File "C:\PROGRA~1\Python21\Pmw\Pmw_0_8_5\lib\PmwEntryField.py", line 72,
in __init__
    sequences = root.bind_class(tag)
  File "C:\PROGRA~1\Python21\lib\lib-tk\Tkinter.py", line 923, in bind_class
    return self._bind(('bind', className), sequence, func, add, 0)
  File "C:\PROGRA~1\Python21\lib\lib-tk\Tkinter.py", line 858, in _bind
    return self.tk.splitlist(self.tk.call(what))
TclError: bad window path name ".45079580.45079148"

Any suggestions will be most appreciated!

Thanks,

              Mark Rivers
              Dept. of Geophysical Sciences, University of Chicago
              GSECARS, Advanced Photon Source, Argonne National Laboratory




From bokr at oz.net  Thu Jun 27 13:04:26 2002
From: bokr at oz.net (Bengt Richter)
Date: 27 Jun 2002 17:04:26 GMT
Subject: "multi" zip
References: <lrGS8.13892$bt.307@nwrddc04.gnilink.net> <NIGS8.326499$352.35968@sccrnsc02>
Message-ID: <affgiq$13n$0@216.39.172.122>

On Thu, 27 Jun 2002 16:05:01 GMT, "Emile van Sebille" <emile at fenx.com> wrote:

>Mark
>> 1) instead of producing tuples, I'd like to produce lists.
>>
>> 2) I'd like to extend this to an arbitrary number of lists.  If it
>were
>> possible, I'd do zip(line[0], line[1], ..., line[n]) but this doesn't
>seem
>> to compile :).
>>
>> Is there a way to make zip() do this, or should I stop trying to
>hammer with
>> a wrench?
>>
>
>
>>>> line = [range(ii*10,11*10+5) for ii in range(5) ]
                         ^^-- assuming you meant ii
>>>> map (list,zip(*line))
>
Nice.

Regards,
Bengt Richter


From wweexxsseessssaa at telusplanet.net  Sun Jun 30 23:12:39 2002
From: wweexxsseessssaa at telusplanet.net (John Hall)
Date: Mon, 01 Jul 2002 03:12:39 GMT
Subject: Q: Status of MySQL & Python?
Message-ID: <69hvhu4otkrcncmu0d4vudfhv74o7maiul@4ax.com>

I'm a python newbie, recently discovered PythonCard, and, perhaps
unwisely with so little experience, am considering PythonCard & MySQL
on Windows 2000 Pro for a small DB aplication.

I have Python22, wxWindows, PythonCard & MySQL (just today) apparently
working OK.

When trying to run the dbBrowser from Samples, MySQLdb/__init__.py
does a traceback in line #27 because _mysql is not found. Sure 'nuff,
I don't have that. I do have souce C source: _mysql.c

Q1: Do I need to do something to build _mysql from something?
I can't find one ready to roll on the 'net.

q2: I get the impression that support for MySQL with Python is less
robust and up-to-date for Windows than Linux et al. Is it feasible and
wise to persue this?


-- 
John W Hall <wweexxsseessssaa at telusplanet.net>
Calgary, Alberta, Canada.
"Helping People Prosper in the Information Age"


From jepler at unpythonic.net  Wed Jun  5 08:37:07 2002
From: jepler at unpythonic.net (Jeff Epler)
Date: Wed, 5 Jun 2002 07:37:07 -0500
Subject: Pop a list from beginning ? and memory saving...
In-Reply-To: <3CFD8747.4AA8D29F@engcorp.com>
References: <adcpdg$vui53$1@ID-146704.news.dfncis.de> <3CFA0676.68F80DE6@engcorp.com> <add289$vud42$1@ID-146704.news.dfncis.de> <Xns9221CE2AF76D0cliechtigmxnet@62.2.16.82> <adhrsl$118kvn$1@ID-146704.news.dfncis.de> <mailman.1023194347.14669.python-list@python.org> <3CFD5B60.90831F21@engcorp.com> <aIeL8.120588$Kp.12235125@bin7.nnrp.aus1.giganews.com> <3CFD86D4.C4F81404@engcorp.com> <3CFD8747.4AA8D29F@engcorp.com>
Message-ID: <20020605073707.B25640@unpythonic.net>

On Tue, Jun 04, 2002 at 11:36:39PM -0400, Peter Hansen wrote:
(speaking about this expression):
>> l[:] = []
> 
> Oh, and in my opinion, that is obscure and deserves a
> comment explaining how it works, and why it is being done.
> I might not feel that way if I write that idiom a few times,
> though. :)

IMO
    del l[:]
is clearer than
    l[:] = []
I'd accept the first one immediately, and puzzle over the second one for
a short (or even long) time.

Jeff




From whisper at oz.net  Thu Jun 13 17:14:49 2002
From: whisper at oz.net (David LeBlanc)
Date: Thu, 13 Jun 2002 14:14:49 -0700
Subject: Type subclassing: bug or feature
In-Reply-To: <aeau48$111$1@panix1.panix.com>
Message-ID: <GCEDKONBLEFPPADDJCOEIEPFDIAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of Aahz
> Sent: Thursday, June 13, 2002 13:09
> To: python-list at python.org
> Subject: Type subclassing: bug or feature
>
>
> Consider the following code:
>
> class MyStr(str):
>     def contains(self, value):
>         return self.find(value) >= 0
>
> s = MyStr("hello, world!")
> s = s.capitalize()
> if s.contains('Hello'):
>     print "Found it!"
>
> It fails with an AttributeError when it calls s.contains(), because
> s.capitalize() returned a str instead of a MyStr.  Anyone want to take a
> whack at defending this as the correct behavior?
> --
> Aahz (aahz at pythoncraft.com)           <*>
> http://www.pythoncraft.com/

Looks like a bug to me. How can self be something else then what it is?
Should MySocket return a socket, thus stripping off any additional
functionality added by my derived class?

David LeBlanc
Seattle, WA USA





From ken at hotmail.com  Tue Jun  4 09:37:32 2002
From: ken at hotmail.com (Ken)
Date: Tue, 4 Jun 2002 23:37:32 +1000
Subject: Automated Email - What did I do wrong here?
References: <adg1do$10nmr5$1@ID-49758.news.dfncis.de>
Message-ID: <adifmg$11i5ks$1@ID-49758.news.dfncis.de>

"Ken" <ken at hotmail.com> wrote in message
news:adg1do$10nmr5$1 at ID-49758.news.dfncis.de...
> How to do automated emailing when results being processed?
>
> Thanks
>

What did I do wrong? It crash the program. Without these code, the cgi
script works fine. Thanks

    fromaddr = 'ken at hotmail.com'
    toaddrs = 'ken at hotmail.com'
    msg = "testing\n"
    server = smtplib.SMTP('smtp.aol.com')
    server.sendmail(fromaddr, toaddrs, msg)
    server.quit()




From kragen at pobox.com  Mon Jun 10 04:16:40 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 10 Jun 2002 04:16:40 -0400
Subject: Looking for a list subclassing example...
References: <adi4dv$118tko$1@ID-146704.news.dfncis.de>
Message-ID: <83bsaj4kgn.fsf@panacea.canonical.org>

"Shagshag13" <shagshag13 at yahoo.fr> writes:
> I'm looking for a list subclassing example, and i can't find it...
> 
> Do you have one to show ?

Why do you think it's a good idea to subclass 'list'?

I don't have one to show, but I think it's generally not a good idea.



From pyth at devel.trillke.net  Sun Jun 30 06:57:05 2002
From: pyth at devel.trillke.net (holger krekel)
Date: Sun, 30 Jun 2002 12:57:05 +0200
Subject: Python needs better error reporting
In-Reply-To: <3D1E7655.29975.167945D@localhost>; from opus@value.net on Sun, Jun 30, 2002 at 03:09:09AM -0700
References: <3D1E5376.32716.DF5C94@localhost>; <20020630095830.C20310@prim.han.de> <3D1E7655.29975.167945D@localhost>
Message-ID: <20020630125705.A10625@prim.han.de>

Opus wrote:
> Interesting idea.  Would these statistics be built from normal usage of the 
> language or from that person's usage?

that'd certainly be a configuration option.  But i guess common usage of 
python-expressions/tokens is not that different.

My personal try would just pump the standard lib modules through the
AST/Token-statistic profiler.

    holger




From kemusoft at hotmail.com  Tue Jun 11 09:49:38 2002
From: kemusoft at hotmail.com (kemu)
Date: Tue, 11 Jun 2002 15:49:38 +0200
Subject: frame.bind ?
Message-ID: <3d05fff1$0$8149$ba620e4c@news.skynet.be>

frame.bind("<Button-1>", callback)

I don't understand what .bind does ?
I know button-1 is the first mouse button and he calls the function callback
but what does bind do ??


__________________________________________________________________ jonas
Geiregat ICQ#: 115695917 Current ICQ status: + More ways to contact me
__________________________________________________________________




From bokr at oz.net  Thu Jun 20 22:19:17 2002
From: bokr at oz.net (Bengt Richter)
Date: 21 Jun 2002 02:19:17 GMT
Subject: Raw String Output
References: <aetnkf$tot$1@news.umbc.edu>
Message-ID: <aeu2f5$lno$0@216.39.172.122>

On Thu, 20 Jun 2002 19:09:41 -0400, "mike" <gringro1 at umbc.edu> wrote:

>If I have a string variable with the value '\n' how do I output it to a file
>with the value of '\012' instead ?
>
First, let me get straight what you mean vs what I think you might mean ;-)
If you want to write a single line feed character with code value 012 (10 decimal),
don't worry about it, just write the string (in binary on windows, or it will become
\r\n in the file). Note that:

 >>> ord('\n'), ord('\012'), ord('\x0a')
 (10, 10, 10)

and

 >>> 'one\ntwo\012three\x0a'
 'one\ntwo\nthree\n'
 >>> print 'one\ntwo\012three\x0a'
 one
 two
 three

To see the string in terms of the character sequence a file write would see, convert
to a list:

 >>> list('one\ntwo\012three\x0a')
 ['o', 'n', 'e', '\n', 't', 'w', 'o', '\n', 't', 'h', 'r', 'e', 'e', '\n']

I.e., the actual internal \n, \012,and \x0a codes are the same single characters.
OTOH, if you need to pass custom escaped strings to some other platform or software
for evaluation there, that would be different.

If you type
    s1 = '\n'
You will get an internal string object with length one bound to the 's1' on the left.

If you type
    s2 = '\012'
or
    s3 = '\x0a'
You will get string objects with the identical value (even sharing the identical immutable single-char
string object in current Python, I believe), but bound to 's2'. The numeric character code will be 10
in decimal for all three, which you can verify with ord():
 >>> s1='\n'; s2='\012'; s3='\x0a'

When you inspect s1 or s2 by just typing the names interactively, the interactive loop will
use the __repr__ methods of the respective strings and print the charaters returned.
 >>> s1,s2,s3
 ('\n', '\n', '\n')
It's the same character, so convential repr shows it as '\n'

 >>> ord(s1),ord(s2),ord(s3)
 (10, 10, 10)

"'\n'" is a string literal, which if you evaluate it will give you a length-1
string whose single character will have a numerical code with value 10 decimal.

If you write (in binary mode) any of the strings s1, s2 or s3 to a file, you will only write
one character, and it will be the same character, with the ascii value 10.

If for some reason you want to write a _representation_ of the internal encoded string, you
have choices. The conventional representation is produced by repr(s) or s.__repr__, or `s`,
and line feed will be represented as '\n' on the screen, which comes from two characters:
'\\' and 'n'. To create an alternate _representation_ for the same internal character, using
four characters (i.e., your '\\' '0','1','2') instead of two, you could produce the conventional
representation and substitute '\\012' for '\\n'. ('\\012' and '\\n' can be written r'\012' and r'\n').

Here is an example of text with two line feeds:
 >>> s="""\
 ... line 1
 ... line 2
 ... """

Triple quotes keeps the line feeds, and the convential string
representation printed shows the \n characters escaped:
 >>> s
 'line 1\nline 2\n'

Backquotes are short for calling for a repr() or __repr__ representation string,
which will have the actual backslash escape characters as such paired with the plain
characters they are escaping. When shown on the screen, _that_ string with its back-slashes
is repr'd and printed, so the backslashes are shown doubled:
 >>> sr1 = `s`
 >>> sr1
 "'line 1\\nline 2\\n'"

If you want to change the _representation_ r'\n' of chr(10) to _representation_ r'\012',
you can split out the former and join in the latter:
 >>> sr1.split(r'\n')
 ["'line 1", 'line 2', "'"]
 >>> sr2 = r'\012'.join(sr1.split(r'\n'))
 >>> sr2
 "'line 1\\012line 2\\012'"

Now compare the conventional representation sr1 and your sr2:

 >>> print sr1
 'line 1\nline 2\n'
 >>> print sr2
 'line 1\012line 2\012'

Note the single quotes printed. They are part of the repr string (though they could have been
double quotes if the text included single quotes, so don't depend on the quoting character's being
"'" or '"'. You might conceivably want to strip the outer quotes off, depending on the destination
of your custom string representation.

And note that if you evaluate the representation strings, you get the original
identical string back:
 >>> eval(sr1)
 'line 1\nline 2\n'
 >>> eval(sr2)
 'line 1\nline 2\n'

as the original
 >>> s
 'line 1\nline 2\n'

And naturally if you print them, they print the same:
 >>> print eval(sr1)
 line 1
 line 2

 >>> print eval(sr2)
 line 1
 line 2

Now the question is, do you want to write a string to a file
where it takes one character to represent EOL (or two if windows
cooks it in text mode), or do you actually want to write a representation
string that has to be eval'd to retrieve the original internal string?
And if the latter, why would you want r'\012' instead of r'\n' ?

Again, the issue of abstractions vs representations that play different
and related roles is the key.

>Thank you in advance!
>
HTH

Regards,
Bengt Richter


From cliechti at gmx.net  Thu Jun 13 17:33:33 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 13 Jun 2002 23:33:33 +0200
Subject: Shrinking Python
References: <mailman.1023999624.27508.python-list@python.org>
Message-ID: <Xns922CF0108507Fcliechtigmxnet@62.2.16.82>

"Mahrt, Dallas" <dallasm at aiinet.com> wrote in
news:mailman.1023999624.27508.python-list at python.org: 

> I am looking into using Python in a small device with limited space
> for binaries. Due to this, I am both investigating creating a shared
> library with the interpreter and shrinking the footprint of this
> library. I have found many threads talking bout shared libraries, so I
> am not too concerned about that aspect. 
> 
> I am concerned about shrinking the library's binary footprint. I did
> find several links to a "Deeply Embedded Python" based on 1.5.1 which
> has since vanished. I have stripped the library to improve size some
> (to ~1.13 MB) but would like to reduce it further. Has anyone tackled
> this or even attempted? Are there any links to resources that may help
> me that I haven't found? 

http://pippy.sourceforge.net/ seems to use 137k (app) + 300k (lib, VM) on 
my palm. its a stripped down 1.5.2

chris

-- 
Chris <cliechti at gmx.net>



From tdelaney at avaya.com  Thu Jun 27 20:24:29 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 28 Jun 2002 10:24:29 +1000
Subject: Most elegant python program to date.
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A3CF@natasha.auslabs.avaya.com>

> From: PoulsenL at capecon.com [mailto:PoulsenL at capecon.com]
> 
> examples of its prose.  Are there any nominations for the most elegant
> example of a python program to date.  Not sure under what 

print 'Hello, world!'

1. Does what it is designed to do.

2. No extra cruft.

3. Does not suffer any performance penalties dues to poor algorithms.

4. Does not complicate itself by attempting to optimise algorithms which
don't need to be.

Of course, this is more a reflection of the elegance of Python itself ...

public class HelloWorld
{
    public static void main (String argv[] )
    {
        System.out.println("Hello, world!");
    }
}

Tim Delaney




From gcordova at hebmex.com  Thu Jun 27 16:41:53 2002
From: gcordova at hebmex.com (Gustavo Cordova)
Date: Thu, 27 Jun 2002 15:41:53 -0500
Subject: How to find out operating system
Message-ID: <F7DB8D13DB61D511B6FF00B0D0F06233015DE85E@mail.hebmex.com>

> 
> C:\>python
> Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit 
> (Intel)] on win32
> Type "help", "copyright", "credits" or "license" for more information.
> >>> import os
> >>> os.environ.has_key('WINOS')
> 1
> >>> os.environ['WINOS']
> 'WIN2000'
> >>>
> 

Why check if maybe "WINOS" is in the environment?
I've personally never seen that variable. Better
check for "WINDIR", or for "COMSPEC"; the first
is, well, the directory under which Windows is
installed, and I've seen that variable in all
the versions I've used: Win95, 98, ME, NT, 2000;
the other one points to the command shell
for that version of windows.

Good luck :-)

-gus




From lewy0lewy at poczta.onet.pl  Mon Jun 17 05:41:35 2002
From: lewy0lewy at poczta.onet.pl (Pawel Lewicki)
Date: Mon, 17 Jun 2002 11:41:35 +0200
Subject: MySQL database location
Message-ID: <aekasd$96t$1@news.onet.pl>

Hi,
Is is possible to get the physical path to the MySQL database file using
MySQLdb or any other pythonous way?

Pawel Lewicki




From anthony at interlink.com.au  Thu Jun 13 01:19:32 2002
From: anthony at interlink.com.au (Anthony Baxter)
Date: Thu, 13 Jun 2002 15:19:32 +1000
Subject: using m2crypto to encrypt credit card numbers 
In-Reply-To: Message from Graham Ashton <gashton@cmedltd.com> 
   of "12 Jun 2002 10:52:48 +0100." <1023875569.603.44.camel@walter> 
Message-ID: <200206130519.g5D5JWO10417@localhost.localdomain>

>>> Graham Ashton wrote
> If you get too
> much fraud (I think the limit was 40% fraud by value in the UK, but my
> memory is hazy) going through your site you can get your merchant ID
> status revoked by the banks (i.e. you get your transaction processing
> account chopped).

The stuff they get really shitty about is charge-backs. If you can detect
the fraud and do the refunds yourself before the end-user notices, you can
get away with it... 

I hate the credit card system. As designed, it's completely insecure, and
the merchants end up copping it in the neck because of the appalling design.
Since the banks and the card networks don't take any losses, they have 
little or no incentive to FIX the bloody problem. Anyone remember SET?
It was going to save all our lives, make the world a good and great thing.
And that was in 1996. Still waiting for it...

Anthony
-- 
Anthony Baxter     <anthony at interlink.com.au>   
It's never too late to have a happy childhood.





From bokr at oz.net  Sat Jun 29 05:16:44 2002
From: bokr at oz.net (Bengt Richter)
Date: 29 Jun 2002 09:16:44 GMT
Subject: I'm an idiot
References: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16>
Message-ID: <afjtts$h1h$0@216.39.172.122>

On Sat, 29 Jun 2002 01:00:22 GMT, David <david_griswold1 at yahoo.com> wrote:

>OK, I am the first to admit it.  I am an idiot.  I have RTFM on this over 
>and over, and I can still not figure out what I am doing wrong.
>
IMHO the key thing to learn from all the responses is not the solution to the
particular problem, but some hints as to how you can figure things out for yourself.

That is one of the nice things about Python. You can effectively ask it what is
happening as you walk through your problem interactively.

>I think the intent of the code is obvious, but just to clarify, I want to 
>read every line in a file and write those lines back out to another file, 
>but with leading and training space removed.  I also want to have some 
>elegant way to determine that I have reached the end of the file and 
>break out of the loop.
>
>And just to explain my stupidity, my reference langauge is BASIC.  I 
>could have written this in BASIC in a minute, but I need to learn 
>something new.  Any help would be appreciated.
>
>David
>
>
>f=open('c:\\temp\\temp.txt', 'r')
>g=open('c:\\temp\\temp1.txt', 'w')
>while 1:
>    try:
>        s=f.readline
>        g.write(s.split())
>    except IOError:
>        break
>
>g.close
>f.close

Ok, why doesn't that work, you probably asked yourself.
First thing, did you have a known temp.txt file? You'd want
one that will demonstrate that things are working, so make one up,
and store it where you are specifying it to be e.g.:

--< temp.txt >--
   Three leading spaces and not trailing.
Second line with four trailing spaces.    
Third line has no spaces around it, and has blank line following.

	Last line, with leading tab and ending with newline.
--

So first thing is opening the file
 >>> f=open('c:\\temp\\temp.txt', 'r')
Now what is f at this point? You can check to make sure:
 >>> f
 <open file 'c:\temp\temp.txt', mode 'r' at 0x007CF3A0>

Well, that seems to have worked.

Now before putting it in a loop, let's just try the read statement:
 >> s=f.readline

You expect a string, so check to make sure:
 >>> s
 <built-in method readline of file object at 0x007CF3A0>
Aha. Not what you expected. Most built-in stuff has doc strings, so here you can
print either s.__doc__ or f.readline.__doc__, since they both refer to the
same thing at this point:
 >>> print f.readline.__doc__
 readline([size]) -> next line from the file, as a string.

 Retain newline.  A non-negative size argument limits the maximum
 number of bytes to return (an incomplete line may be returned then).
 Return an empty string at EOF.

We want the whole line here, and we'll trust that input lines are
reasonable length, so we'll want f.readline(), which we can test:
(and we'll pretend we didn't see that bit about EOF)

 >>> s = f.readline()
What did we get?
 >>> s
 '   Three leading spaces and not trailing.\n'
Looks good so far. Now what about s.split() ? Let's see:

 >>> s.split()
 ['Three', 'leading', 'spaces', 'and', 'not', 'trailing.']

Hm, that doesn't look ready to write out. Must be another string method
that does the job. dir(s) will tell us what other s.xxx methods there are
associated with s (here s is bound to a string, so we'll see string methods):

 >>> dir(s)
 ['__add__', '__class__', '__contains__', '__delattr__', '__eq__', '__ge__', '__getattribut
 e__', '__getitem__', '__getslice__', '__gt__', '__hash__', '__init__', '__le__', '__len__'
 , '__lt__', '__mul__', '__ne__', '__new__', '__reduce__', '__repr__', '__rmul__', '__setat
 tr__', '__str__', 'capitalize', 'center', 'count', 'decode', 'encode', 'endswith', 'expand
 tabs', 'find', 'index', 'isalnum', 'isalpha', 'isdigit', 'islower', 'isspace', 'istitle',
 'isupper', 'join', 'ljust', 'lower', 'lstrip', 'replace', 'rfind', 'rindex', 'rjust', 'rst
 rip', 'split', 'splitlines', 'startswith', 'strip', 'swapcase', 'title', 'translate', 'upp
 er']

Whoa, that line wrapped. lstrip, rstrip, and strip look like the obvious candidates, with
the last likely to do the job. So check it out:

You can print the docstring info:
 >>> print s.strip.__doc__
 S.strip() -> string

 Return a copy of the string S with leading and trailing
 whitespace removed.

Or with >= version 2.1 you can:
 >>> help(s.strip)
 Help on built-in function strip:

 strip(...)
     S.strip() -> string

     Return a copy of the string S with leading and trailing
     whitespace removed.

Ok, let's make sure it works:
 >>> s.strip
 <built-in method strip of str object at 0x007A1E30>
Oops. We know about that one now. Need a () to execute the method ...
 >>> s.strip()
 'Three leading spaces and not trailing.'

s is still there for comparison, since we have not rebound it with s=s.strip() or something:
 >>> s
 '   Three leading spaces and not trailing.\n'

Well, the leading spaces diasppeared ok, but note that the newline also disappeared with .strip().
The virtue of trying things out ;-) So if we want to write out the stripped strings
as lines, we'll have to add the '\n' back on.

Perhaps we are ready to try that:

 >>> g=open('c:\\temp\\temp1.txt', 'w')
 >>> g
 <open file 'c:\temp\temp1.txt', mode 'w' at 0x0083D0A0>

Make sure it's what we think:
 >>> s.strip()+'\n'
 'Three leading spaces and not trailing.\n'
Ok, write it
 >>> g.write(s.strip()+'\n')

No complaint, so we might as well do the rest:

 >>> s = f.readline()
 >>> s
 'Second line with four trailing spaces.    \n'
 >>> s.strip()+'\n'
 'Second line with four trailing spaces.\n'
 >>> g.write(s.strip()+'\n')
 >>> s = f.readline()
 >>> s
 'Third line has no spaces around it, and has blank line following.\n'
 >>> s.strip()+'\n'
 'Third line has no spaces around it, and has blank line following.\n'
 >>> g.write(s.strip()+'\n')
 >>> s = f.readline()
 >>> s
 '\n'
 >>> s.strip()+'\n'
 '\n'
 >>> g.write(s.strip()+'\n')
 >>> s = f.readline()
 >>> s
 '\tLast line, with leading tab and ending with newline.\n'
 >>> s.strip()+'\n'
 'Last line, with leading tab and ending with newline.\n'
 >>> g.write(s.strip()+'\n')
 >>> s = f.readline()
 >>> s
 ''
Oops, what was that?
 >>> help(f.readline)
 Help on built-in function readline:

 readline(...)
     readline([size]) -> next line from the file, as a string.

     Retain newline.  A non-negative size argument limits the maximum
     number of bytes to return (an incomplete line may be returned then).
     Return an empty string at EOF.

Aha. (we should've remembered about EOF from before ;-)
 >>> f.close
 <built-in method close of file object at 0x007A5180>
;-)
 >>> f.close()
 >>> g.close()

Check results:
 >>> h=open('c:\\temp\\temp1.txt', 'r')
 >>> h.read()
 'Three leading spaces and not trailing.\nSecond line with four trailing spaces.\nThird lin
 e has no spaces around it, and has blank line following.\n\nLast line, with leading tab an
 d ending with newline.\n'

That wrapped, but looks ok, so now use the knowledge gained to revise your original.

My point here is just to show that you need not be at a loss when a program as
a whole doesn't work. Just break it down and check on what your code is actually
doing. After a while, just a well-placed print statement or two to check on
intermediate results will usually tell you enough.

To do that, you probably want to use an editor to edit a file representing the program,
and then either run it from the editor, if your editor supports that, or run it from
a separate command line window, and switch back and forth between editor and cmd line window.

When you want to be even more systematic about testing your code, you can set up automatic
testing to verify expected results. This is prudent once things get past trivial and you
need to make sure evolving improvements don't break existing functionality. There are modules
to help with that (see doctest & unittest).

You can always keep an extra console window open with python running to try little snippets
and type out help docs etc. Python makes it easy to explore what's really happening.
Then if you're stumped, you can post an excerpt from a session showing what you tried. ;-)

HTH

Regards,
Bengt Richter


From david.abrahams at rcn.com  Tue Jun 18 18:36:29 2002
From: david.abrahams at rcn.com (David Abrahams)
Date: Tue, 18 Jun 2002 18:36:29 -0400
Subject: SWIG and Callbacks
References: <3d0a8b6e@nntp.server.uni-frankfurt.de> <aee6m9$qop$1@bob.news.rcn.net> <3d0b4fd6@nntp.server.uni-frankfurt.de> <aeftmj$m29$1@bob.news.rcn.net> <3d0c6d67@nntp.server.uni-frankfurt.de> <3D0CE698.1050806@SPAMnwinternet.com> <aeiu3a$4ff$1@bob.news.rcn.net> <3D0D3223.5070500@SPAMnwinternet.com> <aekrd1$4ao$1@bob.news.rcn.net> <3D0E2606.5050001@SPAMnwinternet.com> <aen6ch$d5u$1@bob.news.rcn.net> <3D0F807A.40203@SPAMnwinternet.com>
Message-ID: <aeoco5$adc$1@bob.news.rcn.net>

"Jim" <jbublitzNO at SPAMnwinternet.com> wrote in message
news:3D0F807A.40203 at SPAMnwinternet.com...
> David Abrahams wrote:
> > "Jim" <jbublitzNO at SPAMnwinternet.com> wrote in message
>
> >>Template-based code does require special treatment with sip.
> >>In most cases where I've wrapped template-based types or
> >>classes, I manually write C++ code for conversion C++ <-> Py
> >>(sip has some functions that make that easier) and implement
> >>it as a sip 'mapped type', which lets me refer to the same
> >>conversion code from any return value or arg.
>
> > That's a given with Boost.Python; with v2 conversions are automatically
> > registered in a global repository so they can be re-used automatically
> > across extension modules.
>
> For C++ code with a lot of templates Boost.Python has an
> advantage, particularly if you're early on the learning curve.

Well, yes, but I wasn't talking about that. I was talking about re-usability
of conversions. My point was that if you wrap a class X in one module, you
can can wrap [member] functions that operate on X in another module.

> Sure - you can extend sip in almost any way you want also
> (the bindings generated are C++ code), but it may be a
> little trickier fitting things in the binding's framework.
> It's also possible to embed custom Python or C++ code in
> sip description files and I do that when appropriate (rarely).
> It kind of goes against the design philosophy of sip IMHO,
> so I try to avoid it personally.

Understood.

> >>The real problem wouldn't seem to be C++ types as such, but
> >>finding a useful (to the Python programmer) representation
> >>via PyObject, and still maintaining the range of abilities
> >>built into the C++ code (esp things like subclassing).
>
> > That's built-in to Boost.Python. V1 had to implement a complicated
Python
> > class look-alike itself; V2 uses Python's new-style classes so
subclassing
> > "just works". Making virtual functions overridable in Python requires a
> > small amount of boilerplate on the part of the programmer. I often wish
I
> > had a good C++ parser front-end which I could use to automate that part.
>
> New-style classes lock you into Python 2.2+ though - I've
> got users still at 1.5.2 and place a lot of emphasis on long
> term version support.

That didn't seem to be important to my users. The amount of code I was able
to throw away when I started using new-style classes was really staggering
(and gratifying!). If it was really important to some people, it should be
possible to re-implement all of the Python 2.2+ features used by
Boost.Python very much more cleanly in C++, for 1.5.2 compatibility... but
it's not a job I want to spend time on. As they say, there will be many more
people using future versions than have every used any existing version.
Anyway, those who really need it can always get v1 out of its archive and
dust it off.

> sip does the virtual code boilerplate'
> automatically, unless the method requires handwritten code
> for some other reason (rare). On the stuff I do, I see a lot
> more virtual code than templates, but that obviously varies
> a lot.

Yeah, that's important. Another area that Boost.Python could use some help
from a parser with is in wrapping overloaded functions and those with
default args. C++ is missing the ability to represent an overload set as a
single object ;-)

> > OK. The C++ code you're wrapping still has to get by with #ifdef, I
> > suppose...
>
> Depends on the code I suppose. Qt and KDE just issue all
> new files - no versioning in the C++ code.

Ah. I guess if they were wrapped with Boost.Python, the wrapping code would
be treated the same, then?

> > There is no "resorting to C++" with Boost.Python -- Boost.Python
*embraces*
> > C++! Since you're coding the wrappers in C++ anyway, it's trivial to
make a
> > "thin wrapper function" which calls the function being wrapped if you
want
> > to change the interface somehow. Then you just wrap the thin wrapper.
>
> I finally went back to your website - I took your 'like an
> IDL' comments a little too literally. IMHO, the fact that
> Boost.Python *is* C++ is a big advantage for people who
> want/need that. Personally, I don't want to write C++,
> which is how I got into this in the first place.

<smile> A lot of people seem to feel that way.

> Thin wrappers are easily implementable with sip too. In
> contrast to the very large packages like PyQt and PyKDE,
> they tend to be very lightweight and quick to build.

Sorry, I think you misunderstood what I meant by "thin wrapper". A classic
trivial example might be used to wrap a function with default args:

    // function to wrap
    void f(int a, char* b = expression-1, Foo const& c = expression-2);

Thin wrappers:

    void f1(int a, char* b) { f(a,b); }
    void f2(int a) { f(a); }

wrapping code:

    module("my_module")
        .def("f", f)
        .def("f", f1) // overloads handle
        .def("f", f2) // default args
    ;

Obviously you can do arbitrary pre- and post- processing inside a thin
wrapper function.

> >>Yes, but I don't write it - I just edit it mostly to correct
> >>mistakes my description file generator creates and to handwrite
> >>code where the generator has flagged that. Crude but effective (TM)
> >>Even without automation, the process is easiest if you start by
> >>saving the h file with a .sip extension and then (mostly) delete
> >>the things that sip doesn't want/need, so it's still mostly
> >>editing (sometimes A LOT of editing).
>
> > That seems to be the key difference. It's not clear wether it's really
an
> > advantage, but I suppose that time will tell.
>
> I doubt that it's ever clearly an advantage.

Really? It seems like it should lower the barrier to entry and get people
started with an iterative development process, at least.

> It works very
> well for PyQt and PyKDE because of the C++ style those projects
> use generally - very consistent,

That uniformity tends to be important to parser-based systems. I learned a
lot about the wide range of code people needed to wrap during v1
development, and I'm trying hard to lift restrictions imposed by that kind
of assumption for v2.

> relatively small amount of
> template usage, lots of enums and of course sip has features to
> support Qt. sip has been used outside of the Qt/KDE area, but
> perhaps not on anything available publicly yet.

Probably the docs are the main obstacle there. It's amazing how many
different wrapping tools there seems to be a market for; people will try
anything once <wink>!

> There's certainly plenty of opportunities for something like
> "benchmarketing".

Let's not, please!

> Either one of us could contrive examples to
> make one or the other look clearly better, but like most things
> it's probably more rational to look at what the user actually
> needs to accomplish. I've avoided actual code comparisons,
> because the simple cases tend to be one-sided one way or the
> other.

Well, code comparisons are important in showing the system's user-interface.

> >>I should probably go back and review Boost. I had thought that
> >>with Boost it was possible to wrap as much or as little of a
> >>C++ library as you want to.
>
> > That's true.
>
> Big selling point IMHO.

:-)

> >>I had the impression SWIG was supposed to do most of that in a single
>  >>step. Just wondering if I could implement something like that with
> Boost,
>  >>or need to code a line manually for each method/class/whatever.
>
> > If you have something which can mostly parse your headers, you can
generate
> > much of the wrapping code automatically.
>
> Yes - it looks like that should be possible. Virtual code
> might require a little more work in generating Boost.Python
> code automatically, but the tradeoff is probably less things
> flagged as requiring handwritten code.

Not sure what you mean here about the tradeoff. Example?

> > Yep. BTW, Boost is much more than just my Python binding library, so
please,
> > for the sake of the other authors: "Boost.Python" when discussing my
lib.
>
> Noted - I've tried to do better in this post.

Thanks.

> > Not sure that you'll find much to steal; they're so different in
approach...
>
> I can be very creative when stealing other people's code :)
>
> Actually "under the hood" sip and Boost.Python have similarities.

Yeah, well that page is slightly outdated, and doesn't really show much of
what's going on other than the class data structure. There really aren't
many other practical ways to make extension class wrappers.

-Dave




From cliechti at gmx.net  Tue Jun 25 14:43:29 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 25 Jun 2002 20:43:29 +0200
Subject: import precedence
References: <mailman.1025011774.28523.python-list@python.org>
Message-ID: <Xns9238D343DD4AFcliechtigmxnet@62.2.16.82>

"Mark McEahern" <mark at mceahern.com> wrote in
news:mailman.1025011774.28523.python-list at python.org: 

> Suppose I have a package like this:
> 
>   spam/
>     __init__.py
>     setup.py
>     foo.py
> 
> For cygwin, setup.py creates:
> 
>   foo.dll
> 
> For linux2, setup.py creates:
> 
>   foo.so
> 
> But for win32, setup.py is basically a no-op because I can do what I
> need to do in foo.py using pythoncom.
> 
> Here's the question:
> 
> Can I just assume that:
> 
>   from spam import foo
> 
> will only import the foo.py module if the foo.dll and foo.so are not
> found--that is, if the import statement is executed on win32 (where
> those extensions won't exist)?
> 
> I looked at the documentation for import and it's not clear to me that
> it addresses this issue.  Perhaps because noone in their right mind
> would consider this approach?  ;-)

why don't you add an underline to the C extension, it's pretty common to do 
that. then in foo.py you can do:

try:
    	from _foo import *
except ImportError:
    	#get python implementation...

you could also choose the correct import by looking at os.name.
example from pyserial:

if os.name == 'nt': #sys.platform == 'win32':
    from serialwin32 import *
elif os.name == 'posix':
    from serialposix import *
elif os.name == 'java':
    from serialjava import *
else:
    raise "Sorry no implementation for your platform available."

chris

-- 
Chris <cliechti at gmx.net>



From gerhard at bigfoot.de  Wed Jun 19 01:10:04 2002
From: gerhard at bigfoot.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: Wed, 19 Jun 2002 07:10:04 +0200
Subject: how i use =?iso-8859-15?Q?'=E3?=
	=?iso-8859-15?Q?'?= in python???
In-Reply-To: <3D0F6623000017AA@www.zipmail.com.br>
References: <3D0F6623000017AA@www.zipmail.com.br>
Message-ID: <20020619051004.GA8897@lilith.my-fqdn.de>

* jubafre at zipmail.com.br <jubafre at zipmail.com.br> [2002-06-19 01:38 -0300]:
> i?m brazilian, and in my laguage(Portguese) there are many other caracters
> how ?, ?, ? and a string in python doesn?t support this
> how i can use??? have a import module for this???
> 
> BRASILEIROS DA LISTA COMO VCS CONSEGUEM USAR ACENTO NO PYTHON??
> 
> for exmple:
> s='?'
> doesn?t work?why??

You need to give more details than "doesn't work". Most probably there
is no problem with the statement

    s='?'
    
but a problem occurs when you try to print a string that contains a
non-ASCII character, right?

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))




From printers at sendme.cz  Fri Jun 21 03:34:05 2002
From: printers at sendme.cz (A)
Date: Fri, 21 Jun 2002 09:34:05 +0200
Subject: How to find all emails (on the same line)
Message-ID: <3D12F30D.2071.204353@localhost>

Hi,
I need to find ALL emails in a file. I can use re module but I do not 
know how to find all emails if there are more emails on the same 
line because re module ( as far as I know )finds only one/first 
occurrence on that line.
Thanks for help.
Ladislav






From whisper at oz.net  Fri Jun 28 22:30:12 2002
From: whisper at oz.net (David LeBlanc)
Date: Fri, 28 Jun 2002 19:30:12 -0700
Subject: I'm an idiot
In-Reply-To: <Xns923BB72B1BAF7davidgriswold1netsca@204.127.202.16>
Message-ID: <GCEDKONBLEFPPADDJCOEAEONDMAA.whisper@oz.net>

> -----Original Message-----
> From: python-list-admin at python.org
> [mailto:python-list-admin at python.org]On Behalf Of David
> Sent: Friday, June 28, 2002 18:00
> To: python-list at python.org
> Subject: I'm an idiot
>
>
> OK, I am the first to admit it.  I am an idiot.  I have RTFM on this over
> and over, and I can still not figure out what I am doing wrong.
>
> I think the intent of the code is obvious, but just to clarify, I want to
> read every line in a file and write those lines back out to another file,
> but with leading and training space removed.  I also want to have some
> elegant way to determine that I have reached the end of the file and
> break out of the loop.
>
> And just to explain my stupidity, my reference langauge is BASIC.  I
> could have written this in BASIC in a minute, but I need to learn
> something new.  Any help would be appreciated.
>
> David
>
>
> f=open('c:\\temp\\temp.txt', 'r')
> g=open('c:\\temp\\temp1.txt', 'w')
> while 1:
>     try:
>         s=f.readline
>         g.write(s.split())
>     except IOError:
>         break
>
> g.close
> f.close

Here's my version that also strips out blank lines:

#lineio.py
f=open('j:/python22/lineio.py', 'r')
g=open('j:/python22/lineio.txt', 'w')

for s in f:
        if s == '\n':
            continue
        g.write(s.strip() + '\n')

g.close()
f.close()

If you want to keep blank lines:

#lineio.py
f=open('j:/python22/lineio.py', 'r')
g=open('j:/python22/lineio.txt', 'w')

for s in f: g.write(s.strip() + '\n')

g.close()
f.close()

Dave LeBlanc
Seattle, WA USA





From uwe at rocksport.de  Fri Jun  7 05:35:07 2002
From: uwe at rocksport.de (Uwe Schmitt)
Date: 7 Jun 2002 09:35:07 GMT
Subject: Distributing Packages / os independent installation folder
References: <adpsrg$m02$03$1@news.t-online.com>
Message-ID: <adpuob$e29eb$3@hades.rz.uni-sb.de>

Achim Domma <achim.domma at syynx.de> wrote:
| Hi,

| - where should my package and shared librarie go to? On windows I put the
| dll into python22/DLLs and the rest into a folder
| python22/Lib/site-packages/ImageMagick. What to do on other plattforms? I
| don't expect a DLLs folder on Linux. ;-)

import sys
echo sys.path

... this might help you.

| - In my __init__.py I have to change one environment variable. Currently I
| do it like this:

| import os
| os.environ['MAGICK_HOME'] = r'C:\Python22\Lib\site-packages\ImageMagick'
| from _ImageMagick import *

| Is there an plattform independent way to get the installation folder of my
| package? On windows I can get the python folder from registry, but this will
| also break on linux.

maybe  os.getcwd()  in __init__.py does the job...

Greetings, Uwe


-- 
Dr. rer. nat. Uwe Schmitt      ICQ# 159647634   Uwe.Schmitt at num.uni-sb.de


From sholden at holdenweb.com  Tue Jun  4 16:57:00 2002
From: sholden at holdenweb.com (Steve Holden)
Date: Tue, 4 Jun 2002 16:57:00 -0400
Subject: Thought on PEP 204 and 276
References: <ckp4fuc4fskjcpilnn2mrt3665fqvld92g@4ax.com> <mailman.1022528744.19859.python-list@python.org> <gm17fu40rh6r675g3a6d1ovl49tb9pd822@4ax.com> <mailman.1022594566.27265.python-list@python.org> <slrnaf7mo4.dfe.philh@comuno.freeserve.co.uk> <6w4L8.229234$%u2.120502@atlpnn01.usenetserver.com> <j36qfu4r84j8320iegrujjf8dpvdvs20bm@4ax.com>
Message-ID: <YU9L8.240282$%u2.115403@atlpnn01.usenetserver.com>

"Stephen Horne" <steve at lurking.demon.co.uk> wrote ...
> On Tue, 4 Jun 2002 10:49:19 -0400, "Steve Holden"
> <sholden at holdenweb.com> wrote:
>
> >"phil hunt" <philh at comuno.freeserve.co.uk> wrote in message
> >news:slrnaf7mo4.dfe.philh at comuno.freeserve.co.uk...
>
> >> I think they are essentially the same. A sequence tpye is just a
> >> mapping type where the keys are restricted to the integers from
> >> 0 to however many elements there are minus 1.
> >>
> >No matter what you think, they are essentially different. Insertion into
a
> >mapping doesn't alter the key to which other elements are mapped, but
> >insertion into a sequence (potentially) does that.
>
> What is normally referred to as insertion in dictionaries is actually
> more like overwriting a list element. This is why handling both with
> the 'container [key] = value' syntax is natural.
>
aarrgghh. So we can completely ignore the fact that the len() of a
dictionary will change after an insertion while the len() of a list won't
after an element is overwritten?

> Insertion with modification of existing keys to make room (ie list
> insertion) simply does not exist in lists. The usefulness of list
> insertion derives from the fact that a list is a very specialised type
> of key-to-value mapping - in this respect, the type 'list' should
> clearly inheret from and extend the more general key-to-value type
> called a dictionary.
>
It might be clear to you. It certainly isn't to me. I believe you are
looking to unify two fundamentally different semantics.

Neither does it make sense in implementation terms. List insertion's
usefulness is in fact limited to situations where performance isn't unduly
degraded by the need to modify the representation too radically. Of course,
where performance doesn't matter, it doesn't matter.

> In essence, you are confusing the operations on the underlying data
> structures with the operations on the conceptual types. These are two
> separate concepts. The implementation is distinct from the
> user-visible abstraction.
>
Fine. You write your programs to delete lists element-by-element from the
front, and don't complain about performance -- that's just an aspect of the
implementation.

While I admire your attempt to maintain theoretical purity here, Python is a
pragmatist's language. I don't think I'll be bothering to conceptualise
dictionary insertion as overwriting a non-existent element, thank you very
much.

pragmatical-ly y'rs  - steve
--
-----------------------------------------------------------------------
Steve Holden                                 http://www.holdenweb.com/
Python Web Programming                http://pydish.holdenweb.com/pwp/
-----------------------------------------------------------------------








From ken at hotmail.com  Sat Jun  1 07:10:39 2002
From: ken at hotmail.com (Ken)
Date: Sat, 1 Jun 2002 21:10:39 +1000
Subject: Python and HTML frames
References: <ada6hf$vico3$1@ID-49758.news.dfncis.de> <r08ada.b7b.ln@10.0.0.1>
Message-ID: <ada9ur$vcigh$1@ID-49758.news.dfncis.de>

"Jiri Baum" <jiri at baum.com.au> wrote in message
news:r08ada.b7b.ln at 10.0.0.1...
> Ken:
> > How do I output HTML onto frames? This code below doesn't work. Can
> > someone help me?
>
> This isn't really a python question, it's a HTML question (or possibly
> HTTP; are you ouputting the right headers?). However, I should note that
> setting scrolling to "no" is a really, really bad idea unless the frame
> contains just an image, because not all people have the same-sized fonts.
>
> I trust you've read the guidelines for using frames at www.useit.com ?
> Briefly: don't, and if you have to, make all links TARGET="_top".

Why you think it isn't python question? I want that HTML frame output from
the python script. Can you tell me how?

Thanks




From K.Rdt at TU-Berlin.DE  Fri Jun 28 01:12:09 2002
From: K.Rdt at TU-Berlin.DE (Klaus Reinhardt)
Date: Fri, 28 Jun 2002 07:12:09 +0200
Subject: crypt for windows?
In-Reply-To: <slrnahmisf.3eo.gerhard.haering@lilith.my-fqdn.de>
Message-ID: <53VT85VUA6GBXVTQOOKPOIDWRJEQKTQ.3d1bf029@FRITZweb>

Am 27.06.02 19:29:52, schrieb Gerhard H?ring <gerhard.haering at gmx.de>:

>> Is there a crypt for windows-python, so
>> I can manage some password-files?
>
>Probably not. I believe Cygwin's Python has this module. Is that an
>option for you?
---------------------------------------------------------------------
Hi

No, I dont like cygwin. I tried pgp, but it's to
complicated. So I decided to aks me for
password each session. For this I can use
python in 'the next script'.

		K at Rdt
---------------------------------------------------------------------







From peter at engcorp.com  Sat Jun 22 20:33:34 2002
From: peter at engcorp.com (Peter Hansen)
Date: Sat, 22 Jun 2002 20:33:34 -0400
Subject: position of matches
References: <653cf871.0206221452.6cb17ef2@posting.google.com>
Message-ID: <3D15175E.74934C43@engcorp.com>

les wrote:
> 
> i am new to python and want to do the following:
> given a string str='abcABCDefsgRSTVderDae'
> i would like to find all the internal lowercase strings between 2 caps
> i.e. pattern=re.compile(r'[A-Z]([a-z]+)[A-Z]')
> 
> match_obj=pattern.search(str)
> begin,end=match_obj.span()
> 
> however i would like to get all the begining and end positions
> of the pattern,
> i.e.
> efsg  begin=7 end=10
> der   begin=15 end 17

Try this:
>>> for x in m.finditer(s):
...   print '%s\tbegin=%s end=%s' % ((x.group(1),) + x.span(1))
...
efsg    begin=7 end=11
der     begin=15 end=18

Note that you might want to adjust the "end" index to fit Python's
view of the world.  Python generally uses the index of a "slice"
made to the string *between* two characters rather than referring
to the index of one of the characters.  This lets you easily use
the slice notation as below.  (This explains why my code shows 11
and 18 instead of 10 and 17 for the end position.)  A very recent 
thread explained this in more detail.

>>> s[7:11]
'efsg'
>>> s[15:18]
'der'

(Note: I also used 's' above instead of 'str' since 'str' is
a builtin in 2.2.  Safer not to use it.)

-Peter


From tim.one at comcast.net  Mon Jun  3 16:32:02 2002
From: tim.one at comcast.net (Tim Peters)
Date: Mon, 03 Jun 2002 16:32:02 -0400
Subject: encoding and decoding repr'd version of strings
In-Reply-To: <1YNK8.9622$cE6.591635@wagner.videotron.net>
Message-ID: <LNBBLJKPBEHFEDALKOLCIECJPKAA.tim.one@comcast.net>

[Daniel Parisien]
> is there an codec that bundles with python so you can transform strings
> into the repr'd (escaped) version and back?

Not now; there may be in 2.3.  Getting the escaped version is trivial, i.e.
that's what repr(string) has always done.  It's the other direction that's
missing.

> If not, how easy would it be to code?

It's a matter of rearranging code that already exists (at C level).


> I don't want to use eval(...) because I don't trust the  source of the
> string and I might mistakenly eval an expression like "' '*(2**30)" which
> would allocate about 1 GB of data (eek!)

Use eval anyway, but first pass the putative string literal through
tokenize.py to ensure that it really is a string literal.





From theller at python.net  Wed Jun 26 06:57:19 2002
From: theller at python.net (Thomas Heller)
Date: Wed, 26 Jun 2002 12:57:19 +0200
Subject: 'parent object'?
References: <3d18228c$0$221$4d4ebb8e@news.nl.uu.net>
Message-ID: <afc6mm$d7gs1$1@ID-59885.news.dfncis.de>

> Is there, like 'self, also a 'parent' object of some sort?
>
> The problem i'm having is that i have a class, which contains a list of
> instances of another class. How can i let this 'childclass' call a function
> which is on the 'parent class'?
>

You could check out Environmental Acquisition:
http://www.zope.org/Members/acquisition/

Thomas




From tim.one at comcast.net  Tue Jun 25 11:14:22 2002
From: tim.one at comcast.net (Tim Peters)
Date: Tue, 25 Jun 2002 11:14:22 -0400
Subject: Thread Memory Leak
In-Reply-To: <pan.2002.06.25.12.37.12.535288.17223@pobox.com>
Message-ID: <BIEJKCLHCIOIHAGOKOLHMEELDFAA.tim.one@comcast.net>

[Marco Catunda]
> I've developed a python program with some threads.
> This program is a daemon server. I have seen the
> memory of that program growing very fast in a short
> period of time. So I decided to make a little program
> to see what's happen.
> ...

Which version of Python, which OS, and which thread package?  There's no
leak evident when running your program on Win2K using any of Python 2.1.3,
2.2.1, or current CVS, so whatever you're seeing is specific to something
you haven't told us.





From jdhunter at nitace.bsd.uchicago.edu  Wed Jun 12 17:49:32 2002
From: jdhunter at nitace.bsd.uchicago.edu (John Hunter)
Date: Wed, 12 Jun 2002 16:49:32 -0500
Subject: map -> class instance
Message-ID: <m2r8jcp3pv.fsf@video.paradise.lost>

I want to convert dictionary instances to instances of a class that
has the keys of the map as attributes.  I naively tried this:

class Map2Class:
    def __init__(self,m):
        self.m = m

    def __getattr__(self, key):
        return self.m[key]

    def __setattr__(self, key, val):
        self.m[key] = val

    def __delattr__(self, key):
        if self.m.has_key(key):
            del self.m[key]

m = {'first' : 'John',
     'last' : 'Hunter',
     'age' : 34}

c = Map2Class(m)
print c.first, c.last, c.age

But got an infinite recursion:

~/python/test $ python map_to_class.py 
Traceback (most recent call last):
  File "map_to_class.py", line 20, in ?
    c = Map2Class(m)
  File "map_to_class.py", line 3, in __init__
    self.m = m
  File "map_to_class.py", line 8, in __setattr__
    self.m[key] = val
  File "map_to_class.py", line 5, in __getattr__
    return self.m[key]
  File "map_to_class.py", line 5, in __getattr__
    return self.m[key]
        [snip ... lots more line fivers ]

Is there a better/right way to do what I want?

Thanks,
John Hunter


From chris.gonnerman at newcenturycomputers.net  Thu Jun 13 10:09:20 2002
From: chris.gonnerman at newcenturycomputers.net (Chris Gonnerman)
Date: Thu, 13 Jun 2002 09:09:20 CDT
Subject: Poll: Readable languages
In-Reply-To: <Pine.LNX.4.44.0206130853001.20311-100000@adams.patriot.net>
Message-ID: <200206131409.JAA05450@scotland.k12.mo.us>

On Thu, 13 Jun 2002 09:36:06 -0400 (EDT), Attila Horvath wrote:

> In my years of experience I have come across VERY readable assembler and
> very unreadable COBOL, PASCAL, FORTRAN, ADA, C/C++, JAVA, etc. So I
> suggest that a langauge does not inherently make resultant code readable.

True, and false.  It is possible to write bad code in any language; but in my
opinion, any programmer who has mastered Python (takes what, a couple of
weeks? :-) *tends* to write more readable code than an equivalently 
accomplished programmer in any other language.

Those of us who have been here a while have learned from experience of the
ease of writing Python code, and reading it again later.  I have to admit, most
of the Python I write is little stuff to do jobs I would have done with Bourne
shell in the past... but I tend to save them for reuse as I would not have done
with a command-line pipeline.




From R.Barrett at ftel.co.uk  Mon Jun 24 10:51:52 2002
From: R.Barrett at ftel.co.uk (Richard Barrett)
Date: Mon, 24 Jun 2002 15:51:52 +0100
Subject: I'd give up Perl tomorrow if only...
In-Reply-To: <3D17214F.5000902@destiny.com>
Message-ID: <5.1.0.14.2.20020624152755.00ad2d88@pop.ftel.co.uk>

At 09:40 24/06/2002 -0400, Michael Chermside wrote:
>Michael Chermside wrote:
> > Where can I find the pyperl module?
>
>Gerhard H?ring responded:
>>Google - Enter "pyperl" - then hit "I feel lucky" :-)
>>This will bring you to
>>http://www.cpan.org/modules/by-module/LWP/GAAS/pyperl-1.0.readme which 
>>looks to
>>a non-Perl guy like me that its available as a CPAN module.
>
>Well, I had tried that, but (not being a perl kind of guy) I'm not really 
>sure what a CPAN module is or how to obtain or download it. I DID find 
>some version of pyperl at ActiveState (yeah, ActiveState!), but couldn't 
>(in the brief amount of time I spent on it) figure out how to set it up.
>
>What I WAS able to figure out has been written up at 
>http://www.faqts.com/knowledge_base/view.phtml/aid/17202/fid/1102
>If anyone else can help more, let me know.
>
>-- Michael Chermside

I was following this thread with interest. As a result I downloaded 
http://downloads.activestate.com//Zope-Perl/pyperl-1.0.1.tar.gz last Friday.

Having unpacked the .tar.gz I did just what it said in the README:

<quote>
Build instructions:

   - make sure your PATH is set up so that 'perl' and 'python'
     reference the versions of the language interpreters that you
     want to use.

   - If you are using Python-1.5.2, then you need to install
     the Distutils package version 0.9 or better first.

   - Run these commands:

       (cd Python-Object; perl Makefile.PL; make install)
       python setup.py install

   - You should now be able to run the test.py test script.

       python test.py

</quote>

The only snag was quickly fixed by renaming the file MULTI_PERL to 
xMULTI_PERL to indicate that my Perl installation hadn't been build with 
-Dusethreads option.

All working in less than half an hour on my Linux desktop with Python 2.2.1 
and Perl 5.6.1.

Say 'import perl' in your Python code and you have a pretty seamless way of 
using Perl modules from Python. Wonderful for using that legacy code and 
particularly for using a Perl module for which you cannot find a Python 
equivalent ready to hand.

As an old Perl hand who these days much prefers Python but cannot always 
escape history, I am mightily impressed. That said both Python and more 
recently Perl are pretty good at embedding and extending so I should not be 
as impressed as I am that it works so nicely.

While I've not given it a real thrashing yet, the implementation seemed 
pretty solid to me.

Many thanks to the folks at ActiveState who I gather did the embedding work.

Definitely recommended for the compleat script-writer's use. All I have to 
do now is see how it all interacts with Tk/Tcl :)





From gumuz at looze.net  Tue Jun 25 04:01:53 2002
From: gumuz at looze.net (Guyon Morée)
Date: Tue, 25 Jun 2002 10:01:53 +0200
Subject: 'parent object'?
Message-ID: <3d18228c$0$221$4d4ebb8e@news.nl.uu.net>

Hi,

Is there, like 'self, also a 'parent' object of some sort?

The problem i'm having is that i have a class, which contains a list of
instances of another class. How can i let this 'childclass' call a function
which is on the 'parent class'?

maybe i am thinking wrong and is there a better way to setup this class
structure, please tell me if it's so.

thanx,

Guyon




From rchopra at calypso.cs.brandeis.edu  Mon Jun 17 00:16:36 2002
From: rchopra at calypso.cs.brandeis.edu (Rajat Chopra)
Date: Mon, 17 Jun 2002 00:16:36 -0400
Subject: System Configuration
Message-ID: <Pine.LNX.4.44.0206170010400.15076-100000@calypso.cs.brandeis.edu>

I am looking to create a utility program that will query various aspects 
of my system, such as the hardware configuration and the software programs 
installed. 

I have looked into the Python module list but could not find much to do
with system analysis. At the moment, I would like my program to function
only on Windows 2000 systems. What modules can I use to determine such
things as:

-How much memory is installed in the system?
-How much disk space is available on the system?
-What is the processor speed of the system?
-What is the bus speed?
-Is Micrsoft Office installed on the system?
-What version of Internet Explorer is on the system?
-Certain registry actions, such as inserting a key, querying a key, etc.

Thank you in advance for your help.

Rajat



From mdtorre at freemail.it  Mon Jun 24 04:49:08 2002
From: mdtorre at freemail.it (Matteo)
Date: 24 Jun 2002 01:49:08 -0700
Subject: What language - platform use to data acquisiton/numerical work?
References: <1c849ea9.0206130130.6cdaa36f@posting.google.com> <1c849ea9.0206210457.36d494db@posting.google.com> <yfsk7osfszk.fsf@black132.ex.ac.uk>
Message-ID: <1c849ea9.0206240049.472b0d4a@posting.google.com>

Alexander Schmolck <a.schmolck at gmx.net> wrote in message news:<yfsk7osfszk.fsf at black132.ex.ac.uk>...

> If you are interested I could send you
> the code, but you'll need python2.2.1 and it might still have rough edges.

I wold be grateful if you can send me the code, 

Thank you very much

Matteo Della Torre


From dev at newdeal.ch  Fri Jun 28 09:19:11 2002
From: dev at newdeal.ch (erreur)
Date: Fri, 28 Jun 2002 15:19:11 +0200
Subject: How to write accent in WinXP Command Prompt ?
Message-ID: <3d1c628a@news.swissonline.ch>

How to write accent in WinXP Command Prompt ?

It's ok with the Python interpretor, but if I run sript <<print "???">>
write <<???>>



thanks





From news at nospam.majid.fm  Wed Jun 26 20:32:43 2002
From: news at nospam.majid.fm (Fazal Majid)
Date: Wed, 26 Jun 2002 17:32:43 -0700
Subject: Installation problems DCOracle2 & W2000
References: <mailman.1025115815.15029.python-list@python.org>
Message-ID: <3D1A5D2B.1070906@nospam.majid.fm>

brueckd at tbye.com wrote:
> Aha! Here's what I think the problem is: the DCOracle2 bundle off the Zope 
> site comes all wrapped up in the Zope database adapter (ZOracleDA), so 
> your above script is actually trying to import the DA and not the 
> DCOracle2 package itself (I can see why this is confusing - both 
> directories are named DCOracle2 but one is a subdirectory of the other!).

I just reinstalled Oracle 9.2.0.1 + Python 2.2 + DCOracle2 on a new 
computer so I can reproduce the steps to a correct install. I think Dave 
is right on track here, there is a name conflict between the module 
DCOracle2.py, which is standalone without Zope and that you want to use, 
and DCOracle2 the Zope database adapter package directory which is used 
only from within Zope.

Here is my setup:

PYTHONPATH=D:\Etc\DCOracle2

D:\Etc\DCOracle2>python
Python 2.2.1 (#34, Apr  9 2002, 19:34:33) [MSC 32 bit (Intel)] on win32
Type "help", "copyright", "credits" or "license" for more information.
 >>> import DCOracle2
 >>> d=DCOracle2.connect('scott/tiger at database')
 >>> d.execute('select SYSDATE from dual')
1
 >>> d.fetchall()
[[OracleDate("2002-06-26 17:31:57")]]
 >>> ^Z
D:\Etc\DCOracle2>dir
  Volume in drive D is MAJID
  Volume Serial Number is B8C2-4135

  Directory of D:\Etc\DCOracle2

2002-06-11  14:47       <DIR>          .
2002-06-11  14:47       <DIR>          ..
2002-06-05  15:00               73,728 dco2.pyd
2002-06-05  12:16               56,957 DCOracle2.py
2002-06-05  15:00              180,304 win32-python-1.5.2-805-dco2.pyd
2002-06-05  15:00               73,728 win32-python-1.5.2-dco2.pyd
2002-06-05  15:00              180,298 win32-python-2.1-805-dco2.pyd
2002-06-05  15:00               73,728 win32-python-2.1-dco2.pyd
2002-06-05  15:00               73,728 win32-python-2.2-dco2.pyd
2002-06-05  12:16                1,725 __init__.py
2002-06-26  17:21               46,293 DCOracle2.pyc
                9 File(s)        760,489 bytes
                2 Dir(s)   6,778,052,608 bytes free



From gerhard.haering at gmx.de  Thu Jun 27 13:33:32 2002
From: gerhard.haering at gmx.de (Gerhard =?iso-8859-15?Q?H=E4ring?=)
Date: 27 Jun 2002 17:33:32 GMT
Subject: How to find out DNS ?
References: <mailman.1025197535.3041.python-list@python.org>
Message-ID: <slrnahmj3b.3eo.gerhard.haering@lilith.my-fqdn.de>

A wrote in comp.lang.python:
> Hi,
> Is there a way how to find out, from Python , what primary or 
> secondary DNS I use when connecting to internet?

I don't think so. There might be platform specific ways to get that
info, though. On Unixen, for example, you can parse /etc/resolv.conf
for lines starting with "nameserver".

On Windows, your best bet is to read MSDN on how to do it and try to
get that info with the win32 extensions.

Gerhard
-- 
mail:   gerhard <at> bigfoot <dot> de       registered Linux user #64239
web:    http://www.cs.fhm.edu/~ifw00065/    OpenPGP public key id AD24C930
public key fingerprint: 3FCC 8700 3012 0A9E B0C9  3667 814B 9CAA AD24 C930
reduce(lambda x,y:x+y,map(lambda x:chr(ord(x)^42),tuple('zS^BED\nX_FOY\x0b')))


From mbelanger_NOSPAM at NOSPAM_pixelsystems.com  Tue Jun 25 12:05:53 2002
From: mbelanger_NOSPAM at NOSPAM_pixelsystems.com (Mathieu Belanger)
Date: Tue, 25 Jun 2002 12:05:53 -0400
Subject: Switch from 2.1 to 2.2
References: <f29c2586.0206250600.48e0a8e3@posting.google.com> <3D188302.1030109@thomas-guettler.de>
Message-ID: <zv0S8.3036$%N4.731526@news20.bellglobal.com>

> Plese send the stracktrace, too
>

Yeah, I know it will be easier, but I can't have a complete one for any
reason !?!

So here it is:

Traceback (most recent call last):
  File "D:\WINNT\system32\RxScriptFile.py", line 551, in DCMgmtScript
  File "D:\Python22\Lib\site-packages\_xmlplus\dom\ext\__init__.py", line
82, in PrettyPrint
    Printer.PrintWalker(visitor, root).run()
  File "D:\Python22\Lib\site-packages\_xmlplus\dom\ext\Printer.py", line
385, in run
    return self.step()
  File "D:\Python22\Lib\site-packages\_xmlplus\dom\ext\Printer.py", line
381, in step
    self.visitor.visit(self.start_node)
  File "D:\Python22\Lib\site-packa

Thanks




From garrett at bgb.cc  Wed Jun 19 01:27:30 2002
From: garrett at bgb.cc (Don Garrett)
Date: Wed, 19 Jun 2002 05:27:30 GMT
Subject: What If.....    Strong Types
References: <3D0C0F0D.4030900@bgb.cc> <20020616.175421.861584889.1250@localhost.localdomain>
Message-ID: <3D1015B9.207@bgb.cc>

Lamy Jean-Baptiste wrote:
> I've done really nice hack by changing object class, adding method or
> properties, and so on... !

   Can you give some examples? I really would just like to understand what 
people use this stuff for. I'm not trying to imply that people shouldn't. I'd 
just like to have some concrete examples to help me understand.

-- 
Don Garrett                             http://www.bgb.cc/garrett/
BGB Consulting                                      garrett at bgb.cc



From tdelaney at avaya.com  Thu Jun 20 23:00:57 2002
From: tdelaney at avaya.com (Delaney, Timothy)
Date: Fri, 21 Jun 2002 13:00:57 +1000
Subject: GOTO w/ Python?
Message-ID: <B43D149A9AB2D411971300B0D03D7E8BF0A3AC@natasha.auslabs.avaya.com>

> From: John Roth [mailto:johnroth at ameritech.net]
> 
> "Max M" <maxm at mxm.dk> wrote in message news:3D11A21D.9010809 at mxm.dk...
> >
> > What do you mean???
> >
> > Goto's are simple ;-)

[Clever, silly, not real GOTO example snipped]

> Those aren't goto's in the classical sense. For a real goto,
> you need to think assembler: if it's a statement anywhere,
> you can branch to it, and who cares about the execution
> history! Having the compiler manage your flow control is
> for sissys!

Methinks someone missed something ...

Tim Delaney




From SBrunning at trisystems.co.uk  Fri Jun 28 07:57:45 2002
From: SBrunning at trisystems.co.uk (Simon Brunning)
Date: Fri, 28 Jun 2002 12:57:45 +0100
Subject: GUI-Control for HTML
Message-ID: <31575A892FF6D1118F5800600846864DCBD462@intrepid>

> From:	Thomas Guettler [SMTP:zopestoller at thomas-guettler.de]
> Is there are GUI-control which can display HTML?
> 
> The HTML that will be used will be simple, no javascript or
> form support is needed.
> 
> Is there are cross-plattform solution? (Unix + Win32)
 
See
<http://www-106.ibm.com/developerworks/linux/library/l-wxpython/?open&l=968,
t=gr>.

Cheers,
Simon Brunning
TriSystems Ltd.
sbrunning at trisystems.co.uk




-----------------------------------------------------------------------
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorised. If you are not the intended recipient, any disclosure,
copying, distribution, or any action taken or omitted to be taken in
reliance on it, is prohibited and may be unlawful. TriSystems Ltd. cannot
accept liability for statements made which are clearly the senders own.




From dfackrell at DELETETHIS.linuxmail.org  Thu Jun 20 11:46:26 2002
From: dfackrell at DELETETHIS.linuxmail.org (Daniel Fackrell)
Date: Thu, 20 Jun 2002 09:46:26 -0600
Subject: Function to add commas in numbers?
References: <Xns923372EFBBC5Dmdkmdkmdkmdk@130.133.1.4>
Message-ID: <3d11f8d2$1_1@hpb10302.boi.hp.com>

mdk,

The following thread is about formatting numbers for Danish, but it should
give the info you need (paste it as one line):

http://groups.google.com/groups?hl=en&lr=&ie=UTF8&oe=UTF8&threadm=3D0A1791.3
060307%40mxm.dk&rnum=5&prev=/groups%3Fas_q%3Dnumber%2520format%26ie%3DUTF8%2
6oe%3DUTF8%26as_ugroup%3Dcomp.lang.python%26lr%3D%26hl%3Den

--
Daniel Fackrell (dfackrell at linuxmail.org)
When we attempt the impossible, we can experience true growth.

"mdk" <mdk at mdk.com> wrote in message
news:Xns923372EFBBC5Dmdkmdkmdkmdk at 130.133.1.4...
> Hello,
>
> Is there a function that will take, for example, 1234567 and return
> 1,234,567?
>
> Thanks




From paul at boddie.net  Tue Jun 25 06:57:01 2002
From: paul at boddie.net (Paul Boddie)
Date: 25 Jun 2002 03:57:01 -0700
Subject: ? and %s placeholders, help?
References: <af5oq7$sj1$1@news8.svr.pol.co.uk> <m2k7oolcrm.fsf@video.paradise.lost> <3D178542.604@mxm.dk> <af87cj$n5b$1@newsg3.svr.pol.co.uk>
Message-ID: <23891c90.0206250257.23d38d30@posting.google.com>

"Duncan Smith" <buzzard at urubu.freeserve.co.uk> wrote in message news:<af87cj$n5b$1 at newsg3.svr.pol.co.uk>...
> Thanks both,
>                    But I'm still having problems with this.  I want to use
> prepared statements to avoid unnecessary parsing, and I get the impression
> (from what information I can find, eg Python Programming on Win32) that I
> should (need to?) use the '?' placeholder to achieve this.  But I cannot
> come up with anything that works.

First, check the paramstyle of the MySQLdb module.

  # This generally works with database modules.

  print MySQLdb.paramstyle

The parameter notation to be used - informally known as the paramstyle
- should be printed. Only if "?" is printed can you actually use that
particular notation. (There's been some discussion about the merits of
having loads of paramstyles on the DB-SIG mailing list, but it would
arguably be a lot easier if "?" were in use throughout, just like with
JDBC.)

> >>> query = 'INSERT INTO %s (%s) VALUES (%s)' % (tblname, string.join(vars,
>  ', '), string.join(['%s']*len(vars), ', '))
> >>> query
>  'INSERT INTO tbl (var1, var2, var3) VALUES (%s, %s, %s)'
> >>> curs.execute(query % ("'a'", "'b'", "'c'"))
> 1L
> 
> #Hurray, but does this avoid parsing the statement on each INSERT?

This isn't what you ought to be doing, since you're effectively doing
Python string substitution to supply the values. Moreover, bad data
could be used to "attack" your application or to "exploit" it, should
you employ these methods.

> >>> query = 'INSERT INTO %s (%s) VALUES (%s)' % (tblname, string.join(vars,
>  ', '), string.join(['?']*len(vars), ', '))
> >>> query
>  'INSERT INTO tbl (var1, var2, var3) VALUES (?, ?, ?)'

You're doing the right thing when building the query string, since the
table and column names cannot be passed in as parameters - therefore
you are right to substitute them in first. The only potential problem
is whether the parameter notation is correct - check out the
paramstyle to be sure, and modify the '?' to '%s' in the statement
above if the paramstyle turns out to be '%s'.

> >>> curs.execute(query, ("'a'", "'b'", "'c'"))
> Traceback (most recent call last):
>   File "<interactive input>", line 1, in ?
>   File "C:\Python22\Lib\site-packages\MySQLdb\cursors.py", line 70, in
> execute
>     raise ProgrammingError, m.args[0]
> ProgrammingError: not all arguments converted
> >>>

Be aware that by supplying "'a'", the actual inserted value will be
"'a'" and not "a" - you don't need to quote parameter values. Apart
from this, the above looks correct, and I suspect that the cause of
the error is the paramstyle that you've chosen being incorrect.

To summarise:

  1. Check the paramstyle.
  2. Build the query (or statement, in this case). Put the table and
     column names in here, along with the parameter "placeholders"
     according to the paramstyle.
  3. Execute the query (or statement) supplying the parameter values,
     remembering that you don't need to quote the parameter values.

Paul


From tom at huno.net  Fri Jun 28 15:42:56 2002
From: tom at huno.net (thomas)
Date: Fri, 28 Jun 2002 21:42:56 +0200
Subject: feeding a prompt with python
References: <3D1C5341.9070001@huno.net> <3D1C7CA0.7060201@ivs.tu-berlin.de>
Message-ID: <3D1CBC40.6090604@huno.net>

Andreas Ulbrich wrote:
> 
> What you actually want to do is using pipes. See os.popen and related 
> stuff for documentation.
> 

thx! exactly what i was looking for.



From pixie888 at hotmail.com  Thu Jun  6 08:36:04 2002
From: pixie888 at hotmail.com (pixie888 at hotmail.com)
Date: Thu, 06 Jun 2002 12:36:04 GMT
Subject: check on numerical value?
Message-ID: <3cff56b4.511834453@news.skynet.be>

Hi,

Is there any way to discover when a value is numerical or not in
Python? I want a check in my code on a given value. If somebody enters
or gives me something like "akjlvq?irq" this is surely not a numerical
value.

Thanks,

Henk


From johnroth at ameritech.net  Sat Jun 29 18:28:38 2002
From: johnroth at ameritech.net (John Roth)
Date: Sat, 29 Jun 2002 18:28:38 -0400
Subject: private
References: <pan.2002.06.29.17.38.51.789527.440@myself.com> <3D1DE132.4A890D9D@engcorp.com> <afl3e4$f8abc$1@ID-71831.news.dfncis.de>
Message-ID: <uhsd24138ahh67@news.supernews.com>

"James Kew" <james.kew at btinternet.com> wrote in message
news:afl3e4$f8abc$1 at ID-71831.news.dfncis.de...
> "Peter Hansen" <peter at engcorp.com> wrote in message
> news:3D1DE132.4A890D9D at engcorp.com...
> > Rhymes wrote:
> > >
> > > Is there any possibility to build up a _real_ private attribute?
> > > Such as C++ private data members...
> >
> > You realize that it's quite possible to get at private data members
> > in C++, don't you?
>
> I can't help feeling this is overly dismissive!
>
> Yes, of course one can circumvent private in C++. But for me the main
value
> of the public and private specifiers is that they _document_ to the
client
> of the class what it should access and what it should not.
>
> I don't -- in my admittedly very limited experience -- find Python
classes
> as immediately self-documenting. It's not obvious from the class
definition
> what is intended to be called or modfied by the client and what is
internal
> to the implementation.
>
> I suspect I'm still in the crossover period from strictly declarative
C/C++
> and the "anything goes" Pythonic way and that, given time, I'll find
my
> groove and feel comfortable.
>
> > This has been discussed about 479 times in the last decade.
>
> I'll bet. I'll shut up about it now.
>
> James
>
>
>




From fredrik at pythonware.com  Sun Jun 16 16:06:16 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Sun, 16 Jun 2002 20:06:16 GMT
Subject: PIL and miscolored images
References: <3D09D25A.4C04E03@ipm.fhg.de>
Message-ID: <Yc6P8.12489$p56.3857317@newsb.telia.net>

Markus von Ehr wrote:
> I'd like to show a greylevel image as miscolored image,
> anyone has an idea how i can do it and maybe adjust
> the color scale?

do you mean pseudocoloring?  something like this should
work:

    from PIL import Image

    im = Image.open("some grayscale image")
    im.load() # make sure it's loaded into memory

    assert im.mode == "L"

    # create a lookup table (r, g, b, r, g, b, r, g, b, ...)
    lut = []
    for i in range(256):
        lut.extend([255-i, i/2, i])
    im.putpalette(lut)

    assert im.mode == "P" # now has a palette

    im.save("out.gif")

the "load" call is a workaround for a bug in 1.1.3 and earlier.

without it, you will sometimes get a ValueError exception when
attempting to save (or further process) the image.

</F>




From fperez528 at yahoo.com  Thu Jun 20 23:52:01 2002
From: fperez528 at yahoo.com (Fernando =?ISO-8859-1?Q?P=E9rez?=)
Date: Thu, 20 Jun 2002 21:52:01 -0600
Subject: newbie string question
References: <aeu40s$9jg2s$1@ID-145503.news.dfncis.de> <3D129BB7.C7EA15D2@engcorp.com> <aeu76k$9brq1$1@ID-145503.news.dfncis.de>
Message-ID: <aeu7t3$58f$1@peabody.colorado.edu>

Don Low wrote:

> So if I understand, pystr[3] refers to the subscript index, but pystr[3:]
> refers to the slice index. pystr[2:5] means slice at slice index 2 and 5,
> not start at subscript index 2 and slice at slice index 5.
> 
> I just want to understand this once and for all.

The easiest way to understand this is to build a list which is a range. There 
you have a one-to-one mapping between list indices and elements. Then slice 
it to your heart's content and the behavior of : and friends will become 
second nature:

In [1]: x=range(10)

In [2]: x
Out[2]: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]

In [3]: x[5]
Out[3]: 5

In [4]: x[3:7]
Out[4]: [3, 4, 5, 6]

In [5]: x[:]
Out[5]: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]

In [6]: x[5:]
Out[6]: [5, 6, 7, 8, 9]

In [7]: x[-1]
Out[7]: 9

In [8]: x[-5]
Out[8]: 5

In [9]: x[-5:]
Out[9]: [5, 6, 7, 8, 9]


cheers,

f


From xah at xahlee.org  Sat Jun 15 04:03:33 2002
From: xah at xahlee.org (Xah Lee)
Date: 15 Jun 2002 01:03:33 -0700
Subject: html parser etc help
References: <7fe97cc4.0206111445.43b196de@posting.google.com> <ae63pm$47b95$1@ID-114079.news.dfncis.de>
Message-ID: <7fe97cc4.0206150003.65fe8821@posting.google.com>

Thanks Jeremy Yallop for help. Some follow up questions...

in the following code:

import os 
def visit(arg, dirname, fnames): 
    for file in fnames: 
        if file.endswith('.html'): 
            print file 
            print arg 
os.path.walk('/export/home/xah/unixnotes/',visit,'---')


* in the os.path.walk, what's the third argument for?
* what if i want to print the full path?
* what if i want to print a particular type of file, for example only
directories.

* how to find out about a particular function or method, or what
modules are available?
in perl i'd do "perldoc -tf functionname" to find out about a
particular function. Or
, or in Java there's the API doc.
http://java.sun.com/j2se/1.3/docs/api/index.html

Thanks.

 Xah
 xah at xahlee.org
 http://xahlee.org/PageTwo_dir/more.html


From ken at hotmail.com  Sun Jun  2 06:01:19 2002
From: ken at hotmail.com (Ken)
Date: Sun, 2 Jun 2002 20:01:19 +1000
Subject: How to call functions inside a cgi script from the html interface?
References: <adc4sv$vjrdc$2@ID-49758.news.dfncis.de> <2bkcda.c25.ln@10.0.0.1>
Message-ID: <adcq91$vmqig$1@ID-49758.news.dfncis.de>

"Jiri Baum" <jiri at baum.com.au> wrote in message
news:2bkcda.c25.ln at 10.0.0.1...
> Ken:
> > How to call functions inside a cgi script from the html interface?
>
> > Eg. If there is a link on the html display call "Search", how do I get
it
> > to call the function call "def search( )" inside the cgi script?
>
> You need to get yourself a CGI programming reference, or even a good HTML
> book, and refer to them before you post questions on the (wrong)
newsgroup.
>
> Also read the docs for the python cgi module (eg at www.python.org).
>
> HTTP doesn't have any facilities for doing this directly. Every time a
user
> selects a link, a request is made to the server for that page. This can
> either be a completely different cgi script, or it can be the same script
> with different parameters. Either way, it's a separate connection, so
don't
> expect variables to stay.

Where can I find good references with both python and HTML explained in
detail on how they work together?




From cliechti at gmx.net  Wed Jun 26 15:32:00 2002
From: cliechti at gmx.net (Chris Liechti)
Date: 26 Jun 2002 21:32:00 +0200
Subject: printing (again!!)
References: <afcusp$66r$1@newsg2.svr.pol.co.uk> <mailman.1025117834.9485.python-list@python.org>
Message-ID: <Xns9239DB7E46291cliechtigmxnet@62.2.16.82>

Martin Franklin <mfranklin1 at gatwick.westerngeco.slb.com> wrote in
news:mailman.1025117834.9485.python-list at python.org: 
> Get hold of the excelent win32 extension for python (url????)

http://starship.python.net/crew/mhammond/

-- 
Chris <cliechti at gmx.net>



From kragen at pobox.com  Mon Jun 17 01:44:49 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 17 Jun 2002 01:44:49 -0400
Subject: Args not set in extension class
References: <aefcav$8ef$1@reader05.wxs.nl> <aefs7n$4cc$1@reader09.wxs.nl> <83sn3otlkf.fsf@panacea.canonical.org> <aehngd$22l$1@reader06.wxs.nl>
Message-ID: <83bsaampbi.fsf@panacea.canonical.org>

"Gorny" <gorny0 at zonnet.nl> writes:
> That doesn't seem to work as well for some vague reason... It complains
> it's not a valid tuple

I suspect you are overlooking something significant in your source
code, but I don't know what it is without seeing the code.


From kwokng at earthlink.net  Sat Jun  1 12:37:57 2002
From: kwokng at earthlink.net (Billy Ng)
Date: Sat, 01 Jun 2002 16:37:57 GMT
Subject: Copying a database
Message-ID: <FL6K8.771$NG1.217@newsread2.prod.itd.earthlink.net>

I need you guys help on mysql again.  I need to push my python program to
production now, but I don't want to recreate the mysql database on the
production box.  I want to copy the database from the development box to
production box.. Would anybody please tell me how to do it, thanks!

Billy Ng





From kragen at pobox.com  Sat Jun  1 15:59:41 2002
From: kragen at pobox.com (Kragen Sitaker)
Date: 01 Jun 2002 15:59:41 -0400
Subject: symbolic python
References: <3CEFAE4C.38A8022B@doc.ic.ac.uk> <3CEFD881.D7992095@doc.ic.ac.uk> <3cf8ce86$1@giga.realtime.net>
Message-ID: <83sn46vkgi.fsf@panacea.canonical.org>

"Ira Baxter" <idbaxter at semdesigns.com> writes:
> If you want program transformation capability,
> you can get it in the form the of DMS Software Reengineering
> Toolkit.
> See http://www.semdesigns.com/Products/DMS/DMSToolkit.html.

It sounds very interesting.  How can I get a copy?



From fredrik at pythonware.com  Thu Jun 20 03:50:23 2002
From: fredrik at pythonware.com (Fredrik Lundh)
Date: Thu, 20 Jun 2002 07:50:23 GMT
Subject: SNIP IT: tk.Canvas Drag Items Around
References: <63604d2.0206031518.2da29d3b@posting.google.com> <63604d2.0206050957.60a47f1@posting.google.com>
Message-ID: <3PfQ8.44244$n4.10307657@newsc.telia.net>

"Phlip" wrote:

>     points = event.widget.coords (tk.CURRENT)
>     anchors = copy.copy(points[:2])
>     print points
>
>     for idx in range(len(points)):
> #        print idx, xy[idx % 2], anchors[idx % 2]
>         mouse = xy[idx % 2]
>         zone = anchors[idx % 2]
>         points[idx] = points[idx] - zone + mouse
>
>     print points
>     apply(event.widget.coords, [tk.CURRENT] + points)

footnote: the canvas provides a nice little method called "move"
that adds X and Y offsets to all coordinates of all matching items:

    dx = mouse[0] - anchors[0]
    dy = mouse[1] - anchors[1]
    event.widget.move(tk.CURRENT, dx, dy)

for more info, see

http://www.pythonware.com/library/tkinter/introduction/canvas.htm
=> Methods

</F>